Routine Screening for CYP2C19 Polymorphisms for Patients being Treated with Clopidogrel is not Recommended

Abstract

Recent efforts directed at potential litigation in Hawai‘i have resulted in a renewed interest for genetic screening for cytochrome P450 2C19 (CYP2C19) polymorphisms in patients treated with clopidogrel. Clopidogrel is an antiplatelet agent, frequently used in combination with aspirin, for the prevention of thrombotic complications with acute coronary syndrome and in patients undergoing percutaneous coronary interventions.

Cytochrome P-450 (CYP) 2C19 is an enzyme involved in the bioactivation of clopidogrel from a pro-drug to an active inhibitor of platelet action. Patients of Asian and Pacific Island background have been reported to have an increase in CYP2C19 polymorphisms associated with loss-of-function of this enzyme when compared to other ethnicities. This has created an interest in genetic testing for CYP2C19 polymorphisms in Hawai‘i.

Based upon our review of the current literature, we do not feel that there is support for the routine screening for CYP2C19 polymorphisms in patients being treated with clopidogrel; furthermore, the results of genetic testing may not be helpful in guiding therapeutic decisions. We recommend that decisions on the type of antiplatelet treatment be made based upon clinical evidence of potential differential outcomes associated with the use of these agents rather than on the basis of genetic testing.

Introduction

Clopidogrel is used in the treatment of patients with atherosclerotic vascular disease and percutaneous coronary interventions (PCI) to prevent thrombotic complications associated with enhanced platelet activation. Patients undergoing PCI or with acute coronary syndromes (ACS) are usually placed on dual antiplatelet therapy. The treatment of these patients with aspirin and clopidogrel has been well documented to be related to improved clinical outcomes.^1,2

Clopidogrel is a pro-drug that is activated in a two-step process in the liver by several cytochrome pathways. Clopidogrel is converted to 2-oxyclopidogrel by hepatic enzymes: cytochrome P-450 (CYP) 2C19, CYP1A2 and CYP2B6; 2-oxyclopidogrel is converted into an active metabolite by CYP3A4/5, CYP2B6, CYP2C19 and CYP2C9. This active metabolite irreversibly inhibits activation of the platelet GIIb/IIIa complex by binding to the P2Y12 receptor.^3–5

CYP2C19 polymorphisms associated with decreased rates of clopidogrel activation have been identified in approximately 12-13% of the Caucasian population.^6,7 Polymorphisms in the CYP2C19*2 and CYP2C19*3 alleles have been ascribed to decreased antiplatelet response to clopidogrel and potentially increased rates of thrombotic events. The reduced-function alleles have been found in frequency ranges of 25–50% among patients from East Asian and South Asian backgrounds.^6,7 It has been proposed that genetic testing for CYP2C19 polymorphisms might be useful in identifying patients with reduced-function alleles.^3–5 It has been further suggested that patients with reduced-function genetic profiles may benefit from the routine conversion from clopidogrel to agents such as ticagrelor or prasugrel, which may not require bioactivation by CYP2C19 enzymes.

In 2010, an expert panel published recommendations addressing potential genetic testing for CYP2C19 polymorphisms in patients treated with clopidogrel. This panel was convened by the American College of Cardiology Foundation (ACCF) and American Heart Association (AHA) to address a “boxed warning” that had been issued by the US Food and Drug Administration concerning the possible need for pharmacogenomics testing to identify patients with potentially altered clopidogrel metabolism, who may be at increased risk for adverse clinical outcomes.⁸ It was the conclusion of this panel that while clinicians should be aware that genetic variability in CYP enzymes may alter clopidogrel metabolism and in turn affect the drug's inhibition of platelet function, the evidence, at that time, was insufficient to recommend that either routine genetic testing for CYP polymorphisms or routine platelet function testing be performed. Furthermore, the panel recommended against basing clinical decisions upon the results of genetic testing.⁸

Nevertheless, on March 20, 2014, David Louie, the Attorney General of the state of Hawai‘i, charged Bristol-Meyer Squibb and Sanofi with failing to disclose that a significant portion of patients in Hawai‘i had genetic polymorphisms in the cytochrome P450 2C19 (CYP2C19) gene.⁹ This finding was proposed to be associated with the potential loss of function of clopidogrel in a significant portion in Asians and Pacific Islander ethnicities. It was asserted that patients of Asian and Pacific Island ethnicity may be both less responsive to the actions of clopidogrel and at persistent risks for adverse reactions to this medication including gastrointestinal bleeding. This announcement created a renewed interest in the role of genetic testing for CYP2C19 polymorphisms in Hawai‘i.

