Noninvasive estimation of left ventricular filling pressures in patients with heart failure after surgical ventricular restoration and restrictive mitral annuloplasty

Abstract

Objective: Doppler echocardiography, including tissue Doppler imaging, is widely applied to assess diastolic left ventricular function using early transmitral flow velocity combined with mitral annular velocity as a noninvasive estimate of left ventricular filling pressures. However, the accuracy of early transmitral flow velocity/mitral annular velocity in patients with heart failure, particularly after extensive cardiac surgery, is debated. Global diastolic strain rate during isovolumic relaxation obtained with 2-dimensional speckle-tracking analysis was recently proposed as an alternative approach to estimate left ventricular filling pressures.

Methods: We analyzed diastolic function in patients with heart failure after surgical ventricular restoration and/or restrictive mitral annuloplasty. Echocardiography, including tissue Doppler imaging and speckle-tracking analysis, was performed to determine early transmitral flow velocity/atrial transmitral flow velocity, isovolumetric relaxation time, deceleration time, early transmitral flow velocity/mean mitral annular velocity, strain rate during isovolumic relaxation, and early transmitral flow velocity/strain rate during isovolumic relaxation. These noninvasive indices were correlated with relaxation time constant Tau, peak rate of pressure decline, and left ventricular end-diastolic pressure obtained in the catheterization room using high-fidelity pressure catheters.

Results: Twenty-three patients were analyzed 6 months after restrictive mitral annuloplasty (n = 8), surgical ventricular restoration (n = 4), or a combined procedure (n = 11). The strongest correlation with invasive indices, in particular left ventricular end-diastolic pressure, was found for strain rate during isovolumic relaxation (r = -0.76, P < .001). Early transmitral flow velocity/mean mitral annular velocity did not correlate significantly with any of the invasive indices. Strain rate during isovolumic relaxation (cutoff value < 0.38 s(-1)) accurately predicted left ventricular end-diastolic pressure of 16 mm Hg or more with 100% sensitivity and 93% specificity.

Conclusions: In a group of patients with heart failure who were investigated 6 months after cardiac surgery, early transmitral flow velocity/mean mitral annular velocity correlated poorly with invasively obtained diastolic indexes. Global strain rate during isovolumic relaxation, however, correlated well with left ventricular end-diastolic pressure and peak rate of pressure decline. Our data suggest that global strain rate during isovolumic relaxation is a promising noninvasive index to assess left ventricular filling pressures in patients with heart failure after extensive cardiac surgery, including restrictive mitral annuloplasty and surgical ventricular restoration.