Systematic review and meta-analysis of adverse cardiovascular events associated with proton pump inhibitors used alone or in combination with antiplatelet agents. (16th March 2019)

• Record Type: Journal Article
• Authors: Farhat, Nawal; Fortin, Yannick; Haddad, Nisrine; Birkett, Nicholas; Mattison, Donald R.; Momoli, Franco; Wu Wen, Shi; Krewski, Daniel
• Is Part Of: Critical reviews in toxicology. Volume 49:Number 3(2019)
• Page Start: 215
• Page End: 261
• Publication Date: 2019-03-16
• Subjects: Systematic review -- meta-analysis -- proton pump inhibitors -- clopidogrel -- platelet aggregation inhibitors -- cardiovascular diseases -- myocardial infarction -- drug interactions; Toxicology -- Periodicals; Poisons -- Physiological effect -- Periodicals; Toxicology -- Periodicals; 615.9
• DOI: 10.1080/10408444.2019.1583167
• Languages: English
• ISSNs: 1040-8444

Abstract

Abstract: The potential association between major adverse cardiovascular events (MACE) and concomitant treatment with proton pump inhibitors (PPIs) and clopidogrel has been debated since 2009. Recent reports, however, suggest that PPIs may increase the risk of MACE independently of clopidogrel. This review evaluates epidemiological findings relevant to the association between PPIs, taken alone or concomitantly with antiplatelets, and the risk of MACE. A systematic review and meta-analysis were conducted. Relevant studies were identified from MEDLINE, EMBASE, and the Cochrane Central Register of Controlled Trials and then screened. Included studies were categorized into three groups: Group A: PPIs versus no PPIs; Group B: combined PPIs and clopidogrel versus clopidogrel alone; Group C: combined PPIs and other drugs versus other drugs. Pooled risk ratios (RRs) were calculated for each outcome of interest in each comparison group. Of the 1667 studies identified, 118 were included in the systematic review, of which 66 were included in the meta-analyses. Among Group A observational studies, RRs for MACE outcomes were statistically significant for some patient populations but not others. Pooled RRs from Group A RCTs were not statistically significant for any outcome. Pooled RRs for Group B observational studies were statistically significant for all-cause mortality and MI, but were diminished in magnitude when pooling was restricted to propensity score matched studies or post hoc analyses of RCTs. Group C studies did not demonstrate an association with MACE. Findings do not consistently support an association between MACE and PPIs when taken alone, or concomitantly with antiplatelets.