Since the publication of the ACCF/AHA expert panel consensus in 2010,⁸ there have been several additional studies that have looked at the testing of platelet reactivity and clinical outcomes in patients with different CYP2C19 polymorphisms. This paper reports on the state of the literature on CYP2C19 screening and targeted treatment with clopidogrel, to determine if recent studies continue to support the panel's recommendations.

Methods

A systemic search of the literature between January 2009 and June 2014 was conducted using PubMed. PubMed is a publication search engine supported by the U.S. National Library of Medicine and the National Institutes of Health. An additional review was performed using the Google search engine. The search terms used were “coronary artery disease,” “CYP,” “myocardial infarction,” “cytochrome P 450,” and “clopidogrel.”

A retrospective review of references provided by the 2010 ACCF-AHA expert consensus was also performed to provide a selective review of historical references. Finally, the National Institutes of Health Clinical Research Trial database was queried to identify potential trials that may have addressed the use of genetic testing for CYP polymorphisms in assessing clinical outcomes but may have not been published.

Studies specifically outlining clinical outcomes and genetic testing were reviewed by the first two authors.

Results

Studies and Clinical Trials Conducted Preceding the 2010 Consensus Statement

The internet-based literature search generated several articles published before the 2010 consensus statement; 53 articles were reviewed and 18 are included in this review. Another 16 articles published after the 2010 consensus statement were separately reviewed, out of which 8 have been included in this review

It has been long recognized that there may be variation in platelet function inhibition, as assessed by platelet aggregometry, while on PGY12 inhibitors. In 2002, Järemo and colleagues reported individual variations in platelet inhibition after loading doses of clopidogrel in a group of 18 patients with stable angina undergoing PCI.¹⁰ This finding was reproduced in a group of 96 patients in 2003 by Gurbel, et al.¹¹

To explain these findings, a linkage between the intensity of clopidogrel-induced inhibition of platelet reactivity and clinical outcomes was also suggested.

In 2004, Matetzky described 60 patients with clopidogrel resistance who were at an increased risk of recurrent atherothrombotic events while being treated with antiplatelet therapy after acute myocardial infarctions. ¹² Schuldiner, et al, and Mega, et al,^4,5 in separate publications, linked nonresponsiveness to clopidogrel to CYP2C19 polymorphisms associated with a potential loss-of-function (LoF) of this enzyme required for the bioactivation of clopidogrel.^4,5 Between 2009 and 2011, several reviews of CYP2C19 polymorphisms and outcomes were performed.^13–16 They contended that clopidogrel non-responsiveness may result in an increased risk for major cardiac events while on antiplatelet treatment. The described studies were performed using retrospective analyses.

This opinion was confirmed by some clinical trials conducted during the time period. In the results of the Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition with Prasugrel-Thrombolysis in Myocardial Infarction (TRITON TIMI 38), published in 2007,¹⁷ 13,608 patients undergoing treatment for ACS or ST segment elevation myocardial infarction were randomized to treatment with clopidogrel versus prasugrel. A genetic study was performed assessing CYP2C19 allele status. Carriers of the LoF allele treated with clopidogrel had higher rates of the primary outcomes, including death from cardiovascular causes, non fatal MI, or non fatal stroke (12.1% versus 8.0%; hazard ratio = 1.53; P = .01) and definite/ probable stent thrombosis (2.6% versus 0.8%; hazard ratio = 3.09; 95% confidence interval = 1.19–8.00; P = .02) compared with non-carriers. However among prasugrel treated patients, LoF carrier status was unrelated to outcomes. Higher rate of events occurred predominantly in the group of patients treated with PCI.¹⁷ This study suggested that genetic testing for CYP2C19 polymorphisms might be used to predict clinical outcome while on antiplatelet therapy.

The Platelet Inhibition and Patient Outcomes (PLATO) study, published in 2010, was a double-blinded randomized clinical trial involving 10,285 subjects, comparing the effects of clopidogrel and ticagrelor in patients presenting with ACS.¹⁸ In this study, 60% of patients underwent PCI. Routine genetic testing for CYP2C19 polymorphisms was performed, which identified a subgroup of 27% of the study population with the LoF of at least one CYP2C19 allele. In this subgroup, patients treated with clopidogrel had a higher risk of adverse cardiovascular events, with a hazard ratio of 1.37 (CI 1.04–1.82; P = .028). In patients treated with ticagrelor, an antiplatelet agent that does not require bioactivation by CYP2C19, no clear relationship between genotype and clinical outcome was established. Based upon these findings, the authors suggested that the routine use of ticagrelor obviated the need for genetic testing and that the presence of at least one CYP2C19 allele associated with potential LoF resulted in decreased cardioprotection in patients treated with clopidogrel. The validity of this interpretation was limited by the fact that these results were based upon a retrospective analysis and changes in the definitions of the intention to compare groups, with the original intention to compare six phenotype groups which was later changed to compare just two genotype groups.

