Survival in Patients With Severe Ischemic Cardiomyopathy Undergoing Revascularization Versus Medical Therapy

Abstract

The value of assessment of viability as a predictor of surgical revascularization benefit in ischemic cardiomyopathy has recently been questioned in a large trial. We sought to determine whether the contribution of viability as myocardial scar burden (SB) to predict revascularization outcomes could be modulated by end-systolic volume index (ESVi). Delayed hyperenhancement-MRI was obtained in 450 patients with ≥70% stenosis in ≥1 epicardial coronary artery (75% men; median age, 62.8 ± 10.7 years; mean left ventricular ejection fraction, 23 ± 9%; mean ESVi, 115 ± 50 mL) from 2002 to 2006. SB was quantified as scar percentage (infarcted mass/total left ventricular mass). Subsequent surgical revascularization was performed in 245 (54%) patients and subsequent percutaneous coronary interventions were performed in 28 (6%) patients. A propensity score was developed for revascularization. Cox proportional hazards models of all-cause mortality were used for risk adjustment. Over a mean follow-up of 5.8 ± 2.7 years, 186 (41%) deaths occurred. After adjusting for prior revascularization, sex, diabetes, age, use of cardiac resynchronization therapy, implantable cardioverter defibrillator, mitral regurgitation, and mitral valve procedures; an interaction between scar percentage and ESVi (P=0.016) and an interaction between post-MRI revascularization and ESVi (P=0.0017) were independently associated with mortality. ESVi demonstrated a significant interaction with revascularization and female sex, such that enhanced survival was associated with ESVi. ESVi also showed an interaction with SB; better survival was associated with lower volumes and less scar. ESVi and SB provide independent, incremental prognostic value in patients with severe ischemic cardiomyopathy. The risk associated with SB should not be assessed in isolation.