Observed Versus Predicted Mortality After Mitraclip Treatment in Patients with Symptomatic Heart Failure and Significant Functional Mitral Regurgitation

Patients with chronic heart failure (HF) commonly develop functional (or secondary) mitral regurgitation (FMR) that takes place as a result of a ‘ventricular secondary’ mechanism, due to dilatation and spherical remodelling of the left ventricle.1 In HF patients, the presence of severe FMR is associated with HF symptoms, increased hospitalization rates and worse long-term prognosis despite optimal medical therapy (OMT).2 Transcatheter edge-to-edge mitral valve repair using the MitraClip device (Abbott Vascular, Santa Clara, CA, USA) has evolved as a new tool for treatment of high-risk or inoperable patients with FMR. Several studies based on real-world registries (ACCESS-EU, TRAMI, EVEREST II HRR and REALISM HR) showed how MitraClip is a safe treatment that could induce improvement in symptoms, favourable left ventricular remodelling, and reduced HF hospitalizations in high-risk patients with severe mitral regurgitation (MR). Despite these evidences, it is still unclear how surgical or percutaneous interventions on FMR may change long-term survival of HF patients. Therefore, we retrospectively evaluated a consecutive series of HF patients (n = 238) with significant (3+ moderate and 4+ severe) ischaemic (n = 167) and non-ischaemic (n = 71) FMR treated with MitraClip, between October 2008 and April 2017, at San Raffaele Hospital of Milan (n = 133), at Spedali Civili of Brescia (n = 59), and at University Heart Center of Zurich (n = 46), to compare the observed all-cause mortality with that predicted by the 3C-HF, MAGGIC and SHFM scores, three well-validated scoring systems used to predict the probability of survival in patients with HF.3-5 All included patients had symptomatic systolic chronic HF [New York Heart Association (NYHA) class III–IV] despite OMT (including cardiac resynchronization therapy). The 3C-HF, MAGGIC and SHFM scores were calculated for each patient on the date of the MitraClip implantation. The observed overall survival was obtained using Kaplan–Meier analysis. Observed and predicted survivals were compared at the 1-, 2-, 3- and 5-year time-point after MitraClip implantation using 2x2 matrix and χ2 test. Procedural success was observed in 89.5% of patients (Mitral Valve Academic Research Consortium definition). In-hospital death occurred in 7 patients (3%) due to cardiac causes (n = 6) and pneumonia (n = 1). After discharge, we observed 89 deaths over a median (range) follow-up of 655 (739–912) days, 68 of them were cardiac. Data on patients (n = 44) who were lost to follow-up were censored at the time of the last contact. The latter may have led us to underestimate the observed mortality. The observed overall survival was consistent with that predicted by SHFM (1-year: 79% observed vs. 80% predicted, Δ = 1%, P = 0.785; 2-year: 66% observed vs. 67% predicted, Δ = 1%, P = 0.815; 5-year: 41% observed vs. 40% predicted, Δ = 1%, P = 0.822), higher than that predicted by MAGGIC (1-year: 79% observed vs. 73% predicted, Δ = 6%, P = 0.124; 3-year: 55% observed vs. 48% predicted, Δ = 7%, P = 0.125) and significantly higher than that predicted by 3C-HF (1-year: 79% observed vs. 51.5% predicted, Δ = 27.5%, P < 0.0001) (Figure 1). Currently, there are limited data regarding observed and predicted mortality by the available HF scores after MitraClip treatment. Schau et al.6 already investigated short-term mortality following MitraClip treatment in patients with MR and moderate-to-severe symptomatic HF in comparison to mortality predicted by the SHFM and MAGGIC scores. First of all, our study provided a longer follow-up and, secondly, the use of the 3C-HF risk calculator allowed us to weight the impact of mitral disease. We observed rates of mortality comparable with those predicted by the SHFM score at 1, 2 and 5 years, higher than those predicted by the MAGGIC risk score at 1 and 3 years, and significantly higher compared with that predicted by the 3C-HF score. Nevertheless, the authors need to highlight the fact that the SHFM and MAGGIC scores were validated in HF patients without predominance of severe MR. Presently, it is known that severe MR is an independent risk factor for mortality in patients with HF despite OMT.2-7 Therefore, assuming that SHFM and MAGGIC scores would have underestimated the predicted mortality reported in Figure 1, we can speculate that a benefit can be conferred by the MitraClip procedure in reducing the adverse effects of volume overload induced by FMR6. On the contrary, 3C-HF weighs the degree of functional mitral insufficiency in patients with HF. Therefore, the greater observed 1-year survival after MitraClip could reflect the benefit that this procedure conferred to patients with HF by reducing MR, improving NYHA functional class and inducing favourable left ventricular remodelling. To further confirm these data, we calculated the 3C-HF score for each patient using the degree of residual MR after MitraClip implantation, observing a higher predicted 1-year survival (51.5% baseline vs. 66.4% post-MitraClip, Δ = 14.9%, P = 0.001), which highlights once again the impact of severe mitral valve disease on the prognosis of HF patients. Although intriguing, the conclusions of this study remain purely speculative and hypothesis-generating, because they result from indirect comparisons and an analysis performed in the absence of a real control group receiving OMT. Therefore, only the results of prospective randomized controlled trials, such as the ongoing RESHAPE, COAPT and MITRA-FR studies, will provide definite answers to this issue. Conflict of interest: none declared.