Nevertheless, conflicting information regarding the utility of identifying LoF CYP2C19 polymorphisms in the prediction of associated cardiovascular events while on antiplatelet therapy was obtained from other prospective clinical studies. In the Clopidogrel in Unstable Angina to Prevent Recurrent Events (CURE) trial, published in 2010, 5,059 patients were treated for ACS with either placebo or clopidogrel and underwent genetic testing. ¹⁹ CYP2C19 polymorphisms associated with a potential LoF were noted in 19.8% of patients. These investigators found that the clinical outcomes of patients on clopidogrel were not significantly different based upon metabolizer status. Of note, in this study, only18% of the patients underwent PCI. The authors concluded that CYP2C19 LoF allele status did not preclude the use of clopidogrel in patients with ACS who were being managed conservatively.¹⁹

Studies and Clinical Trials Conducted Following the 2010 Consensus Statement

Table 1 summarizes the results of the 6 clinical trials initiated between 2010–2014 addressing genetic testing and treatment strategies. Since the publication of the ACCF AHA expert consensus in 2010,⁸ there have been no randomized trials that have demonstrated that treatment strategies based upon genetic testing for CYP2C19 polymorphisms have affected clinical outcomes.

Table 1.

Summarizing Six Clinical Trials Initiated between 2010–2014 Addressing Genetic Testing and Treatment Strategies

Study	Outcomes
TARGET-PCI (NCT01177592) Sponsor: LifeBridge Health 2010. Prospective, single-center, randomized trial including 1500 patients with stable coronary artery disease (CAD) and elective PCI. Patients within 24 hrs of PCI were randomized to either Control (remain on clopidogrel) or Guided therapy (CYP2C19*2 carriers and/or if Platelet reactivity unit (PRU)>230, reload with prasugrel). Major Adverse Cardiac Events (MACE) including death, myocardial infarction (MI) and target vessel revascularization were recorded in a follow up (FU) period of 6 months.	The study was suspended in October 2010 after it was concluded that the sample size was underpowered to address the initial hypothesis
TRIGGER-PCI (NCT00910299) Sponsor: Eli Lilly 2009–2012. A randomized, active-control, double blind, phase II, multicenter trial enrolled 2150 patients with stable CAD and elective PCI. Selection criteria of patients 24 hour post-drug eluting stent (DES) and 2 to 7 hour post clopidogrel and HRPA randomized to either Prasugrel 60 mg load/10 mg daily; or Clopidogrel 75 mg daily. Cardiovascular (CV) death or nonfatal MI was recorded for a FU period of 6 months.	In 212 patients assigned to prasugrel, PRU decreased from 245 to 80 at 3 months, whereas in 211 patients assigned to clopidogrel, PRU decreased from 249 to 241 (P < .001 vs prasugrel). The primary efficacy endpoint at 6 months occurred in no patient on prasugrel versus 1 on clopidogrel.25 Thus Switching from clopidogrel to prasugrel in patients with high on treatment platelet reactivity (HTPR) did not demonstrate clinical utility given low rate of adverse ischemic events after PCI with contemporary drug eluting stents (DES) in stable CAD
CLOVIS-2 (NCT00822666) PI: J.-P. Collet PI: G. Montalescot 2012. A randomized, open-label, phase III, crossover Pharmacodynamic, PD/ Pharmacokinetic, PK study) trial including 120 patients who were Post-MI, 45-years-old and enrolled in AFIJI registry. Comparison of 2 loading strategies of clopidogrel (300 mg vs 900 mg) in 2 genetic profiles: wild-type 2C19*1 and carriers of 2C19*2 was performed. Inhibition of Residual platelet activity (IRPA) by optical aggregometry; measurement of active metabolites, was studied at 6 hour post loading dose.	Carriers of CYP2C19*2 display significantly lower responses to clopidogrel with a gene-dose effect. Clopidogrel resistance can be overcome by increasing the dose in heterozygous carriers but not in homozygous carriers26 Clinical outcomes were not studied.
ACCEL-2C19, (NCT01012193) PI: Y.-H. Jeong 2011. A randomized, active control, single-blind (PD study) of 80 patients with Stable CAD with type 2 DM and elective PCI. Patients genotyped for CYP2C19 variants randomized to high dose clopidogrel (150 mg) and ASA 200 mg vs cilostazol 100 mg bid, 75 mg clopidogrel and ASA 200 mg (triple therapy) and Inhibition of maximum platelet aggregation (MPA) by optical aggregometry; VerifyNow, was studied at 30day FU.	Among T2DM patients, adding cilostazol achieves greater platelet inhibition compared with clopidogrel (150 mg/d), which is not influenced by genetic polymorphisms. Clinical outcomes were not studied with genetic polymorphisms
ACCELAMI2C19 (NCT00915733) PI: I.-S. Kim 2009. Randomized, active control, open-label (PD study) of 60 patients with acute MI, post PCI. Patients genotyped for CYP2C19 variants randomized to high dose clopidogrel (150 mg) and ASA 100 mg vs. cilostazol, 100 mg bid, 75 mg clopidogrel and ASA 100 mg. Inhibition of MPA studied at 30 day follow up.	Adjunctive cilostazol reduces the rate of high post treatment platelet reactivity (HPPR) and intensifies platelet inhibition as compared with a high-maintenance dose (MD) clopidogrel of 150 mg/day. Clinical outcomes were not studied with genetic polymorphisms
ACCEL-2C19 (NCT00891670) PI: Y.-H. Jeong 2009. Randomized, active control, open-label (PD study) of 126 patients with stable CAD and elective PCI. Patients genotyped for CYP2C19 variants randomized to high dose clopidogrel (150 mg) ASA 100 mg vs. cilostazol, 100 mg bid, 75 mg, clopidogrel ASA 100 mg (triple therapy). Maximum platelet aggregation was studied at 30 day follow up.	Among PCI-treated patients receiving high-MD clopidogrel, carriage of CYP2C19 variant relates to increased PR and predicts risk of HPPR. Clinical outcomes based on prespecified genetic testing were not studied

