The study by Gaglia et al shows an association between proton pump inhibitors (PPIs) and cardiovascular morbidity and mortality in patients with drug-eluting stents who are taking clopidogrel. However, the prescription of PPIs could be related to certain factors, such as a history of peptic ulcer or reflux disease. Peptic and cardiovascular disease have common risk factors, such as low socioeconomic status and unhealthy habits. Thus, patients with peptic disease could also be at greater risk for cardiovascular disease. Peptic disease could act as a confounding factor that was uncontrolled in this multivariate analysis.

The problem is the same for Ho et al’s cohort study, which showed an association between PPIs and cardiovascular disease. The Clopidogrel and the Optimization of Gastrointestinal Events Trial (COGENT) found no association, but it ended early and analyzed few cardiovascular events, so it is not solid enough to detect a relevant difference. However, a third cohort study with the Trial to Assess Improvement in Therapeutic Outcomes by Optimizing Platelet Inhibition With Prasugrel (TRITON)–Thrombolysis In Myocardial Infarction (TIMI) 38 trial population analyzed 6,795 patients receiving clopidogrel therapy. Peptic ulcer and statin use were more frequent in patients taking PPIs (10.4% vs 4.1%, p <0.001). Moreover, 12.8% more patients were treated with statins among PPI users (p = 0.008), an observation compatible with cardiovascular and peptic disease linkage by common risk factors. Peptic ulcer and statin use were controlled by multivariate analysis. This cohort study found no difference for cardiovascular events between PPI users and PPI nonusers (hazard ratio 0.94, 95% confidence interval 0.80 to 1.11). Another cohort study recently showed an increased prevalence of previous peptic and cardiovascular disease in PPI users compared with nonusers (e.g., 28.8% vs 7.8% for esophageal disease and 40.9% vs 29.4% for cardiovascular disease hospitalizations, both p values <0.001). Controlling these outcomes, the investigators found a hazard ratio for cardiovascular disease of 0.99 (95% confidence interval 0.82 to 1.19).

Thus, we think that previous or current peptic disease acts as a confounding factor and is at least partially responsible for increasing cardiovascular events in PPI-treated groups. It would be important to control and quantify this effect, because the underuse of PPI prophylaxis or substitution by less effective anti-H ₂ agents could be dangerous in patients at risk for upper gastrointestinal bleeding. Future cohort multivariate studies on this topic should control peptic ulcer or reflux history. On the basis of the data available, we would recommend that PPIs continue to be prescribed to patients at high risk for upper gastrointestinal bleeding, particularly pantoprazole, because it has the lowest affinity for cytochrome P450 2C19, as noted by Gaglia et al in their discussion.