Discrepancies in Left Ventricular Mass Calculation Based on Echocardiography and Cardiovascular Magnetic Resonance Measurements in Patients with Left Ventricular Hypertrophy.

Abstract

Increased left ventricular (LV) mass is associated with adverse cardiovascular outcomes, and its accurate assessment is important. The aim of this study was to analyze the degree of difference among various methods of LV mass calculation based on transthoracic echocardiographic (TTE) measurements and cardiovascular magnetic resonance (CMR) measurements, especially in patients with aortic stenosis with varying degrees of LV hypertrophy (LVH). The mechanism underlying this disagreement was also investigated. Ninety-nine patients with moderate to severe aortic stenosis and 33 control subjects matched for age, sex, body weight, and height were enrolled in this prospective observational cohort study. All patients underwent TTE and CMR imaging. LV mass index (LVMI) was calculated using three formulas on the basis of TTE measurements (the Penn-cube, American Society of Echocardiography [ASE], and Teichholz methods) and compared with measurements obtained using CMR, the reference method. Although all methods calculated using TTE measurements showed good correlations with CMR measurements, LVMI measured using the Penn-cube and ASE methods tended to be larger than LVMI measured using CMR (difference in LVMI by the Penn-cube and ASE methods, 59.3 ± 29.7 and 30.6 ± 22.3 g/m², respectively). This tendency decreased with the Teichholz method (difference in LVMI by the Teichholz method, 22.9 ± 19.1 g/m²). The degree of LVMI overestimation was significantly different among the three methods (P < .001 by one-way analysis of variance), which was more significant in patients with LVH, especially with the Penn-cube method (differences between CMR and TTE measurements in patients with aortic stenosis and LVH, 66.3 ± 34.8 vs 31.2 ± 26.6 vs 15.5 ± 20.9 g/m² for the Penn-cube, ASE, and Teichholz methods, respectively; P < .001 with post hoc Tukey analysis). There was a good correlation between LVMI and LV diameter-to-length ratio (r = 0.468, P < .001), which suggested that the left ventricle takes on a more globular shape with the increase of LVMI, resulting in a significant deviation from the basic assumptions on which the Penn-cube and ASE methods were built. Current methods of calculating LVMI from echocardiographic measurements carry a tendency to measure LVMI larger than methods based on CMR measurements, which was more significant in patients with LVH. The change of the left ventricle's shape with LVH may be a plausible explanation for this, and a correction method may be needed when calculating LVMI from echocardiographic measurements, especially in patients with LVH and smaller body size.