LAUSR

Inherent to its geometry, echocardiographic imaging of the systemic right ventricle (sRV) may not always be feasible and/or reproducible in daily clinical practice. Here, we aim to validate the usefulness of all published echocardiographic measurements of systemic RV function in a longitudinal cohort by serial assessment of their correlations with cardiac magnetic resonance (CMR)-derived sRV ejection fraction (sRVEF). A single-center, retrospective cohort study was performed. Adult patients with a sRV who underwent both CMR and echocardiography, at two different points in time, were included. Offline analysis of echocardiographic images was blinded to offline CMR analysis and vice versa. Measurements were repeated by a second observer blinded to the results of the first observer. Correlations between echocardiographic and CMR measures were assessed with Pearson's correlation coefficient and inter-observer variability was quantified with intraclass-correlation coefficients. Fourteen patients were included, of which 4 with congenitally corrected transposition of the great arteries (TGA) and 10 patients with TGA late after atrial switch. Eight patients (57%) were female. There was a mean of 8 years between the first and second imaging assessment. The average CMR-derived sRVEF was 39±6% at t=1. Only global systemic RV function (r=−0.77, p=0.002 at t=1, r=−0.63, p=0.024 at t=2), fractional area change (FAC; r=0.79, p=0.001 at t=1, r=0.67 and p=0.018 at t=2) and global longitudinal strain (GLS; r=−0.73 and p=0.005 at t=1 and r=−0.70 and p=0.011 at t=2) correlated consistently with CMR-sRVEF at both time points. The intraclass-correlation coefficient of GLS was consistently more favorable than that of FAC. Out of the published echocardiographic parameters, GLS is the most robust measurement of sRV function in terms of feasibility, interobserver variability and correlation with CMR sRVEF. Figure 1. Echocardiographic assessment Type of funding source: None