Comparison of two- and three-dimensional unenhanced and contrast-enhanced echocardiographies versus cineventriculography versus cardiac magnetic resonance for determination of left ventricular function

Hoffmann R, von Bardeleben S, Barletta G, Pasques A, Kasprzak J, Greis C, Becher H

Am. J. Cardiol. 2014 Jan;113(2):395-401

PMID: 24188892

Abstract

Contrast enhancement has been shown to improve detection of regional wall motion abnormalities (RWMA) in 2-dimensional (2D) echocardiography. This study determined the use of contrast enhancement in the setting of 3-dimensional (3D) echocardiography for definition of left ventricular RWMA compared with 2D echocardiography, cineventriculography, and cardiac magnetic resonance (CMR). In 63 patients, unenhanced and contrast-enhanced (SonoVue; Bracco Imaging S.p.A., Milan, Italy) 2D and 3D echocardiographies, CMR, and cineventriculography were performed. Hypokinesia in ≥1 segment defined the presence of RWMA. Interreader agreement (IRA) between 2 blinded off-site readers on presence of RWMA was determined within each imaging technique. Intermethod agreement among imaging techniques was analyzed. A standard of truth for the presence of RWMA was obtained by an independent expert panel decision. IRA on presence of RWMA expressed as Cohen’s κ coefficient was 0.27 for unenhanced 3D echocardiography, 0.40 for unenhanced 2D echocardiography, 0.57 for CMR, and 0.51 for cineventriculography. The use of contrast increased IRA on RWMA to 0.42 for 3D echocardiography and to 0.56 for 2D echocardiography. Agreement with CMR on RWMA increased for 3D echocardiography when contrast enhancement was used (κ 0.40 vs 0.22 for unenhanced 3D echocardiography). Similarly, agreement of 2D echocardiography with CMR on RWMA increased with contrast enhancement (κ 0.50 vs 0.32). Accuracy to detect expert panel-defined RWMA was highest for CMR (84%) followed by 2D contrast echocardiography (78%) and 3D contrast echocardiography (76%). It was lesser for 2D and 3D unenhanced echocardiographies. In conclusion, analysis of RWMA is characterized by considerable interreader variability even using high-quality imaging techniques. IRA on RWMA is lower with 3D echocardiography compared with 2D echocardiography. IRA on RWMA and accuracy to detect panel-defined RWMA improve with contrast enhancement irrespective of the 2D or 3D echocardiography use.