Left ventricular mass and systolic function in children with chronic kidney disease-comparing echocardiography with cardiac magnetic resonance imaging.

BACKGROUND Increased left ventricular mass (LVM) is an important risk marker of uremic cardiovascular disease. Calculation of LVM by echocardiography (Echo) relies on geometric assumptions and in adults on hemodialysis overestimates LVM compared to cardiac magnetic resonance (CMR). We compare both techniques in children with chronic kidney disease (CKD). METHODS Concurrent Echo and CMR was performed in 25 children with CKD (14 after kidney transplantation) aged 8-17 years. RESULTS Compared to normal children, CMR-LVM was increased (standard deviation score (SDS) 0.39 ± 0.8 (p = 0.03)), stroke volume and cardiac output decreased (SDS -1.76 ± 1.1, p = 0.002 and -1.11 ± 2.0, p = 0.001). CMR-LVM index but not Echo-LVMI correlated to future glomerular filtration rate (GFR) decline (r = -0.52, p = 0.01). Mean Echo-LVM was higher than CMR-LVM (117 ± 40 vs. 89 ± 29 g, p < 0.0001), with wide limits of agreement (-6.2 to 62.8 g). The Echo-CMR LVM difference increased with higher Echo-LVMI (r = 0.77, p < 0.0001). Agreement of classifying left ventricular hypertrophy was poor with Cohen's kappa of 0.08. Mean Echo and CMR-ejection fraction differed by 1.42% with wide limits of agreement (-12.6 to 15.4%). CONCLUSIONS Echo overestimates LVM compared to CMR, especially at higher LVM. Despite this, CMR confirms increased LVM in children with CKD. Only CMR-LVMI but not Echo-LVMI correlated to future GFR decline