Advertisement

Transcatheter or surgical repair for degenerative mitral regurgitation in elderly patients: A propensity-weighted analysis

Open ArchivePublished:January 22, 2019DOI:https://doi.org/10.1016/j.jtcvs.2019.01.023

      Abstract

      Objective

      To compare the outcomes of MitraClip and surgical mitral repair in low-intermediate risk elderly patients affected by degenerative mitral regurgitation (DMR).

      Methods

      We retrospectively selected patients aged ≥75 years, with Society of Thoracic Surgeons Predicted Risk Of Mortality (STS-PROM) <8%, submitted to MitraClip (n = 100) or isolated surgical repair (n = 206) for DMR at 2 centers between January 2005 and May 2017. To adjust for baseline imbalances, we used a propensity score model for average treatment effect on survival.

      Results

      After weighting, MitraClip showed fewer postoperative complications (P < .05) but increased residual mitral regurgitation (MR) ≥2 (27.0% vs 2.8%, P < .001) compared with surgery. One-year survival was greater after MitraClip compared with surgery (97.6% vs 95.3%, hazard ratio [HR], 0.09; confidence interval [CI], 0.02-0.37, P = .001), whereas 5-year survival was lower (34.5% vs 82.2% respectively, HR, 4.12; CI, 2.31-7.34, P < .001). Greater STS-PROM (HR, 1.18; CI, 1.12-1.24, P < .001) and MR ≥3+ recurrence (HR, 2.18; CI, 1.07-4.48, P = .033) were associated with reduced survival. 5-year MR ≥3+ was more frequent after MitraClip compared with surgery: 36.9% versus 3.9%, odds ratio, 11.4; CI, 4.40-29.68, P < .001.

      Conclusions

      In elderly patients affected by DMR and STS-PROM <8%, the average effect of MitraClip resulted in lower acute postoperative complications and improved 1-year survival compared with surgery. However, MitraClip was associated with greater MR recurrence and reduced survival beyond 1 year. Long-term survival was impaired by patients' greater risk profile and MR recurrence. Early results are promising, but in the setting of operable patients with life expectancy beyond 1 year, the quality bar for transcatheter mitral repair needs to be raised.

      Graphical abstract

      Key Words

      Abbreviations and Acronyms:

