Correlation of tissue Doppler and two-dimensional speckle myocardial velocities and comparison of derived ratios with invasively measured left ventricular filling pressures

Abstract

Background: There are a paucity of data comparing spectral and color tissue Doppler (TD) with non-Doppler, speckle-based myocardial velocities, and it is unknown how early transmitral diastolic velocity/mitral annular velocity (E/Ea) calculated using speckle velocities compares with TD-derived E/Ea.

Methods: We measured systolic (Sa), Ea, and late diastolic (Aa) myocardial velocities using these 3 methods and compared calculated E/Ea with invasively measured left ventricular (LV) hemodynamics. Consecutive patients referred for cardiac catheterization were imaged, and LV pre-A contraction pressure was measured by retrograde aortic cardiac catheterization.

Results: Fifty patients (22 women, 44%) were studied with a mean age of 54 +/- 10 years and a mean LV ejection fraction of 48% +/- 19%. Speckle and color TD Sa, Ea, and Aa were significantly lower than spectral TD velocities, resulting in higher E/Ea values compared with spectral TD E/Ea. Spectral TD E/Ea (R = 0.65, P < .001), color TD E/Ea (R = 0.69, P < .001), and speckle E/Ea (R = 0.76, P < .001) all correlated with LV pre-A pressure. Different cutoff values were needed for spectral TD, color TD, and speckle E/Ea to predict LV pre-A pressure >or= 15 mm Hg; spectral E/Ea > 14 (area under the curve [AUC] = 0.87, P < .001, sensitivity = 83%, specificity = 77%), color E/Ea > 16 (AUC = 0.89, P < .0001, sensitivity = 79%, specificity = 81%), and speckle E/Ea > 18 (AUC = 0.87, P < .0001, sensitivity = 88%, specificity = 74%; P = not significant for comparisons between the groups).

Conclusion: Spectral TD, color TD, and speckle imaging measure different velocities, and consequently different cutoff values are needed for E/Ea to predict invasively measured LV filling pressure.