AVOID PPIs WITH CLOPIDOGREL

April 28, 2010 (Irvine, California) — The clopidogrel–proton-pump inhibitor (PPI) interaction saga just keeps rumbling on, with yet another study suggesting an adverse effect of taking these two drugs together published this week [1].

The latest study, published in the April 26, 2010 issue of the Archives of Internal Medicine, suggests almost a doubling in the risk of rehospitalization for MI in patients taking clopidogrel plus a PPI vs those taking clopidogrel alone.

In addition, in this study this increased risk was found with pantoprazole as well as other PPIs, whereas previous studies have suggested that pantoprazole does not interact with clopidogrel as much as some of the other PPIs.

Commentators Cautious

But commentators reviewing the study for heartwire have pointed out that this is another retrospective analysis and that such studies cannot control for all differences between the two groups of patients. Therefore, although this paper adds more provocative information to the debate, it doesn't really shed much new light on what is really going on, they say.

Dr Paul Gurbel (Sinai Center for Thrombosis Research, Baltimore, MD) commented to heartwire : "This study suffers from all of the problems of retrospective analyses based on insurance-claims databases. The authors listed many of the limitations in their discussion. However, there was no mention of potential unbalance in clopidogrel nonresponders and in genetic variables between the groups. Other concerns are the lack of compliance recording, lack of information on smoking status, and lack of information on aspirin therapy. I would regard these data as provocative, but we will never know the truth about the PPI-clopidogrel interaction in the absence of a well-conducted prospective trial that measures both pharmacodynamics and clinical events."

Dr Bernard Gersh (Mayo Clinic, Rochester, MN) had a similar opinion. "This remains an area of some confusion. There certainly is an interaction in vitro, but the magnitude of this from a clinical standpoint is likely to be small, and in some studies no effect has been found. This study changes nothing for me. The first step is to see whether patients really need the PPI. If they do, then I would use pantoprazole on theoretical grounds," he said.

Propensity Scoring Used to Match Groups

In the paper, the authors, led by Dr Karen Stockl (Prescription Solutions, Irvine, CA), explain that previous retrospective analyses that have suggested interactions between clopidogrel and PPIs have been limited by the fact that patients receiving PPIs had a greater number of cardiovascular risk factors and comorbidities compared with patients using clopidogrel alone, and statistical adjustments may not have been adequate to control for unmeasured baseline differences between the two treatment groups. In addition, previous studies involved limited patient populations (eg, commercial insurance only, elderly only, and Veterans Affairs populations).

They therefore conducted the current study, which, although still retrospective, used propensity scoring to match patients according to their baseline cardiovascular risk factors in an effort to obtain more comparable patients across treatment groups.

The authors used electronic pharmacy and medical claims from commercial and Medicare members of a health-insurance plan with annual membership of around two million people. Patients were eligible if they had filled a prescription for clopidogrel between 2004 and 2006 and had an inpatient hospitalization with a primary diagnosis code for acute MI or a procedure code for coronary stent placement during the 30 days before the identification date.

The two study cohorts were based on PPI prescription records. The clopidogrel-alone cohort consisted of patients who did not fill a prescription for a PPI during the 90 days before or the 90 days after the index date. The clopidogrel-PPI group was made up of those who filled a prescription for a PPI supply that overlapped the index date and had at least one refill of the PPI during the 90 days after the index date.

Patients in the clopidogrel-alone cohort were matched 1:1 with patients in the clopidogrel-PPI cohort using the propensity-scoring method, which balanced covariates between the two groups. Patients were followed up for a 360-day period for hospitalizations with a primary diagnosis code of acute MI or a procedure code for coronary stent placement.

A total of 6008 patients met the criteria for the clopidogrel-alone group, and 1041 patients met the criteria for the clopidogrel-PPI group. After matching, there were 1033 patients in each cohort; the cohorts were similar in terms of demographics and preperiod cardiovascular medical history, but the distribution of the Charlson comorbidity index scores was different between the groups, with a higher mean score in the clopidogrel-PPI group.

Results showed higher rates of rehospitalization for MI or MI/stent placement in the clopidogrel-PPI group.

Rehospitalization Rates for MI or Stenting in Patients Receiving PPI and Matched Controls
Outcome Clopidogrel + PPI: events per 100 person-years, n Clopidogrel alone: events per 100 person-years, n Adjusted HR* (95% CI) p
Hospitalization for MI 9.7 4.1 1.93 (1.05–3.54) 0.03
Hospitalization for MI or

coronary stenting
27.6 14.3 1.64 (1.16–2.32) 0.005

*Adjusted for Charlson comorbidity index score

Pantoprazole Just the Same as Other PPIs?

Because pantoprazole has been suggested as having less of an effect on clopidogrel than other PPIs, the authors conducted a subanalysis to investigate whether there was a differential effect on risk with pantoprazole, which was the most-used PPI in this patient population, accounting for 63% of PPI use. This showed similar results to the main analysis. The authors say that this result "suggests that the potential interaction between PPIs and clopidogrel is not specific to omeprazole." Noting that platelet-reactivity studies have found that use of pantoprazole (or esomeprazole [Nexium, AstraZeneca]) was not associated with impaired response to clopidogrel, they say their results also raise questions about the potential mechanism of a PPI-clopidogrel interaction.

Limitations

Also commenting on the study for heartwire , Dr Anthony Gershlick (University Hospitals of Leicester, UK) said it had some merits--namely, the use of propensity analysis to match patients. But he added that its conclusions that there is a serious potential interaction must be tempered by methodological flaws.

He said the interaction with aspirin was particularly noteworthy. "They were unable to track aspirin use, and one might expect that those patients needing PPIs may be less able to tolerate this drug, in addition to clopidogrel."

Other limitations highlighted by Gershlick included a lack of information about the timing and severity of follow-up events, clopidogrel compliance, and platelet responsiveness to clopidogrel. "So while this study keeps the question alive, it probably adds little to helping us make clinical decisions on individual patients. There are too many unknowns," he said.

He repeated the call made by many others for an appropriately powered, robust prospective randomized trial to look at this issue properly.