Myocardial scar location as detected by cardiac magnetic resonance is associated with the outcome in heart failure patients undergoing surgical ventricular reconstruction

OBJECTIVES: Post-infarction myocardial scar causes adverse left ventricular remodelling and negatively affects the prognosis. We sought to investigate whether scar extent and location obtained by cardiac magnetic resonance may affect the reverse remodelling and survival of heart failure patients undergoing surgical ventricular reconstruction. METHODS: From January 2011 to December 2015, 151 consecutive patients with previous myocardial infarction and left ventricular remodelling underwent surgical ventricular reconstruction at our Institution, of which 88 (58%) patients had a preoperative protocolstandardized late gadolinium enhancement (LGE)-cardiac magnetic resonance examination during the week before surgery. We excluded 40 patients with devices (26%), 15 patients with irregular heart rhythm (permanent atrial fibrillation, 10% not included in the device group) or mixed contraindications (severe claustrophobia or presence of material magnetic resonance not compatible). Among the 145 survivors, 11 patients received an implantable cardioverter defibrillator after surgery (mostly for persistent low ejection fraction) and were excluded as well, yielding a total of 59 patients (48 men, aged 65 +/- 9 years) who repeated a protocol-standardized LGE-cardiac magnetic resonance examination even 6 months postoperatively and therefore represent the study population. Patients were grouped according to the presence of LGE in the antero-basal left ventricular segments (Group A) or the absence of LGE in the same segments (Group B). The postoperative left ventricular end-systolic volume index was considered the primary end-point. RESULTS: After surgery, left ventricular end-systolic volume index and end-diastolic volume index significantly decreased (P < 0.001, for both), while diastolic sphericity index and ejection fraction significantly increased (P = 0.015 and P < 0.001, respectively). The presence of LGE in the antero-basal left ventricular segments (10 patients, Group A) was the only independent predictor of outcome (P = 0.02) at multivariate analysis, being the postoperative left ventricular end-systolic volume index significantly higher compared to that of patients of Group B (49 patients) (78 +/- 26 ml/m(2) vs 55 +/- 20 ml/m(2), P = 0.003). Furthermore, patients with a postoperative left ventricular end-systolic volume index >60 ml/m(2) showed a higher risk of cardiac events (hazard ratio = 3.67, P = 0.02). CONCLUSIONS: In patients undergoing surgical ventricular reconstruction, LGE scar location affects the left ventricular reverse remodelling, which in turn might limit the survival benefit.