Abstract 17417: Complexity of Scar Assessed with Late Gadolinium Enhancement Cardiovascular Magnetic Resonance as an Independent Specific Predictor of Appropriate Implantable Cardioverter Defibrillator Therapy Delivery in Dilated Cardiomyopathy
Purpose: to assess late gadolinium enhancement (LGE) cardiovascular magnetic resonance (CMR) imaging characteristics yelding a high risk of ventricular arrhythmias and appropriate implantable cardioverter defibrillator (ICD) therapy delivery in patients implanted for sudden cardiac death (SCD) primary or secondary prevention.
Methods: from Jan. 1st, 2006 to Dec. 31st, 2010, 96 consecutive patients receiving an ICD (84 for primary prevention and 12 for secondary prevention) and previously evaluated with CMR were enrolled. LV and RV volumes and ejection fraction (EF) were evaluated, as well as late gadolinium enhancement imaging. LGE was defined “complex” (Cx-LGE) in presence of one of the following: 1) ischemic pattern, involving ≥ 2 different coronary territories; 2) epicardial distribution; 3) “diffuse” subendocardial; 4) presence of ≥ 2 different patterns. Patients were regularly followed and appropriate ICD interventions were detected. The primary end-point was the occurrence of a sustained ventricular arrhythmia requiring an ICD therapy. A composite secondary end-point of cardiovascular death, cardiac transplantation or ventricular assist device implantation was also considered.
Results: 47 and 49 patients affected respectively by ischemic and non ischemic cardiomyopathy were enrolled. During a median follow-up of 15,9 months, 10 and 13 patients reached the primary and secondary end-point, respectively. Only one patient experienced an ICD shock before heart transplantation. At univariate analysis, Cx-LGE was highly predictive of primary end-point occurrence (HR=6.96 [C.I. 95%:1.47-33.00], p=0.023), as well as implant indication in primary or secondary prevention (HR=4.06 [C.I.95%:1.13-14.63], p=0.032) and NYHA class (HR=0.19 [C.I.95%:0.08-0.42], p<0.001); at a multivariate analysis, only Cx-LGE was confirmed as an independent predictor of ICD interventions (HR=6.59 [C.I. 95%:1.15-37.70], p=0.034). Cx-LGE complexity was not associated with secondary end-point (p=ns for univariate and multivariate analysis).
Conclusions: Cx-LGE is a new, powerful, independent, specific, easily determined risk factor for ventricular arrhythmias and SCD in dilated cardiomyopathy of any etiology.
- © 2012 by American Heart Association, Inc.