Transcatheter Edge-to-Edge Mitral Valve Repair: What is the Measure of Success?

This article refers to ‘Left ventricular reverse remodelling predicts long-term outcomes in patients with functional mitral regurgitation undergoing MitraClip therapy: results from a multicentre registry’ by M. Adamo et al., published in this issue on pages xxx. Although a decade has passed since the CE Mark approval of the MitraClip percutaneous edge-to-edge mitral valve repair system (Abbott, Santa Clara, CA, USA), its impact on the survival of patients with heart failure (HF) with reduced ejection fraction and associated functional mitral regurgitation (FMR) is still a matter of debate, with both unresolved procedural and imaging issues. The accuracy of the conventional echocardiographic methods, as well as the cut-off value to define FMR severity, are not convincing.1 Besides, grading of post-MitraClip residual regurgitation emerged as one of the most challenging and unsettled topics in contemporary echocardiography.2 As a consequence of these uncertainties, the cardiologist needs to carefully evaluate surrogate indexes, including structural (i.e. chamber dimensions), haemodynamic (i.e. cardiac output, pulmonary pressure, filling pattern), and biochemical (i.e. B-type natriuretic peptide, renal function indexes) markers of effective relief from left ventricular (LV) volume overload to confirm procedural success over time. The adoption of an objective measure of post-procedural LV reverse remodelling (LVRR) (i.e. change in LV volumes) might be particularly helpful in this context. Left ventricular reverse remodelling has already been considered as a marker of important prognostic significance for several years. For example, after both pharmacological3 and cardiac resynchronization4 treatment for HF, surgical studies have shown that effective mitral valve repair eliminates LV volume overload, leading to LVRR and improved clinical outcomes over time and, furthermore, these pathophysiological notions were already well established at the moment of MitraClip introduction.5 However, LVRR clinical usefulness in patients undergoing percutaneous edge-to-edge repair took a few years to gain relevance6 with progressive accumulation of data, albeit conflicting, in patients presenting FMR associated with advanced HF.7, 8 In this issue of the Journal, Adamo et al.9 report on a multicentre retrospective registry of 184 consecutive patients with severe FMR who underwent MitraClip procedure. The demonstration of a significant survival benefit related to LVRR, defined as a decrease of ≥ 10% in LV end-systolic volume at 6 months after the index procedure, is the most relevant finding of the study. LVRR occurred in 43% of cases and was significantly associated with a decrease in mortality, HF hospitalizations and their composite at 2-year follow-up. Importantly, this difference in prognosis could not be explained by the degree of residual mitral regurgitation (MR) (comparable between LVRR and non-LVRR patients, P = 0.093). Of note, the study reported a higher, albeit non-significant, improvement in MR severity immediately post-procedure in the non-LVRR group. The authors justified this apparent inconsistency by pointing out the higher rate of pharmacological and mechanical support required in these patients during the procedure and the consequent possible underestimation of MR severity post-MitraClip. However, the difference in MR severity disappeared at 30-day follow-up, and the authors suggested that LVVR occurred independently of MR degree and that even a minimal post-procedural reduction (one degree) in MR severity could be associated with reverse remodelling. Of note, any semiquantitative or non-homogeneous evaluation of residual MR (often absent/mild and sometimes moderate) might not be able to identify the burden of haemodynamic LV overload.10 Thus, as supported by the results of this paper, a comprehensive echocardiographic evaluation of the procedural results, not limited to the identification of residual MR, plays a crucial role during post-procedural follow-up. Adamo et al.9 also showed that female gender, absence of diabetes, freedom from HF hospitalization within 6 months before the procedure, non-ischaemic aetiology of MR and a LV end-diastolic diameter < 75 mm were determinants of LVRR. As pointed out, female gender remains the less explained predictor of LVRR. A recent small report11 on reverse remodelling post-MitraClip, although not focusing exclusively on FMR, did not find female gender to be a predictor of outcome. A recent large series showed the effect of sex on reverse remodelling in patients with chronic HF12; thus larger studies are warranted to investigate this finding in the context of percutaneous edge-to-edge mitral valve repair. The authors suggested that a key factor leading to a favourable