Brain natriuretic peptide level on admission predicts recurrent stroke after discharge in stroke survivors with atrial fibrillation

Abstract

ObjectiveWe investigated whether brain natriuretic peptide (BNP) levels could be used as a marker to predict recurrent stroke in ischemic stroke survivors. MethodsFrom April 2007 to March 2011, consecutive patients within 24h of onset of ischemic stroke were prospectively enrolled, and admission plasma BNP levels were measured. Survivors were followed up to 12 months after stroke onset. Patients were divided into two groups: the recurrence group and the non-recurrence group. Factors associated with stroke recurrence were investigated by multiple logistic regression analysis. ResultsA total of 793 patients who were alive at hospital discharge were included; 42 (5%) patients had recurrent stroke. There were no differences in BNP levels between groups. With respect to 257 patients with atrial fibrillation (AF), BNP levels were significantly higher in the recurrence group than in the non-recurrence group (426.0 vs. 192.0pg/mL, P=0.0007). The optimal cutoff level, sensitivity, and specificity of BNP levels to distinguish the recurrence group from the non-recurrence group were 300.0pg/mL, 80%, and 73%, respectively. After adjustment for age and sex, plasma BNP≥300.0pg/mL (OR, 9.2; 95% CI, 1.87–45.01, P=0.0062) was found to be independently associated with recurrent stroke in stroke survivors with AF. ConclusionAdmission BNP levels can predict recurrent stroke in stroke survivors with AF.