Comparison of interstudy reproducibility of cardiovascular magnetic resonance with two-dimensional echocardiography in normal subjects and in patients with heart failure or left ventricular hypertrophy.

Fast breath-hold cardiovascular magnetic resonance (CMR) shows excellent results for interstudy reproducibility of left ventricular (LV) volumes, ejection fraction, and mass, which are thought to be superior to results of 2-dimensional echocardiography. However, there is no direct comparison of the interstudy reproducibility of both methods in the same subjects. A total of 60 subjects (normal volunteers [n = 20], or patients with heart failure [n = 20] or LV hypertrophy [n = 20]) underwent 2 CMRs and 2 echocardiographic studies for assessment of LV volumes, function, and mass. The interstudy reproducibility coefficient of variability was superior for CMR in all groups for all parameters. Statistical significance was reached for end-systolic volume (4.4% to 9.2% vs 13.7% to 20.3%, p <0.001), ejection fraction (2.4% to 7.3% vs 8.6% to 19.4%, p <0.001), and mass (2.8% to 4.8% vs 11.6% to 15.7% p <0.001), with a trend for end-diastolic volume (2.9% to 4.9% vs 5.5% to 10.5%, p = 0.17). The superior interstudy reproducibility resulted in considerably lower calculated sample sizes (reductions of 55% to 93%) required by CMR compared with echocardiography to show clinically relevant changes in LV dimensions and function. Thus, CMR has excellent interstudy reproducibility in normal, dilated, and hypertrophic hearts, and is superior to 2-dimensional echocardiography.