Abstract 15462: Risk of Stroke in Chronic Heart Failure Patients without Atrial Fibrillation: Analysis of the CORONA and GISSI-HF Trials

Introduction: The contemporary incidence of stroke in patients with heart failure (HF) but not in atrial fibrillation (AF) is uncertain. Furthermore, there may be a subset of such patients with a risk of stroke sufficient to justify anticoagulant prophylaxis. We report the incidence of stroke in HF patients with and without AF. We also describe predictors of stroke in patients without AF. These identify subjects with a risk of stroke as high as in patients with AF and who might merit anticoagulation. Methods: We determined the incidence of stroke in HF patients in the Controlled Rosuvastatin in Multinational Trial Heart Failure trial (CORONA) and the Gruppo Italiano per lo Studio della Sopravvivenza nell’Insufficienza cardiac- Heart Failure trial (GISSI-HF). Multivariable (MV) Cox regression models for stroke in patients without AF were built using 10 routine clinical and laboratory variables derived from univariable analysis of 25 baseline variables and variables identified from review of prior studies of stroke predictors. Results: A total of 9,585 patients were included. Of these, 6,054 had no AF. The median overall follow-up time was 3.10 (IQR: 2.38-3.82) years and 371 patients experienced a stroke (rate 13.1 per 1000 patient-years). The stroke rates in patients with and without AF were, 16.8 per 1000 patient-years and 11.1 per 1000 patient-years, respectively. In MV analysis, age, body mass index, NYHA class, insulin-treated diabetes, and history of previous stroke were the only independent predictors of stroke. Using the risk score formulated from this model, patients in the upper tertile had a rate of stroke that exceeded that of the patients with AF (19.3 per 1000 patient-years). Conclusions: In a large contemporary HF population, we found that a third of patients without AF have a risk of stroke that is at least as great as in HF patients with AF. This high-risk subset can be identified using simple clinical variables. A clinical trial of prophylactic anticoagulation is merited in these patients.