December 2010
Feature Story

The move to personalize antiplatelet therapy got a few boosts this year and a jolt, leaving many to sort out the implications.

Beginning in March, the FDA required Plavix (clopidogrel) to carry a black-box warning advising clinicians to consider using alternative treatment or treatment strategies for poor metabolizers of CYP2C19, the liver enzyme known to transform the drug into its clot-preventing form.

A meta-analysis published Oct. 27 in the Journal of the American Medical Association showed that heart stent patients with even one reduced-function CYP2C19 allele had a dramatically heightened risk of negative cardiovascular outcomes.

Also in October, the FDA cleared AutoGenomics’ automated Infiniti CYP2C19 assay with *2, *3, and *17 mutations, making the test potentially more accessible.

The succession of developments set the stage for genotyping, says Ram Vairavan, senior vice president at AutoGenomics, Vista, Calif.

No wonder then that in November, clinicians and laboratorians were eagerly awaiting the results of the GRAVITAS trial (Gauging Responsiveness with a VerifyNow Assay—Impact on Thrombosis and Safety) at the American Heart Association meeting. The randomized, prospective trial looked at whether new stent patients with high platelet reactivity, referred to as nonresponsiveness, would have fewer ischemic events if they took twice the daily standard 75 mg dose of Plavix.

The double dose, 150 mg, didn’t appear to do the trick.

Cardiologist Matthew Price, MD, principal investigator for GRAVITAS, who presented the study findings at the AHA meeting, says he had the general impression that people were surprised the trial did not show a benefit. People also seemed to be surprised, he told CAP TODAY, at the relatively moderate effect that 150 mg of Plavix had on high platelet reactivity. Patients on the higher dose didn’t show increased bleeding, which he notes is consistent with the lack of clinical response.

To identify those with high residual platelet reactivity, the 5,000-plus patients in the study underwent platelet function testing using Accumetrics’ VerifyNow P2Y12. Patients were tested 12 to 24 hours after receiving a stent, at 30 days, and at six months, which concluded the study’s followup period.

Still to come is a genetic subanalysis of the GRAVITAS data (the GIFT trial). Researchers will examine whether reduced-function alleles for CYP2C19 may have played a role in the weak effect on platelet reactivity despite the heftier Plavix dose. (Dr. Price does not know when those results will be available.)

Dianah Schmidt, executive director of marketing at Accumetrics, sponsor of the trial, says GRAVITAS reinforced the notion of individualized therapy for patients with high residual platelet reactivity. “A one-size-fits-all treatment strategy of doubling the Plavix dose did not improve outcomes in those patients,” she says, adding that platelet function testing can help physicians make more informed treatment decisions for each patient.

Anand Vairavan, marketing manager at AutoGenomics, called the GRAVITAS findings “rather disappointing,” but said, “If anything, the conclusions will tip people in the direction of 2C19 genetic testing.” That’s why the latter continues to look promising, he says, and adds: “The conclusions of the GIFT trial will hopefully confirm its role and importance.”

Until the GIFT genetic sub-study outcome is known, Alan Wu, PhD, says the GRAVITAS trial findings raise a “now what?” question in terms of what to do with information regarding a patient having clopidogrel resistance.

“We remain in a situation where the FDA has a warning out there” about the genetic effect on clopidogrel metabolism, says Dr. Wu, chief of clinical chemistry and toxicology and professor of laboratory medicine at San Francisco General Hospital and the University of California. “I don’t think the FDA is going to withdraw the warning, but we don’t have a countermeasure” in terms of increasing clopidogrel dosing.

Paul Gurbel, MD, director of cardiovascular research at the Center for Thrombosis Research at Sinai Hospital of Baltimore and a co-author of the Oct. 27 JAMA meta-analysis, says GRAVITAS “didn’t answer the question of whether more potent platelet inhibition in clopidogrel nonresponders improves outcomes because double-dose clopidogrel only modestly reduced platelet function.”

Might a higher Plavix dose have helped?

Dr. Price, director of the cardiac catheterization laboratory at Scripps Clinic Green Hospital, La Jolla, Calif., says some physicians in Europe today prescribe higher doses—225 mg to 300 mg a day—“but when we started the study, 150 mg was considered a high dose.” And only since more data have emerged are people even comfortable with a 150 mg dose. Thus, at the study’s outset two years ago, “it wasn’t considered feasible to give more.”

