Download this article's Factoid (PDF & PPT for Gold Subscribers)

In more than one-quarter of patients undergoing percutaneous coronary intervention (PCI), levels of on-clopidogrel platelet reactivity change over the first month. But an algorithm integrating baseline platelet reactivity, genetic polymorphisms affecting clopidogrel metabolism, and key clinical variables can predict 1-month platelet reactivity, itself a strong determinant of long-term clinical outcomes.

The findings appear in the June 21, 2011, issue of the Journal of the American College of Cardiology.

Investigators led by Gianluca Campo, MD, of Azienda Ospedaliero-Universitaria Sant’Anna (Ferrara, Italy), looked at on-clopidogrel platelet reactivity levels at multiple time points as well as genotype and outcomes in 300 NSTEMI patients undergoing PCI between December 2008 and May 2009.

Blood samples were taken at baseline (just before PCI) and at 1 and 6 months. Platelet reactivity was measured using the VerifyNow P2Y12 assay (Accumetrics, San Diego, CA). In addition, gene polymorphisms affecting clopidogrel absorption (ABCB1) and metabolism (CYP2C19*2 and *17, and CYP3A5*3) were identified for each patient.

From Poor to Full Response

On-clopidogrel platelet reactivity varied significantly over time. Eighty-three patients (most of whom presented with NSTE ACS) had a change in responsiveness status over the first month, the great majority (90%) switching from poor response (defined as ≥ 235 platelet reaction units [PRU]) to full response. On the other hand, only 2 patients changed status between 1 month and 6 months. Viewed as a continuous variable, PRU was 190 ± 97 at baseline and fell to 147 ± 85 at1 month (P < 0.01) and then just slightly to 146 ± 85 at 6 months (P = 0.9 for 6 months vs. 1 month). Interestingly, although the 1-month decline was more pronounced in NSTE ACS patients (from 209 ± 96 to 159 ± 87; P < 0.01), it also occurred in stable patients (from 161 ± 91 to 128 ±78; P < 0.01).

Both clinical and genetic factors influenced platelet reactivity. Among the former, the independent predictors of on-clopidogrel variation over time were:

• Age
• Diabetes
• Creatinine clearance
• Admission for NSTE ACS

Not surprisingly, ACS had a stronger impact at baseline (about 15%; P = 0.07) than at 1 month (about 10%; P for difference = 0.07) or 6 months (about 7%; P for difference = 0.02) as patients’ hypercoagulability “cooled down” over time.

The influence of the 2 CYP2C19 variants was constant over time, accounting for about 18% of the variation in platelet reactivity, while that of ABCB1was highest at baseline (9%) and decreased gradually over time (6% at 1 month and 5% at 6 months; P = 0.04 for difference between baseline and later times).

In terms of ischemic events, 21 patients (7%) experienced the composite of death, MI, or stroke between 1 month and 1 year, while 4 suffered Academic Research Consortium-defined definite or probable stent thrombosis (1.3%). After multivariable analysis, only admission for NSTE ACS (HR 3.5; 95% CI 1.2-9.6; P = 0.04) and high on-clopidogrel platelet reactivity were independent predictors of ischemic events. The latter was a stronger predictor at 1 month (HR 28.5; 95% CI 8-104; P < 0.01 as a categorical variable) than at baseline (HR 3.1; 95% CI 1.3-7.3; P = 0.02). Importantly, in survival analyses, poor clopidogrel responders at 1 month had suboptimal outcomes, whereas poor responders at baseline who became full responders at 1 month had an excellent prognosis with a very low event rate.

In addition, there were 19 TIMI bleeding events (6.3%), 4 of which were major. Multivariable analysis yielded 3 independent predictors of composite bleeding endpoints:

• Age (HR 1.03; 95% CI 1.02-1.04; P = 0.04, as a continuous variable)
• CYP2C19*17 gain-of-function genotype (HR 2.3; 95% CI 1.03-5.3; P = 0.03)
• On-clopidogrel platelet reactivity measured at 30 days (HR 0.94; 95% CI 0.93-0.95; P = 0.04, as a continuous variable)

Defining a Therapeutic Window

The range of 1-month platelet reactivity determined to produce the lowest combined incidence of ischemic and bleeding events was defined by a PRU between 86 and 238. In addition, because a poor response at 1 month was identified as the strongest predictor of adverse outcomes, a risk score of baseline characteristics was constructed to predict both 1-month reactivity and outcomes. In this model, each of the following factors was assigned 1 point:

• Carriage of an ABCB1 and CYP2C19*2 allele (carriers)
• Baseline on-clopidogrel platelet reactivity (> 258 PRU)
• Creatinine clearance (< 52 ml/min)

For example, based on this algorithm, patients with a score of 0 were predicted to have a 1-month PRU of 109 ± 77 and no adverse events, while those with a score of 3 were predicted to have a 1-month PRU of 248 ± 87 and experience a 35% rate of adverse events.

