Analysis of upper gastrointestinal adverse events among patients given dabigatran in the RE-LY trial

Background & Aims: Dabigatran is an oral and direct inhibitor of thrombin. In a study of patients with atrial fibrillation (the RE-LY trial), twice as many subjects given dabigatran reported dyspepsia-like symptoms compared with those given warfarin (controls). We analyzed data from this trial to quantify upper gastrointestinal nonbleeding adverse events (NB-UGI AEs). Methods: We analyzed the AE database from the RE-LY trial (18,113 subjects) and assigned NB-UGI AEs to 4 groups: those associated with gastroesophageal reflux (GERD), upper abdominal pain and dyspepsia, dysmotility, or gastroduodenal injury. We analyzed frequency, timing, and severity, and clinical variables associated with NB-UGI AEs. Results: NB-UGI AEs occurred in 16.9% of subjects given dabigatran and in 9.4% of controls (relative risk [RR], 1.81; 95% confidence interval [CI], 1.66%-1.97%; P < .001). Rates of AEs were not associated with the dose of dabigatran. Among subjects with any UGI symptom who were given dabigatran (n = 2045), symptoms were rated as mild in 46.3%, moderate in 44.8%, and severe in 8.9%; these values were similar to those of controls. GERD-associated NB-UGI AEs were most frequent among the 4 groups (compared with controls, RR, 3.71; 95% CI, 2.98%-4.62%; P < .001). Four percent of subjects stopped taking dabigatran because of NB-UGI AEs (most within 3 months of starting therapy), compared with 1.7% of controls (RR, 2.34; 95% CI, 1.90%-2.88%; P < .001). NB-UGI AEs slightly increased risk of major GI bleeding among subjects given dabigatran and controls (6.8% vs 2.3%, P < .001). Conclusions: Among patients given dabigatran for atrial fibrillation, NB-UGI AEs are generally mild or moderate; 4% stopped taking the drug over a median of 21.7 months. The greatest increase was in GERD-type NB-UGI AEs. These observations should guide management and prevention strategies.