      ASMD (absolute standardized mean differences), ATE (average treatment effect), CI (confidence interval), DMR (degenerative mitral regurgitation), GEE (generalized estimating equation), HR (hazard ratio), IPTW (inverse probability of treatment weight), MR (mitral regurgitation), OR (odds ratio), STS-PROM (Society of Thoracic Surgeons Predicted Risk Of Mortality)
      Figure thumbnail fx2
      ATE-weighted Kaplan–Meier survival estimates after MitraClip or surgical repair for DMR.
      In low-intermediate risk elderly patients with DMR, MitraClip was associated with survival improvement within but impairment beyond 1-year compared with surgery. Greater risk and MR recurrence reduced survival.
      Whether MitraClip should be performed in low-intermediate risk patients is debated. Improvement in early mortality compared with surgery is promising, but the impairment of long-term survival with high MR recurrence warrants improved efficacy, through refined patient selection and procedure execution. In the setting of operable patients, the quality bar for transcatheter mitral repair needs to be raised.
      See Commentaries on pages 95 and 97.
      The elderly patient represents an ever-growing subset of patients affected by degenerative mitral regurgitation (DMR). In these patients, the best balance between the acute risk of the operation and the long-term benefit is particularly difficult to ascertain. Indeed, compared with younger patients, they suffer per se a heavier burden after surgery
      • Mehta R.H.
      • Eagle K.A.
      • Coombs L.P.
      • Peterson E.D.
      • Edwards F.H.
      • Pagani F.D.
      • et al.
      Influence of age on outcomes in patients undergoing mitral valve replacement.
      • Detaint D.
      • Sundt T.M.
      • Nkomo V.T.
      • Scott C.G.
      • Tajik A.J.
      • Schaff H.V.
      • et al.
      Surgical correction of mitral regurgitation in the elderly: outcomes and recent improvements.
      • Chikwe J.
      • Goldstone A.B.
      • Passage J.
      • Anyanwu A.C.
      • Seeburger J.
      • Castillo J.G.
      • et al.
      A propensity score-adjusted retrospective comparison of early and mid-term results of mitral valve repair versus replacement in octogenarians.
      while having a naturally reduced life expectancy and concomitant comorbidities that can frustrate the benefit of the operation.
      Despite the good results provided by mitral surgery for DMR, even in the setting of selected elderly patients,
      • Detaint D.
      • Sundt T.M.
      • Nkomo V.T.
      • Scott C.G.
      • Tajik A.J.
      • Schaff H.V.
      • et al.
      Surgical correction of mitral regurgitation in the elderly: outcomes and recent improvements.
      • Badhwar V.
      • Peterson E.D.
      • Jacobs J.P.
      • He X.
      • Brennan J.M.
      • O'Brien S.M.
      • et al.
      Longitudinal outcome of isolated mitral repair in older patients: results from 14,604 procedures performed from 1991 to 2007.
      until 10 years ago up to 50% of all patients affected by severe mitral regurgitation (MR) were not submitted to treatment because of a perceived excessively high procedural risk, one of the main reasons being advanced age.
      • Mirabel M.
      • Iung B.
      • Baron G.
      • Messika-Zeitoun D.
      • Detaint D.
      • Vanoverschelde J.L.
      • et al.
      What are the characteristics of patients with severe, symptomatic, mitral regurgitation who are denied surgery?.
      In this setting, transcatheter mitral repair with the MitraClip System (Abbott Vascular, Santa Clara, Calif) provided over the last decade excellent results in terms of safety, despite the high-risk profile of the treated patients, at the price of a suboptimal MR reduction.
      • Maisano F.
      • Franzen O.
      • Baldus S.
      • Schafer U.
      • Hausleiter J.
      • Butter C.
      • et al.
      Percutaneous mitral valve interventions in the real world: early and 1-year results from the ACCESS-EU, a prospective, multicenter, nonrandomized post-approval study of the MitraClip therapy in Europe.
      • Buzzatti N.
      • Maisano F.
      • Latib A.
      • Taramasso M.
      • Denti P.
      • La Canna G.
      • et al.
      Comparison of outcomes of percutaneous MitraClip versus surgical repair or replacement for degenerative mitral regurgitation in octogenarians.
      • Lim D.S.
      • Reynolds M.R.
      • Feldman T.
      • Kar S.
      • Herrmann H.C.
      • Wang A.
      • et al.
      Improved functional status and quality of life in prohibitive surgical risk patients with degenerative mitral regurgitation after transcatheter mitral valve repair.
      Today, MitraClip is recognized as a viable option to treat high-risk or inoperable symptomatic patients after Heart Team discussion.
      • Baumgartner H.
      • Falk V.
      • Bax J.J.
      • De Bonis M.
      • Hamm C.
      • Holm P.J.
      • et al.
      2017 ESC/EACTS guidelines for the management of valvular heart disease.
      • Nishimura R.A.
      • Otto C.M.
      • Bonow R.O.
      • Carabello B.A.
      • Erwin III, J.P.
      • Fleisher L.A.
      • et al.
      2017 AHA/ACC focused update of the 2014 AHA/ACC guideline for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association task force on clinical practice guidelines.
      In the daily practice, however, many elderly patients do not actually show specific high-risk features but rather fall into the “low-to-intermediate” risk category. Whether the access to MitraClip should be opened to these “gray” patients, with MitraClip being preferred to surgery as the first-line option, is already debated but not yet supported by any specific data.
      The Endovascular Valve Edge-to-Edge Repair Study (EVEREST) II randomized patients between MitraClip and surgery but, as per study design, information on the real-world elderly DMR population is difficult to infer, because of the study cohort's young age, mixed etiologies, and strict inclusion/exclusion criteria.
      • Feldman T.
      • Kar S.
      • Elmariah S.
      • Smart S.C.
      • Trento A.
      • Siegel R.J.
      • et al.
      Randomized comparison of percutaneous repair and surgery for mitral regurgitation: 5-year results of EVEREST II.
      Subsequently, the High and Intermediate Risk Degenerative Mitral Regurgitation Treatment (HiRiDe) trial was specifically designed to randomize elderly patients affected by DMR but was prematurely terminated due to severe enrolling difficulties. Up until now therefore, only few retrospective unbalanced data were available on this topic.
      • Buzzatti N.
      • Maisano F.
      • Latib A.
      • Taramasso M.
      • Denti P.
      • La Canna G.
      • et al.
      Comparison of outcomes of percutaneous MitraClip versus surgical repair or replacement for degenerative mitral regurgitation in octogenarians.
      Given this lack of knowledge, we sought to compare the outcomes following MitraClip or surgical mitral repair in the typical, real-world, low-to-intermediate risk elderly population currently referred to mitral repair for DMR at 2 European leading mitral centers.

