22. European Stroke Conference 263 Etiology of stroke and risk factors NT-proBNP shows diagnostic accuracy for atrial fibrillation A.C. Fonseca1, D. Brito2, J. Sampaio Matias3, T. Pinho e Melo4, R. Geraldes5, P. Canhão6, L.R. Ca-plan7, 426 © 2013 S. Karger AG, Basel Scientific Programme J.M. Ferro8 Neurology Department, Hospital de Santa Maria, Lisboa, PORTUGAL1, Cardiology Depart-ment, Hospital de Santa Maria, Lisboa, PORTUGAL2, Clinical Pathology Department, Lisboa, PORTUGAL3, Neurology Department, Hospital de Santa Maria, Lisboa, PORTUGAL4, Neurology Department, Hospital de Santa Maria, Lisboa, PORTUGAL5, Neurology Department, Hospital de Santa Maria, Lisboa, PORTUGAL6, Beth Israel Deaconess Medical Ctr, Boston, USA7, Neurology Department, Hospital de Santa Maria, Lisboa, PORTUGAL8 Background: The diagnosis of paroxysmal atrial fibrillation (pAF) in patients with stroke can be dif-ficult. Aim: To determine if cut-off points of serum NT-proBNP at the time of stroke onset can help identi-fy pAF in cryptogenic stroke patients. Methods: Among 264 ischemic stroke patients admitted to a stroke unit NT-proBNP serum levels were measured within 72 hours of stroke onset. Etiology was classified using TOAST criteria. In pa-tients with cryptogenic stroke, 24-Holter was done to look for pAF within the first week, three and six months after admission. First, patients with a defined etiology were used to construct a Receiver Operating Characteristic (ROC) curve for the diagnosis of AF. From this curve, the sensitivity and specificity of pre-established cut-off points was calculated. A logistic regression was done to assess the independent relationship of lnNT-proBNP with AF controlling for confounders. In a second phase the cut-off points previously established were evaluated, in patients with cryptogenic stroke, for the diagnosis of pAF. Results: In 184 patients a stroke etiology could be established. Fifty five patients had the diagno-sis of cardioembolic stroke related to AF. Using multivariate analysis the lnNT-proBNP was inde-pendently associated to AF(OR –2.7, 95%IC 1.6-4.5; p<0.0001). The Area Under the Curve (AUC) of the ROC curve of lnNT-proBNP for the diagnosis of AF was excellent -0.91, 95%IC (0.87-0.95). The previously defined cut-off of 265.5 pg/ml had a sensitivity of 100% and specificity of 70.6%. The cut-off point of 912 pg/ml had a specificity of 87.5%. 80 patients had a cryptogenic stroke. In 17 (21%) pAF was found during follow-up. The cut-off point of 265.5 pg/ml had a sensitivity of 88.2% and a specificity of 61.9% for the diagnosis of pAF. The cut-off point of 912 pg/mL had a specificity of 88.9%. Conclusion: NT-proBNP had good accuracy to predict pAF in patients with cryptogenic stroke and can help evaluate these patients. 264 Etiology of stroke and risk factors Withdrawn!
Karger_ESC London_2013
To see the actual publication please follow the link above