In a 2-part series, TCTMD will explore how the GRAVITAS trial is influencing the field of functional platelet testing. Part 1 summarizes the trial and its aftermath, including explanations for the results and remaining issues. Next up, part 2 will address the risks of using platelet function tests to guide therapy in lieu of randomized trial data and discuss their future in clinical practice.

At the American Heart Association Scientific Sessions 2010 this past November in Chicago, one of the most keenly awaited late-breaking trials delivered a blow to the attractive concept of adjusting antiplatelet therapy following percutaneous coronary intervention (PCI) based on functional measurement of platelet reactivity.

GRAVITAS (Gauging Responsiveness with A VerifyNow assay—Impact on Thrombosis And Safety) promised to answer the question of whether a relatively straightforward strategy for managing the almost one-third of patients who have high platelet reactivity despite receiving standard dual therapy after PCI—and thus are at substantially increased risk of thrombotic events—was in fact effective.

The trial had the potential to change practice. After mixed signals from several small observational studies, here were data from the first large randomized trial to close the circle between pharmacodynamic and clinical data by rigorously testing whether adjusting clopidogrel dose based on results of a point-of-care functional test would improve patient outcomes.

The answer, reported by lead investigator Matthew J. Price, MD, of Scripps Clinic (La Jolla, CA), was no, bringing with it larger implications for the future of this nascent field.

Disappointing Results—and Plenty of Explanations

In the targeted cohort with high on-treatment reactivity, at 6 months rates of the primary endpoint, a composite of cardiovascular death, MI, or stent thrombosis, were identical for the high- and standard-dose clopidogrel groups: 2.3%.

For the trial, almost 5,500 patients with stable or unstable CAD underwent DES implantation and received a 600-mg loading dose of clopidogrel (if they were drug naïve). Testing 12 to 24 hours later using the VerifyNow P2Y12 assay (Accumetrics, San Diego, CA) identified 2,214 patients (41%) with high residual reactivity, defined as a P2Y12 reaction unit (PRU) value equal to or above 230. That cohort was then randomized to continue on the standard 75-mg maintenance dose or receive another 600-mg loading dose and then 150 mg daily, both for 6 months.

While GRAVITAS is clearly not the final word on the use of functional testing to guide therapy, several commentators noted, at the least it invalidated the particular strategy tested and put routine point-of-care platelet testing on hold. Meanwhile, they said, it is important to try to understand why the trial was negative in order to move forward. And there has been no shortage of proposed explanations, with attention directed at nearly every aspect of the trial design, from the patients studied to the antiplatelet regimen used and the reactivity cutpoint chosen.

Analysis confirmed that the more aggressive antiplatelet regimen did in fact have a pharmacodynamic impact. At 30 days after PCI, platelet reactivity had decreased in both arms, but the decline was steeper in those who received the double clopidogrel dose, and the difference remained largely unchanged at 6 months. At 1 month, high residual reactivity persisted in only 40% of the high-dose group compared with 62% of the standard-dose group (P < 0.001), for an absolute reduction of 22%.

Significantly, however, even in the high-dose arm, the absolute reduction in PRU was modest, declining from a median of about 280 at baseline to about 200, Dr. Price reported. Moreover, the relatively small absolute difference in median platelet reactivity between the 2 arms—about 40 PRU—was reflected in the fact that rates of GUSTO severe and moderate bleeding were similar (1.4% in the high-dose arm vs. 2.3% in the standard-dose arm; P = 0.10). 

Weighing in on the Trial Design

Interestingly, when cardiovascular events were plotted according to PRU, all events were seen to occur at values above 180 PRU. “That suggests that the [reactivity] threshold concept is real,” said Paul A. Gurbel, MD, of the Sinai Center for Thrombosis Research (Baltimore, MD), in a telephone interview with TCTMD. “Once you get above a certain cutpoint, risk really goes up.

“The problem [in GRAVITAS] was that even with 150-mg clopidogrel [maintenance] a lot of patients remained over that cutpoint,” he continued. “If you have a low event rate and you’re using a therapy that only gets about half of patients with high reactivity down below the cutpoint, the chances of seeing a clinical effect are scant. If they had chosen a lower cutpoint and used a more effective drug, they may have seen a signal.”

William A. Gray, MD, of Columbia University Medical Center (New York), NY, made a similar point in a video review of the trial for TCTMD. Parsing possible reasons why GRAVITAS was negative, he suggested that the investigators might have done better to look not at patients who received double-dose clopidogrel but specifically at those whose PRU actually fell below the 230 cutpoint.