      Methods

      We retrospectively selected all consecutive elderly patients submitted to MitraClip or surgical mitral repair at 2 mitral centers between January 2005 and May 2017: San Raffaele Scientific Institute (Milan, Italy) and Zurich University Hospital (Zurich, Switzerland). Timeframe was on purpose selected to include also patients treated surgically before MitraClip's introduction in clinical practice to have more similar patients in the surgical arm: patients with a greater-risk profile who would get shifted to percutaneous treatment after it became available.

      Study Design

      Figure 1 describes the flowchart for patients' inclusion into the study. Patients submitted to mitral repair aged ≥75 years at the time of operation and affected by severe MR due to leaflet prolapse with or without flail (DMR) were included into the study.
      Figure thumbnail gr1
      Figure 1Flowchart of study patients' selection. MR, Mitral regurgitation; TR, tricuspid regurgitation; STS, Society of Thoracic Surgeons.
      Patients with MR etiologies other than DMR were excluded. Patients who underwent immediate mitral replacement as intention-to-treat were excluded, as well as patients who received associated procedures as tricuspid valve, arrhythmia ablation, coronary artery bypass grafting, aortic valve, ascending aorta, and myectomy procedures. Ten patients receiving concomitant left atrial appendage closure were included, 4 in the MitraClip and 6 in the surgical group. Patients with associated preoperative tricuspid regurgitation ≥3+ were excluded.
      All patients treated with MitraClip underwent a comprehensive Heart Team discussion, which included the surgical risk assessment using the Society of Thoracic Surgeons Predicted Risk Of Mortality (STS-PROM). To focus on low and intermediate risk, patients with preoperative STS-PROM ≥8% were excluded.
      Although EVEREST II eligibility criteria for MitraClip were used as a reference,
      • Mauri L.
      • Garg P.
      • Massaro J.M.
      • Foster E.
      • Glower D.
      • Mehoudar P.
      • et al.
      The EVEREST II Trial: eesign and rationale for a randomized study of the evalve MitraClip system compared with mitral valve surgery for mitral regurgitation.
      a number of patients outside these was also treated. Patients with target leaflet calcifications and leaflet perforations were excluded.
      The study complied with the Declaration of Helsinki and it was approved by the institutions' ethics committees.

      Data Collection

      All patients underwent a preoperative transthoracic and transoesophageal ecocardiographic mitral examination. Patients in the MitraClip group underwent an in-hospital prospective data collection pathway. Data of surgical patients were retrospectively retrieved from Institutional databases.
      After the operation, patients in the MitraClip group were enrolled in a dedicated echocardiographic outpatient clinic, undergoing a transthoracic echocardiogram at every visit (1, 6, and 12 months and every year thereafter). Surgical patients and patients missing follow-up visits were contacted via a telephone call and asked to provide their follow-up echocardiograms.
      MR was graded according to European Association of Echocardiography and Mitral Valve Academic Research Consortium recommendations as: 1+ (mild), 2+ (mild-to-moderate), 3+ (moderate-to-severe), and 4+ (severe).
      • Lancellotti P.
      • Moura L.
      • Pierard L.A.
      • Agricola E.
      • Popescu B.A.
      • Tribouilloy C.
      • et al.
      European Association of Echocardiography recommendations for the assessment of valvular regurgitation. Part 2: mitral and tricuspid regurgitation (native valve disease).
      • Stone G.W.
      • Adams D.H.
      • Abraham W.T.
      • Kappetein A.P.
      • Genereux P.
      • Vranckx P.
      • et al.
      Clinical trial design principles and endpoint definitions for transcatheter mitral valve repair and replacement: part 2: endpoint definitions: a consensus document from the Mitral Valve Academic Research Consortium.
      Because of the potential inaccuracy of conventional parameters for MR quantification in the setting of double-orifice mitral valve, the jet-related parameters (vena contracta, effective regurgitant orifice) were used only in presence of a single-orifice regurgitant jet, whereas the regurgitant fraction was used in multiple-jet MR.
      • Zoghbi W.A.
      • Enriquez-Sarano M.
      • Foster E.
      • Grayburn P.A.
      • Kraft C.D.
      • Levine R.A.
      • et al.
      Recommendations for evaluation of the severity of native valvular regurgitation with two-dimensional and Doppler echocardiography.
      Residual MR was captured at the end of the procedure in the operating room.
      Among surgical patients, all patients treated at San Raffaele Scientific Institute (n = 173, 84.0%) underwent mid-sternotomy, whereas all patients treated at Zurich University Hospital (n = 33, 16.0%) underwent a minithoracotomy approach.

      Outcomes

      The main endpoints of the study were the comparisons of survival and MR recurrence between MitraClip and surgical patients.

