By Turna Ray
CYP2C19 alleles ⢀ which various studies have shown to limit response to the anti-platelet drug Plavix in carriers of the polymorphisms ⢀ did not have the same impact on patients' ability to metabolize the drug in a study published in the New England Journal of Medicine last week.
The authors of the study said they believe the data raises doubts on the clinical utility of genotyping for CYP2C19 loss-of-function alleles as a way to direct Plavix therapy in patients with acute coronary syndrome and atrial fibrillation.
However, this finding comes up against a body of clinical evidence claiming that carriers of the CYP2C19 alleles • *2 and *3 • impair carriers' ability to convert Plavix to its active metabolite, and respond to the drug as well as non-carriers.
After the study was presented at the European Society of Cardiology Congress last week, Eric Topol, a prominent cardiologist and supporter of PGx testing for Plavix dosing in stent patients, identified a flaw in the paper. He told Pharmacogenomics Reporter that the findings were confounded by the limited number of participants who had undergone drug-eluting stenting procedures. Plavix is given to patients after a coronary artery stent procedure to reduce the risk of sudden death from blockages in the artery.
The NEJM study genotyped patients from two previously conducted randomized trials: the Clopidogrel in Unstable Angina to Prevent Recurrent Events trial, known as CURE, and the Atrial Fibrillation Clopidogrel Trial with Irbesartan for Prevention of Vascular Events, called ACTIVE A. Both trials reported that Plavix reduced the rate of cardiovascular events in ACS and atrial fibrillation patients.
After pharmacogenetically analyzing patients from those studies, Guillaume Pare, assistant professor of pathology and molecular medicine at McMaster University, and colleagues concluded in the NEJM paper that "among patients with ACS and AF, the effect of Plavix as compared with placebo is consistent, irrespective of CYP2C19 loss-of-function carrier status."
In addition to the data on loss-of-function alleles, the study, which was funded by Plavix sponsors Bristol-Myers Squibb and Sanofi Aventis, reported in line with previously published data that carriers of the gain-of-function CYP2C19 *17 allele derived greater benefit from Plavix compared to placebo. However, it is the study's conclusions about the loss-of-function CYP2C19 alleles that have been raising eyebrows among researchers and industry observers.
The NEJM paper comes at a time when many personalized medicine stakeholders are attempting to kick-start PGx testing for Plavix to hone in on the subpopulations most likely to respond from the drug and to better position the drug against competing treatments. Proponents of genetic testing in this setting say that there is ample evidence supporting the notion that performing CYP2C19 genotyping to guide Plavix treatment improves outcomes, makes healthcare more precise, and may even save money by avoiding costly adverse reactions in poor metabolizers.
Earlier this year, pharmacy-benefit manager Medco and CVS Caremark announced competing personalized medicine programs that include PGx testing for Plavix. Both Medco and CVS offer PGx testing for various drugs to their PBM customers, which include large employers and health insurance companies (PGx Reporter 06/02/10).
Medco is also studying the utility and cost-effectiveness of PGx dosing for Plavix, a drug slated to go off patent in 2012. This is being done as part of Medco's Genetics for Generics, a program that aims to apply PGx data to increase the use of generic drugs and drive down the cost of healthcare. Medco's study will examine whether the 70 to 75 percent of patients who are "extensive metabolizers" of Plavix will have "comparable outcomes" with patients taking Lilly/Daiichi Sankyo's Effient (PGx Reporter 10/21/09). The findings of this study may have significant financial implications for Medco, as the PBM garners the bulk of its revenues from the dispensing of generic drugs.
[ pagebreak ]
Other proponents of genetic testing to guide Plavix therapy are PGx-savvy hospitals, such as Scripps Green Hospital, where under the leadership of Topol, cardiologists have adopted genetic testing to guide treatment with the anti-platelet drug (PGx Reporter 10/28/09).
