Predicting stroke in heart failure and reduced ejection fraction without atrial fibrillation.

Abstract

Patients with heart failure with reduced ejection fraction (HFrEF) are at significant risk of stroke. Anticoagulation reduces this risk in patients with and without atrial fibrillation (AF), but the risk-to-benefit balance in the latter group, overall, is not favourable. Identification of patients with HFrEF, without AF, at the highest risk of stroke may allow targeted and safer use of prophylactic anticoagulant therapy. In a pooled patient-level cohort of the PARADIGM-HF, ATMOSPHERE, and DAPA-HF trials, a previously derived simple risk model for stroke, consisting of three variables (history of prior stroke, insulin-treated diabetes, and plasma N-terminal pro-B-type natriuretic peptide level), was validated. Of the 20 159 patients included, 12 751 patients did not have AF at baseline. Among patients without AF, 346 (2.7%) experienced a stroke over a median follow up of 2.0 years (rate 11.7 per 1000 patient-years). The risk for stroke increased with increasing risk score: fourth quintile hazard ratio (HR) 2.35 [95% confidence interval (CI) 1.60-3.45]; fifth quintile HR 3.73 (95% CI 2.58-5.38), with the first quintile as reference. For patients in the top quintile, the rate of stroke was 21.2 per 1000 patient-years, similar to participants with AF not receiving anticoagulation (20.1 per 1000 patient-years). Model discrimination was good with a C-index of 0.84 (0.75-0.91). It is possible to identify a subset of HFrEF patients without AF with a stroke-risk equivalent to that of patients with AF who are not anticoagulated. In these patients, the risk-to-benefit balance might justify the use of prophylactic anticoagulation, but this hypothesis needs to be tested prospectively.