Preoperative left ventricular dimensions predict reverse remodeling following restrictive mitral annuloplasty in ischemic mitral regurgitation

Objective: Ischemic mitral regurgitation can be treated with a restrictive mitral annuloplasty, with or without coronary revascularization. In this study, the extent of reverse remodeling of the left ventricle following this strategy is assessed, as well as the factors that influence it. Methods: Eighty-seven consecutive patients with ischemic mitral regurgitation and a mean ejection fraction of 32±10% underwent restrictive mitral annuloplasty (downsizing by two ring sizes, median ring size 26), with additional coronary revascularization in 75 patients. All underwent transthoracic echocardiography 18 months after surgery to assess residual mitral regurgitation, mitral valve gradient and left ventricular end-systolic and end-diastolic dimensions. Univariate and multivariate analysis was performed to identify predictors for reverse remodeling, defined as a 10% reduction in left ventricular dimension. Receiver-operating characteristic analysis was used to identify cut-off values for preoperative left ventricular dimensions in predicting reverse remodeling. Results: Early mortality was 8.0% (seven patients, three non-cardiac), late mortality was 7.5% (six patients, four non-cardiac). There were two reoperations (redo annuloplasty), and four readmissions for heart failure. At 29 months follow-up, NYHA class improved from 3.0±0.9 to 1.3±0.5 (P≪0.01). Mitral regurgitation grade decreased from 3.1±0.5 to 0.6±0.6 at 18 months, left ventricular end-systolic dimension decreased from 52±8 to 44±11 mm (P≪0.01), and end-diastolic dimension from 64±8 to 58±10mm (P≪0.01). Multivariate analysis identified preoperative left ventricular end-diastolic dimension as the single best factor in predicting occurrence of reverse remodeling. For end-systolic dimension, 51mm was the optimal cut-off value to predict reverse remodeling (specificity and sensitivity 81%, area under curve 0.85); for end-diastolic dimension, the cut-off value was 65mm (specificity and sensitivity 89%, area under curve 0.92). Conclusions: Stringent restrictive mitral annuloplasty with or without revascularization provides excellent clinical results with acceptable mortality. At 18 months follow-up, there is no significant residual mitral regurgitation. Reverse remodeling occurs in the majority of patients, but is limited by preoperative left ventricular dimensions. In patients with a left ventricular end-diastolic dimension exceeding 65mm, additional surgical procedures are necessary to try and obtain revers