Dr. Wu doesn’t think clinicians in the United States would be willing to give clopidogrel doses that run three or four times higher than the FDA-approved dose.

Another option is for clinicians to use Eli Lilly’s FDA-approved antiplatelet drug, Effient (prasugrel), for patients shown to have high residual platelet reactivity or reduced-function alleles for CYP2C19. AstraZeneca’s ticagrelor could also soon be on the alternative drug menu for antiplatelet therapy. The drug is awaiting FDA approval; a decision was expected at CAP TODAY press time, according to company spokesperson Kate Childress. Genetic variations of CYP2C19 have not been shown to affect the efficacy of prasugrel or ticagrelor.

Jeff Riesmeyer, MD, medical director of Eli Lilly, reports that the drug-maker has an ongoing clinical trial, TRIGGER-PCI, which is similar to GRAVITAS. Instead of 150 mg of clopidogrel, however, study participants with high platelet reactivity on 75 mg of clopidogrel are receiving 10 mg of prasugrel, the standard dose. “This will give us a chance to look at differences between prasugrel 10 mg and clopidogrel 75 mg in nonresponders,” he says. Like GRAVITAS, the Lilly trial is using VerifyNow to measure subjects’ platelet reactivity.

Prasugrel isn’t suitable for everyone. In the prasugrel trial, TRITON-TIMI 38, some people had a markedly increased risk of major bleeding compared with clopidogrel use, Dr. Riesmeyer says. The drug is thus contraindicated for individuals with a history of transient ischemic attack or stroke. The TRITON-TIMI 38 clinical trial showed that those age 75 and older and those who weigh less than 60 kilograms had a higher risk of TIMI major bleeding. The elderly and low-weight patients had more bleeding on clopidogrel as well, but “more so on prasugrel,” he says.

In the trial, patients with ST-elevation myocardial infarction (STEMI) or diabetes mellitus didn’t show different rates of major bleeding on the two drugs, Dr. Riesmeyer says. So the risk-benefit ratio for prasugrel is favorable for people with those two conditions.

Dr. Gurbel of Sinai Hospital agrees that “STEMI and diabetic patients appear to derive a particular clinical benefit without significantly increased bleeding.”

Dr. Price of Scripps urges caution, however, noting that the TRITON subgroup analysis for these groups of patients was somewhat small. “While STEMI and diabetic patients had particular ischemic benefit with prasugrel, whether there is no price to pay with respect to bleeding in these patients is less clear, though there’s definitely a signal there,” he says—but not one he views as conclusive.

Cost will also figure into the drug selection equation, predicts Mayo Clinic cardiologist Allan Jaffe, MD, who gave a presentation on antiplatelet therapy at the AACC meeting last July. Based on “Mayo prices,” he said in his talk, clopidogrel costs $4.65 a day while prasugrel is $5.40 a day.

But Plavix is slated to become generic in 2011. Once that happens, it could become “very cheap—pennies per pill,” Dr. Jaffe says.

Most studies don’t take into account the large number of people who can’t afford to refill their clopidogrel after hospitalization, Dr. Jaffe says. “It may turn out that the frequency of re-thrombosis goes down very substantially if one had a cheaper drug. That in and of itself could be a major breakthrough,” he says. He did note, however, that a pub­lished cost-effectiveness analysis comparing clopidogrel to prasugrel still suggested cost-effectiveness of the latter agent.

David Flockhart, PhD, MD, chief of clinical pharmacology at Indiana University in Bloomington, says one could view the work being done on genetic testing for Plavix as “Custer’s last stand for the drug.”

“If genetic testing can identify groups of people who are better treated by clopidogrel than prasugrel, then the testing could keep clopidogrel around as a cheaper, more effective antiplatelet therapy than prasugrel in those cases,” he says, noting that genetic testing for CYP2C19 is a “one-time deal” that costs about $200 to $300.

On one point, experts appear to agree: CYP2C19 genotyping can predict risk of negative cardiovascular outcomes for Plavix takers who have coronary artery stents.

