Predicting Major Bleeding in Ischemic Stroke Patients With Atrial Fibrillation

Area of Science:

• Cardiology
• Neurology
• Pharmacology

Background:

• Assessing bleeding risk in atrial fibrillation (AF) patients with prior stroke/TIA is crucial for oral anticoagulant (OAC) therapy.
• Existing risk scores' performance in this specific population is not well-established.

Purpose of the Study:

• To validate established bleeding risk scores in AF patients post-cerebral ischemia on OACs.
• To explore the net clinical benefit of OACs across different bleeding risk categories.

Main Methods:

• Analysis of 3623 patients from the RE-LY trial with AF and history of TIA/stroke.
• Performance assessment of HEMORR2HAGES, Shireman, HAS-BLED, ATRIA, and ORBIT scores using C statistics and calibration.
• Evaluation of net benefit by comparing ischemic stroke risk reduction with major bleeding risk increase on warfarin.

Main Results:

• Major bleeding occurred in 3.8 per 100 person-years.
• C statistics for risk scores ranged from 0.62 to 0.67, with ATRIA showing poor calibration and others moderate.
• Warfarin demonstrated a greater reduction in recurrent ischemic strokes than the increase in major bleeding risk.

Conclusions:

• Bleeding risk prediction models in AF patients with cerebral ischemia have modest, comparable performance to those without.
• Risk scores may help identify modifiable bleeding factors but cannot solely guide OAC treatment decisions.
• The ORBIT score shows potential clinical utility due to its simplicity and reasonable performance.