Abstract

Atrial and brain natriuretic peptides (ANP and BNP) are produced by the heart, and their plasma concentrations are increased in human chronic congestive heart failure. Although separate studies have suggested that circulating levels of the biologically active C-terminal ANP, the biologically inactive N-terminal ANP, and BNP may have diagnostic utility in the detection of left ventricular systolic dysfunction or left ventricular hypertrophy, no studies have directly assessed the relative value of these peptides prospectively. We therefore designed this study to compare the relative ability of the different natriuretic peptides to detect abnormal left ventricular systolic and diastolic function and left ventricular hypertrophy. Using a prospective study design, we investigated 94 patients referred for cardiac catheterization and 15 age-matched normal subjects. The diagnostic abilities of elevated plasma C-terminal ANP, N-terminal ANP-(1-30), and BNP concentrations to identify systolic dysfunction (ejection fraction <45%), diastolic dysfunction (time constant of left ventricular relaxation >55 milliseconds, left ventricular end-diastolic pressure >18 mm Hg), and left ventricular hypertrophy (left ventricular mass index >120 g/m 2 ) were objectively compared by receiver operating characteristic analysis. The areas under the receiver operating characteristic curve of BNP for detecting each of these abnormalities ranged from 0.715 to 0.908 and were significantly greater than those of C-terminal ANP or N-terminal ANP-(1-30). The sensitivity and specificity of an elevated plasma BNP, which we defined as greater than the mean+3 SD of the 15 age-matched normal subjects, were 0.83 and 0.77, respectively, for detecting ejection fraction less than 45%, 0.85 and 0.70 for detecting the time constant of left ventricular relaxation greater than 55 milliseconds, 0.63 and 0.76 for detecting left ventricular end-diastolic pressure greater than 18 mm Hg, and 0.81 and 0.85 for detecting left ventricular mass index greater than 120 g/m 2 . The use of BNP and one other peptide increased sensitivity (0.90 to 0.96), albeit with lower specificity (0.56 to 0.71). An elevated plasma BNP was a more powerful marker of left ventricular systolic dysfunction, left ventricular diastolic dysfunction, and left ventricular hypertrophy than C-terminal ANP or N-terminal ANP-(1-30) in this population of patients with suspected cardiac disease. Measurement of BNP alone or in combination with C-terminal ANP or N-terminal ANP-(1-30) has potential utility for the detection of altered left ventricular structure and function in a patient population at risk for cardiovascular disease.