BACKGROUND:Peak oxygen consumption is a cornerstone for prognostic determination in patients with congestive heart failure. The purpose of this study was to assess whether plasma B-type natriuretic peptide (BNP) provided any additional prognostic information. METHODS:Plasma concentrations of atrial natriuretic peptide, N terminal pro-atrial natriuretic peptide, BNP, endothelin-1, norepinephrine, and peak VO2 were measured in 250 consecutive outpatients with mild to moderate heart failure (96% in New York Heart Association [NYHA] class II or III) and left ventricular ejection fraction (LVEF) <45%. RESULTS:During a median follow-up of 584 days, 42 patients died (19 from sudden death) and 5 underwent urgent heart transplantation. Multivariate stepwise regression analysis showed that, among 13 variables including NYHA and LVEF, plasma BNP (chi2 = 11.9, P =.0001) was the strongest independent predictor of death or urgent transplantation, followed by serum sodium (chi2 = 8, P =.0046), resting heart rate (chi2 = 7.5, P =.0062), plasma endothelin-1 (chi2 = 7.2, P =.007), and peak VO2 (chi2 = 6.2, P =.012). Patients with plasma BNP above the upper quartile value (260 pg/mL) had a 1-year rate of death or urgent transplantation of 31%. The 1- and 2-year survival rates without urgent transplantation in patients with a peak VO2 < or =14 mL x kg(-1) x min(-1) were 71% and 59%, respectively, when plasma BNP was >137 pg/mL (median value), compared with 100% and 89%, respectively, when plasma BNP was < or =137 pg/mL (P =.008). Furthermore, plasma BNP was the only independent predictor of sudden death (chi2 = 19.9, P =.00001). CONCLUSIONS:Plasma BNP provides additive independent prognostic information compared to peak VO2 alone in outpatients with mild to moderate heart failure.

We compared the value of plasma neurohormones and cardiopulmonary exercise testing for predicting long-term prognosis in patients with moderate congestive heart failure (CHF). We studied 264 consecutive patients with CHF due to left ventricular systolic dysfunction. Plasma atrial natriuretic peptide (ANP), norepinephrine, and endothelin-1 were measured at rest in all patients, who also underwent a symptom-limited maximal exercise with oxygen consumption (VO(2)) determination. After a median follow-up of 789 days, 52 deaths and 31 heart transplantations occurred, of which 4 were urgent. In an univariate analysis, New York Heart Association functional class, systolic blood pressure at rest, left ventricular end-diastolic diameter, left ventricular ejection fraction, peak VO(2), percent of predicted peak VO(2), plasma ANP, plasma norepinephrine, and plasma endothelin-1 were associated with survival without urgent heart transplantation. In a multivariate stepwise regression analysis, only plasma ANP (p = 0.0001), left ventricular ejection fraction (p = 0.007), and plasma norepinephrine (p = 0.035), but neither peak VO(2) nor percentage of predicted peak VO(2), were independent predictors of death or urgent heart transplantation. Determination of plasma ANP and norepinephrine provides additional independent information for long-term prognostic determination compared with exercise testing alone. Measurement of plasma neurohormones should therefore be considered routinely as a complementary or alternative tool for identifying high-risk patients with moderate CHF.