Abstract
A B-type natriuretic peptide (BNP)-guided strategy is being widely used as a superior management technique for heart failure (HF). However, the optimal target level of BNP to improve the prognosis of HF in clinical practice remains unclear. Several studies have recently demonstrated that the existence of atrial fibrillation (AF) affects plasma BNP levels. We evaluated the prognostic value of BNP assay for HF management and found the optimal target level under the BNP-guided HF management according to the basal cardiac rhythms: AF or sinus rhythm (SR).
Patients hospitalized for HF exacerbation between 1996 and 2002 were stratified into SR (n = 129) and chronic AF (CAF, n = 58) groups as basal cardiac rhythms during hospitalization. Cardiac events including death and re-admission for HF exacerbation after discharge were analyzed in relation to the plasma BNP levels at predischarge. Receiver-operating characteristic (ROC) analysis demonstrated that the cut-off values for predischarge BNP, which predict cardiac events at 36 months after discharge, were 125 pg/mL in the SR group and 165 pg/mL in the CAF group. The area under the ROC curve was 0.72 and 0.82, respectively. Stratified subgroup analysis using the Kaplan-Meier method demonstrated that the risk of a cardiac event decreased in a stepwise fashion across a decreasing predischarge BNP range above these cut-off levels, while the minimum decreased risk was recognized at a BNP range below these cut-off levels in each group.
In conclusion, the optimal target levels of plasma BNP at predischarge to improve the prognosis of HF should be different and distinguishable depending on with or without AF.
