Analysis of left ventricular volumes and function: a multicenter comparison of cardiac magnetic resonance imaging, cine ventriculography, and unenhanced and contrast-enhanced two-dimensional and three-dimensional echocardiography

Hoffmann R, Barletta G, von Bardeleben S, Vanoverschelde JL, Kasprzak J, Greis C, Becher H

J Am Soc Echocardiogr 2014 Mar;27(3):292-301

PMID: 24440110

Abstract

BACKGROUND: Contrast echocardiography improves accuracy and reduces interreader variability on left ventricular (LV) functional analyses in the setting of two-dimensional (2D) echocardiography. The need for contrast imaging using three-dimensional (3D) echocardiography is less defined. The aim of this multicenter study was to define the accuracy and interreader agreement of unenhanced and contrast-enhanced 2D and 3D echocardiography for the assessment of LV volumes and ejection fraction (EF).

METHODS: A multicenter, open-label study was conducted including 63 patients, using intrasubject comparisons to assess the agreement of unenhanced and contrast-enhanced 2D and 3D echocardiography as well as calibrated biplane cine ventriculography with cardiac magnetic resonance for the determination of LV volumes and EF. Each of the imaging techniques used to define LV function was assessed by two independent, off-site readers unaware of the results of the other imaging techniques.

RESULTS: LV end-systolic and end-diastolic volumes were underestimated by 2D and 3D unenhanced echocardiography compared with cardiac magnetic resonance. Contrast enhancement resulted in similar significant increases in LV volumes on 2D and 3D echocardiography. The mean percentage of interreader variability for LV EF was reduced from 14.3% (95% confidence interval [CI], 11.7%-16.8%) for unenhanced 2D echocardiography and 14.3% (95% CI, 9.7%-18.9%) for unenhanced 3D echocardiography to 8.0% (95% CI, 6.3%-9.7%; P < .001) for contrast-enhanced 2D echocardiography and 7.4% (95% CI, 5.7%-9.1%; P < .01) for contrast-enhanced 3D echocardiography and thus to a similar level as for cardiac magnetic resonance (7.9%; 95% CI, 5.4%-10.5%). A similar effect was observed for interreader variability for LV volumes.

CONCLUSIONS: Contrast administration on 3D echocardiography results in improved determination of LV volumes and reduced interreader variability. The use of 3D echocardiography requires contrast application as much as 2D echocardiography to reduce interreader variability for volumes and EF.