In an accompanying editorial, Dominick J. Angiolillo, MD, PhD, of the University of Florida College of Medicine-Jacksonville (Jacksonville, FL), writes that while the study raises concerns about the value of functional platelet testing in the periprocedural period, assessment of reactivity following hospital discharge, especially after 1 month, may not be practical in the real world. For that reason, “future investigations are warranted to better understand if a narrower time frame following PCI can be identified to minimize false-positive connotations of poor-responder status,” he observes. On the other hand, “prompt identification of true poor responders is also important, given the higher risk of recurrent events in the earlier phases after PCI,” he adds.

Although the research is a step forward in defining a therapeutic window for platelet reactivity with the VerifyNow assay, “larger studies are needed to provide a better determination of the optimal cutoff values . . . as well as their prognostic value, which may potentially vary according to a specific clinical setting.”

Linking Reactivity, Genes and Clinical Outcome a First

The research represents the first time that the parameters of changing platelet reactivity, genetics, and clinical outcomes have been analyzed together, Udaya S. Tantry, PhD, of Sinai Hospital of Baltimore (Baltimore, MD), told TCTMD in both e-mail and telephone interviews.

However, a marked decline in platelet reactivity from baseline over the first month has been observed in several studies, Dr. Tantry pointed out, adding that this pattern is likely due less to changes in clopidogrel response per se than to the diminution of the traumatic effect of stenting or the ‘cooling down’ of an ACS event.

In part for that reason, Dr. Tantry commented, the difference between European and US practices makes comparisons of studies done in the different settings difficult. In Europe, clopidogrel is usually loaded and platelet reactivity measured before treatment, whereas in the United States, a clopidogrel load usually is administered after stenting and platelet reactivity assessed 6 to 24 hours later. Moreover, he said, the failure of the analysis to include periprocedural adverse events was quite surprising, given the high risk during that period.

It is important to understand that the optimal platelet reactivity target may vary over time, Dr. Tantry said. For example, he explained, “lower on-treatment platelet reactivity may be optimal in the early period following ACS and/or PCI, whereas the same low level may not provide the same clinical advantage 6 months later due to excessive bleeding.”

Multiple Testing the Best Strategy

The typical evolution of reactivity over time suggests a 2-stage strategy, starting with more potent inhibition, eg, a higher clopidogrel dose or prasugrel, and then scaling back if further testing shows a decline in residual reactivity, Dr. Tantry acknowledged. But on one hand, as GRAVITAS showed, the first stage may not be clinically effective while on the other a proportion of patients may over-respond, increasing bleeding risk. However, in the future, testing will likely be performed at multiple time points to target a therapeutic window similar to the International Normalized Ratio (INR) for warfarin, he predicted.

Dr. Tantry counseled against giving much weight to the observed influence of the various genetic polymorphisms on reactivity and clinical outcome, noting that the trial was small and the findings may not be replicated. “I will give this more importance only after seeing the results of GIFT [the genetic substudy of GRAVITAS],” he added. Moreover, the study ignored any possible impact from drug-drug interactions, he noted, emphasizing that the effect of genetic variants on platelet reactivity is very difficult to predict except in patients with 2 loss-of-function alleles.

In addition, the risk score integrating functional testing with genetic and clinical factors is mainly a down payment on further research, Dr. Tantry commented.

Study Details

All patients were treated with 300 mg of aspirin at hospital admission followed by 100 mg daily. A 600-mg loading dose of clopidogrel was given at least 12 hours before PCI and 75 mg a day was prescribed for 12 months. Anticoagulant and glycoprotein IIb/IIIa inhibitors were administered at the interventionalist’s discretion.

2. Angiolillo DJ. Unraveling myths of platelet function and genetic testing: The road to making tailored antiplatelet therapy a reality. J Am Coll Cardiol. 2011;57:2484-2486.

Related Stories:

• Some Clopidogrel Poor Responders Helped by Higher Dosing
• Genotype, Clinical Factors Unreliable Predictors of Clopidogrel Response