Doppler echocardiographic transmitral peak early velocity does not directly reflect hemodynamic changes in humans:...

Doppler echocardiographic transmitral peak early velocity normalized to the time-velocity integral during diastole is equivalent to volumetric peak filling rate normalized to stroke volume. To compare the pathophysiologic validity of normalized and nonnormalized peak early flow velocity, pulsed Doppler echocardiography with simultaneous high fidelity left ventricular pressure measurements was performed in 52 patients with coronary artery disease. Left ventricular loading conditions were changed by intravenous administration of norepinephrine in 15 patients and synthetic atrial natriuretic polypeptide in 15 others. Norepinephrine increased nonnormalized and normalized peak early flow velocities in association with significantly elevated end-diastolic, peak systolic and mitral valve opening pressures and decelerated the time constant of left ventricular isovolumetric pressure decline. Atrial natriuretic polypeptide did not change either nonnormalized or normalized peak early flow velocity, despite significant reductions in end-diastolic, peak systolic and mitral valve opening pressure and an accelerated time constant. Normalized peak early flow velocity showed the highest univariate correlation with long-term change in mitral valve opening pressure (n = 52, r = 0.67, p less than 0.0001). It provided a modest univariate correlation (n = 30, r = 0.74, p less than 0.0001) with immediate change in mitral valve opening pressure during norepinephrine infusion, whereas this correlation was lower (n = 30, r = 0.57, p less than 0.001) during polypeptide infusion. However, multivariate regression analysis relating normalized peak velocity with long- and short-term changes in end-diastolic, peak systolic and mitral valve opening pressures, time constant and constant of left ventricular chamber stiffness improved the correlation coefficients (r = 0.80 to 0.85, all p less than 0.0001). In contrast, neither univariate nor multivariate correlations of nonnormalized velocity with long- and short-term changes in these hemodynamic variables were satisfactory. Thus, nonnormalized peak early flow velocity does not directly reflect underlying hemodynamic changes in humans. Normalization to mitral stroke volume clarifies the dependence of peak early flow velocity on the determinants of early diastolic filling. When left ventricular early diastolic filling is evaluated by Doppler echocardiography, normalized peak early flow velocity should be taken into consideration.