Outcomes for that population as reported in the literature, says cardiologist Eric Topol, MD, “speak for themselves about people who are carriers or homozygotes for loss-of-function alleles [for CYP2C19].” Dr. Topol, director of the Scripps Translational Science Institute and chief academic officer for Scripps Health, points to the Oct. 27 meta-analysis in JAMA as the latest validation of the data.

The meta-analysis included nine studies consisting of 9,685 patients, 91.3 percent of whom had stents. Carriers of one reduced-function allele had a 167 percent increased risk of stent thrombosis (HR of 2.67); those with two such alleles showed an almost 300 percent heightened risk compared with normal metabolizers (HR 3.97). “Overall, 71.5% were noncarriers, 26.3% had 1 reduced-function CYP2C19 allele, and 2.2% had 2 reduced-function CYP2C19 alleles,” the authors wrote (Mega J, et al. JAMA. 2010; 304:1821–1830).

When you look at the stent population, says Jay Wohlgemuth, MD, VP of science and innovation at Quest Diagnostics, San Juan Capistrano, Calif., “the clinical validation of mutation testing is very, very strong and has never been refuted.” Placing a metal stent in an artery ranks as “the riskiest thing you can do to activate platelets,” he says.

As for how the CYP2C19 genotype affects clopidogrel response, a normal metabolizer or “wild type” for CYP2C19 is *1, *1, Dr. Wohlgemuth says. The most common genetic variant rendering someone an intermediate (heterozygote) or poor metabolizer (homozygote) is *2. “It falls off after that to *3 and *4,” he says.

Quest is offering the AccuType CP test to genotype patients for CYP2C19. The test includes the *1 through *5 alleles and *17. The latter, *17, represents an ultra-rapid metabolizer. Quest is planning by the end of this year to add *6 through *10 and *12. It’s making the upgrade based on an analysis that identified the frequency of CYP2C19 alleles and their functional impact on clopidogrel response. The analysis will be submitted for publication in a peer-reviewed journal.

Gwen McMillin, PhD, co-director of pharmacogenomics at ARUP Laboratories, Salt Lake City, uses AutoGenomics’ expanded panel of reagents for its CYP2C19 testing. (AutoGenomics’ FDA-cleared Infiniti CYP2C19 assay tests for *2, *3, and *17.)

ARUP’s test “covers from *2 to *10 and *17,” says Dr. McMillin, who notes ARUP saw a 40-fold increase in CYP2C19 testing after the Plavix black-box warning was issued in March. The testing remained at that level for several months before dropping about 30 percent and plateauing.

In theory, says Dr. McMillin, *17 would produce more active metabolite of clopidogrel. Thus, “if you were carrying two *17s, you might need a dose reduction.”

Guillaume Paré, MD, MSc, director of the genetic and molecular epidemiology laboratory at McMaster University in Hamilton, Ontario, says the CURE (Clopidogrel in Unstable Angina to Prevent Recurrent Events) trial found that the gain-of-function allele (*17) helped improve outcomes.

“What is really curious,” Dr. McMillin says, “is how to handle a *2, *17 genotype” in terms of how the combination of a reduced-function and ultra-metabolizer allele would affect clopidogrel metabolism. ARUP doesn’t have the numbers for how often the combination is seen, she says, but *2 tends to occur with about a 20 percent frequency, on average, and *17, 30 percent. “Compound heterozygotes with *17 exist, and we can’t predict the phenotype,” she says.

As for other genes that could affect Plavix response, Dr. Wu says a study published in Lancet recently found that ABCB1, a gene involved in drug transport, was also associated with negative outcomes for patients taking clopidogrel (Mega J, et al. Lancet. 2010;376[9749]:1312–1319).

Quest is developing a test for genotyping of ABCB1, which it intends to offer as a separate test code from the AccuType CP panel with CYP2C19 alleles. “We do believe,” says Dr. Wohlgemuth, “that it does impact drug absorption” and can thus affect how patients respond to clopidogrel. “More studies will be needed to look at ABCB1,” he says.

Other patients, Dr. Wu says, may produce enough of the active metabolite of clopidogrel and be able to transport it to the P2Y12 platelet receptor, which is the target. But the patients have polymorphisms that enable them to bypass that receptor and clot. Those patients “will still show resistance even though that receptor [P2Y12] is blocked by Plavix.” (VerifyNow measures only inhibition of that receptor.)

