Magnetic resonance imaging quantification of left ventricular mechanical dispersion and scar heterogeneity optimize risk stratification after myocardial infarction.

Abstract

Background: Left ventricular (LV) myocardial contraction patterns can be assessed using LV mechanical dispersion (LVMD), a parameter closely associated with electrical activation patterns. Despite its potential clinical significance, limited research has been conducted on LVMD following myocardial infarction (MI). This study aims to evaluate the predictive value of cardiac magnetic resonance (CMR)-derived LVMD for adverse clinical outcomes and to explore its correlation with myocardial scar heterogeneity.

Methods: We enrolled 181 post-MI patients (median age: 55.7 years; 76.8% male) who underwent CMR examinations. LVMD was calculated using the CMR-feature tracking (CMR-FT) technique, defined as the standard deviation (SD) of the time from the R-wave peak to the negative strain peak across 16 myocardial segments. Entropy was quantified using an algorithm implemented with a generic Python package. The primary composite endpoints included sudden cardiac death (SCD), sustained ventricular arrhythmias (VA), and new-onset heart failure (HF).

Results: Over a median follow-up of 31 months, LVMD and border zone (BZ) entropy demonstrated relatively high accuracy for predicting the primary composite endpoints, with area under the curve (AUC) values of 0.825 and 0.771, respectively. Patients with LVMD above the cut-off value (86.955 ms) were significantly more likely to experience the primary composite endpoints compared to those with lower LVMD values (p < 0.001). Multivariable analysis identified LVMD as an independent predictor of the primary composite endpoints after adjusting for entropy parameters, strain, and left ventricular ejection fraction (LVEF) (hazard ratio [HR]: 1.014; 95% confidence interval [CI]: 1.003-1.024; p = 0.010). A combined prediction model incorporating LVMD, BZ entropy, and LVEF achieved the highest predictive accuracy, with an AUC of 0.871 for the primary composite endpoints. Spearman rank correlation analysis revealed significant linear correlations between LVMD and entropy parameters (p < 0.001 for all).

Conclusions: Myocardial heterogeneity, as assessed by LVMD and BZ entropy, represents reliable and reproducible parameters reflecting cardiac remodeling following MI. LVMD has independent prognostic value, and the combination of LVMD and BZ entropy with the guideline-recommended LVEF as a unified model enhances the accuracy of forecasting the risk of primary combined endpoints in patients after MI.