Concurrent use of proton-pump inhibitor (PPI) and clopidogrel decreases the risk of hospitalisation for gastro-duodenal bleeding, and does not significantly increase the risk of adverse cardiovascular events, according to a cohort study published in the Annals of Internal Medicine.

Clopidogrel increases the risk of gastro-duodenal bleeding, hence prophylaxis with PPI has been suggested in patients at increased risk from such bleeds. The overall benefits of this approach have been questioned by studies suggesting that PPI may reduce the cardiovascular benefits of clopidogrel by reducing its metabolic activation, however the evidence so far is conflicting. The authors of this retrospective cohort study used data from US healthcare databases to examine outcomes in patients treated with clopidogrel, with or without concurrent PPI. Participants were patients in the US Tennessee Medicaid Program who were prescribed clopidogrel after hospitalisation for coronary heart disease between January 1999 and December 2005 inclusive. Comprehensive prescription and outcome data were available for this population. Users and non-users of PPI were identified, and outcomes for the two groups compared: primary outcomes were hospitalisation for gastro-duodenal bleeding, and serious cardiovascular disease (defined as fatal or nonfatal myocardial infarction or sudden cardiac death, stroke, or other cardiovascular death).

There were 20,956 eligible patients, of whom 7,593 were PPI users. The PPI most often prescribed was pantoprazole (62%), with omeprazole the next most used (9%). Users were more likely to have a history of previous cardiovascular and gastro-intestinal disease.

Current clopidogrel users who were not PPI users had an overall risk of hospitalisation for gastro-duodenal bleeding of 12.2 per 1,000 patient-years (py) of follow-up. The risk increased according to number of GI bleeding risk factors from 3.2 per 1,000 py for those with no risk factors to 46.7 per 1,000 py for those with 3 or more. The overall risk was reduced by 50% in PPI users (hazard ratio [HR], 0.50; 95% CI, 0.39 to 0.65), and in those at highest risk (3 or more risk factors) the absolute rate of hospitalisations was reduced by 28.5 (CI, 11.7 to 36.9) per 1,000 py.

There was no significant difference in the rate of serious cardiovascular disease between current clopidogrel users who did and did not take concurrent PPI (overall HR 0.99; 95% CI, 0.82 to 1.19). There was also no significant difference in the pre-specified subgroup who were taking clopidogrel following stent placement (HR 1.01; 95% CI, 0.76 to 1.34). Analysis by PPI used and by PPI dose also showed no significant differences.

The authors conclude that in this patient cohort, concurrent use of clopidogrel and PPI reduced hospitalisation for gastro-duodenal bleeding without significantly increasing risk for serious cardiovascular events. The benefit increased according to baseline bleeding risk, and even in the worst-case scenario, PPI use was beneficial overall in those with high bleeding risk and low to moderate cardiovascular risk. They note some of the limitations to their analysis, and suggest reasons why the results may have differed from other studies. As it is likely that PPI use was mainly in higher risk patients, the results may underestimate the gastro-duodenal benefits and overestimate the cardiovascular risk. They note that the most widely used PPI in their cohort was pantoprazole, which has limited effects on the cytochrome P450 enzyme involved in activating clopidogrel.