PCV25 - Evaluation Of Bleeding Outcomes Linked To New Oral Anticoagulants: Review Of The Fda Adverse Event Reporting System (Faers)

Abstract

Clinical trials comparing rivaroxaban or dabigatran to warfarin show similar results for major and non-major clinically relevant bleeding. Post-marketing data regarding bleeding rates for these new oral anticoagulants (NOACs) is lacking. Our goal was to evaluate bleeding related outcomes with NOACs using a disproportionality analysis of spontaneous adverse event reports. We evaluated the FAERS database from 10/2011 – 12/2012 to compare frequency of events in reports of rivaroxaban, dabigatran, and warfarin related to fatal and non-fatal bleeding. Reports were included if the anticoagulant was the primary suspected agent for the event. Reporting odds ratio (ROR) and proportional reporting ratio (PRR) were calculated to estimate risk for each anticoagulant. Of 22,244 eligible adverse drug reports for the anticoagulants, 7661 (34%) bleeding reports were submitted, with 1868 (24%) deaths, 5028 (66%) hospitalizations and 1040 (14%) life-threatening events. Dabigatran was the most commonly reported anticoagulant exposure among bleeding cases (n=5203, 68%). Of the bleeding cases associated with dabigatran, death was reported in 29%, hospitalization in 67%, and life-threatening events in 15% of cases. The odds of exposure for fatal bleeding was significantly higher with dabigatran (ROR 2.28, 95% confidence interval [CI] 2.02-2.56) and significantly lower with rivaroxaban (ROR 0.75, CI = 0.69-0.8) and warfarin (ROR 0.53, CI = 0.44-0.63). PRR indicated significantly increased risk of non-fatal bleeding with dabigatran (PRR = 2.12, CI = 1.99-2.24) and warfarin (PRR = 1.27, CI = 1.17-1.38) with a significant decrease associated with rivaroxaban (PRR 0.8, CI = 0.75-0.86). Among the NOACs, patients having fatal and non-fatal bleeding were found to have higher odds of dabigatran exposure and significantly lower odds of rivaroxaban exposure. Though FAERS is subject to significant bias, the results suggest dabigatran-related bleeding is higher in clinical practice and rivaroxaban may be a safer alternative.