Is cardiac magnetic resonance necessary for prediction of left ventricular remodeling in patients with reperfused ST-segment elevation myocardial infarction?

Abstract

As cardiac magnetic resonance imaging (CMR) has become widely used for evaluation of myocardial viability after acute myocardial infarction, the additional value of CMR parameters for prediction of left ventricle (LV) remodeling has been receiving interest. The aim of the study was to investigate the additional predictive value of CMR parameters for LV remodeling after successful reperfusion of ST-segment elevation myocardial infarction (STEMI) using multiple predictive models. LV remodeling was defined as ≥20% increase in end-diastolic volume at 6 month follow-up echocardiography. Using multiple stepwise regression analysis, conventional risk model was classified as following; model 1 (clinical factors), model 2 (model 1 + angiographic factors), model 3 (model 2 + echocardiographic factors) and CMR-added model; model 4 (model 3 + CMR factors). Among 262 enrolled patients, 25.1% showed LV remodeling. There were significant increments of c-statistics from the predictive model 1 to model 3 (AUC; 0.675 [0.60–0.75], 0.708 [0.64–0.78], 0.756 [0.69–0.82], respectively. all p < 0.05). However, model 4, which added the CMR variables, did not show any increase in predictive value compared with model 3 (AUC; 0.763 [0.70–0.83] versus 0.756 [0.69–0.82], p = 0.11). During the 28.2 months of median follow up, the incidence of hospitalization for heart failure was significantly higher in the patients with LV remodeling (6.1% vs. 0.5%, p = 0.02). CMR parameters did not provide incremental predictive value above the assessment by conventional echocardiography-based risk model in patients with STEMI.