Use of Biomarkers to Predict Specific Causes of Death in Patients With Atrial Fibrillation : Insights From the ARISTOTLE Trial

Background: Atrial fibrillation is associated with an increased risk of death. High-sensitivity troponin T, growth differentiation factor-15, NT-proBNP (N-terminal pro-B-type natriuretic peptide), and interleukin-6 levels are predictive of cardiovascular events and total cardiovascular death in anticoagulated patients with atrial fibrillation. The prognostic utility of these biomarkers for cause-specific death is unknown. Methods: The ARISTOTLE trial (Apixaban for the Prevention of Stroke in Subjects With Atrial Fibrillation) randomized 18201 patients with atrial fibrillation to apixaban or warfarin. Biomarkers were measured at randomization in 14798 patients (1.9 years median follow-up). Cox models were used to identify clinical variables and biomarkers independently associated with each specific cause of death. Results: In total, 1272 patients died: 652 (51%) cardiovascular, 32 (3%) bleeding, and 588 (46%) noncardiovascular/nonbleeding deaths. Among cardiovascular deaths, 255 (39%) were sudden cardiac deaths, 168 (26%) heart failure deaths, and 106 (16%) stroke/systemic embolism deaths. Biomarkers were the strongest predictors of cause-specific death: a doubling of troponin T was most strongly associated with sudden death (hazard ratio [HR], 1.48; P<0.001), NT-proBNP with heart failure death (HR, 1.62; P<0.001), and growth differentiation factor-15 with bleeding death (HR, 1.72; P=0.028). Prior stroke/systemic embolism (HR, 2.58; P>0.001) followed by troponin T (HR, 1.45; P<0.0029) were the most predictive for stroke/ systemic embolism death. Adding all biomarkers to clinical variables improved discrimination for each cause-specific death. Conclusions: Biomarkers were some of the strongest predictors of cause-specific death and may improve the ability to discriminate among patients' risks for different causes of death. These data suggest a potential role of biomarkers for the identification of patients at risk for different causes of death in patients