PON1 Q192R Variant and Clopidogrel Efficacy: Fact or Fiction?

Clopidogrel, in combination with aspirin is the standard of care for preventing ischemic cardiovascular events in coronary artery disease (CAD) patients, especially those who undergo percutaneous coronary intervention (PCI). Despite its widespread use, significant inter-individual variability in clopidogrel response is consistently observed, and recent studies suggest that as much as 70% of this variability can be attributed to genetic factors^{1, 2}.

Clopidogrel is a thienopyridine prodrug that once activated exerts its anti-platelet effect by irreversibly binding to the P2Y12 receptor on the surface of platelets inhibiting adenosine diphosphate (ADP)-stimulated platelet aggregation. Clopidogrel activation requires a two-step conversion by hepatic enzymes, most notably CYP2C19, while esterases (e.g. carboxylesterase 1 and carboxylesterase 2) lead to the production of biologically inactive carboxylic acid derivatives³. Common loss-of-function variants in CYP2C19 are associated with lower active metabolite levels, higher residual on-clopidogrel ADP-stimulated platelet aggregation, and cardiovascular outcomes³. However, this variant explains only about 12% of the variation in clopidogrel response leaving most of the variation due to genes unidentified.

Recently, Bouman and colleagues observed, through in vitro studies, that paraoxonase 1 (PON1) was the major enzyme in converting 2-oxo-clopidogrel to the active thiol metabolite⁴. Moreover, they reported that a common missense variant in PON1, Q192R, was a major determinant of clopidogrel efficacy, accounting for nearly 70% of the variability in ADP-stimulated platelet aggregation post-clopidogrel treatment. 192Q PON1 had lower hydrolysis efficiency for 2-oxo-clopidogrel compared to 192R PON1 in microsomal preparations (V_max/K_m = 0.36 vs. 1.36, respectively), and clopidogrel-treated PCI patients homozygous for the 192Q-allele had a significantly higher rate of stent thrombosis compared to patients carrying at least 1 copy of the 192R-allele (odds ratio [OR] = 3.6, 95% CI 1.6 – 7.9, P = 0.003). Curiously, no effect of the well-described CYP2C19*2 loss-of-function variant on clopidogrel response was observed in the same patient sample.

In the past year, our group and several others have attempted to replicate these striking results, with little success^5-12. In this issue of Circulation Cardiovascular Genetics, Paré and colleagues report efficacy and safety data regarding clopidogrel among PON1 Q192R genotypes in two of the largest prospective placebo-controlled trials studied to date: the Clopidogrel in Unstable Angina to Prevent Recurrent Events (CURE) trial¹³ and the Atrial Fibrillation Clopidogrel Trial with Irbesartan for Prevention of Vascular Events (ACTIVE A)¹⁴.

Consistent with the parent CURE trial report, in the subset of 5,059 acute coronary syndrome (ACS) patients in this genetic substudy (2,549 randomized to clopidogrel and 2,510 randomized to placebo), clopidogrel-treated patients had lower cardiovascular event rates than placebo-treated patients (HR = 0.70, 95% CI 0.59 – 0.84, P < 0.001). There was no evidence of association between 192Q PON1 and the composite primary cardiovascular outcome in clopidogrel-treated patients (HR = 0.93 per Q-allele, 95% CI 0.76 – 1.13, P = 0.46). Interestingly, placebo-treated patients had significantly increased incidence of the primary cardiovascular outcome (HR = 1.23 per Q-allele, 95% CI 1.03 – 1.47, P = 0.03). These findings are consistent with a number of other (although not all) studies showing association of PON1 variants with increased cardiovascular events presumably as a result of their association with lower HDL-cholesterol levels and increased underlying atherosclerotic burden¹⁵. Thus when comparing cardiovascular event rates in patients treated with clopidogrel to patients treated with placebo stratified by PON1 Q192R genotype to evaluate effect modification, clopidogrel-treated 192Q homozygotes had significantly decreased risk of experiencing the primary outcome compared to placebo-treated 192Q homozygotes (8.5% vs. 13.9%, respectively, P < 0.001); a similar trend was apparent in 192Q heterozygotes. Clearly, this effect was primarily driven by the increased incidence of cardiovascular events in placebo-treated patients with the 192Q-allele. In any event, these findings are in contrast with the results of Bouman and colleagues of increased risk of experiencing a cardiovascular event in clopidogrel-treated 192Q homozygotes.

In 570 patients treated with clopidogrel and 586 patients who received placebo in the ACTIVE A genetic substudy, clopidogrel-treated patients had lower cardiovascular event rates than placebo-treated controls. Like in the CURE trial, no association was observed between PON1 Q192R genotype and cardiovascular events in clopidogrel-treated patients, and there tended to be higher cardiovascular event rates in 192Q-allele carriers in the placebo-treated group. Thus, there was a non-significant trend of decreased cardiovascular events in the clopidogrel-treated group with increasing copies of the PON1 192Q-allele, consistent with results of the CURE trial, and again in the opposite direction of the initial report of Bouman.

In this study by Paré, none of the treatment group × PON1 genotype interaction terms were statistically significant. Clearly, this investigation does not support a role for the 192Q-allele of PON1 in clopidogrel therapy response and is consistent with a number of recent studies showing no effect of this allele on clopidogrel active metabolite levels, on-clopidogrel ADP-stimulated platelet aggregation, or cardiovascular outcomes^5-12. The reason(s) for discrepant findings between the initial report by Bouman and subsequent investigations is unclear.

