This article is more than 5 years old. Information may no longer be current.
Future unclear for platelet function and genetic testing for patients with ACS, stents
Click Here to Manage Email Alerts
Clopidogrel is often the antiplatelet agent of choice for patients with ACS or those receiving a stent because, as a generic, it is inexpensive and, as a drug that has been around for a long time, it has shown consistent efficacy in most patients. However, 20% to 30% of patients have a genetic variant, CYP2C19*2, that is associated with reduced activation of clopidogrel and the antiplatelet response to clopidogrel.
It has been hypothesized, based on observational data and knowledge of the mechanism of action of clopidogrel, that testing patients for the CYP2C19*2 variant and giving those without the variant clopidogrel and those with the variant a newer, more expensive antiplatelet agent such as prasugrel (Effient, Daiichi Sankyo/Eli Lilly) or ticagrelor (Brilinta, AstraZeneca) would lead to better outcomes. It also was hypothesized that testing for platelet function to guide the choice of antiplatelet agent and dosing would improve outcomes.
Paul A. Gurbel, MD, discussed results of
recent trials that studied personalized
antiplatelet therapy.
Image: Scott Wendler, LifeBridge Health
However, recent randomized prospective trials yielded results that did not confirm the hypotheses. The GRAVITAS trial data, published in 2010, showed that outcomes were not affected when patients with high platelet reactivity who underwent PCI were switched to a higher dose of clopidogrel. Results from the ARCTIC trial, published in 2012, did not show that platelet function monitoring affected ischemic outcomes. Findings from ARCTIC-GENE, a substudy of the ARCTIC population, presented at the European Society of Cardiology Congress in 2013, did not show that genotyping for the CYP2C19*2 variant led to better outcomes for death, MI, stroke, stent thrombosis and urgent revascularization at 1 year after stent implantation.
“That is confounding the picture,” Robert Roberts, MD, FRCPC, MACC, FAHA, president and CEO of the University of Ottawa Heart Institute, professor of medicine at the University of Ottawa and director of the Ruddy Canadian Cardiovascular Genetics Centre in Ottawa, Ontario, Canada, told Cardiology Today.
The outlook for genetic profiling and platelet function testing for response to clopidogrel may come down to whether the cardiology community is willing to use testing based on observational and pharmacological data, or whether it insists on randomized, prospective trial results for proof of concept.
ARCTIC-GENE
For the ARCTIC-GENE trial, Jean-Philippe Collet, MD, PhD, professor of cardiology at the Pitié-Salpêtrière Hospital in Paris, and colleagues analyzed DNA samples from 1,420 patients who were genotyped for CYP2C19 genes related to platelet reactivity. Patients were stratified by whether genotyping predicted them as slow or fast metabolizers of clopidogrel.
The rate of the primary composite endpoint (death, MI, stroke, stent thrombosis and urgent revascularization) at 1 year was 32.7% in slow metabolizers compared with 32.2% in rapid metabolizers (HR=0.99; 95% CI, 0.81-1.2). The rate of death or MI was 30.5% in slow metabolizers compared with 29.3% in rapid metabolizers (HR=0.97; 95% CI, 0.79-1.18). The rate of stent thrombosis or urgent revascularization was 4.4% in slow metabolizers compared with 5.7% of rapid metabolizers (HR=1.31; 95% CI, 0.79-2.2), according to data presented at ESC 2013.
There was no difference between the groups in bleeding outcomes. All types of bleeding were observed in 3.5% of slow metabolizers compared with 3.6% of rapid metabolizers (HR=1.05; 95% CI, 0.58-1.9), major bleeding occurred in 3.1% of slow metabolizers compared with 2.2% of rapid metabolizers (HR=0.74; 95% CI, 0.38-1.45) and minor bleeding occurred in 0.9% of slow metabolizers compared with 1.7% of rapid metabolizers (HR=1.98; 95% CI, 0.66-5.93).
Paul A. Gurbel, MD, director of the Sinai Center for Thrombosis Research at Sinai Hospital in Baltimore, professor of medicine at Johns Hopkins University School of Medicine, and adjunct professor of medicine at Duke University School of Medicine, cautioned against interpreting the ARCTIC-GENE findings as definitive proof that personalized antiplatelet therapy has no benefit.
“The limitations of ARCTIC are the patient population, which was overall low risk, and the protocol used to address high platelet reactivity,” Gurbel said. “Choice of therapy was left up to the doctor, and was not specified by rigorous protocol, and most patients [with high platelet reactivity] did not get prasugrel but double-dose clopidogrel, which we know is a suboptimal strategy to treat high platelet reactivity. The study was very similar to GRAVITAS with respect to patient risk and therapy for high platelet reactivity. I am not surprised that ARCTIC-GENE did not show any positive relation between genotype and outcomes. This genetic substudy of ARCTIC is also not earth-shattering because we’ve already known that the genotype is far from a perfect surrogate for the phenotype. This is not new information here.”
TRANSLATE-POPS
Findings from another recent study showed that access to platelet function testing led to increased changes in antiplatelet therapy, but it did not affect short-term outcomes.
