Problem of implicit assumptions about left ventricular geometry in the calculation of myocardial mass by echocardiography – a validation study with cardiac magnetic resonance

Abstract Background/Introduction Cardiac magnetic resonance imaging (CMR) is regarded as the reference method in assessing left ventricular (LV) myocardial mass. However, 2-dimensional echocardiography (Echo) is still used due to availability and practicability. Purpose We sought to assess measurements of LV myocardial mass from Echo exams and validate the results with CMR on a large cohort of patients. Methods We analyzed 357 patients who underwent clinically indicated CMR examinations and performed a standard Echo at the same day. In Echo, the commonly used Devereux formula (CUBE) was assessed for the calculation of LV mass: 0.8 × {1.04 × [([LV end-diastolic diameter + diastolic septum-thickness + posterior wall-thickness]3 − LV end-diastolic diameter3)]} + 0.6. Mass was indexed on body surface area (g/m2). Correlation (r), regression (R2) and Bland-Altman analysis were performed in order to assess bias between the two methods. Significance was defined as a 2-tailed P value <0.05. Results LV myocardial mass was measured in 345 patients with CMR and Echo. The median age was 61 years [19–79], 44% were female and CMR was performed due to coronary artery diseases (45%), suspected or florid myocarditis (31%) or further diagnosis of non-ischemic heart failure (24%). LV ejection fraction (EF) ranged between 13 and 71%. We found a significant relation of measurements for myocardial mass between Echo and CMR (R2=0.6, p<0.001) (Figure 1, left). However, there was a significant proportional bias (R2=0.47, p<0.001) depending in the amount of LV mass measured (Figure 1, right). As expected, the measurement bias correlated not only with the amount of mean wall-thickness (septum & posterior wall) from Echo, but also with LV end-diastolic diameter (r=−0.55 and −0.41, p<0.001 for all). Interestingly, the amount of LV EF correlated reversely with the proportional bias (r=0.34, p<0.001). Conclusion The results provided by our study, highlight the problem of implicit assumptions of LV geometry in the calculation of LV myocardial mass by Echo (CUBE) compared to CMR. Furthermore, the strongly biased differences between these two measurement methods are not negligible and warrant further investigation. Funding Acknowledgement Type of funding sources: None.