Keywords: left atrial function; strain; end-stage renal disease; hemodialysis
Introduction
The occurrence and development of end-stage renal disease (ESRD) is complex, which causes different degrees of damage in cardiac structure and function. Cardiovascular complication is by far the most important reason for death in patients with ESRD on hemodialysis (HD) therapy [1]. In the past, the focus of clinical research has been on the left ventricular morphology and function, but the change of left atrial (LA) function was often overlooked. Previous studies have shown that LA dilatation is closely related to the sustained volume and pressure overload, which can predict left ventricular hypertrophy, high filling pressure, and even mortality. [2] Thus, LA parameters have important clinical value in assessing the risk and prognosis of cardiovascular complications.
In clinical application, conventional echocardiography is the most common noninvasive method of evaluating cardiac structure and function in almost all patients with ESRD. However, it may not be sufficiently sensitive to detect early and subtle changes in ESRD with preserved ejection fraction, especially before serious cardiac involvement occurs. The EACVI NORRE study recently recommended the use of LA deformation imaging based on two-dimensional speckle tracking echocardiography (2D-STE)[3]. 2D-STE has been used in previous studies on LA function of this triphasic nature to document subclinical dysfunction.
Four-Dimensional Auto Left Atrial Quantification (LAQ) is a new method using three-dimensional(3D) data, which has a simple workflow and offers a new diagnosis method, evaluating the value of LA volume and emptying fraction aside from strain parameters analysis in the ESRD. LAQ can directly show LA endocardial contraction and provide sensitive and reproducible indices of myocardial dysfunction; thus, myocardial deformation was assessed directly and angle-independently [4].
The LA cavity volume is measured from a semi-automated segmentation algorithm during the cardiac cycle. The method calculates the deformation of the 3D model by solving a state estimation problem, which was constructed from an extended Kalman filter combined with LA geometry, a motion model, and edge detection algorithms. The strain is calculated based on the length change of different lines in each anatomical direction. The longitudinal strain is calculated by eight longitudinal lines from an automatically constructed triangular grid, which is connected to 2 opposite LA basal points (Figure 1). To calculate circumferential strain, 7 circumferential lines are equidistantly distributed between the LA base and apex (Figure 2).Then, strain is calculated for each frame time according to the following formula: s(t)=(L(t)-L(tref))/L(tref) X 100%, where L(t) is the line length at time t, and tref is set as the end diastolic time point of left ventricular. Global strain in each direction is obtained by calculating the average strain of the respective directional lines.