Background: Echocardiographic diagnosis of radial fractional shortening or global ejection fraction fails to diagnose all patients with heart failure; about 40%, with apparently normal global systolic function, will have elevated brain natriuretic peptide (BNP) concentrations and "pure" diastolic dysfunction. Screening methods do not include assessment of left ventricular (LV) longitudinal function, however, which is a more sensitive marker of subclinical disease.
Aims: We investigated the diagnostic potential of assessment of LV longitudinal function in suspected heart failure, in a comparison against BNP.
Methods: Fifty consecutive subjects (aged 65+/-12 years; 30 men) referred to a heart failure clinic with unexplained breathlessness were examined by echocardiography and had plasma BNP measured by fluorescence immunoassay.
Results: Global systolic function (ejection fraction) correlated moderately with log-transformed BNP (r=-0.54), as did global diastolic function (r=0.55 for estimated LV filling pressure, and r=-0.51 for transmitral flow propagation velocity), and radial systolic function (r=-0.60) (all p<0.001). The echocardiographic parameter that correlated best with BNP, however, was LV longitudinal systolic function (r=-0.78, p<0.001). By stepwise multiple regression analysis, BNP was predicted by longitudinal systolic velocity, in association with LV mass index and radial systolic velocity (r=0.81, r(2)=0.66, p<0.0001). The sensitivity and specificity of a longitudinal systolic velocity of </=5.5 cm/s (mean of 4 basal segments) to diagnose heart failure (defined as an elevated BNP) were 94% and 85%, respectively; the negative predictive value was 97%.
Conclusion: Non-invasive diagnosis or exclusion of suspected heart failure, whether systolic or diastolic, can be performed accurately by echocardiographic measurement of LV longitudinal systolic function.