In fact, Dr. Price has reported that a subanalysis is now underway to examine whether patients who had a more robust pharmacodynamic response to the higher dose also obtained better outcomes. 

Clearly, “the 230 cutpoint “is not [written] in stone,” Dr. Gray commented. “It’s predicated on prior historical event rates and observational studies. And as we saw from this trial, all the events occurred [in patients with a PRU] over 200. So maybe that should be the [cutpoint].” Other commentators concurred that lowering the cutpoint would be a reasonable approach.

The investigators’ power assumptions—a 5% event rate in patients on standard clopidogrel and a 50% risk reduction with high-dose drug—also came under fire as being overly optimistic. The low event rate observed in the standard-dose arm of only 2.3% set a near-impossible bar for double-dose patients to surpass, several commentators suggested. They added that increasing the sample size and/or lengthening follow-up might boost the event rate, improving the odds of seeing a difference between the treatment groups, but it would also expose even more patients to the possibility of bleeding when there was no hint of ischemic benefit.

“I suspect that we’re just not seeing enough events to make [greater inhibition] critical,” concluded Neal S. Kleiman, MD, of the Methodist DeBakey Heart and Vascular Center (Houston, TX), in a telephone interview with TCTMD.

In general, outcomes in elective PCI patients—which most GRAVITAS subjects were—are excellent, and as a result it is difficult to show a meaningful clinical improvement with a more intensive antiplatelet regimen, observed Sunil V. Rao, MD, of Duke University Medical Center (Durham, NC) in a telephone interview with TCTMD.

Building on GRAVITAS

Negative trials are very important because they help sharpen the focus of future research, Dr. Rao continued. “GRAVITAS answered a broad question, but now we need to get a little more sophisticated about what that question is,” he explained. “Who will benefit more [from test-guided escalation of therapy]? It’s probably high-risk patients, like ACS and maybe STEMI patients. And how will they benefit? It’s probably going to [require] more intensified platelet inhibition.”

In fact, after GRAVITAS, many experts were pinning their hopes on the more aggressive strategy of the TRIGGER-PCI trial. It not only set a PRU threshold of 208 but was testing the more potent antiplatelet agent prasugrel in elective PCI patients with high platelet reactivity on clopidogrel.

Recently, however, after an interim analysis showed that the event rate was even lower than in GRAVITAS, the trial was halted. “I think for stable CAD, this settles the question,” said principal investigator Gregg W. Stone, MD, of Columbia University Medical Center (New York, NY) in a telephone interview with TCTMD for a previous story. “In this population, on the basis of both GRAVITAS and TRIGGER-PCI, a personalized approach to platelet function analyses does not appear to be necessary.”

Nonetheless, other, ongoing randomized trials are exploring different approaches to test-guided therapy and may provide insights into the question of whether functional testing holds any future role. These include:

• ARCTIC, in which clopidogrel dosage will be adjusted for elective patients with a suboptimal response to initial therapy based on a VerifyNow cutpoint of 235 PRU
• DANTE, in which ACS patients will be randomized to a standard or double maintenance dose of clopidogrel based on VerifyNow-measured residual reactivity
• TARGET PCI, in which PCI patients determined by genotyping or serial functional testing (230 PRU cutpoint) to have high residual platelet reactivity will be switched to prasugrel

Unresolved Issues

One problematic aspect of functional testing in general is its timing. In GRAVITAS, tests were performed within 24 hours of PCI. Yet platelet reactivity declined considerably over 30 days even in patients who received standard clopidogrel, potentially diluting any clinical advantage for the double dose. Assessment at a single time point does not take account of the evolution of platelet reactivity after PCI, observed GRAVITAS coauthor Dominick J. Angiolillo, MD, PhD, of the University of Florida College of Medicine-Jacksonville (Jacksonville, FL) in a telephone interview with TCTMD.

In fact, Dr. Gurbel called the questions of when and how often to monitor platelet reactivity a “black hole” of research. “Disease state varies over time, and the reactivity cutpoints [associated with ischemic risk] may be different around the time of an ACS event than they are for stable CAD,” he pointed out.

Dr. Kleiman agreed, commenting, “In my view, if somebody has just had an acute coronary syndrome or a sent thrombosis, that’s absolutely the wrong time to test them.”