Then there are regulatory drivers. As data from PGx studies have rolled in, the US Food and Drug Administration has updated the label for Plavix several times, each time issuing slightly stronger language in support of the notion that genotyping for loss-of-function CYP2C19 alleles can hone in on those unlikely to respond to Plavix. In its latest move in March, the FDA issued a black box warning for the drug cautioning healthcare providers that patients with diminished CYP2C19 function, or poor metabolizers of the drug, are at a greater risk of cardiovascular adverse events after an ACS or a percutaneous coronary intervention than normal metabolizers of Plavix (PGx Reporter 03/17/10).
Pare, the lead author of the NEJM paper, said he believes the findings in his study are strong enough to potentially affect additional changes to Plavix's label. Although he acknowledged that the results of this study need to be investigated further in trials involving stented ACS patients, Pare maintained the data "show ... that genetic testing will have limited utility in certain patient populations."
Pieces of a PGx Puzzle
When it comes to using genetics to guide medical care, the data often arrives in pieces, and the body of genetic knowledge on subpopulations then become refined through subsequent prospective genotyping studies, retrospective genome-wide association analyses, and targeted sequencing approaches. In the case of Plavix, the findings of the NEJM study can be considered another piece of the PGx puzzle that cardiologists will have to consider when prescribing the drug.
For their study, Pare and his team genotyped more than 5,000 patients from the CURE project and more than 1,100 patients from the ACTIVE A trial. They used Life Technologies/Applied Biosystems' TaqMan assay to test for the presence of three polymorphisms: CYP2C19 *2, CYP2C10*3, and CYP2C19*17. The researchers then randomized these patients to receive either Plavix or placebo.
Analysis of the resulting data revealed that ACS patients treated with Plavix experienced a lower rate of cardiovascular adverse events compared to those on the placebo arm, but this did not change when patients were evaluated based on genotype.
"The effect of Plavix in reducing the rate of the primary efficacy outcome was similar in patients who were heterozygous or homozygous for loss-of-function alleles and in those who were not carriers of the alleles," the researchers reported. "In contrast, gain-of-function carriers derived more benefit from Plavix treatment as compared with placebo than did non-carriers."
[ pagebreak ]
Additionally, the study revealed that the effect of Plavix on bleeding events did not change based on genotypic subgroups. "Among 1,156 genotyped patients with atrial fibrillation, there was no evidence of an interaction with respect to either efficacy or bleeding between the study treatment and the metabolizer phenotype, loss-of-function carrier status, or gain-of-function carrier status," the authors wrote in the paper.
However, Pare explained that in ACTIVE A, data on the impact of loss-of-function alleles was challenged by the small sample size, and as such "we can't entirely rule out a role of loss-of-function alleles in atrial fibrillation patients."
Despite the flaws of the study, as far as Pare is concerned, the leads are still significant for future clinical trials exploring Plavix pharmacogenetics.
"We did not see any effect of loss-of-function variants in stented patients in CURE, but our sample size was limited and therefore results must be interpreted cautiously," Pare said. "Data from randomized trials incorporating genetic information will be needed to fully understand the role of genetics, if any, in guiding Plavix treatment in ACS patients with stents."
Topol, in reviewing Pare's study, cautioned that the study should not be used to dismiss the need for genotyping to guide Plavix treatment. "The data from CURE/ACTIVE are off the mark since there was virtually no, or very little, stenting in CURE and ACTIVE as an atrial fibrillation trial, which only showed minimal to no benefit for clopidogrel to start with," he said.
"The CYP2C19 variant interaction with Plavix is predominantly a stent story," Topol explained. "That is where there is extraordinary support of this interaction."
The other side of the argument • the component that gives credence to Pare's findings • is the fact that the analysis compared Plavix's impact on cardiovascular events to placebo, which is necessary to isolate the impact of CYP2C19 genes on Plavix response.
"CYP2C19 lies within a cluster of CYP genes including CYP2C9, CYP2C8 and CYP2C18. It is therefore not unreasonable to hypothesize that the variants genotyped could have affected the metabolism of other drugs," Pare said. "Only a placebo-controlled study can rule out this possibility. In our study, we did not see any significant effect in placebo-treated individuals."
