Abstract
BACKGROUND: Transthoracic echocardiography (TTE) plays a fundamental role in the management of patients supported with extra-corporeal membrane oxygenation (ECMO). In light of fluctuating clinical states, serial monitoring of cardiac function is required. Formal quantification of ventricular parameters and myocardial mechanics offer benefit over qualitative assessment. The aim of this research was to compare unenhanced (UE) versus contrast-enhanced (CE) quantification of myocardial function and mechanics during ECMO in a validated ovine model. METHODS: Twenty-four sheep were commenced on peripheral veno-venous ECMO. Acute smoke-induced lung injury was induced in 21 sheep (3 controls). CE-TTE with Definity using Cadence Pulse Sequencing was performed. Two readers performed image analysis with TomTec Arena. End diastolic area (EDA, cm2), end systolic area (ESA, cm2), fractional area change (FAC, %), endocardial global circumferential strain (EGCS, %), myocardial global circumferential strain (MGCS, %), endocardial rotation (ER, degrees) and global radial strain (GRD, %) were evaluated for UE-TTE and CE-TTE. RESULTS: Full data sets are available in 22 sheep (92%). Mean CE EDA and ESA were significantly larger than in unenhanced images. Mean FAC was almost identical between the two techniques. There was no significant difference between UE and CE EGCS, MGCS and ER. There was significant difference in GRS between imaging techniques. Unenhanced inter-observer variability was from 0.48-0.70 but significantly improved to 0.71-0.89 for contrast imaging in all echocardiographic parameters. CONCLUSION: Semi-automated methods of myocardial function and mechanics using CE-TTE during ECMO was feasible and similar to UE-TTE for all parameters except ventricular areas and global radial strain. Addition of contrast significantly decreased inter-observer variability of all measurements.