Failure of Another Trial Testing Guided Antiplatelet Therapy Prompts Debate 4

Updated, Saturday, April 7

Yet another trial has failed to prove the hypothesis that guided antiplatelet therapy with platelet function testing or genetic testing improves outcomes. Although there has been no public announcement so far, the TARGET-PCI (Thrombocyte Activity Reassessment and GEnoTyping for PCI) was suspended last October, according to the TARGET-PCI page on ClinicalTrials.Gov.

TARGET-PCI was a large study in which 1,500 PCI patients were randomized to guided antiplatelet therapy with genotyping and platelet function tests or standard antiplatelet therapy. The primary endpoint was major adverse cardiovascular events at 6 months. The principal investigator was Paul Gurbel of Sinai Hospital of Baltimore, where the study was performed.

Here is what one cardiologist who wished to remain anonymous commented about the trial:

It was an ambitious study that was probably underpowered to look at efficacy outcomes. The positive predictive value of the CYP2C19*2 allele is reported as being < 20%. My guess is that the study was halted due to futility, and they discovered that it would be unlikely that any difference in outcome would be apparent with either personalized guided therapy or with routine care, even if the study was carried to completion.

This mirrors what we saw with the TRIGGER-PCI study and GRAVITAS, with the addition of the genetic testing by the Verigene assay.   It seems that both the Verify-now and Verigene assays are solutions still looking for problems to solve.

Sanjay Kaul offered the following comment:

The prognostic utility for genotype testing is not substantial or consistent. The outcome most strongly associated with CYP2C19 loss of function genotype is stent thrombosis (hazard ratio of 3 to 4), a rare but potentially life-threatening event. However, given the rare occurrence of stent thrombosis, the positive predictive value of genotype is only 3-4%, too small to be clinically directive. A randomized controlled trial to definitively address the role of genotype is currently lacking. Given the paucity of thrombotic events, a trial with a sample size of 20,000-30,000 would be required to adjudicate the clinical utility of genotype testing, a number that might be understandably viewed as being prohibitive. With a sample size of around 1500 and a follow-up up to 6 months, the TARGET-PCI is woefully underpowered to address this. Not surprisingly, and given the null results of GRAVITAS and premature stopping of TRIGGER-PCI for futility, the decision to suspend TARGET-PCI does not seem unwarranted.

Based on the totality of evidence, personalized antiplatelet therapy using genotype testing cannot be routinely recommended at the present time. The CYP2C19 polymorphisms affecting clopidogrel metabolism offers a cautionary tale about the unlimited power of genetic technology to generate ‘bountiful’ information on one hand, and our limited ability to properly analyze, reliably interpret, and judiciously translate it into clinical practice on the other hand. The CYP2C19 polymorphisms were heralded with much fanfare to catapult genomics to prominence in personalized cardiovascular medicine. However, the cold hard facts are that not only is the use of genetic testing to improve risk stratification premature, but whether the power of genomics can be harnessed to improve decision making and clinical outcomes also remains to be defined. The unbridled (and uncritical) enthusiasm for genomics revolutionizing personalized medicine has often blurred the distinction between hope, hype and reality. While the progress in DNA sequencing technology and bioinformatics is undeniably breathtaking, the translation of this knowledge to clinical practice has lagged behind. Genomics will eventually get there, but it cannot be expected to move faster than the speed of science.

Eric Topol offered this response to Kaul’s statement:

With at least 30% of individuals carrying DNA loss-of-function sequence variants for metabolizing clopidogrel, it should be standard of care to genotype individuals who are undergoing coronary stenting. The recent RAPID GENE publication in the Lancet (from last week), the first randomized trial of rapid, point of care genotyping , demonstrated unequivocal and marked improvement in platelet suppression using genotype-guided information as compared with no DNA data. The resistance to use genomic data to improve patient care is indefensible, and does not require “a trial with a sample size of 20,000-30,000 would be required to adjudicate the clinical utility of genotype testing.” That trial will never be done, is not even warranted, and while the debate for its need ensues, too many patients will experience stent thrombosis unnecessarily. And the stakes are high, since most stent thrombosis results in MI or death. In next week’s JAMA correspondence, there will be important perspectives published on this matter.

Update, Saturday, April 7:  Sanjay Kaul provided the following response to Topol’s statement. CardioBrief has invited Topol to respond:

The message I take away from the RAPID GENE trial is that point-of-care genetic testing is clinically feasible and might facilitate rapid personalization of anti-platelet therapy. However, genomic data is best utilized to “improve patient care”, not to “improve platelet suppression” – a surrogate marker of unclear clinical relevance. Although the trial was not powered to address clinical outcomes, bleeding episodes occurred in 5/91 of the rapid genotyping patients compared with 2/96 of standard therapy patients.

Individualizing antiplatelet therapy should ideally be based on a biomarker that precisely measures drug responsiveness, accurately characterizes low and high risk patients, is readily available and affordable, and reliably guides treatment decisions to optimize outcomes in a cost-effective manner. Unfortunately the quest for such a biomarker remains elusive as none of the currently available genetic tests (or platelet function tests) exhibits these desirable attributes.

Despite the inordinate attention being given to genomic and personalized medicine, the body of evidence continues to be insufficient to justify the clarion calls for “standard of care” use of genotype testing in the care of patients undergoing coronary stenting.

CardioBrief will be happy to provide a forum for additional discussion and debate about this topic.



  1. -The resistance to use genomic data to improve patient care is indefensible, and does not require “a trial with a sample size of 20,000-30,000 would be required to adjudicate the clinical utility of genotype testing.”: Genomic testing is no different from any other test which needs to undergo the rigor of being evaluated for its impact on the outcomes that the pathway it quantifies is related to.
    – “That trial will never be done, is not even warranted”: Not clear to me why not-with an effect size in the 30% range fro RAPIDGENE,this should be doable. also not clear why not warranted.
    -“while the debate for its need ensues, too many patients will experience stent thrombosis unnecessarily”: the current rate of stent thrombosis with newer generation stents is in the 0.1-0.2% per year.
    -If RAPIDGENE was an outcomes study and one of the arms simply involved switching to prasugrel/ticagrelor, I would imagine the outcomes would be comparable to the genotyping based switch to prasugrel-this then comes down to a cost comparison between genotyping+clopidogrel versus global use of prasugrel/ticagrelor.

  2. It is important that the results from this trial are published promptly to enable physicians to have full access to all the available evidence, especially when there is such uncertainty surrounding the clinical use of CYP2C19 to guide clopidogrel in any clinical setting.

    A clinical trial is the gold standard to prove efficacy and to claim that it is not warranted or feasible is to ignore the lack of evidence base ( surrounding genotype testing in this setting.

  3. This very interesting debate highlights the profound need we have for an improvement upon the decades old and notoriously variable and unreliable platelet aggregation assay. A biologically meaningful and reproducible thrombosis assay preferably one that incorporates both the coagulation system as well as the endothelium is a great unmet need!

  4. Kaul’s point that the number of endpoints for the problem most associated with the genotype in question (stent thrombosis) is going to be small and the necessary sample size large is not refuted at all by Topol. The practicality of doing such a trial is a separate (and important matter). Instead Topol cites a recent trial (RAPID GENE) that used surrogate endpoints. Should we have skipped ILLUMINATE? How would that have worked out?

What do you think?

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s