Log in to MyACC
Risk of Stroke in Heart Failure Patients Without AF | Journal Scan
Study Questions:
What are the incidence of and risk factors for stroke in patients with heart failure (HF) without atrial fibrillation (AF)?
Methods:
This analysis was performed using pooled data from the CORONA (Controlled Rosuvastatin Multinational Study in Heart Failure) and GISSI-HF (Gruppo Italiano per lo Studio della Sopravvivenza nell’Insufficienza Cardiac-Heart Failure) trials. CORONA enrolled patients >60 years old, with New York Heart Association (NYHA) class II-IV HF and HF with reduced ejection fraction (HFrEF) due to ischemic heart disease. GISSI-HF enrolled stable chronic NYHA class II-IV HF irrespective of age and etiology and included both HFrEF and HF with preserved EF (HFpEF). Incident stroke was evaluated in both studies whereas AF was collected prospectively only in GISSI-HF. AF was analyzed retrospectively as an adverse event in CORONA. N-terminal pro–B-type natriuretic peptide (NT-proBNP) was assessed in a subset of patients in both studies. A total of 9,585 patients were included, 6,054 of which did not have AF. Patients with AF were defined as those with AF confirmed on baseline electrocardiogram or a history of AF. Uni- and multivariable predictors of risk for stroke were assessed using Cox proportional hazards regression analysis. Multivariable analysis was performed in two steps: Step 1 used a best clinical model and variables found to be significant on univariate analysis. In Step 2, (loge)NT-proBNP was added to the independent variables. Cumulative incidence for stroke was estimated using competing risk analysis. The predictive model was validated in the CHARM (Candesartan in Heart Failure: Reduction in Mortality and Morbidity)-Alternative and CHARM-Added trials.
Results:
A total of 9,585 patients were included, 3,531 with AF and 6,054 without. NT-proBNP was available in 4,381 patients, 1,749 of whom had AF. In general, patients without AF were younger, had lower EF, better NYHA class, higher glomerular filatration rate, and lower NT-proBNP than those with AF. Patients without AF but who had experienced stroke were older, had higher NYHA class, higher creatinine, and were more likely to have had prior cerebrovascular accident, myocardial infarction, peripheral arterial disease, hypertension, and diabetes than those without stroke. The rate of stroke was higher in patients with AF compared to those without AF, 16.8 per 1,000 patient-years (pt-yrs) vs. 11.1 per 1,000 pt-yrs, respectively. In addition, the 1-, 2-, and 3-year cumulative incidence function rates of stroke were higher in patients with AF (1.7, 2.8, and 4.2 respectively) compared to those without AF (1.2, 2.2, and 3.1, respectively). The rate of stroke was also higher in patients who were not receiving anticoagulation. In those patients with AF who had NT-proBNP measurements, the rate of stroke was 20.3 per 1,000 pt-yrs compared to 13.5 per 1,000 pt-yrs in those without AF. The predictors of stroke in patients without AF were age (hazard ratio [HR], 1.34), NYHA class (HR, 1.6), diabetes treated with insulin (HR, 1.87), body mass index (HR, 0.74), and prior stroke (1.81). When BNP was added to the independent predictors, log NT-proBNP (HR, 1.32), diabetes treated with insulin (HR, 2.09), and prior stroke were significant predictors (HR, 1.92). Patients in the highest risk tertile had an overall stroke rate of 22.9 per 1,000 pt-yrs.
Conclusions:
HF patients without AF have a lower risk of stroke than those with AF. This risk-predictive model, using a small number of clinical and demographic variables, identified a subset of HF patients without AF whose risk of stroke approximated that of patients with AF.
Perspective:
This predictive model is a feasible method of identifying HF patients without AF who are at high risk for stroke. While these provocative findings raise the question of anticoagulation in a high-risk subgroup, prospective studies are required to determine its benefit.
Keywords: Anticoagulants, Atrial Fibrillation, Body Mass Index, Creatinine, Diabetes Mellitus, Heart Failure, Hypertension, Incidence, Insulin, Myocardial Infarction, Myocardial Ischemia, Natriuretic Peptide, Brain, Peptide Fragments, Peripheral Arterial Disease, Regression Analysis, Retrospective Studies, Risk Factors, Stroke
< Back to Listings
