Ethan J. Halpern MD, Presenter: Research grant, Koninklijke Philips Electronics NV Research grant, Lantheus Medical Imaging, Inc Equipment support, Toshiba Corporation

Left ventricular (LV) mass is an independent predictor of clinical outcome for cardiovascular disease, but its utility is limited by the variability of echocardiographic LV mass calculation. Quantification of LV mass by transthoracic echocardiography (TTE) is most frequently accomplished with a cubed formula that incorporates only short axis measurements of the LV cavity and assumes an ellipsoid geometry. We compared LV mass based upon short axis TTE measurements with volumetric measurements obtained by coronary CT angiography (CCTA) during systole and diastole. Additionally, we evaluated whether short axis TTE quantification could be improved by introducing a correction factor for ventricular length as measured during CCTA.

TTE and ECG-gated CCTA were performed within 30 days on each of 41 patients. LV mass was computed from short axis TTE images based upon the cubed formula advocated in the standards of the American Society of Echocardiography. CCTA was performed with a Brilliance CT scanner (Philips Medical Systems). LV mass on CCTA was computed at both end-diastole and end-systole as the difference in calculated volumes within the epicardial and endocardial outlines of the left ventricle. Pearson correlation was computed between TTE and CCTA calculations of LV mass. In an attempt to improve this correlation we introduced a correction factor into the echocardiographic calculation based upon the LV long axis dimension of each patient as measured on CCTA, and calculated a “corrected” LV mass.

Excellent correlation was obtained between systolic and diastolic LV mass computed from CCTA (correlation coefficient = 0.87), suggesting that CCTA provides a reproducible measure of LV mass. Correlation between LV mass obtained by TTE and CCTA was fair (correlation coefficients: 0.56-0.64). Addition of LV long axis dimension to compute a corrected LV mass by echocardiography produced only minimal improvement in agreement between TTE and CCTA (correlation coefficients: 0.61-0.69; p > 0.3).

Gated CCTA provides reproducible estimates of LV mass during systole and diastole. Estimates of LV mass by TTE differ from LV mass estimates by CCTA because of complex geometric factors that are not corrected by measurement of LV length.

Estimates of LV mass by CCTA provide a reliable measurement that is independent of the phase of the cardiac cycle.

Halpern, E, Estimation of Left Ventricular Mass during Coronary CTA as a Predictor of Clinical Outcome: Comparison with Conventional Echocardiography.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11034204.html