4175Five-year outcome after successful transcatheter mitral valve repair in surgical high-risk patients

Abstract

Introduction: MitraClip (MC) has become the most widely used system for transcatheter mitral valve repair and has proven to be successful in mid-term outcomes. Yet, long-term data are scarce to date. Purpose: We intended to evaluate outcome measures at 5-year (yr) follow-up (FU). Thus, we divided our real-life cohort according to the etiology of mitral regurgitation (MR) leading to MC therapy. Methods: At our center, 693 patients (pts; 75.3±8.8 yrs; 60.8% men; 67.1% functional MR [FMR]) were treated with MC from September 2008 to May 2016 (success rate 90.6% [MR≤2+ at discharge]). Follow-up consisted of site visits including echocardiography following standard operating procedures at 6 months and annually for up to 5 years. Results: Patients with degenerative MR (DMR) were older and suffered of higher MR grading (p=0.02). In contrast, pts with FMR presented with a greater comorbidities. Furthermore, adverse echocardiographic findings were observed in FMR pts, expressed most noticeably in wider left ventricular (LV) diameters and lower LV ejection fraction (FMR: 36±14% vs DMR: 55±12%, p<0.0001). Consequently, the preprocedural risk, as determined by logistic EuroScore was deemed higher in these pts (DMR: 18.5% [10.8-28.31] vs FMR: 22.0% [12.7-35.8], p=0.007). Yet, no difference in procedural success could be determined. Grading of MR≤2+ at 2-yr FU was 89.3% (FMR) and 84.1% (DMR, p=0.19). At 5-yr FU MR≤2+ was observed in 80.0% (FMR) and 68.8% (DMR, p=0.47). While no improvement in LV ejection fraction over time could be shown (DMR: 51.5±10.9% vs FMR: 32.2±17.8%, p=0.019 at 5-yr FU), LV end diastolic diameters decreased in both groups (mean -10.0% [FMR] vs -6.95% [DMR] at 5-yr FU). This was reflected by a great reduction of NT-proBNP levels with an equalization of levels over time (DMR vs FMR: 2867ng/l [1309-6027] vs 4417ng/l [2161-8569], p=0.003 at baseline [BL] and 1429ng/l [524-4645] vs 3896ng/l [1226-9729], p=NS at 5-yr FU). A remarkable reduction in NYHA classification was achieved displaying no inter-group differences during FU time ([FMR] vs [DMR]; NYHA class ≤2. BL: 4.7% vs 4.4%, p=0.32; 2-yr FU: 62.0% vs 53.3%, p=0.23; 5-yr FU: 52.3% vs 58.6%, p=0.49). The Minnesota Living with Heart Failure Questionnaire demonstrated comparable results (40.8±18.4 [FMR] vs 37.6±15.6 [DMR], p=NS at BL; 31.1±21.0 vs 32.0±21.5, p=NS at 2-yr FU; 37.7±20.7 vs 29.2±20.7, p=NS at 5yr-FU). Kaplan- Meier curves revealed no differences for the endpoints mortality (p=0.32) or rehospitalization for heart failure (p=0.39). Conclusions: Despite adverse baseline findings in FMR pts, no differences in success or hard endpoints were observed. A reduction of LV end diastolic diameters and NT-proBNP values with stable LV ejection fraction parameters might be due to an improvement of LV geometry in all pts following successful MC therapy. Our data suggest beneficial long-term results of MC therapy in a real-life cohort, disregarding the underlying MR etiology.