Evaluation of plasma natriuretic peptides as markers for left ventricular dysfunction

Abstract

To test the hypothesis that elevated plasma levels of natriuretic peptides may serve to identify patients with left ventricular (LV) dysfunction, we assessed the predictive diagnostic value of natriuretic peptide levels, in addition to clinical and electrocardiographic risk factors, as noninvasive indicators of cardiac dysfunction. Plasma levels of atrial natriuretic peptide (cANP) (99-126), N-terminal fragment of proANP (nANP) (26-55), nANP(80-96), brain natriuretic peptide (BNP-32), proBNP(22-46), and C-type natriuretic peptide (CNP-22) were measured in 211 subjects before cardiac catheterization. The strongest correlations with parameters of LV function were found for nANP(80-96) (up to r = –0.55, p < 0.0001), whereas there was no significant correlation with proBNP(22-46) or CNP-22. In patients with LV ejection fractions (LVEF) ≤45% ( n = 38) nANP(26-55), nANP(80-96), cANP(99-126), and BNP-32 were significantly increased ( p < 0.001). Partition values for elevated versus normal natriuretic peptide levels were obtained from normal controls and used to separate subjects with and without LV dysfunction. Receiver operating characteristic analysis for LVEF ≤45% indicated a significantly better diagnostic accuracy for high levels of nANP(80-96), nANP(22-56), cANP(99-126), and BNP-32 than for proBNP and CNP-22. Multivariate analysis by logistic regression identified Q waves and bundle branch block in the electrocardiogram as well as elevated plasma levels of cANP, nANP(80-96), and nANP(26-55) as the strongest independent predictors of low ejection fractions. The relative risk of LV dysfunction was raised up to tenfold in subjects with high natriuretic peptide levels ( p < 0.001). The addition of nANP(80-96) and nANP(26-55) to the combination of clinical and electrocardiographic risk factors did not further improve the diagnostic sensitivity for the detection of LVEF ≤45%, but it markedly increased the overall accuracy (59% to 81%, p < 0.001) and specificity (55% to 81%, p < 0.001). Among natriuretic peptides, elevated nANP(80-96) and nANP(26-55) levels have the strongest impact on the detection of LV dysfunction. They add to the diagnostic information contained in clinical and electrocardiographic factors. Plasma levels alone or in combination with clinical factors seem to be of value for a refined identification of abnormal LV function in the individual patient. (Am Heart J 1997;134:442-9.)