A Meta-Analysis of Impact of Proton Pump Inhibitors on Antiplatelet Effect of Clopidogrel


Rabi Yacoub, M.D., University at Buffalo, Department of Internal Medicine, David K. Miller Building, 462 Grider ST, Buffalo, NY 14215, USA. Tel.: (716) 898-4226; Fax: (716) 898-3279; E-mail: rabiyaco@buffalo.edu

I read with great interest the article by Chen et al. [1] evaluating the impact of proton pump inhibitors (PPIs) on antiplatelet effect of clopidogrel by conducting a meta-analysis of comparative concomitant use of clopidogrel with PPIs versus clopidogrel without PPIs, with the endpoint being cardiovascular events. The authors conducted a comprehensive search using multiple databases. Thirteen articles were included in their article and only seven were analyzed in the final module. Heterogeneity was found across the study and they concluded that there is an obvious discrepancy between the randomized control trials (RCTs) and the observational trials studied, and recommended large-scale RCTs to ascertain the true effects of PPIs on clopidogrel with respect to cardiovascular events.

Careful review of this interesting article reveals many weaknesses and flaws.

  • 1Search criteria and articles included in their final analysis: No obvious inclusion and exclusion criteria were described in the study. It is unclear why they evaluated only seven articles out
    of the 13 articles they found through their search. A recent meta-analysis conducted by Siller-Matula et al. [2] studied 25 trials regarding the concomitant use of PPIs and clopidogrel; these trials included one RCT, two post hoc, 20 observational study, and two case-control studies. The endpoints in their meta-analysis were the same as in this article.
  • 2Statistical analysis and methods used: Heterogeneity was found across the study when each group was studied separately. However, they still used fixed effect method to analyze the data along with the random effect method that they should have only used, making the interpretation of their results even more confusing.

I congratulate the authors on their important contribution and I agree with their final statement that a well-designed RCT that confers a favorable or unfavorable outcome to this treatment modality must be conducted. I look forward to their feedback about the impact of the concomitant use of clopidogrel and PPIs on cardiovascular events.