Treatment with clopidogrel and PPI increased risk for adverse outcomes following acute coronary syndrome

Issue: April 10, 2009

In early 2009, the FDA announced an ongoing safety review of clopidogrel after receiving reports that the use of proton pump inhibitors in combination with the drug would reduce its efficacy. The FDA said there was a need for additional studies in order to obtain more information about the two drugs interactions.

To address this, researchers conducted a retrospective study of adverse outcomes among patients assigned clopidogrel with or without PPI after hospitalization for ACS.

The study included 8,025 patients; 63.9% received PPI at discharge, follow-up or both and 36.1% did not receive PPI. Patients prescribed to both clopidogrel and PPI had a higher incidence of death or rehospitalization compared with patients prescribed clopidogrel alone (29.8% vs. 20.8%).

Patients prescribed to both clopidogrel and PPI also had higher rates of recurrent hospitalizations for ACS, revascularization procedures and death, compared with patients prescribed clopidogrel alone. After 1,080 days of follow-up, the cumulative incidence rates of death or rehospitalization for ACS according to each medication exposure group were: 0.62 for neither, 0.55 for PPI alone, 0.47 for clopidogrel plus PPI and 0.33 for clopidogrel alone.

“Pending further studies to confirm these results and prospectively assess cardiovascular outcomes for patients taking clopidogrel plus PPI vs. clopidogrel without PPI, the results of this study may suggest that PPIs should be used for patients with a clear indication for the medication, rather than routine prophylactic prescription,” the researchers wrote. – by Stacey L. Adams

Ho PM. JAMA. 2009;301:937-944.

If a patient truly requires a PPI plus an important platelet inhibitor, I suggest consideration of the new ADP-agonist inhibitor, prasugrel, as a substitution for clopidogrel.

– Harry S. Jacob, MD

HemOnc Today Chief Medical Editor

This is a very well done analysis, though importantly it’s an observational analysis, not a randomized clinical trial. Like other recent reports, it suggests that there might be some association between proton pump inhibitor use and some sort of interaction with clopidogrel. The only way to really prove that would be with a randomized clinical trial. An observational study, of course, can adjust for some differences between patients on and not on PPIs, and this study does a nice job of that, but it can’t adjust for everything, and that’s one of the limitations of all observational studies. This study adds to the observational evidence that there might be a clinical interaction, but whether we can say that with certainty still remains unknown.

However, what the authors concluded is the appropriate interpretation of their study, which is, if a patient needs clopidogrel there’s nothing about their study that indicates that they shouldn’t use it. And likewise, if a patient needs a PPI on top of clopidogrel and there’s a good indication, then that’s fine to use it. One area to be cautious though is if a doctor in practice has been overly liberal in their use of PPIs. For example, using it in patients where there really isn’t any good indication, but just on the presumption that it might reduce gastrointestinal bleeding in a patient who’s otherwise at low risk for gastrointestinal bleeding; then it’s probably a good idea not to use the PPI. Though one could make the argument even prior to this research, that if you don’t need to be on a drug it’s probably best not to be on it, in terms of minimizing potential for any interactions, minimizing costs, increasing compliance, and all the other good reasons to do that anyway.

This is an area that’s been of intense interest these past several months with these reports coming out, but we still don’t know whether there is a clinically significant interaction or not, it’s certainly biologically plausible. There might be an interaction simply because there have been cytochrome p450 mediated genetic polymorphisms described that appeared to affect the response of the platelet to clopidogrel. Potentially, medications that affect those same pathways could affect clopidogrel's activity and therefore affect the clinical protection that it provides. But on the other hand, we did go through a similar sort of exercise a few years back with statins where there was a lot of concern based on in vitro and observational analyses that clopidogrel and statins interact in an adverse way. Ultimately, the overall conclusion after a series of different studies was that there wasn’t a clinically significant interaction. As clinicians in general, we should keep an eye on the potential for interaction with clopidogrel and PPIs, but I myself haven’t changed my practice yet, other than doing what the authors suggested, which is good cautious medicine.

One other option physicians have asked about is whether they should switch, if they use a PPI with clopidogrel, to a PPI that’s not metabolized by the same pathway that metabolizes clopidogrel. That’s not an irrational thing to do, and certainly makes biological sense, though personally I haven’t gone to that step because I am keeping an open mind about whether this interaction is clinically relevant. Even if PPIs do slightly affect the metabolism of clopidogrel, there are so many other things that potentially affect its metabolism: whether a patient comes in with acute coronary syndrome, whether they have diabetes or not, whether they’re obese or not and of course genetic polymorphisms. When you think of all the things that potentially affect clopidogrel’s activity, this might not be such a big proportion of the overall variability. The only way to know [if there is an interaction] would be with future studies that look at patients that are randomized to a PPI or not on top of clopidogrel.

– Deepak Bhatt, MD, MPH

Director, Integrated Interventional Cardiovascular Program
Brigham and Women’s Hospital, Boston