Supporting that view are results of a recent Italian trial (Campo G, et al. J Am Coll Cardiol. 2011;57:2474-2483) in which high on-clopidogrel platelet reactivity (defined as a  PRU of ≥ 235) at  baseline frequently fell below that threshold by 1 month. Moreover, 1-month reactivity levels were found to be stronger predictors of long-term outcomes.

Given this temporal variability, a proposed way to rescue a test-guided antiplatelet strategy is to treat patients to a target level of reactivity, adjusting therapy based on the results of serial testing. A hint that such an approach might work comes from a small, nonrandomized study of ACS patients undergoing PCI (Bonello L, et al. J Am Coll Cardiol. 2010;56:1630-1636). Among poor responders to a 600-mg loading dose (about one-third of whom carried at least 1 loss-of-function CYP2C19 allele), the loading dose was repeated up to 4 times—guided by reactivity monitoring—until 88% of them reached ‘adequate’ platelet inhibition (defined as a VASP [vasodilator stimulated phosphoprotein] index < 50%).

A serial testing approach certainly has intuitive appeal, said Deepak L. Bhatt, MD, MPH, of Brigham and Women’s Hospital (Boston, MA), in a telephone interview with TCTMD. “After all, if you start a patient on antihypertensive medications, you have to check to make sure they’re working,” he noted. But the strategy is not evidence-based, he cautioned, adding that not only does repeated testing entail considerable cost but it may have unintended consequences. “It could end up increasing bleeding risk without decreasing ischemic outcomes,” he observed. “This approach might ultimately be validated, but we need to wait for positive data from a large, well-designed, well-executed clinical trial.”  

At This Point, Why Test?

Extrapolating from GRAVITAS, Dr. Bhatt asked rhetorically, “If [at this point] you can’t do anything useful with the knowledge that a functional test provides, why would you test at all?” Dr. Rao also emphatically discouraged testing, saying, “Physicians who routinely use VerifyNow to tailor therapy are practicing in an evidence-free zone.” Citing current guidelines, Dr. Angiolillo agreed.

Nonetheless, GRAVITAS does not in itself sound the death knell for functional testing, Dr. Bhatt said, and in fact all the experts supported use of functional testing as a research tool.   

For now, it may also have a limited but practical role in the clinic. To the extent that physicians perform functional testing today, the main reason is to check up on patient compliance, reported Etisham Mahmud, MD, of UCSD Medical Center (San Diego, CA). “If a patient measures zero [platelet inhibition], I may give them a big loading dose of clopidogrel and then have them come back the next day and measure again to find out if they are a true nonresponder or were just not taking their medication,” he told TCTMD in a telephone interview.

The larger hope of using testing to guide therapy, however, faces many hurdles. “I’ve always felt there will ultimately be some role for testing, but we need to wait for the evidence rather than just act on gut feelings,” Dr. Bhatt stressed, offering as a cautionary example the discredited therapy for premature ventricular contractions (PVCs). Suppressing these dangerous arrhythmias seemed to make sense, he explained. The only problem was that the drug used for that purpose ended up causing even worse arrhythmias.

Similarly, “it’s logical to think that responding to [inadequate] antiplatelet inhibition on a point-of-care assay [with higher antiplatelet doses or a more potent drug] should improve patient outcomes,” Dr. Bhatt said. “But it’s tricky, because more isn’t always better. There may also be more bleeding.” Furthermore, the risk-benefit ratio could vary widely based on patient presentation—for example, elective vs. urgent PCI with a DES, he added.

The second half of this feature, which will appear Monday, July 25, will discuss how these debates play out in clinical practice.

 

---

Sources:
1. Price MJ. Standard versus high-dose clopidogrel according to platelet function testing after PCI: Results of the GRAVITAS trial. Presented at: American Heart Association Scientific Sessions; November 16, 2010; Chicago, IL.

2. Price MJ, Berger PB, Teirstein PS, et al. Standard- vs high-dose clopidogrel based on platelet function testing after percutaneous coronary intervention. JAMA. 2011;11:1097-1105.

3. Gurbel PA and Tantry US. An initial experiment with personalized antiplatelet therapy. JAMA. 2011;11:1136-1137.

 

 

Related Stories:

• GRAVITAS: ‘Personalized’ Clopidogrel Dosing Yields No Benefits in PCI Patients
• High-Dose Maintenance Clopidogrel Lowers Platelet Reactivity, Other Risk Factors
• Dose Escalation Overcomes Poor Clopidogrel Response in Most Patients