The TRANSLATE-POPS study results were presented at TCT 2013. Researchers evaluated whether routine availability of platelet function testing altered physician selection and dosing of antiplatelet therapy in patients treated with PCI after acute MI, and whether access to testing affected bleeding complications or major adverse cardiac events at 30 days. The answer to the first question was yes; the answer to the second was no.
There were 2,013 patients in the device arm, in which clinicians had free and routine use of platelet function testing, and 1,853 patients in the usual care arm, in which clinicians did not have free and routine use of the test but could order the test if desired. Platelet function testing was performed in 66% of patients in the device arm compared with 1.4% of those in the usual care arm. Patients in the device arm were more likely to have an in-hospital therapeutic adjustment of their antiplatelet regimen (15.9% vs. 11.6%; P=.01) and had a higher rate of switching antiplatelet agents (14.5% vs. 10.6%; P=.02).
At 30 days, there was no difference between the groups in rates of major adverse cardiac events (device arm, 4.5%; usual care arm, 5.1%; OR=0.93; 95% CI, 0.66-1.31) or in rates of bleeding events (device arm, 4.2%; usual care arm, 4.3%; 95% CI, 0.5-1.26), according to data reported by Tracy Wang, MD, MHS, MS, of the Duke Clinical Research Institute in Durham, N.C.
Matthew J. Price
This study, too, has limitations, some experts said.
“This trial … showed that people used [platelet function testing], but the problem is that there was no prescription about how to use it,” Matthew J. Price, MD, director of the cardiac catheterization laboratory at Scripps Clinic in La Jolla, Calif., told Cardiology Today. “If you give a doctor a test but you don’t tell them how to use or interpret it, and you have a lot of publicity saying that the test doesn’t work, I don’t think a lot of people are going to use it.”
The ‘right’ populations
The issue, experts said, may be that the treatment effect is real in certain patients, but the trials are not high-powered enough to detect it.
“If you look at the current trials, the endpoint is bleeding as well as thrombotic effect,” said Roberts, a Cardiology Today Editorial Board member. “In ARCTIC … the incidence of thrombosis was 1%. We will not expect to see a big difference in that group because 1% is so low. We would need several thousand people to determine whether the incidence of thrombosis is lower in someone who is found to have the genetic defect and then given a new drug such as prasugrel, which is not dependent on that defect. The reason [these studies] have a combined endpoint is because you can decrease the sample size.”
To get a true answer, Roberts said the endpoints of thrombosis and bleeding will have to be separated out because “if you increase the dose, even if you decrease incidence of thrombosis, you’re always going to increase bleeding. The endpoints may balance out for that reason.”
Gurbel, a member of the Cardiology Today’s Intervention Editorial Board, agreed. “The bottom line is that these studies are underpowered for post-discharge event rates,” he said. “They’re just not large enough in a low-risk population like this to refute the utility of platelet function testing.”
In a letter to The New England Journal of Medicine, Raffaele De Caterina, MD, PhD, of Gabriele d’Annunzio University, Chieti, Italy, and colleagues estimated that it would take 17,540 patients in each study group.
“Who’s going to fund that?” Gurbel asked. “The NIH? No. Who in private industry can really afford this? I don’t know. I would say that we may never get it.”
Since a prospective trial large enough to accurately assess the utility of platelet function testing and/or genotyping for antiplatelets in all patients with ACS or stents may not happen, it might be a better strategy to focus on subpopulations that have shown improved outcomes, Roberts said. The most obvious is those who have two copies of the CYP2C19*2 variant.
“It would be good to see that there is that suggestion in that small homozygous group,” he said. “If there was no difference there, that would suggest that clopidogrel is not working by the action that we think it is. The data are overwhelming that it is working through activation, which depends on that CYP transformation.”
By their nature, trials of all types of patients who have ACS or are undergoing PCI are not equipped to identify these subpopulations, Dan Roden, MD, assistant vice chancellor for personalized medicine and professor of medicine and pharmacology at Vanderbilt University School of Medicine, told Cardiology Today.
Dan Roden
“When you do personalized medicine in a population, what ends up happening is that if there are small subsets that benefit, it’s hard to identify them,” said Roden, a Cardiology Today Editorial Board member. “Either the variability in clopidogrel or other antiplatelet response is largely unexplained by genetic or implementation factors, or there are lots of other factors that contribute to variability across the broad population that we’re not understanding or not capturing in these trials.”
Clinical utility of testing
The question of whether platelet function tests or genetic profiling should be performed to assess potential response to clopidogrel remains unanswered. There are those who will not do so because the data from prospective trials do not support it, but there are those who will do so because of observational data or knowledge about the pharmacology of clopidogrel.
“A lot of people are fixated on the lack of prospective evidence,” Gurbel said. “If it’s not a Class I recommendation in the guidelines, a lot of people won’t do it. When I give lectures on this subject, the majority of people that I speak to do not genotype or phenotype the patients. They are waiting for a positive trial. The absence of robust prospective information is the rate-limiting step right now.”
Without that, widespread adoption may never happen, Roberts said.
