Three-dimensional speckle-tracking echocardiography as a tool to predict subclinical left ventricular dysfunction before percutaneous closure of a patent ductus arteriosus.

Early detection of left ventricular dysfunction in patients diagnosed with hemodynamic significant PDA: application of Three-dimensional speckle-tracking echocardiography versus conventional echocardiography. Comparison of study parameters beteen PDA patients (n=13) and control patients with normal echocardiograms (n=27) Percutaneous closure of the ductus arteriosus: The procedure was performed under general anesthesia, systemic heparinization, anti-microbial prophylaxis, with at least an arterial approach (femoral artery) and often a venous approach when a Duct Occluder prosthesis was implanted. An angiography of the descending aorta allows accurate sizing of the PDA as well as a shape description helping device selection for closure. Pulmonary artery pressure was measured. Once the position was found adequate, repeated angiograms were performed, before and after the release of the device. Conventional echocardiographic parameters In PDA patients, primary echocardiogram was performed 24 hours before cardiac catheterization followed by a second ultrasound heart scan 24 hours after the procedure. In control patients, a complete echocardiogram was carried out at the Outpatient Cardiac Clinic including conventional methods, as well as 3D speckle tracking measurements during the course of one visit. A Philips ultrasound system (Philips IE33) with pS8, S5 or X5 probes was used according to age and weight. Echocardiographic measurements included: LVEDD (left ventricular end-diastolic diameter), LVESD (left ventricular end-systolic diameter), EF (ejection fraction), FS (fractional shortening), LVOT VTI (left ventricular outflow velocity time integral), E early diastolic transmitral flow velocity, A late diastolic transmitral flow velocity, Ea early diastolic mitral annular tissue velocity. Three-Dimensional speckle tracking echocardiography (3D strain): Three D video loop was performed in apical 4-chambers view with the acquisition of 4 consecutive cardiac cycles subsequently digitally stored for offline analysis. Then, the best acquisition was selected and post-treated offline, by a single sonographer. The frame rate was > 50Hz. The analysis was performed using a dedicated quantification software (4D LV-3.0 Analysis, TomTec Imaging Systems, Unterschleissheim, Germany). The measurements were made using the dataset with the best choice in terms of image quality. The 3D acquisition makes it possible to obtain 3 apical views: 4 cavities, 3 cavities and 2 cavities as well as a small axis orthogonal to the first 3 views. Adjustment of the aortic valve, mitral and apex position was made. Then, the surface of the 3D endocardium was automatically reconstructed and followed in a 3D space during the cardiac cycle. The endocardial surface was manually adjusted when needed until a better match was obtained with the actual position of the endocardium visualized in the different points of view. Subsequently, the software displays a 3D model of static left ventricle automatically calculated by the application. Finally, the left ventricle was automatically divided into sixteen 3-D segments using the standard segmentation of the American Society of Echocardiography. The curves and maps of the global (circumferential and longitudinal) strain and the analysis of all LV 3D strain segments (circumferential mean, radial and longitudinal strain) were calculated. The following global analysis parameters retrieved from 4D LV-analysis software were evaluated: - EDV (ml): end-diastolic volume, (maximum volume of the time curve minus overall volume) - ESV (ml): end-systolic volume (minimum volume of the time curve minus overall volume) - SV (ml): systolic flow (EDV minus ESV) - EF (%): ejection fraction ((SV / EDV) x 100) - GLS (%): global longitudinal deformation (global shortening of the longitudes of the surface model) - GCS (%): global circumferential deformation (global shortening of the latitudes of the surface model) - Longitudinal strand, GLS mean segmental (%): the longitudinal strain values of all segments are summed to calculate their average - Circumferential strain, GCS mean segmental (%): the circumferential deformation values of all segments are summed to calculate their mean - Radial strain, GRS average segmental (%): Radial deformation values of all segments are summed to calculate their average - Torsion: the LV twist defined as the angle of rotation of the apex with respect to the base Negative deformation values reflect a shortening, whereas uniquely positive values represent elongation or thickening.