Predictors of left ventricular reverse remodeling after percutaneous therapy for mitral regurgitation with the MitraClip system

Abstract

Percutaneous edge-to-edge repair with MitraClip (MC) system has revolutionized care of high-risk, otherwise severely undertreated patients with both degenerative and functional mitral regurgitation (MR). Left ventricular reverse remodeling (LVRR) has been consistently linked to survival benefit after various treatments including medical, device therapy or surgical valve repair. Transcatheter techniques have been shown to induce left ventricular reverse remodeling in patients with severe MR, yet specific data on predictors of LVRR and their prognostic impact after MitraClip therapy are limited. Since degenerative mitral regurgitation (DMR) and functional mitral regurgitation (FMR) constitute two different disease entities, the present study was specifically designed to explore differences between DMR and FMR regarding reverse remodeling. This retrospective study analyzed 164 patients successfully treated by MC implantation from January 2010 to December 2016 undergoing clinical and echocardiographic evaluation at baseline, 6 months and 12 months follow-up. LVRR was defined as decrease of ≥10% of the left ventricular end-diastolic diameter at 12 months follow-up. Patients were dichotomized according to the absence (no-LVRR group, n=83) and presence of LVRR (LVRR group, n=81) and were analyzed regarding their baseline characteristics as well as their clinical and echocardiographic outcome to identify predictors of LVRR and determine its impact on clinical outcome. Echocardiography revealed significant LVRR in 49% of the patients. Patients with FMR developed reverse ventricular remodeling to the same extent as patients with DMR (47.7% vs 52.8%). In the overall population MC implantation resulted in significant symptomatic improvement of New York Heart Association (NYHA) functional class and MR reduction (3.0±0.2 to 1.5±0.6, p<0.001). Despite clinical amelioration and significant reduction in the end-diastolic and end-systolic dimensions at follow-up, only a modest improvement in the global left ventricular ejection fraction (LVEF) was observed: 44.5±16.3% to 46.5±16.0%, p=0.14. Recurrent MR was evident in 14.0% of the total cohort 12 months after the MC procedure (20.5% in no-LVRR vs. 7.4% in LVRR group, p=0.01). Recurrent MR up to 12 months follow-up was more frequent in the DMR group compared to the FMR group (24.5% vs 9.0%, p=0.007), despite comparable postprocedural results in both etiologies (residual severe MR 4.5% for FMR vs. 5.7% in DMR, p=0.75). In the total cohort, multivariate logistic regression analysis identified severe recurrent MR after 12 months (p=0.01, odds ratio 0.26, confidence interval= 0.094 - 0.72), male gender (p=0.05, odds ratio 0.49, confidence interval= 0.24 - 1.0) and LVEF <20% (p=0.046, odds ratio 0.24, confidence interval= 0.061 - 0.97) as independent predictors of absence of LVRR. In the subgroup analysis according to the etiology, multivariate logistic regression analysis identified severe recurrent MR after 12 months (p=0.03, odds ratio 0.2, confidence interval= 0.047 - 0.867) as independent predictor for the non-occurrence of LVRR only in the DMR subgroup. In FMR, only residual severe tricuspid regurgitation (TR) inversely predicts LVRR (p=0.03, odds ratio 0.361, confidence interval= 0.142 - 0.916). Patients who experienced reverse ventricular remodeling showed lower rehospitalization rates in the first year and lower rehospitalization and mortality rates in the second year, leading to significantly reduced MACCE rates in the LVRR group compared to the no-LVRR group at 2 years follow-up (p=0.049). In conclusion, reverse ventricular remodeling occurred in nearly half of the patients treated by MitraClip and was associated with significant improvement in clinical outcome up to 2 years of follow-up. Recurrent severe MR after 12 months, poor baseline LVEF and male gender are independent predictors for absence of LVRR after MitraClip therapy in the total cohort. In patients with DMR, recurrent severe mitral regurgitation was the only independent predictor for the absence of LVRR, whereas in patients with FMR only severe residual TR after 12 months inversely predicts LVRR after MitraClip therapy. A better understanding of reverse remodeling predictors is essential for the process of risk stratification and decision making about potential need for care escalation in patients receiving percutaneous mitral valve repair. Poor ejection fraction has been proven to be an independent predictor for non-occurrence of LVRR, underlining the importance of prompt interventional treatment before myocardial dysfunction becomes irreversible. Nevertheless, the present study suggests that a more combative approach towards residual tricuspid regurgitation should be promoted in FMR patients in order to maintain the benefits achieved with isolated MitraClip implantation and to improve outcome in this high-risk population.