Baseline natriuretic peptides as a predictor of atrial fibrillation recurrence after radiofrequency-based pulmonary vein isolation in a non-heart failure population: A subanalysis from DECAAF II.

Abstract

Brain natriuretic peptide (BNP) is a marker of myocardial stretch and may have prognostic significance in patients with atrial fibrillation (AF) without heart failure (HF). We investigated the association between baseline BNP levels and arrhythmia recurrence following pulmonary vein isolation (PVI) among patients with persistent AF without HF. We analyzed 125 patients with persistent AF without HF who had baseline BNP measured from the DECAAF II trial. The primary outcome was arrhythmia recurrence following ablation. The baseline characteristics across the two groups were compared using Chi-square test and Wilcoxon rank test accordingly. Cox regression analysis was used to analyze the association between baseline BNP levels and the primary outcome. Across the entire cohort, 64 (51%) patients experienced arrhythmia recurrence. When comparing patients who experienced arrythmia recurrence to patients who did not, patients with recurrent arrhythmia had higher levels of pre-ablation BNP, as evidenced by differences in means (330.05pg/mL) compared to patients without recurrent arrhythmia (182.39 pg/mL) (p<.05). A cut-off BNP value of 300pg/mL provided the largest area under curve (AUC) of receiver-operating characteristic (ROC) curve on univariate logistic regression. On unadjusted Cox analysis, for every 100 unit increase in BNP, the hazard ratio for the primary outcome increased 1.09 (1.026-1.158) times (p=.004). After adjusting for sex, hypertension, and stroke, the results remained significant (HR=1.8516, CI 95% [1.0139 - 3.381], p=.045). In the non-heart failure population, BNP levels predict AF recurrence following PVI in persistent AF patients.