Efficacy and Harms of Direct Oral Anticoagulants in the Elderly for Stroke Prevention in Atrial Fibrillation and Secondary Prevention of Venous Thromboembolism: Systematic Review and Meta-Analysis

Abstract

Background: Evidence regarding the use of direct oral anticoagulants (DOACs) in the elderly, particularly bleeding risks, is unclear despite the presence of greater comorbidities, polypharmacy, and altered pharmacokinetics in this age group.

Methods and results: We performed a systematic review and meta-analysis of randomized trials of DOACs (dabigatran, apixaban, rivaroxaban, and edoxaban) for efficacy and bleeding outcomes in comparison with vitamin K antagonists (VKA) in elderly participants (aged ≥75 years) treated for acute venous thromboembolism or stroke prevention in atrial fibrillation. Nineteen studies were eligible for inclusion, but only 11 reported data specifically for elderly participants. The efficacy in managing thrombotic risks for each DOAC was similar or superior to VKA in elderly patients. A nonsignificantly higher risk of major bleeding than with VKA was observed with dabigatran 150 mg (odds ratio, 1.18; 95% confidence interval, 0.97-1.44) but not with the 110-mg dose. Significantly higher gastrointestinal bleeding risks with dabigatran 150 mg (1.78, 1.35-2.35) and dabigatran 110 mg (1.40, 1.04-1.90) and lower intracranial bleeding risks than VKA for dabigatran 150 mg (0.43, 0.26-0.72) and dabigatran 110 mg (0.36, 0.22-0.61) were also observed. A significantly lower major bleeding risk in comparison with VKA was observed for apixaban (0.63, 0.51-0.77), edoxaban 60 mg (0.81, 0.67-0.98), and 30 mg (0.46, 0.38-0.57), whereas rivaroxaban showed similar risks.

Conclusions: DOACs demonstrated at least equal efficacy to VKA in managing thrombotic risks in the elderly, but bleeding patterns were distinct. In particular, dabigatran was associated with a higher risk of gastrointestinal bleeding than VKA. Insufficient published data for apixaban, edoxaban, and rivaroxaban indicate that further work is needed to clarify the bleeding risks of these DOACs in the elderly.

Systematic review registration: http://www.crd.york.ac.uk/PROSPERO. Unique identifier: PROSPERO CRD42014007171/.

Keywords: aged; anticoagulants; atrial fibrillation; hemorrhage; meta-analysis; systematic review; venous thromboembolism.

Figure 1. Summary of risk of bias assessment

Green(+) =Low bias risk; Red (−)=High bias risk; Yellow(?)=Unclear bias risk

Figure 2. Funnel Plot Comparison for risk of major bleeding in the elderly (top) and total population (below)

*Note: Y axis scales differ between plots above.

Figure 3. Risk of stroke or systemic embolism in atrial fibrillation studies in elderly (left) and total population (right)

*Event numbers for Engage-AF-Timi 48 in elderly have been estimated from published confidence intervals.

Figure 4. Risk of recurrent venous thromboembolism in venous thromboembolism studies in elderly (left) and total population (right)

Figure 5. Risk of major bleeding in elderly (left) and total population (right)

*Event numbers for Engage-AF-Timi-48 and J-Rocket AF in the elderly have been estimated from published confidence intervals.

Figure 6. Risk of secondary outcomes in elderly (left) and total population (right)

GIB= Gastrointestinal bleeding; ICB = Intracranial Bleeding; CRB= Clinically Relevant Bleeding; FB= Fatal Bleeding *CRB estimate was only estimate derived using a random effects model. **Note: Full Forest Plots for each estimate above are available in the supplementary appendix.