Background: This study aimed to depict strain parameters derived from 2-dimensional (2D)- and 3-dimensional (3D) speckle tracking echocardiography and to explore which may best reflect myocardial fibrosis (MF) in dilated cardiomyopathy with advanced heart failure by comparing with histologic fibrosis.
Methods and results: We analyzed 75 patients with dilated cardiomyopathy with advanced heart failure who underwent echocardiographic examination before heart transplantation. Strain parameters derived from 2D- and 3D speckle tracking echocardiography were as follows: left ventricular global longitudinal strain (GLS), global circumferential strain (GCS), global radial strain (GRS) and tangential strain (TS). The degree of MF was quantified using Masson's staining in left ventricular myocardial samples obtained from all patients. Seventy-five patients were divided into 3 groups according to the tertiles of histologic MF (mild, moderate, and severe MF groups). Patients with severe MF had lower 3DGLS, 3DGRS, 3DTS, and 2DGLS than those with mild and moderate MF. MF strongly correlated with 3DGLS (r = 0.72, P < .001), weakly with 3DGRS (r = -0.39, P = .001), 3DGCS (r = 0.30, P = .009), 3DTS (r = 0.47, P < .001), and 2DGLS (r = 0.44, P < .001), but did not correlate with 2DGCS and 2DGRS. Receiver operating characteristic analysis revealed that the area under the curve of 3DGLS for detecting severe MF was significantly larger than that of other strain parameters (0.86 vs 0.59-0.70, P < .05 for all). The multivariate linear regression models using 3DGLS (R2 = 0.76; Akaike information criterion = 331) was found to be a more accurate indicator to predict MF than that with 3DTS (R2 = 0.65, Akaike information criterion = 354) and 2DGLS (R2 = 0.66, Akaike information criterion = 352).
Conclusions: Three-dimensional GLS may be an optimal surrogate marker for reflecting MF in patients with dilated cardiomyopathy with advanced heart failure.
Keywords: Dilated cardiomyopathy; Myocardial fibrosis; Speckle tracking echocardiography; three-dimensional.
Copyright © 2021 Elsevier Inc. All rights reserved.