Predictors of preserved left ventricular systolic function after surgery for chronic organic mitral regurgitation: a prospective study.

BACKGROUND AND AIM OF THE STUDY: For patients with chronic severe mitral regurgitation (MR), one consideration for the timing of mitral valve surgery is the desire to have a preserved postoperative left ventricular ejection fraction (LVEF). It was hypothesized that the preoperative left ventricular (LV) volume and effective forward LVEF (defined as the ratio of forward aortic stroke volume and LV end-diastolic volume (LVEDV)) would accurately predict the intermediate (three-month) LVEF following surgery. It was also hypothesized that at years after successful surgical correction, there would be ongoing positive ventricular remodeling with improvements in systolic function. A quantitative volumetric cardiovascular magnetic resonance (CMR) system was used prospectively to study these relationships, and to define predictors of postoperative LV systolic function after surgery for chronic severe MR. METHODS: Twenty patients (eight females, 12 males; mean age 52 +/-12 years) with chronic severe MR, who had been referred for elective mitral valve repair or replacement with chordal preservation, were prospectively enrolled. All patients underwent quantitative CMR with assessment of LV volume, LVEF, and effective forward LVEF prior to, and at three months after, surgery. Fourteen of the patients returned at a median of 27 months after valve surgery for repeat CMR. RESULTS: The preoperative LVEF was normal in 90% of patients (0.62 +/- 0.09), but the effective forward LVEF was depressed in all subjects (0.36 +/- 0.09). Compared to preoperative values, at three months after surgery the LVEF had declined by 0.16 +/- 0.08 (p < 0.001), whereas the effective forward LVEF was increased by 0.09 +/- 0.10 (p < 0.001) and the LV end-diastolic volume index (LVEDVI) had declined by 32% (p < 0.001). Multivariate analysis demonstrated a significant correlation between postoperative LVEF and both preoperative LV end-systolic volume index (LVESVI) (p = 0.003) and the preoperative effective forward LVEF (p = 0.05). A depressed postoperative LVEF (< 0.50) during the early period was best predicted by a preoperative effective forward LVEF < 0.40 (sensitivity 100%, specificity 75%). At the 27-month follow up, 12 of the 14 patients (86%) had a normal LVEF, and there was a continuing normalization of other LV parameters, including the LVEDVI and LV mass index. CONCLUSION: Among patients with chronic severe MR, the preoperative LVESVI best correlated with the three-month postoperative LVEF. A preoperative effective forward LVEF < 0.40 accurately predicted a depressed postoperative LVEE. Over the succeeding two years, there will be ongoing positive LV remodeling. If confirmed in a larger series, the effective forward LVEF may serve as a useful clinical guide for the timing of corrective surgery in this population.