Lack of decrease in plasma N-terminal pro-brain natriuretic peptide identifies acute heart failure patients with very poor outcome

Background Optimal risk stratification in heart failure patients surviving an episode of acute decompensation has not yet been established. We investigated whether a lack of significant decrease in plasma levels of N-terminal pro-brain natriuretic peptide (NT-proBNP) during hospital stay can identify patients at high risk of poor outcome. Methods We studied 103 consecutive patients with acute heart failure (86 men, age: 64 ± 13 years, LVEF: 28 ± 8%). The primary end-point was all-cause mortality at 1-year follow-up. Results Median plasma NT-proBNP on admission was 6116 pg/mL (upper/lower quartiles: 3575, 10,958) vs. 2930 pg/mL (1674, 5794) after clinical stabilization (7 ± 3 days after admission). During the 1-year follow-up 29 (28%) patients died. A decrease in plasma NT-proBNP during clinical recovery (expressed as percentage of NT-proBNP on admission) predicted favorable outcome in the single predictor analysis (p < 0.001) and multivariable analyses (p < 0.001). Receiver operating characteristic curve analysis revealed that 65% was the cut-off value for NT-proBNP decrease having best prognostic accuracy for predicting death (sensitivity 90%, specificity 37%, AUC = 0.65, 95% CI: 0.54–0.74). Kaplan–Meier analysis showed that 12-month survival was 92% (95% CI: 81–100%) for patients with ≥ 65% NT-proBNP decrease vs 66% (95% CI: 56–76%) in those with < 65% NT-proBNP decrease (p = 0.02). Conclusions The magnitude of plasma NT-proBNP decrease in patients with acute heart failure is helpful in discrimination of patients at high risk of death. Plasma NT-proBNP level monitoring is important for risk stratification in this group of patients.