Ventricular volume and myocardial viability, evaluated using cardiac magnetic resonance imaging, affect long-term results after surgical ventricular reconstruction

Abstract

OBJECTIVES: We aimed to investigate the outcomes of surgical ventricular reconstruction (SVR) for ischaemic cardiomyopathy and to identify ideal candidates for this procedure using cardiac magnetic resonance imaging (MRI) studies. METHODS: We retrospectively examined 50 patients who underwent SVR and were preoperatively evaluated using cardiac MRI from 2004 to 2014. Cine MRI and gadolinium-enhanced MRI were performed to evaluate left ventricular (LV) volume and scarring. The midterm (median, 2.5 years) risk factors of cardiac death or major adverse cardiac events (MACEs) were analysed. Patients were divided into three groups-those with preoperative indexed LV end-systolic volume (LVESVI) of ≤ 100 ml/m2 (Group 1), those with LVESVI of >100 and ≤ 130 ml/m2 (Group 2) and those with LVESVI of >130 ml/m2 (Group 3)-and examined. RESULTS: In total, 17 patients exhibited MACEs at follow-up. Kaplan-Meier analysis showed that the 5-year rate of freedom from MACEs was 66%. The Cox hazard model indicated that preoperative LVESVI was the only significant predictor for MACEs (P = 0.006; hazard ratio, 1.02; 95% confidence interval, 1.01-1.04). Moreover, a significant reduction in LVESVI and an increase in LV ejection fraction (LVEF) were observed early after SVR (preoperative versus postoperative: LVESVI, 110 ± 44 vs 68 ± 28 ml/m2, P < 0.001; LVEF, 24.3 ± 10.3 vs 32.0 ± 10.4%, P < 0.001) and at follow-up (LVESVI, 61 ± 28 ml/m2, P < 0.001; LVEF, 36.5 ± 11.5%, P < 0.001). Although no difference was observed between the groups at follow-up, LVEF and LVESVI primarily improved in Group 2. LVEF improvement was significantly greater in Group 2 than in Group 1 (%increase in LVEF: Group 1, 6%; Group 2, 18%; P = 0.008). In patients with a preoperative LVESVI of >130 ml/m2, the number of non-viable segments was a significant risk factor for MACEs. CONCLUSIONS: Patients with preoperative LVESVI ranging from 100 to 130 ml/m2 had fairly better outcomes, and the percentage improvement in LVEF and the percentage reduction in LVESVI were more pronounced in these patients. Hence, accurate preoperative assessments of LV volume and viability testing using cardiac MRI studies are essential for better stratification of the SVR procedure.