Use and safety of medications associated with peptic ulcer disease and gastrointestinal bleeding

Abstract

Asian population are shown to be associated with higher risks of peptic ulcer disease (PUD) and/or gastrointestinal bleeding (GIB). As the decrease of Helicobacter Pylori infection led to a decline of the prevalence of PUD/GIB worldwide, medication use became a crucial superimposing factor for PUD/GIB. To date, the research findings are still unclear or inconclusive on the association of calcium channel blockers (CCBs) and non-vitamin K antagonists (NOACs) with PUD/GIB, or the role of gastroprotective agents (GPAs) in the prophylaxis of medication-associated PUD/GIB. This thesis contains four parts of pharmacoepidemiological studies and aims to assess the risk of PUD/GIB associated with the aforementioned medication use, and to investigate the prophylactic role of GPAs including histamine-2 receptor antagonists (H2RAs) and proton pump inhibitors (PPIs). First, we conducted a meta-analysis using Review Manager to investigate the association between CCB use and the risk of GIB. The meta-analysis, including four randomized controlled trials (RCTs) and thirteen observational studies, suggested a modest association with a relative risk (RR) of 1.17 (95%CI 1.01-1.36). However, this is of weak clinical significance as results from neither RCTs nor subgroup analyses supported this association. Unlike CCBs, nonsteroidal anti-inflammatory drugs (NSAIDs) are known to be the most common risk factor for PUD/GIB. To investigate the dosage effect of H2RAs in preventing NSAID-associated PUD, we conducted a population-based retrospective cohort study using the Clinical Data Analysis and Reporting System (CDARS) database on adult patients (≥ 18 years) in Hong Kong between 2009 and 2012. Among the 102 042 NSAID users, 76% were on low-dose H2RA and 24% were on high-dose, and the RR of PUD was 0.32 (95%CI 0.13-0.79) in high-dose compared to low-dose. Besides NSAIDs, NOACs are another class of well-known medications potentially increasing GIB risk. The up-to-date meta-analysis of observational studies in the third part of this thesis suggested that dabigatran was associated with a modest increased risk of GIB as compared to warfarin (RR1.21, 95%CI 1.05-1.39). Lastly, our CDARS-based retrospective cohort study showed that among the 5041 new users of dabigatran from 2010-2013, 124 (2.5%) developed GIB during follow-up (4.2/100 patient-years). Aged ≥75 year (IRR, incident rate ratio 2.47, 95%CI 1.66–3.68), history of PUD/GIB (IRR 2.31, 95%CI 1.54–3.46), and concomitant use of aspirin (IRR1.52, 95%CI 1.03–2.24) were risk factors for GIB. Concomitant use of H2RAs/PPIs reduced the risk of GIB significantly (IRR, 0.52; 95% CI, 0.35–0.77). In summary, our study found that CCB is unlikely to be associated with GIB clinically and current practice should not be affected. It is important to ensure high-dose H2RA is prescribed as it showed greater effectiveness compared to low-dose for the prophylaxis of NSAID-induced PUD. Dabigatran should be used with caution and concurrent GPAs use should be encouraged, especially for patients at high risk of GIB. The heterogeneity between studies (especially observational studies) in the meta-analyses and the inherent limitations of CDARS as electronic medical database are the major limitations of this study.