      Statistical Analyses

      To adjust for imbalance in baseline characteristics between MitraClip and surgical patients, a propensity score to weight the sample with inverse probability of treatment weight (IPTW) was used so that the distribution of measured baseline covariates was independent of treatment assignment.
      • Austin P.C.
      An introduction to propensity score methods for reducing the effects of confounding in observational studies.
      Propensity scores were estimated running multiple logistic models including various combinations of preoperative variables. The best balance was achieved with the following 10 variables: age at operation, sex, body mass index categorized as normal (20-30) or not normal (<20 or >30), serum creatinine, atrial fibrillation, New York Heart Association class ≥III, ejection fraction, systolic pulmonary artery pressure, isolate P2 prolapse, and STS-PROM.
      Weights were designed to estimate the average treatment effect (ATE): the average effect of moving an entire population from surgery to MitraClip. ATE weights were chosen by the odds to estimate the average effect of MitraClip in a pseudo-sample of patient with a covariate distribution similar to that of the overall sample, thus simulating a randomized trial. Weights were trimmed at 95th percentile to ameliorate efficiency and reduce the influence of outlying observations.
      • Lee B.K.
      • Lessler J.
      • Stuart E.A.
      Weight trimming and propensity score weighting.
      Balance of confounders between groups was assessed calculating absolute standardized mean differences (ASMD) for the original and weighted sample according to Austin and Stuart.
      • Austin P.C.
      • Stuart E.A.
      Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies.
      Balance was considered optimal for ASMD <0.10 and sufficient for ASMD <0.20, according to previous literature.
      • Harder V.S.
      • Stuart E.A.
      • Anthony J.C.
      Propensity score techniques and the assessment of measured covariate balance to test causal associations in psychological research.
      • De Bonis M.
      • Lapenna E.
      • Del Forno B.
      • Di Sanzo S.
      • Giacomini A.
      • Schiavi D.
      • et al.
      Minimally invasive or conventional edge-to-edge repair for severe mitral regurgitation due to bileaflet prolapse in Barlow's disease: does the surgical approach have an impact on the long-term results?.

      Endpoint Analysis

      Analyses were conducted using Stata software version 13 (StataCorp LLC, College Station, Tex) and SAS 9.3 (SAS Institute, Cary, NC). Unweighted and weighted means were calculated and differences between groups were compared by nonparametric test based on Somers' D statistic using the somersd package. Numbers in brackets report interquartile range. Crude and weighted rates of endpoints were calculated and compared using the Kaplan–Meier estimates.
      A Cox model weighted on IPTW was used to estimate the hazard ratio (HR) of death between MitraClip and surgical patients, allowing for changes in the proportionality of hazards over time. The proportional hazard assumption for Cox models was assessed using Schoenfeld residuals and global test. To choose variables included into the final Cox model, we tested extensively several variables and selected the most clinically interesting, with a P value <.05 at univariable analysis. The proportional hazard assumption was also evaluated for each variable. The final variables included into the survival model were “Treatment” (MitraClip vs surgery, not proportional, therefore inserted as a time–dependent covariate), “STS-PROM” (proportional over time), and “follow-up MR” (as a binary MR ≥3+ time-varying covariate, splitting the covariate and setting a time of entry and a time of exit from each status). Robust variance estimator was used to account for the estimation of weights. “Institution” (Milan/Zurich) was inserted into the Cox model for survival as a cluster variable.
      To describe the time course of MR, we performed a longitudinal analysis using generalized estimating equations (GEEs) with random intercept for repeated data. The GEE method was applied because of the possible intrapatient correlations. This method allowed to perform a weighted analysis with a binary model for a repeated measure. Longitudinal analysis with GEE model to assess predictors of MR ≥3+ recurrence in the subgroup of MitraClip patients who had MR ≤2+ at the end of the procedure was also conducted. Variables included into this model were institution, valve calcifications, isolated P2 prolapse, and residual 2+ MR. These covariates were chosen based on their clinical importance in the development of MR, their number of events, and the total number of patients/events. Because of the limited number of cases, this subanalysis was limited to within the first 3 years after the procedure. P values <.05 were considered significant.

