SHOULD PATIENTS PRESCRIBED LONG-TERM LOW-DOSE ASPIRIN RECEIVE PROTON PUMP INHIBITORS? SYSTEMATIC REVIEW AND META-ANALYSIS

Purpose: Several clinical guidelines recommend the use of proton pump inhibitors (PPIs) in patients taking low-dose aspirin, but report no or limited supporting data. This study aimed at systematically reviewing the effects of co-administration of PPIs in patients taking low-dose aspirin on the risks of adverse gastrointestinal (GI) and cardiovascular (CV) events, and on patient adherence to aspirin.

Methods:   We searched PUBMED, EMBASE and Cochrane Central Register of Controlled Trials databases and bibliographies of the retrieved articles in November 2013 for relevant studies. We included randomized controlled trials (RCTs) and observational studies in patients taking low-dose aspirin with and without PPIs. We excluded studies in patients concomitantly taking clopidogrel or any non-steroidal anti-inflammatory drug other than aspirin. Two authors independently reviewed and extracted information from the studies. Main outcomes included relative risks (RRs) of peptic ulcer and ulcer complications, myocardial infarction, ischemic stroke and CV-related mortality, and adherence to aspirin prescription associated with co-administration of PPIs. Risk of bias in the observational studies and RCTs was assessed using the Newcastle Ottawa Scale and the Cochrane Collaboration's tool, respectively. Pooled RRs were computed using a random effects model.

Results: We included 13 studies, of which four (cohorts) reported on the effect on peptic ulcer, five (two cohort and three case-control) on GI bleeding, three (two RCTs and one cohort) on both peptic ulcer and GI bleeding, and one (cohort) on both GI bleeding and CV events. No study reported on adherence to aspirin prescription.

Co-administration of PPIs in patients receiving low-dose aspirin was associated with risk reductions of 73% and 50% in the occurrence of peptic ulcer and GI bleeding, respectively (Figure 1). Estimated RRs [95% CI] of myocardial infarction, stroke, and CV death were reported in one study only and were 1.33 [1.13-1.56], 1.20 [0.99-1.46], and 2.19 [1.92-2.49], respectively. There was evidence of bias in publications reporting on the effects on peptic ulcers and GI bleeding.

Conclusions: The practice of co-prescribing PPIs in patients taking low-dose aspirin is supported by some data, but the evidence is rather weak. It currently remains unclear whether the benefits of co-administration of PPIs in users of low-dose aspirin outweigh their potential harms.

Figure 1. Forest plots of relative risk of (A) peptic ulcer and (B) gastrointestinal bleeding.