Background: Atrial fibrillation often leads to ischemic stroke. For secondary prevention, clinicians typically switch from antiplatelet to anticoagulant therapy for patients with confirmed atrial fibrillation. This study examined the predictive value of N-terminal pro-brain natriuretic peptide (NT-proBNP) and brain natriuretic peptide (BNP) for detecting covert paroxysmal atrial fibrillation (PAF) in patients with ischemic stroke (PWIS).
Methods: We enrolled 438 patients with acute stroke in sinus rhythm on admission from July 2021 to March 2023 and measured their NT-proBNP and BNP levels to evaluate their association with PAF detection. Data analysis included logistic regression, receiver operating characteristic curves, and integrated discrimination improvement (IDI).
Results: Among our 438 enrolled participants, 43 (9.8%) were in the PAF group and the remaining were in the non-PAF group. PAF group patients were older than those in the non-PAF group (PAF group vs. non-PAF group; 84 [78-89] vs. 79 [71-85] years) and had higher levels of both NT-proBNP (581.0 [264.5-1,234.5] vs. 168.0 [76.0-412.5] pg/mL) and BNP (186.0 [100.4-313.0] vs. 56.4 [26.9-118.0] pg/mL). The PAF group also had a higher prevalence of chronic heart failure (30% vs. 10%). Both biomarkers were independent predictors of PAF detection, and there was no significant difference in their predictive accuracy for PAF. However, BNP had a slight advantage in the IDI score over NT-proBNP (-0.03 [-0.05 to -0.01]).
Conclusion: NT-proBNP and BNP can both effectively predict covert PAF in PWIS in sinus rhythm. Thus, either biomarker should be incorporated into treatment planning strategies for these patients.