Abstract 16256: Cardiac Magnetic Resonance Imaging of Myocardial Scar and EntropyX of Cardiac Repolarization Independently Predict Appropriate Shock and Cardiac Death in Primary Prevention ICD Recipients
Introduction: Independent of left-ventricular ejection fraction (LVEF) and other conventional risk factors, LV scar characterization by cardiac magnetic resonance imaging (cMRI) predicts adverse events in primary prevention ICD recipients. Similarly, reduced repetition of fluctuation patterns underlying the variability of cardiac repolarization, quantified by the novel nonlinear parameter EntropyXQT, predicts adverse events. Because cMRI characteristics and EntropyXQT are distinct measures, we hypothesized they have independent and incremental prognostic value for potentially lethal ventricular tachyarrhythmias and cardiac mortality.
Methods: In 220 consecutive cardiomyopathy patients in normal sinus rhythm and receiving optimal medical therapy at time of ICD implantation (baseline), we quantified the amount of heterogeneous myocardial tissue (gray zone), dense core, and total scar on cMRI with late gadolinium enhancement, and EntropyXQT from 5-min ECGs. The primary endpoint was appropriate ICD shock or cardiac deaths not aborted by ICD.
Results: Over 4.4±2.8 years, 68 patients experienced appropriate ICD shocks (N=46) or died (N=22). In multivariate analysis, adjusted hazard ratios comparing the 3rd to the 1st tertile of EntropyXQT and cMRI characteristics were 2.93 (95% CI 1.51-5.70) and 1.96 (1.03-3.69), respectively, regardless of demographics, medical history, ischemic or non-ischemic cardiomyopathy, and absence or presence of other risk factors. Addition of EntropyXQT to a model comprised of LVEF and cMRI characteristics improved the ROC curve area from 60.6±4.5% to 68.3±4.2% (p=0.032) and continuous net reclassification by 49.8% (20.5-85.0). Similar improvements in prognostic discrimination were observed when EntropyXQT was added to conventional risk predictors.
Conclusions: EntropyXQT, a novel, inexpensive and easily measured parameter, expands on simpler concepts of repolarization variability and increases the prognostic value of cMRI, above and beyond other risk predictors. This robust and non-invasive personalized approach for risk stratification may be used to better identify those who would and would not benefit from ICD implantation.
Author Disclosures: D. DeMazumder: Research Grant; Significant; NIH NHLBI. B. O’Rourke: None. G.F. Tomaselli: None. S.R. Jones: None. K.C. Wu: None.
- © 2016 by American Heart Association, Inc.