‘There are interesting epidemiologic associations, but the magnitude is weak and findings are inconsistent.’
Adverse effects attributed to proton pump inhibitors, including a risk for adverse interactions with clopidogrel, have probably been overstated, according to an invited speaker at the 2010 Digestive Disease Week.
The professor of medicine and clinical director of the division of gastroenterology at Vanderbilt University Medical Center in Nashville, Tenn., told attendees that although there are “interesting epidemiologic association” that may have “some biologic plausibility,” the associations are weak in magnitude and are based on inconsistent findings from heterogeneous studies with a high potential for confounding.
Issues with clopidogrel
The potential for an interaction with clopidogrel — that is, whether PPIs inhibit the anticoagulation effect of clopidogrel so that it is less effective in preventing cardiovascular injury — has been a “huge issue,” because cardiologists, or patients themselves, are discontinuing proton pump inhibitors (PPIs) that they were using for reflux disease.
A number of studies have suggested harmful interactions. One study showed a reduction in the platelet reactivity index, indicating poor response to clopidogrel, in 61% of patients who received omeprazole vs 26% of patients in a placebo group, a highly significant difference. In a retrospective Department of Veterans Affairs cohort study, all-cause mortality was significantly increased in patients on PPIs plus clopidogrel vs clopidogrel alone.
But a recent meta-analysis of 23 studies on this topic, involving 93,278 patients, showed no excess risk for cardiovascular events for PPIs that were used with clopidogrel in observational studies (odds ratio, 1.15) among propensity-matched or randomized-trial participants. No significant association was found between PPI use and overall mortality.
“The most recent analysis asked the right question: If you are not on a PPI but are on clopidogrel, what is the risk of bleeding?” A recent study examined a database of more than 20,000 patients (including 7,593 concurrent users of clopidogrel and PPIs) who were hospitalized for gastroduodenal bleeding and serious cardiovascular disease. The adjusted incidence of hospitalization for bleeding in concurrent users was 50% lower than it was in nonusers of PPIs who were taking clopidogrel. For patients at highest risk for bleeding, PPI use was associated with an absolute reduction of 28.5 per 1,000 persons-years, the study found.
The authors concluded that in patients with serious coronary heart disease that were treated with clopidogrel, concurrent PPI use was associated with reduced hospitalizations for gastroduodenal bleeding, and the corresponding point estimate for serious cardiovascular disease was not increased.
Hip fracture
PPIs have been associated with a risk or hip fracture, but a nested, case-control study from the UK General Practice Research Database, including 13,556 cases and 135,386 controls, showed an odds ratio of about 2.0 in the crude analysis, and approximately 1.5 in the adjusted analysis. However, when the other analyzed this study and excluded patients with baseline risk factors for fracture, they observed no increased risk among PPI users.
“If anything, patients who were on PPIs the longest were somewhat protected.” He added that other studies have found that the extended use of PPIs, and also use of H2 receptor blockers, is associated with reduced risk — which are interesting findings that are inconsistent with the concept of harm from PPIs.
However, despite the negative association the DFA most recently has decided to revise the warnings and precautions section of the prescription labeling as well as the OTC drug facts label for PPIs to warn about the potential increase risk of fractures of hip, wrist, and spine.
The most recent study, based on the Manitoba Bone Mineral Density Database, showed that 5 years of PPI use posed no risk for adverse effects to the hip or spine.
Analyses of associations with pneumonia and anemia also point to weak effects.
“Conclusion regarding the potential for adverse effects with PPIs is that there are interesting epidemiologic associations, but the magnitude is weak and the findings are inconsistent.” “There is high potential for confounding in these studies, and when you control for this, the effect is small.”