NEW YORK (Reuters Health) – Patients starting on clopidogrel are almost twice as likely to have an MI over the next few months if they are taking a proton pump inhibitor as those not on a PPI, according to a large Dutch cohort study.
While the results are in line with other reports, the association may not be causal, the investigators report in The American Journal of Gastroenterology online August 24. It could be attributable to channeling bias, “i.e., patients using a PPI have increased baseline cardiovascular and GI risk profiles.”
Dr. Peter D. Siersema at the University Medical Center Utrecht, The Netherlands and colleagues note that two studies from the US and Canada found that concomitant therapy with clopidogrel and PPIs was associated with an increased risk of rehospitalization for acute coronary syndrome. However, other studies have not found any clinically relevant interactions between clopidogrel and PPIs.
“These contradicting results have caused a lot of uncertainness as to whether or not to prescribe a PPI concurrent with clopidogrel,” the authors state.
They go on to assert that the effects of clopidogrel and PPI use can only be assessed if both cardiovascular and GI outcomes are considered. They therefore looked at this issue using data from two large health insurers covering 4 million Dutch inhabitants.
The researchers identified 18,139 new clopidogrel users. Nearly a third of them (5734) were taking a PPI concurrently, and these subjects were significantly older with more comorbidities than the non-users.
The data indicated that the primary composite endpoint of MI, unstable angina, stroke or all-cause mortality occurred in 13% of patients on PPIs while taking clopidogrel and in 7% of those not taking a PPI. On univariate analysis the hazard ratio in PPI users for the composite cardiovascular endpoint was 2.03, and remained significant (HR 1.75) after adjustment for possible confounders.
In multivariate analysis, “PPI cotherapy remained significantly associated with a higher risk of myocardial infarction (HR 1.93), unstable angina pectoris (HR 1.79,) and all-cause mortality (HR 1.79),” Dr. Siersema and colleagues report. However, it was not associated with higher stroke risk.
As for GI events, the hazard ratio for PPI users on multivariate analysis was 4.76.
In discussing the results, the authors suggest that “PPI use may be a proxy for decreased general health, which in turn is related to an increased risk for cardiovascular events.”
They argue that residual confounding due to “unmeasured or unknown risk factors could well have an effect on the results presented in this and other observational studies.”
Cardiovascular and Gastrointestinal Outcomes in Clopidogrel Users on Proton Pump Inhibitors: Results of a Large Dutch Cohort Study
Am J Gastroenterol 2010.