long-term prognosis after MitraClip could be the development of LVRR. In a recent analysis performed by Brouwer et al.11 on 79 patients undergoing MitraClip procedure for severe MR (FMR 81%), the authors found a similar proportion of patients (40/79, 51%) showing LVRR [defined as a reduction of > 15% in LV end-diastolic volume (LVEDV) index] at 6 months. Interestingly, among those patients who did not show LVRR, a small group presented with sustained adverse remodelling (defined as an increase in LVEDV index > 10%) with subsequent higher mortality. These data, together with the findings of Adamo et al.9 that a LV end-diastolic diameter < 75 mm is related to a 2.2-fold increase in the probability of developing LVRR, seem to support the authors when speculating that early intervention on FMR might be associated with favourable LV remodelling. Most likely, in the near future, some answers to these intriguing questions will be provided by the echocardiographic sub-analyses of two recently published randomized trials comparing guideline-directed medical therapy alone vs. transcatheter edge-to-edge mitral valve repair for moderate/severe FMR, and leading to apparently conflicting results. The MITRA-FR study,13 designed as a pragmatic trial, showed no significant differences in both mortality and unplanned hospitalization rates between the two treatment strategies. Conversely, the COAPT trial,14 which focused on a highly selected population, showed a significant reduction in mortality and hospitalization rates at 2-year follow-up. Interestingly, the results of the present study resemble, in part, both COAPT and MITRA-FR device groups in terms of LV size at the time of enrolment as well as in volumetric remodelling over time. Indeed, in the current study, patients in the LVRR group presented baseline LV volumes that seem comparable to the COAPT study (average LVEDV 205 mL vs. 196 mL). Of note, in the COAPT trial, patients with LV end-systolic diameter > 70 mm were excluded per protocol. Both the LVRR group in the Adamo et al. study and the COAPT device arm showed a significant outcome improvement after the procedure. On the other hand, the non-LVRR group presented relatively larger baseline LV volumes, closer to MITRA-FR patients (average LVEDV 223 mL vs. > 240 mL, assuming a mean body surface area of 1.8 m2 in MITRA-FR). In both of these larger LV groups, a less favourable survival was observed after the MitraClip procedure. As far as LVRR over time is concerned, COAPT and MITRA-FR trials did not specifically report on this aspect. However, the COAPT patients allocated in the device arm displayed a significant decrease in LVEDV compared to the optimal medical therapy arm (average −3.7 mL vs. +17 mL, P = 0.003), whereas in the MITRA-FR there were no clear trends towards LV volume reduction over time in either the device or control group. This heterogeneity in the remodelling response observed in these trials might be comparable to the one reported in the subgroups from the real-world cohort presented by Adamo et al., where patients potentially represented in both trials were included. In summary, the study by Adamo et al.9 confirms and supports, in a larger population, the beneficial impact of LVRR on clinical outcomes following MitraClip in patients with FMR. The retrospective non-randomized nature of the study, the small sample size of the analysed population and the relatively short follow-up represent limiting factors to the generalizability of the results. Nevertheless, this report stresses two unmet needs and, therefore, intriguing dilemmas for the future. First, a growing body of knowledge seems to suggest that LVRR is a reliable marker of therapeutic success and prognostic benefit across different treatment modalities in patients with HF, including MitraClip. Still, there is an imperative need for a uniform and validated method to assess residual MR after transcatheter edge-to-edge repair, as well as for the standardization of LVRR echocardiographic identification. Second, the larger LV size at baseline and the absence of LVRR during follow-up seem to identify non-responders and, therefore, efforts are needed to investigate whether these patients may benefit from earlier referral, additional/complementary procedures, or they do not benefit at all from FMR correction. In an era of rapid technical and technological progress, the cardiologist is not only challenged to identify patients who may benefit the most from each procedure, but also to constantly refine the indication and deeply understand the interactions among different therapeutic approaches. The final aim of this constant tailoring each indication and prognostic information is to deliver the most beneficial treatment combination to patients, and each one at the most appropriate time. Conflict of interest: none declared.