144 Real-time three-dimensional echocardiography assessment of left ventricular function: left ventricular volumes underestimation increase with the dilatation degree

Abstract

Accurate quantification of left ventricular (LV) volumes and ejection fraction (EF) is important, especially in heart failure patients. Cardiac magnetic resonance imaging (MRI) is considered as the reference and then we sought to compare three-dimensional echocardiography (3DE) for LV volumes and EF estimation, relative to MRI in heart failure patients. We studied 24 patients (17 men, age 58 ± 15 years) with history of heart failure who underwent 3DE (Full volume real-time 3DE – iE33 (Philips Medical system) QLAB workstation) and MRI in a 48-hour delay. No patient was excluded from the study due to poor image quality. The heart failure etiology was: 41,7% (n=10) ischemic cardiomyopathy, 50% (n=12) dilated idiopathic cardiomyopathy and 8,3% (n=2) of patients suffered from heart failure with preserved EF. Mean LV end-diastolic volume (LVEDV) evaluated by MRI was 208 ± 108mL (121 ± 64 mL/m2), mean EF 31 ± 13% and mean LV end-systolic volume (LVESV) was 149 ± 97 mL. 3DE data sets highly correlated with MRI, especially concerning EF (r : 0.86, 0.88, and 0.96 for LVEDV, LVESV, and EF, respectively) with small biases (-55 mL, -44 mL, 1,1%) and acceptable limits of agreement, but LV volumes are significantly underestimated by 3DE particularly in our patients with severe LV dysfunction, and 3DE data sets do not correlate as well as expected. We then determined that with a LVEDV below 120 mL/m2, 3D was more accurate for volumes and EF evaluation. The degree of underestimation of LV volumes is significatively correlated with the LV dilatation degree (r = 0,60 p=0,01). Compared with MRI, 3DE is a good method to evaluate LVEF, but 3DE appears to underestimate significantly LV volumes, with the problem of foreshortened apical views in heart failure patients. As the LVEDV increase, 3D accuracy simultaneously decrease.