Robert Roberts
“If you look at all the data we have at the moment, you have to conclude that this doesn’t seem to make a big difference, even though it goes against what I believe and what I really do think,” he said. “You have to consider the evidence-based medicine. The sample size is probably not large enough, but we don’t have much of a suggestion at the moment that it’s going to make a difference by doing the testing. So I think the real world will go on without it. As much as we [geneticists in cardiology] hoped that our poster child would be antiplatelet therapy, it is not shaping up that way.”
Another issue is no standardization for platelet function testing technology, according to Price, an investigator for the GRAVITAS trial. “Not all tests are created equal; several use different measurements and standards,” he said. “It’s important when we apply a certain platelet function test to make sure its mechanism of action is validated and that there is clinical data to support the utility of the device.”
However, there are doctors who are not deterred by the lack of results from prospective trials. Gurbel said he uses platelet function testing in high-risk patients, including those with complex coronary anatomy undergoing PCI, those with stents in the left main artery and those with poor ventricular function.
“The reason I do this is because of the fundamental evidence that we have in our data set that treatment of these patients with just aspirin, if they undergo a stent implant, is worse than treating them with dual antiplatelet therapy,” he said. “If I have evidence of an absent pharmacodynamic effect from clopidogrel therapy, in effect, the sole antiplatelet influence to protect the patient from stent thrombosis is just coming from aspirin. Despite the fact that we don’t have a positive clinical trial demonstrating that altering therapy based on platelet function testing improves outcomes, I still don’t think it’s a good idea to send patients out of the hospital with a foreign body in their coronary artery if they have essentially just aspirin monotherapy on board.
“Platelet function testing, I believe, can be useful when used selectively; I think the totality of the data supporting it are highly consistent,” Gurbel said. “Most recently, we had the data from the large-scale ADAPT-DES registry supporting what we observed 10 years ago in a much smaller group of PCI patients. ADAPT-DES clearly showed that discharge platelet function is associated with an approximate HR of almost 4 for 30-day stent thrombosis. That is a significant increase. Similar increased risk associated with high platelet reactivity has been shown in numerous studies. It has been shown with VerifyNow (Accumetrics). It has been shown with conventional aggregation. It has been shown with impedance aggregometry. I don’t think this relation of high platelet reactivity to thrombotic risk is debatable anymore. What everyone argues about is, is it a modifiable risk factor?”
Although some may prefer to prescribe a newer antiplatelet instead of clopidogrel if there are concerns about clopidogrel’s efficacy, in some cases it may not be feasible, according to Price, who also is a member of the Cardiology Today’s Intervention Editorial Board.
“We know that particularly in the setting of ACS, non-responsiveness to clopidogrel is strongly associated with increased thrombotic events,” he said. “We have newer agents like ticagrelor and prasugrel, which both have been shown to be superior in those cases of ACS compared with clopidogrel. However, these agents are expensive, and they also have adverse effects, including more bleeding. In the real world, where many patients are struggling to pay for their medicines, we are under increasing demand to use effective and less-expensive strategies for our patients.”
Roden said one thing that has been learned from experience is that it is useful to have genetic or platelet function data on patients who already have stents. “If we are going to try to identify people in whom the genetic information is most valuable, who is that going to be? It’s going to be people who are highly likely to get a stent. And who is highly likely to get a stent? Someone who already has a stent.”
No easy answers
The jury is still out on platelet function testing and genetic profiling, and may continue until new data are available. The next step is to identify the research gaps and to determine whether they are worth pursuing.
“In low-risk patients, the absolute risk rates are low enough that although high reactivity is predictive of outcomes, it’s probably not worth testing and changing the therapy of all these patients,” Price said. “What remains unaddressed by clinical trials and is perhaps a very good area for the application of platelet function testing is to help select which antiplatelet to give in patients presenting with ACS and undergoing PCI.”
However, the nature of pharmacogenomics may not lend easy answers to the problem, Roden said.
“The fundamental problem is that people are looking for black-and-white answers in this domain when the likelihood is that the variability in response to these drugs is going to be multifactorial,” Roden said. “In some patients, single-gene variants will play a large role. In other patients, lots of genetic variants plus environmental factors will contribute to variability in outcomes. The desperate search for a single best marker that will show superiority of one intervention over another is in some ways misguided.”
What is clear is that no amount of data will ever totally replace clinical judgment.
“Should we throw all fundamental core information away because we don’t have a prospective, positive trial? I say no,” Gurbel said. “If a patient’s platelet function is high and they don’t respond to clopidogrel, but you send that patient home with the antiplatelet effect confined to aspirin therapy, I don’t see that as rational.” – by Erik Swain
Disclosure: Gurbel’s institution receives funding from Accumetrics, AstraZeneca, Bayer, Daiichi Sankyo/Eli Lilly, Hemonetics and Sanofi-Aventis. Price has an equity interest in Iverson Genetics and reports receiving consulting honoraria from Accumetrics, AstraZeneca, Daiichi Sankyo/Eli Lilly, Medicure and The Medicines Company, and speaking honoraria from AstraZeneca and Daiichi Sankyo/Eli Lilly. Roberts and Roden report no relevant financial disclosures.