Assessment of Left Ventricular Reverse Remodeling by Cardiac MRI in Patients Undergoing Repair Surgery for Severe Aortic or Mitral Regurgitation.

OBJECTIVE To evaluate left ventricular (LV) reverse remodeling after repair surgery for mitral regurgitation (MR) or aortic regurgitation (AR), aiming at determining optimal preoperative thresholds for normalization of LV volumes and function after surgery. DESIGN Observational prospective cohort study. SETTING Single-center, academic, tertiary care cardiovascular center. PARTICIPANTS Patients and volunteers. INTERVENTIONS Cardiac magnetic resonance with measurement of indexed LV end-diastolic volume (LVEDVi) and end-systolic volume (LVESVi), mass (LVmassi), and ejection fraction (LVEF) was performed preoperatively and postoperatively. MEASUREMENTS AND MAIN RESULTS The authors included 29 patients with AR and 59 patients with MR (46 ± 12 and 56 ± 12 years, follow-up 222 ± 57 days). Both AR and MR repair resulted in a significant reduction of LV volumes and mass (respectively, delta change in LVEDVi -55 mL/m² and -43 mL/m²; in LVESVi -26 mL/m² and -10 mL/m²; and in LVmassi -24 g/m² and -12 g/m²; p < 0.001 for all). Yet despite the absence of perioperative necrosis, 7 (24%) patients with AR had persistent LV dilatation (LVEDVi >106 mL/m²) relative to controls and 16 (27%) patients with MR developed systolic LV dysfunction (LVEF <50%) postoperatively. Binary logistic regression analysis indicated preoperative LV volumes as the most accurate parameter for predicting both incomplete LV reverse remodeling in AR and LV dysfunction in MR. Receiver operating characteristic-determined thresholds were LVEDVi >155 mL/m² for AR and >129 mL/m² for MR. CONCLUSION Although both AR and MR repair allow significant reverse postoperative LV remodeling, persistent LV dilatation after AR correction and systolic LV dysfunction after MR repair are common and best predicted by increased preoperative LV volumes. This highlights the importance of considering LV volumes in the decision-making process.