Working with Imperfect Data
Likewise, when FDA issued the black box warning for Plavix, agency officials acknowledge that it was doing so in the absence of data from ongoing studies. "The data are not perfectly clear about [whether] all [poor metabolizers] have a very poor outcome," Robert Temple, director of the Office of Medical Policy at FDA's Center for Drug Evaluation and Research, said in March when the agency announced the warning. "The data is somewhat mixed and the study designs are mostly not intended to answer that question."
Although the FDA's labeling update shared with healthcare providers the existing body of evidence describing how CYP2C19's alleles may affect Plavix response, the agency stopped short of telling physicians in what settings they should consider genetically testing patients.
In addition to the NEJM study, other recently published papers further complicate the picture by introducing another anti-platelet drug that might be superior to Plavix, and by implicating additional genes that might play a part in Plavix metabolism.
[ pagebreak ]
For instance, a study published in The Lancet on Aug. 29 analyzed data from the PLATelet inhibition and patient Outcomes, or PLATO, trial to gauge how Plavix-treated carriers of CYP2C19 loss-of-function and gain-of-function alleles fared when compared to patients given AstraZeneca's anticoagulant Brilinta (ticagrelor). An FDA advisory committee reviewed data from the PLATO trial in July, and recommended the agency approve Brilinta. FDA is slated to issue its decision in September.
In the PGx analysis of PLATO published in The Lancet, researchers reported that in patients treated with Plavix after an ACS event or stenting, or both, "the presence of any loss-of-function CYP2C19 allele (*2,*3,*4,*5,*6,*7 and *8) is associated with an increased risk of ischemic events and stent thrombosis." Meanwhile, "the presence of any gain-of-function CYP2C19 allele (*17) is associated with a raised risk of bleeding,” the authors wrote.
In this same study, however, the researchers also found that ACS patients treated with Brilinta had better outcomes that those treated with Plavix, regardless of CYP2C19 and ABCB1 polymorphisms.
Yet another study, also published in The Lancet, analyzed data from the Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition with Prasugrel, or TRITON-TIMI 38, trial. In it they found that carriers of the ABCB1 3435 TT genotype with coronary syndromes undergoing percutaneous interventions have reduced platelet inhibition, which puts them at increased risk of recurrent ischemic events during Plavix treatment.
Additionally, researchers reported that "in patients with acute coronary syndromes who have undergone percutaneous intervention, when both ABCB1 and CYP2C19 are taken into account, nearly half of the population carries a genotype associated with increased risk of major adverse cardiovascular events while on standard doses of Plavix."
"There are clearly other genes involved, such as ABCB1, as demonstrated in The Lancet papers," Topol pointed out. "But the CYP2C19 evidence in patients undergoing stenting is incontrovertible and clinically important. The new Lancet large-scale studies further corroborate this key point."
In contrast to Topol's view, Pare interprets the analysis of the PLATO trial as further "challenging the importance of the [CYP2C19] loss-of-function alleles in patient outcomes," since that study found that patients fared better with Brilinta over Plavix, despite CYP2C19 genotyping..
Although he admits he is no regulatory expert, Pare believes that perhaps "it would be appropriate to … update the [black box] warning" for Plavix based on the results of his own study and the PLATO analysis.
In updating Plavix's label, "I believe the FDA made a decision based on the best evidence available at the time and with the best interest of patients in mind," Pare said. "We have not heard from the FDA, but I would be delighted to discuss my results in more details with them."
It seems that for the time being, however, FDA is not planning to change Plavix's labeling based on these studies. "The FDA is aware of the [NEJM] study and there are no plans to update the label language at this time," an agency spokesperson told PGx Reporter.
It is currently unclear whether Medco or CVS Caremark will alter their personalized medicine programs involving PGx testing for Plavix in light of the new findings. Both Medco and Generation Health, a genetics benefits-management company partially owned by CVS Caremark, were unable to comment in time for this publication.
Pare said he will try to obtain "clearer answers" to the questions raised on the loss-of-function alleles by collaborating with other researchers to pool data on a larger number of patients.