Abstract 16635: Real Time 3d Echocardiography is Comparable to Cardiac Mri in Reducing Sample Size Required to Assess Changes in Lv Function

Abstract

Background: Magnetic Resonance (MRI) is the gold-standard for quantifying left ventricle (LV) volumes and ejection fraction (LVEF), allowing for considerably smaller sample sizes relative to conventional 2D echo, for detecting differences in LVEF and measures of LV remodeling in research studies involving myocardial infarction (MI) and LV dysfunction. MRI, however, is quite costly, not widely available, requires lengthy exam times, and is not portable. We hypothesized that some of the superior accuracy of MRI measurements compared to 2D echo, and invasive contrast LV angiography is due to the full 3D volumes measured by cardiac MRI. Therefore, we predicted that using 3D echo would reduce the sample size required to detect relevant differences in measures of LV remodeling, similar to MRI.

Methods: MI was induced in Yorkshire pigs by balloon occlusion of the proximal LAD for 60 minutes, followed by reperfusion. Within 1 month post MI, animals underwent cardiac MRI, 3D-echo, 2D-echo, and LVGram to measure LV volumes and LVEF. The sample sizes required for each of these imaging techniques to detect a 5% change in LVEF with a power of 80% and an α level of 0.05 were calculated.

Results: Post MI, LVEF measured by cardiac MRI was 39.1 ± 4.1%, compared with 38.9 ± 5.1% measured by 3D echo; 42.4 ± 9.1% measured by 2D-echo; and 43.1 ± 8.1% measured by LVGram. The sample sizes required (power of 90% and an α level of 0.05) for detecting a 5% change in LVEF were calculated to be 12 subjects per group for MRI, 17 per group for 3D-echo, 53 per group for 2D-echo and 47 per group for LVGram. The percent reduction in sample size compared with 2D-echo was 77% for MRI and 68% for 3D-echo for detecting a 5% difference in LVEF. The percent reduction in sample size compared with LVGram was 74% for MRI and 68% for 3D-echo for detecting a 5% difference in LVEF.

Conclusions: Our results demonstrate that 3D-echo allows a considerable reduction in sample size necessary to assess LV function in heart failure and MI studies compared with 2D echo and LVgram. Importantly, this is similar to the reduction observed when using cardiac MRI. Our findings indicate that 3D echo is a fast, inexpensive, and reliable alternative to cardiac MRI that can help preserve important resources in research studies.