Abstract 15722: Proton Pump Inhibitors and Risk of Myocardial Infarction: A Systematic Review and Meta-analysis

Introduction: Proton Pump Inhibitors (PPIs) are generally considered safe, but recent evidence suggests otherwise. Objective: This systematic review and meta-analysis assessed the association between PPIs and the risk of myocardial infarction (MI), using both clinical trials and observational studies. Methods: A systematic search was performed in December 2019 to retrieve all potential studies using PubMed, PsycInfo, International Pharmaceutical Abstracts, Web of Science, and Clinicaltrials.gov. This search initially identified any published studies describing any adverse event (AE) or outcomes related to PPI. Records were included in this study if 1) studies were published in English, 2) study design was clinical trials or observational studies, 3) PPI use was the exposure or treatment, and 4) study outcome was the incidence of MI. Two researchers independently reviewed all identified records, performed full article review, extracted data into structured evidence table, and conducted quality assessment using Newcastle-Ottawa Scale and Quality Assessment Tool for Quantitative Studies. Meta-analysis was performed using the RStudio software to assess the risk of MI with PPI use. Results: A total of 4,507 abstracts meeting the inclusion criteria were identified, and 20 full articles were included in this study, among which 10 were cohort, 3 were case-control, 3 were RCT post-hoc analysis, and 4 were RCT studies. The pooled Odds Ratio (OR)=1.40 with 95% CI=1.20-1.63 for all studies indicated the presence of an association between PPI use and increased risk of MI compared to PPI non-users. Meta-analysis found a similar association between PPI use and increased risk of MI in observational studies (OR=1.40; 95% CI=1.20, 1.64) but no association (OR=0.90; 95% CI=0.47, 1.73) in RCT-studies. Heterogeneity was high (I 2 > 75%) for all analyses except for RCTs (I 2 =0%). Conclusion: Although our meta-analysis identified the association between PPI use and increased risk of MI, results from RCTs did not agree with observational studies. Due to the mixed findings by study designs and high variation of heterogeneity among studies, the pharmacovigilance system should evaluate different levels of evidence to support decision making in safety of drug products.