Dr. Topol predicts that more genes involved in clopidogrel response will be discovered. “The holes will be filled in over time,” he says.

What role does platelet function testing play in current efforts to maximize antiplatelet therapy effectiveness?

In a recent Quest-sponsored webinar, Drs. Wohlgemuth and Price pointed to studies showing that platelet function testing and genetic testing have value when used in tandem. “You can identify risk with genetics that you don’t see with platelet function testing—and the other way around,” says Quest’s Dr. Wohlgemuth.

Accumetrics’ director of scientific affairs, Jackie Coleman, PhD, agrees that platelet function testing can pick up hidden risks. She points out, for example, that some medications are known to affect clopidogrel metabolism. There’s also what she calls the “soup of interference” where a patient is taking multiple drugs. “And you don’t know how they are competing with each other for CYP450 metabolism in the liver.” Comorbidities are a consideration not flagged by genetic testing. For example, diabetes is shown to increase platelet reactivity, she says. In the GRAVITAS trial, patients who had ischemic events were at the higher end of platelet reactivity.

Eli Lilly spokesperson Tamara Hull says Daiichi Sankyo and Lilly are collaborating with Accumetrics to educate physicians about the availability of platelet function testing for their patients on antiplatelet medication. “Physicians then use their clinical judgment to see if the patients are getting the platelet inhibition needed.”

Dr. Topol says Scripps, which was the first in the nation to genotype every patient before stent placement, does platelet function testing whenever clinicians have a question about the effectiveness of antiplatelet therapy “at the platelet level.”

In the Quest webinar, Dr. Price shared a clinical anecdote showing how that approach worked well for one patient, a 60-year-old scientist. The man, who complained of chest pain brought on by bicycle riding, had had a stent placed in his circumflex artery several years earlier. Prior platelet function testing when the patient had been on clopidogrel in the past showed the patient to be a nonresponder, and subsequent genotyping showed the patient had one loss-of-function allele for CYP2C19. The man’s angiogram revealed a potentially life-threatening left main coronary lesion. Since the patient declined bypass surgery, Dr. Price was faced with having to perform a complicated stenting procedure.

Dr. Price put the patient on prasugrel and performed the stent procedure successfully. In this situation, Dr. Price concluded, “the patient could have had a risk for very early life-threatening stent thrombosis, but I was able to mitigate that risk by knowing his genotype at the time of angioplasty.”

Dr. Wohlgemuth thinks platelet function testing would be useful in evaluating patients with one reduced-function allele, such as *2, and the *17 ultra-metabolizer allele, whose effects could, in theory, cancel each other out.

Is there merit in doing platelet function testing alone to determine antiplatelet therapy response? No, Dr. Topol says, because the testing “is notoriously variable and dynamic.” Thus, you also need to know a patient’s “genomic liability,” which always remains the same.

Mayo Clinic pathologist Bradley Karon, MD, says Mayo is studying the utility of a number of platelet function tests for Plavix and aspirin. But Mayo doesn’t currently offer a clinical test for monitoring platelet function during antiplatelet therapy.

“Whether we do so depends on the outcomes of our study and other studies,” says Dr. Karon, noting there are “concerns about disagreements between tests.” Mayo Clinic physicians are also waiting for more data from studies about which test might be predictive with the appropriate cutoff for the test, he says.

“The problem with platelet function testing as a field is that there is no good true reference method for platelet aggregation,” Dr. Karon says. “So people measure different components.”

“Some assays intend to measure the overall ability of platelets to aggregate,” he explains. Others “tend to isolate ADP-induced aggregation to monitor what percent of a person’s ADP-induced platelet aggregation is being inhibited by Plavix.” Both approaches can be helpful, depending on the goal. If the aim is to monitor Plavix therapy, then just testing for ADP-inhibition may be appropriate, he says.

On the other hand, “If you want to determine which patients have activated platelets despite inhibition of ADP-induced platelet activation, then you would want to measure overall platelet function.” The latter concept, however, “isn’t well defined and there are no generally accepted methods” for doing that, he says.