TARGET PCI, Thrombocyte Activity Reassessment and Genotyping for PCI; TRIGGER PCI, Testing Platelet Reactivity in Patients Undergoing Elective stent placement on Clopidogrel to Guide Alternative Therapy with Prasugrel; ACCEL-2C19, Adjunctive Cilostazol Versus High Maintenance-dose Clopidogrel According to Cytochrome 2C19 Polymorphism; ACCELAMI2C19, Adjunctive Cilostazol Versus High Maintenance-dose ClopidogreL in Acute Myocardial Infarction Patients According to CYP2C19 Polymorphism; AFIJI, Appraisal of Risk Factors in Young Ischemic Patients Justifying Aggressive Intervention; CLOVIS-2, Clopidogrel and Response Variability Investigation Study 2

In a substudy of the Gauging Responsiveness with A VerifyNow assay- Impact on Thrombosis and Safety (GRAVITAS) study, called the Genotype Information and Functional Testing (GIFT) study, the influence of single nucleotide polymorphisms (SNPs) on the pharmacodynamic (PD) effect of high- or standard-dose clopidogrel after PCI was evaluated.²⁰ DNA samples obtained from 1,028 patients were genotyped for 41 SNPs in 17 genes related to platelet reactivity. The presence of CYP2C19*2 allele was found to be significantly associated with increased on-treatment platelet reactivity (OTR) at 12 to 24 h (R² = 0.07, P = <.001), 30 days (R² = 0.10, P = <.001), and 6 months after PCI (R² = 0.07, P = <.001), but not with significant differences in clinical outcome ( HR 1.07; CI: 0.91 –1.25; P = .42).

In the recently published findings from the Assessment by a double Randomization of a Conventional antiplatelet strategy versus a monitoring-guided strategy for drug-eluting stent implantation and of Treatment Interruption versus Continuation one year after stenting (ARTIC) trial, 2440 patients were placed on different antiplatelet treatment strategies based upon the results of platelet function testing. The authors reported that tailored therapy based upon genetic testing did not result in improved outcomes when compared to standard treatments that were not based on genotyping.²¹

Several other trials were terminated prior to completion because of concerns that the study population would be inadequate to provide significant statistical analysis. As a representative example, it was hoped that the Thrombocyte Activity Reassessment and Genotyping for PCI (TARGET-PCI) trial would bring definitive information demonstrating whether guided antiplatelet therapy using genotyping and platelet function testing would have advantages over standard antiplatelet therapy. Approximately 1,500 patients were randomized in this trial. The trial was suspended in October 2010 after it was concluded that the sample size was underpowered to address the initial hypothesis. It was recognized that the event rates such as stent thrombosis would be so low that effective analysis of significance would be difficult. Similarly, a large scale clinical trial addressing Genotype Guided Comparison of Clopidogrel and Prasugrel treatment outcomes (NCT00995514) had been planned but was terminated prior to analyses due to administrative reasons.

