Abstract 4728: Assessment of Left Ventricular Relaxation by Untwisting Rate Based on Different Algorithms
Some studies using speckle tracking echocardiography reported reduced untwisting rates in patients with heart failure and normal EF; others noted a normal rate. This may be due to different algorithms for calculating untwisting rate. We sought to investigate the relation between LV relaxation and untwisting rate calculated by different algorithms. Real time loops of LV cross sectional images at the apical, middle, and basal levels were acquired in 72 patients simultaneously with cardiac catheterization. LV apical and basal rotations were measured by speckle tracking, and twist calculated as the difference between apical and basal rotations. Untwisting rate was obtained by 3 algorithms. First, the peak negative time derivative of twist during diastole (URmax) was measured using data points from the isovolumetric relaxation (IVRT) and early filling periods. Second, linear regression of untwisting rate versus time was performed during the 64 ms immediately after aortic valve closure, and the slope of the regression or URslope derived. Third, untwisting rate was calculated as: (Twistmax − Twistmvo/Twistmax × 100)/IVRT, where Twistmvo is twist at the mitral valve opening, and Tmax is maximum twist. The latter approach can be considered the average untwisting rate during the IVR period (URmean). Image quality was adequate in 57 patients, and 18 had heart failure with normal EF. URmax was significantly related to tau (r = −0.54, P<0.001), but not URslope nor URmean (r=0.24 and 0.06 respectively, both P>0.1). Peak diastolic untwisting rate (URmax) was significantly lower (P<0.001) in patients with depressed EF (58±30°/s) vs. normal subjects (116±36°/s), but was normal in patients with heart failure and normal EF (126±34°/s). Untwisting rate by linear regression analysis (URslope) was significantly lower in patients (depressed EF: 0.01; normal EF: 0.22%/ms) vs. normal controls (0.46%/ms, P<0.001). UR mean was not statistically different between patients (irrespective of EF) and normal controls (P=0.15). Among the 3 algorithms for calculating untwisting rate, only peak untwisting rate (bURmax) is significantly related to the time constant of LV relaxation, and is normal in patients with heart failure and normal EF.