The population investigated in the study by Bouman and colleagues consisted of Caucasian CAD patients who underwent PCI, with stent thrombosis being the primary outcome evaluated. Indeed, it is this patient population that benefits most from clopidogrel therapy. While most of the participants in the current investigation by Paré et al. were Caucasian, a limitation of this study, as pointed out by the authors, is the limited number of clopidogrel-treated PCI patients (approximately 400 in the CURE trial population). Furthermore, the ACTIVE A trial enrolled patients with atrial fibrillation, an indication for clopidogrel in which we know has limited benefit. In other words, the lower risk patients of the current investigation derive less benefit from clopidogrel compared to the higher risk PCI patients examined in the study by Bouman and colleagues, and therefore, may be less susceptible to the potential detrimental effects of the PON1 Q192R variant on clopidogrel response. If indeed the 192Q-allele is associated with greater underlying atherosclerotic burden (see below), the results of Paré et al showing that clopidogrel-treated subjects with the 192Q-allele have lower cardiovascular event rates than placebo-treated patients with the 192Q-allele is internally consistent with the hypothesis that clopidogrel has its greatest effect in decreasing cardiovascular events in higher-risk patients. Thus, while the current investigation supports a lack of effect of the Q192R variant on clopidogrel response in lower risk patients, it adds limited new information regarding the effect of this variant in higher risk PCI patients. However, it should be noted that other replication efforts that focused primarily or exclusively on clopidogrel-treated PCI patients also did not observe an effect of the Q192R variant on stent thrombosis^{5, 7, 9} or other cardiovascular events^{6-9, 12} (Table 1).

Table 1.

Effect of PON1 Q192R on Cardiovascular Events in Clopidogrel-Treated Patients

Study	Primary Outcome	Number of Clopidogrel- Treated Participants	Effect of PON1 Q192R on CV Endpoint
Bouman et al4	Stent thrombosis	41 cases, 71 controls	Yes
Sibbing et al5	Stent thrombosis	127 cases, 1439 controls	No
Trenk et al7	Composite CV endpoint	24 events, 760 patients	No
	Stent thrombosis*	18 events, 760 patients	No
Lewis et al6	Composite CV endpoint	27 events, 227 patients	No
Simon et al8	Composite CV endpoint	296 events, 2210 patients	No
	Composite CV endpoint	35 events, 371 patients	No
Hulot et al9	Stent thrombosis	18 events, 371 patients	No
Delaney et al12	Composite CV endpoint	225 cases, 468 controls†	No
Paré et al17	Composite CV endpoint	227 events, 2534 patients	No
	Composite CV endpoint	114 events, 570 patients	No

Abbreviations: CV, cardiovascular; PON1, paraoxonase 1

^*Cases were included that met ARC criteria for definite, possible, or probable stent thrombosis

^†Cases and controls were identified by algorithm using an electronic medical record-linked DNA repository

The relationship between PON1 and cardiovascular disease, irrespective of clopidogrel treatment, has been the focus of multiple investigations for over 10 years. Paraoxonase 1 is a 45kDa high-density lipoprotein (HDL)-associated arylesterase that is believed to have anti-atherosclerotic properties through its ability to metabolize oxidized lipids of low-density lipoproteins (LDL). Furthermore, previous studies have shown that paraoxonase activity is significantly reduced in patients with ST elevation myocardial infarction (STEMI), non-ST elevation myocardial infarction (NSTEMI), or unstable angina compared to healthy controls¹⁶. Consistent with these findings, the current investigation by Paré and colleagues shows a significant association between the decreased-function PON1 Q-allele and increased cardiovascular events in placebo-treated patients. However, this association was not observed in patients treated with clopidogrel. This novel finding suggests that somehow clopidogrel can ameliorate the increased risk of cardiovascular events in 192Q-allele carriers. As mentioned above, the explanation for this observation may be that clopidogrel has its greatest effect in decreasing cardiovascular events in higher-risk patients. It is intriguing to speculate that this effect may be independent of clopidogrel inhibition of platelet function since we and others have shown that the 192Q allele is not associated with on-clopidogrel ADP-stimulated platelet aggregation^{5-7, 9, 10}.

Paré and colleagues conclude with a caution regarding interpretation of pharmacogenetic analyses in the absence of non-treated controls. Indeed, without such a control group, it is often difficult to determine whether observed associations are the result of drug-genotype interactions or represent an underlying effect that is independent of drug treatment. While randomized placebo-controlled trials have and continue to be the gold standard in evidence-based medicine, these investigations are costly and inefficient in studies of variable drug response. Therefore, we believe that it is important to also consider complementary study designs and analytical approaches. For example, phenotypic characterization of surrogate endpoints pre- and post-drug treatment (e.g. ex vivo ADP-stimulated platelet aggregation in the case of clopidogrel pharmacogenetics) allows for patients to serve as their own control group.

Given the current evidence base, should patients with an indication for anti-platelet therapy undergo PON1 genetic testing? Probably not. By contrast, there is a large body of converging evidence (pharmacokinetic, pharmacodynamic, clinical outcomes) supporting a role for CYP2C19 genetic testing to guide anti-platelet therapy in PCI patients. Despite this evidence, many cardiologists have resisted changing practice in lieu of a prospective randomized clinical trial of CYP2C19 genotype-directed anti-platelet therapy to demonstrate efficacy and cost-effectiveness. One such trial, the Pharmacogenomics of Anti-Platelet Intervention-2 (PAPI-2) study (http://clinicaltrials.gov/ct2/show/NCT01452152) is ongoing. Such studies as well as large meta analyses will also offer the opportunity to evaluate PON1 and other candidate gene variants as well as larger genome-wide studies for discovery of novel variants that influence clopidogrel response. Ultimately, with the discovery of additional genetic determinants of clopidogrel response, a panel of genetic tests may be used along with clinical data to guide more effective personalized anti-platelet therapy. Currently, the evidence base does not support PON1 for inclusion in such a panel.