Discussion

Genetic testing for CYP2C19 polymorphisms is now readily available and costs $300–$400 which is covered by most insurance companies with varying co-payments.²² At issue, however, is whether all patients on antiplatelet therapy should undergo routine genetic testing for CYP2C19 polymorphisms.

As mentioned previously, in 2010, the U.S. Food and Drug Administration (FDA) published a “Boxed Warning” for clopidogrel noting that patients who were CYP2C19 poor metabolizers exhibited higher cardiovascular event rates than patients with normal CYP2C19 function. This warning also noted that there were tests available to assess CYP2C19 genotypes and that these tests could potentially be used as an aid in determining therapeutic strategies. While the FDA clearly informed physicians and patients about the availability of genetic testing, it did not recommend or mandate testing.²³

The ACCF and AHA responded to this “Boxed Warning” and convened a panel to create a clinical expert consensus document.⁸ The consensus document was published in 2010 and concluded that “the role of genotyping in everyday practice remains unknown at this present time.” This group also noted that at the time of publication, there were “no prospective studies demonstrating a clinical benefit to personalizing antiplatelet therapy based on genotype analysis.” Advocates of genetic testing were present on the panel and argued that “given the magnitude of the potential clinical consequences of suboptimal platelet inhibition based on genetic variation, assessment of genotypes would be justifiable.”

However, an opposing opinion was also noted by panel members who stated there was no definitive proof that intervening on the basis of genotype would improve outcome. Opponents to recommending routine genetic testing also raised the question of whether genotyping should be confined to LoF mutations involving only the CYP2C19*2 and CYP2C19*3 alleles or be extended to other variants, including potential hypermetabolizers of clopidogrel. Opponents of testing also noted that the predictive performance of the presence of CYP2C19*2 and CYP2C19*3 allele variants was low, ranging only from 12–20% in predicting adverse clinical events. It was suggested that the state of knowledge was incomplete and that further studies would be useful to address these issues.⁸

In an editorial in 2011, Nissen described genetic testing for clopidogrel as a case of “irrational exuberance,” a phrase popularized by former chairman of the Federal Reserve Alan Greenspan when describing the exponential increase in stock prices a few years before the “dot-com bubble” burst in March 2000. The enthusiasm for dot-com stocks led to unrealistic expectations for rapid success of companies without documented financial performance. Similarly the success of pharmacogenomics in some fields of medicine has led to unrealistic expectations for many other specialties, including cardiovascular medicine.²⁴

A review of the literature reveals that the findings of recent studies do not support the routine screening for CYP2C19 polymorphisms or targeted antiplatelet therapy based upon the results of genetic testing. Based upon our review we feel that decisions regarding conversion to newer antiplatelet agents such as ticagrelor or prasugrel should not be based upon genetic testing but on evidence of potentially improved clinical outcomes for patients treated with these newer agents but balanced against bleeding risks. With the end of the patent on clopidogrel and its current availability as a generic formulation, it is unlikely that a trial with a sample size of 20,000–30,000 patients, necessary to definitively study potential effects of genetic variation in CYP2C19 correlated with clinical outcomes would be funded or undertaken.

Summary

In 2010, an expert panel convened by the ACCF and AHA concluded that the routine testing for CYP2C19 polymorphisms in patients treated with clopidogrel was not warranted.⁸ In our review of the published data since this recommendation and our overview of clinical trials on this subject registered by the NIH, we have not found any additional supporting evidence for tailored therapy based upon genetic testing. As clopidogrel is now available in a generic formulation, it is unlikely that further studies clarifying the role of genetic testing will occur and we suspect that there will be limited additional information clarifying a potential role of genetic testing targeted-therapy. While we do not recommend the routine testing for CYP polymorphisms as a basis for changing antiplatelet therapies, clear differences in clinical outcomes have been noted with the use of newer antiplatelet agents such as ticagrelor and prasugrel. The choice of antiplatelet agent should not be based upon the results of genetic testing but rather on clinical indications.

Conflict of Interest

None of the authors identify any conflict of interest.