Dr. Gurbel and colleagues at Sinai Hospital reviewed the topic of high platelet reactivity in a recent white paper (Bonello L, et al. J Am Coll Cardiol. 2010;56:919–933), which says in its conclusion: “Currently, platelet function testing may be considered in determining an antiplatelet strategy in patients with a history of stent thrombosis and in patients prior to undergoing high-risk PCI.” The paper does not, however, recommend routine use of platelet function testing “until the results of large-scale trials of personalized antiplatelet therapy are available.” (Dr. Gurbel and colleagues are using the Nanosphere point-of-care genotyper in the cath lab as part of TARGET-PCI, a personalized prospective antiplatelet therapy trial.)

Cardiologist John McPherson, MD, of Vanderbilt University Medical Center, says VUMC is also awaiting findings from more studies like GRAVITAS before deciding where platelet function testing “fits in our routine practice.” But as part of a large quality improvement initiative, VUMC recently began performing genotyping on all patients who may require a stent in the future, with the idea of treating them differently prospectively. (The genotyping may in the future include efforts to predict warfarin response and risk of developing statin-related myopathy, he adds.)

At this point, the value of CYP2C19 genotyping for non-stent patients remains unclear. Dr. Wu notes that a study reported in the Sept. 10 New England Journal of Medicine by Dr. Paré and colleagues showed that in the setting of atrial fibrillation or acute MI, people with a loss-of-function allele for CYP2C19 didn’t experience different outcomes (Paré G, et al. N Engl J Med. 2010;363:930–942). The study examined the efficacy and safety of clopidogrel compared with placebo according to genotype among patients in two trials: CURE and Active A (Atrial Fibrillation Clopidogrel Trial with Irbesartan for Prevention of Vascular Events).

Those findings do not exactly come as a shock to Quest’s Dr. Wohlgemuth. In patients who just had a heart attack or near heart attack and aren’t getting stents, “the incremental benefit of antiplatelet therapy on top of everything else going on is not huge. The patient comes in and is likely on low-molecular-weight heparin and aspirin and multiple [cardiac medications], such as beta blockers and ACE inhibitors.”

And the atrial fibrillation population probably involves clotting that’s “more thrombin-driven as opposed to platelet-driven,” he adds.

Dr. Paré agrees that the need for antiplatelet therapy may differ based on cardiovascular condition.

“We don’t have a full understanding of the antiplatelet dose-response relationship and the level of platelet inhibition required to prevent different types of ischemic events,” he says. The possibility exists that for some ischemic events, genetic variations causing a slight decline in clopidogrel metabolism won’t produce a “marked effect on the number of ischemic events.”

“In other clinical situations,” Dr. Paré says, “such as perhaps stent thrombosis, individuals might be on a steeper part of that curve [where] differences in clopidogrel metabolism due to genetic variants might lead to more pronounced differences in efficacy.”

Moving away from the genetic causes of stent rethrombosis, Dr. Jaffe argued in his AACC presentation and in a followup interview with CAP TODAY that there are many other reasons why people could be at risk for that problem. Stent mechanics is one. “If one doesn’t have a stent that is adequately expanded, the stent will develop procoagulant activity that won’t have much to do with the polymorphisms.”

“It’s also well known that there are patients who have increased thrombin activity in circulation, which would stimulate platelet aggregation,” Dr. Jaffe says. “And one might confuse that with a primary platelet problem.” While clinicians may not have aggressively pursued that possibility in the past due to lack of oral medications to counter increased thrombin activity, the FDA recently approved such a drug (dabigatran) for atrial fibrillation, Dr. Jaffe says.

Another drug that looks promising but is not yet FDA approved, he adds, is rivaroxaban.

Currently, however, there are no good tests for thrombin, he says. Clinicians usually order APTT for that purpose. “There is a thrombin time test, but none of the testing is particularly good or specific for thrombin activity.” There are research assays and others in development that could improve thrombin measurement.

Dr. Jaffe hypothesizes that “over time, as more data accumulate,” clinicians will measure a “heterogeneity of mechanisms,” including thrombin activity, to figure out what’s going on with individual patients. “And we will pick and choose what we use to treat patients.”

As for CYP2C19 genotyping, “time will tell,” says Scripps’ Dr. Topol, whether clinicians will switch from clopidogrel to alternative antiplatelet drugs that do not require the testing. And those drugs, he adds, will have different genomic stories.