After the approval of dabigatran in 2010 for the prevention of stroke and blood clots in patients with atrial fibrillation (AF), the FDA received many post-marketing reports of bleeding. As a result, the agency began to analyze real-world rates of gastrointestinal (GI) bleeding and intracranial hemorrhage (ICH) in dabigatran patients in comparison to similar bleeding rates in new users of warfarin. On November 2, the FDA released results of this analysis, stating that “bleeding rates associated with new use of dabigatran do not appear to be higher than bleeding rates associated with new use of warfarin, which is consistent with observations from the large clinical trial used to approve dabigatran (RE-LY trial).”

The FDA’s analysis is based on insurance claims and administrative data from their Mini-Sentinel pilot of the Sentinel Initiative. The Sentinel Initiative is an FDA-sponsored active surveillance system using preexisting electronic healthcare data from multiple sources in order to evaluate the safety of approved products.

The FDA states that bleeding rates associated with new use of dabigatran “do not appear to be higher” than bleeding rates associated with new use of warfarin. Despite its use of specific language, the FDA’s own data appear to reflect a significantly lower bleeding rate with dabigatran compared with warfarin. In the Mini-Sentinel assessment, the combined incidence rate of ICH and GI bleeding events per 100,000 days at risk was 1.8 to 2.6 times higher for new users of warfarin than for new users of dabigatran. The incidence rate of GI bleeding events per 100,000 days at risk was 1.6 to 2.2 times higher for warfarin new users than for dabigatran new users; for ICH events, the incidence rate was 2.1 to 3.0 times higher with warfarin than with dabigatran.

The dabigatran phase III trial (2009) data provide valuable context. In the RE-LY (Randomized Evaluation of Long-Term Anticoagulant Therapy) trial, the rate of major bleeding events with dabigatran was similar to that of warfarin (3.11% per year versus 3.36% per year; relative risk=0.93). While the rate of intracranial bleeding was substantially lower with dabigatran than with warfarin (0.3% versus 0.74% per year; relative risk=0.40), the rate of GI bleeding was substantially higher with dabigatran than with warfarin (1.51% versus 1.02% per year; relative risk=1.50).

The FDA’s Mini-Sentinel analysis had limitations, however. First, it did not account for potential differences in patient populations that could affect bleeding outcomes, such as age and the presence of comorbidities. Also, it is unknown whether the codes for ICH, GI bleeding, and AF are accurate reflections of the existence of those conditions in patients using anticoagulants. Note, too, that the FDA states that it “believes that a simple comparison between dabigatran and warfarin with respect to the numbers of post-marketing reports of bleeding in the FDA’s Adverse Event Reporting System (AERS) database is misleading because bleeding events associated with warfarin (a well-recognized consequence of warfarin use) are likely underreported compared with events occurring with the more recently available dabigatran.” As part of its ongoing safety review, the FDA is also carrying out two planned, protocol-based observational assessments of bleeding events in patients taking dabigatran.

The FDA’s Mini-Sentinel analysis represents an important advance that clinicians have been waiting for. These data may serve to reassure both physicians and patients alike as to the safety profile of dabigatran and the overall value of stroke prevention in AF.

Published: 11/26/2012

References:

###

By Mark Abrahams, MD
Reviewed by Philip Green, MD, Assistant in Clinical Medicine, Division of Cardiology, Department of Medicine, Columbia University Medical Center, New York