Brain Natriuretic Peptide and Discovery of Atrial Fibrillation After Stroke A Subanalysis of the Find-AFRANDOMISED Trial

Background and Purpose-Diagnosing paroxysmal atrial fibrillation (pAF) can be challenging after acute ischemic stroke. Enhanced and prolonged Holter-ECG monitoring (EPM) improves the detection rate but is not feasible for all patients. We hypothesized that brain natriuretic peptide (BNP) may help to identify patients with stroke at high risk for pAF to select patients for EPM more effectively. Methods-Patients with acute cerebral ischemia >= 60 years presenting in sinus rhythm and without history of AF were included into a prospective, randomized multicenter study to receive either EPM (3x 10-day Holter-ECG) or usual stroke care diagnostic work-up. BNP plasma levels were measured on randomization and 3 months thereafter. Levels were compared between patients with and without pAF detected by means of EPM or usual care. Furthermore, the number needed to screen for EPM depending on BNP cut offs was calculated. Results-A total of 398 patients were analyzed. In 373 patients (93.7%), BNP was measured at baseline and in 275 patients (69.1%) after 3 months. pAF was found in 27 patients by means of EPM and in 9 patients by means of usual care (P=0.002). Median BNP was higher in patients with pAF as compared to patients without AF in both study arms at baseline (57.8 versus 28.3 pg/mL in the EPM arm, P=0.0003; 46.2 versus 27.7 pg/mL, P=0.28 in the control arm) and after 3 months (74.9 versus 31.3 pg/mL, P=0.012 in the EPM arm, 99.3 versus 26.3 pg/mL, P=0.02 in the control arm). Applying a cut off of 100 pg/mL, the number needed to screen was reduced from 18 by usual care to 3 by EPM. Conclusions-BNP measured early after ischemic stroke identifies a subgroup of patients with stroke at increased risk for AF, in whom EPM is particularly efficacious.