PARADOX Published: Smokers’ Paradox Seen with Clopidogrel But not Prasugrel

Print  |  
Key Points:
  • Study measures influence of smoking on platelet inhibition by clopidogrel, prasugrel in stable CAD patients
  • Clopidogrel less effective in nonsmokers vs. smokers, less potent overall than prasugrel
  • Prasugrel seems unaffected by smoking, but outside sources question study’s clinical relevance

By Caitlin E. Cox
Friday, May 03, 2013

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

Cigarette smoking appears to augment the antiplatelet effect of clopidogrel but not prasugrel, according to a study of patients with stable coronary artery disease (CAD) published online April 24, 2013, ahead of print in the Journal of the American College of Cardiology. Prasugrel, however, is more potent across the board, irrespective of smoking status.

Findings from the PARADOX study were previously presented at the 2012 Transcatheter Cardiovascular Therapeutics scientific symposium in Miami Beach, FL.

Paul A. Gurbel, MD, of Sinai Hospital of Baltimore (Baltimore, MD), and colleagues randomized 110 patients (n = 56 nonsmokers, 54 smokers) in a double-blind fashion to receive a 10-day regimen of clopidogrel 75 mg or prasugrel 10 mg. Patients then crossed over to the other medication after a 14-day washout period.

For the primary comparisons, platelet function was measured using the VerifyNow P2Y12 assay (Accumetrics, San Diego, CA).

With clopidogrel, device-reported inhibition of platelet aggregation trended lower in nonsmokers than in smokers (coprimary endpoint; least squares mean treatment difference [LSMTD] of 7.7%; P = 0.062). Other approaches to quantifying platelet function also showed a disadvantage for nonsmokers vs. smokers; these included calculated inhibition of platelet aggregation, P2Y12 reaction units (PRUs), and vasodilator-stimulated phosphoprotein phosphorylation (VASP; BioCytex, Marseille, France) platelet reactivity index (P = 0.043, 0.005, and 0.042, respectively).

Among smokers, device-reported inhibition was lower with clopidogrel vs. prasugrel (coprimary endpoint; LSMTD of 31.8%; P < 0.0001).

On the other hand, with prasugrel, device-reported inhibition was similar between nonsmokers and smokers (LSMTD of 4.8%; P = 0.244). PRU values within this cohort were numerically but nonsignficantly lower for smokers vs. nonsmokers (LSMTD of -21.2; P = 0.0924).

‘You Have to Use the Data You Have’

In a telephone interview, Dr. Gurbel told TCTMD that the pharmacodynamic and pharmacokinetic effects of smoking on clopidogrel could have real-world consequences. Previous research has shown a 4% increase in risk of ischemic events per 10-unit increase in PRU, he noted, pointing out that in the current study the difference was 36 PRUs.

“[PARADOX] should put to rest whether there’s a smoking/clopidogrel interaction,” Dr. Gurbel commented. “Pharmacodynamically, there shouldn’t be any question about it anymore. What it means clinically is debatable, but we do have evidence from the studies that [looked at aspirin with and without clopidogrel], and the benefit of clopidogrel was really confined to the people who smoked.”

The exact mechanism is still unclear, Dr. Gurbel stressed, suggesting that CYP1A2 and CYP2B6 induction by cigarette smoking, which leads to an increase in clopidogrel active metabolite, may play roles.

As to whether these findings should influence the choice between clopidogrel and prasugrel, Dr. Gurbel stressed that the “effect of clopidogrel is beholden to multiple influences, smoking status being one of them. So yes, it may lead doctors who have high-risk patients who need a P2Y12 inhibitor and don’t smoke [to] use a more potent and predictable drug like prasugrel.”

Despite being aware of the smokers’ paradox, “I think [many clinicians] still question what the clinical relevance of it is, as do I. But you have to use the data you have available to treat the patient,” Dr. Gurbel concluded, predicting that a randomized trial of this question is unlikely.

Findings May Not Have Clinical Utility

But Neal S. Kleiman, MD, of Methodist DeBakey Heart and Vascular Center (Houston, TX), and Kishore J. Harjai, MD, of Columbia University Medical Center (New York, NY), both remained unconvinced of the paradox’s existence.

In a telephone interview with TCTMD, Dr. Kleiman said that the phenomenon “has been observed in some clinical trials but not all. Basically, the story is that the [link] is confounded by differences between smokers and nonsmokers.”

PARADOX provides useful information about mechanism—especially that CYP2C19 is not the only influence on clopidogrel metabolism—he noted. “And it looks like they did a careful trial. The question is what you do with this. You don’t tell people to smoke because you have a better chance you can treat them. What counts in life is your overall probability of having a disease, not how well the disease can be treated.”

Dr. Harjai, meanwhile, said that some of the difficulty lies in the population being studied. ACS, as the approved indication for prasugrel, is much more relevant to this question than stable CAD, he asserted.

Notably, the PARADOX findings do not absolve prasugrel, because the difference in PRUs by smoking status for prasugrel was of borderline significance “likely due to inadequate sample size, as acknowledged by the authors,” he told TCTMD in an e-mail communication. “Second, the degree of attenuation of platelet inhibition . . . by clopidogrel among nonsmokers was only 8%, much less than the 15% that the investigators assumed in their statistical power calculation.

“Although interesting from a mechanistic point of view, I am not sure that these findings have any clinical utility,” Dr. Harjai continued.

In a follow-up telephone interview, he emphasized, “The fact of the matter is that in stable coronary disease patients, these numbers have really no value, and the relevant population of ACS patients has already been studied [by the TRITON, TRILOGY, and PLATO trials]. I would’ve thought that those studies laid the matter to rest.”

Dr. Harjai concluded that “[i]n the face of these 3 large-scale studies, I am inclined to believe that the clinical relevance of the smokers’ paradox is quite negligible.”


Gurbel PA, Bliden KP, Logan DK, et al. The influence of smoking status on the pharmacokinetics and pharmacodynamics of clopidogrel and prasugrel: The PARADOX study. J Am Coll Cardiol. 2013;Epub ahead of print.



  • The study was funded by Daiichi-Sankyo and Eli Lilly
  • Dr. Gurbel reports associations with several pharmaceutical companies.
  • Dr. Kleiman reports serving as a consultant for AstraZeneca and Eli Lilly.
  • Dr. Harjai reports no relevant conflicts of interest.


Related Stories:

Click here for a listing of companies that provide support to the Cardiovascular Research Foundation, owner and operator of TCTMD.

Related Content:


caitlinecox said:
Good catch! Thank you. Will edit accordingly
Conflict of Interest:
5/7/2013 12:30:46 PM

asikhs said:
The summary should say that smoking POTENTIATES the anti platelet effect
Conflict of Interest:
5/7/2013 12:17:36 PM