Clopidogrel and PPIs: Comparing Results from the Growing Body of Evidence

Evidence has been accumulating over the past year that suggests proton pump inhibitors (PPIs) reduce clopidogrel (Plavix®) effectiveness in preventing cardiovascular events after percutaneous coronary interventions (PCI) and acute coronary syndrome (ACS). Enough evidence has accumulated for the U.S. Food and Drug Administration (FDA) to advise Bristol-Myers Squibb that the Plavix label should be modified accordingly. A recent Lancet publication of a post hoc analysis from the TRITON-TIMI 38 study, however, shows no impact on cardiovascular events when clopidogrel is used in combination with PPIs, which consequently casts some doubt on recommendations to avoid these combinations. But any doubts should be based on evaluations of all the studies conducted thus far and not just the TRITON-TIMI 38 analysis.

Interest in PPI effects on cardiovascular outcomes in patients receiving clopidogrel was initially generated by concerns that CYP450 2C19 inhibitors could reduce the amount of clopidogrel converted to its active form and to ex vivo studies showing that the effectiveness of clopidogrel in inhibiting platelet aggregation is diminished in patients taking PPIs. Thus far, in addition to the Medco analysis, four studies evaluating the clinical outcomes of combining clopidogrel with PPIs have been published. Four of them show that patients taking clopidogrel with PPIs are more likely to experience a cardiovascular event than patients taking only clopidogrel. Furthermore, the increased risk among patients receiving the combination is similar in magnitude across these studies. The TRITON-TIMI 38 study showing no such effect is thus the exception among them. (See table)

The four studies showing an effect are all observational in form and compare cohorts receiving clopidogrel with PPIs and cohorts receiving only clopidogrel. These studies generally include all patients, although there can be exclusions based on use of the drugs and how long they are taken.  By and large, these cohorts are fairly representative of the broad range of patients who are likely to receive clopidogrel to prevent bad cardiovascular events following ACS and PCI. The population drawn from the TRITON-TIMI 38 study differs from the populations of the other studies in an important way, and in a way that could explain its different outcome. TRITON-TIMI 38 randomized patients into treatment groups based on specific inclusion and exclusion criteria. Meeting these criteria produced cohorts that were generally younger and in better health compared with the cohorts used in the other studies. If effects of older age and comorbidities excluded from TRITON-TIMI 38 contribute to the attenuation of clopidogrel effectiveness in patients receiving PPIs, then the different outcomes seen between TRITON-TIMI 38 and the other studies could be expected. In addition, the variability inherent in observational and post hoc comparisons can produce different findings among similar studies as well.

At best, then, the analysis of TRITON-TIMI 38 data shows that the combination of clopidogrel and PPIs may not be a problem for younger patients with few comorbidities. Still, the published experience overall shows that there may indeed be serious consequences resulting from the combined use of clopidogrel and PPIs. Based on the sum of this evidence, plus the absence of any prospective research on the safety and effectiveness of combining PPIs and clopidogrel, physicians and patients should be particularly judicious with the use of PPIs for patients requiring clopidogrel.

Rob Epstein, M.D., M.S.