      Results

      The final number of patients included into the study was 306, among whom 100 (32.7%) had undergone MitraClip and 206 (67.3%) were submitted to surgical repair. The number of patients per each Institution was as follows: at San Raffaele Scientific Institute, n = 227, 54 (23.8%) MitraClip procedures; at Zurich University Hospital n = 79, 46 (58.2%) MitraClip procedures.
      Major baseline patient characteristics before and after propensity-score weighting are summarized in Table 1. At baseline, patients in the MitraClip group were older, at greater risk, and affected by more comorbidities compared with surgical patients, but propensity weighting allowed adequate balance of the 2 groups.
      Table 1Preoperative patients' characteristics
      UnweightedWeighted
      Surgery (n = 206)MitraClip (n = 100)P valueASMD (%)Surgery (SoW = 305.08)MitraClip (SoW = 247.49)ASMD (%)
      Clinical
       Age, y78.8 ± 3.1382.9 ± 3.5<.00112480.1 ± 6.8981.2 ± 6.2715
       Male sex118 (57)55 (55).7065175.3 (57)141.0 (57)1
       BMI
      <20 or >3025 (12)20 (20)47.5 (16)43.7 (17)
      20-30181 (88)80 (80).30522257.6 (84)203.8 (83)3
       Creatinine, mg/dL0.99 [0.81; 1.19]1.08 [0.89; 1.34].018271 [0.82; 1.2]1.01 [0.86; 1.28]2
       CAD41 (20)28 (28).1121962.6 (20)73.2 (29)12
       COPD14 (6.8)19 (19).0013730.7 (10)47.1 (19)14
       CVD8 (3.9)9 (9).0672111.4 (4)25.0 (10)14
       PVD14 (6.8)13 (13).0732119 (6)38.2 (15)16
       Atrial fibrillation25 (12)33 (33)<.0015286.8 (28)83.1 (34)6
       Previous cardiac surgery2 (1.0)9 (9)<.001372.2 (1)31.6 (12)20
       NYHA class III-IV81 (39)66 (66)<.00155143.7 (47)133.7 (54)8
      Echocardiographic
       EDD, mm54.6 ± 6.1855.3 ± 7.05.4091055.0 ± 9.7154.8 ± 13.132
       ESD, mm32.7 ± 6.5533.9 ± 8.59.2031725.9 ± 21.1726.1 ± 29.091
       EF, %60 [58; 66]60 [55; 66.5].1851960 [58; 65]60 [58; 68]2
       sPAP, mm Hg37 [30; 45]45 [30.5; 54.5].0073440 [30; 47]42 [30; 54]4
       Isolated posterior leaflet prolapse164 (80)67 (67).01629233.5 (77)175.3 (71)7
       Isolate P2 prolapse136 (66)53 (53).02829193.3 (63)150.2 (61)3
      STS scores
       PROM, %1.64 [1.30; 2.41]2.99 [2.29; 4.38]<.0011031.91 [1.43; 3.24]2.48 [1.91; 3.49]12
       MM, %14.88 [12.32; 18.06]19.21 [15.47; 22.9]<.0017516.2 [13; 21.2]17.3 [14.6; 21.1]8
      Numbers in brackets report interquartile range.
      ASMD, Absolute standardized mean difference; SoW, sum of weights; BMI, body mass index; CAD, coronary artery disease; COPD, chronic obstructive pulmonary disease; CVD, cerebrovascular disease; PVD, peripheral vascular disease; NYHA, New York Heart Association; EDD, end-diastolic diameter; ESD, end-systolic diameter; EF, ejection fraction; sPAP, systolic pulmonary artery pressure; STS, Society of Thoracic Surgeons; PROM, Predicted Risk Of Mortality; MM, morbidity and mortality.
      Intraoperative details are reported in Table 2. Postoperatively, 4 deaths were observed: 2 patients in the MitraClip group (1 multiorgan failure after emergent conversion to surgery, 1 electromechanical dissociation) and 2 surgical patients (1 hemorrhagic shock due to heart rupture, 1 multiorgan failure). At weighted analysis, the risk of all postoperative complications was significantly reduced after MitraClip compared with surgery (Table 3). In contrast, greater degrees of residual MR were observed after MitraClip (Table 3, Figure E1).
      Table 2Intraoperative data
      Surgery

      SoW = 305.08
      MitraClip

      SoW = 247.49
      Total clip time, min89.3 ± 46.82
      CPB time, min88.5 ± 37.12
      Crossclamp time, min63.2 ± 26.11
      Number of clips, n (%)
       134.9 (14)
       2119.1 (48)
       351.1 (21)
      Associated LAA closure6.7 (2.6)9.9 (4.8)
      Surgical access, n (%)
       Sternotomy138.4 (45)
       Minithoracotomy166.7 (55)
      Surgical techniques, n (%)
       Resection150.0 (49)
       Neochordae70.9 (23)
       Edge to edge105.2 (34)
       Cleft closure42.3 (14)
      Ring implantation, n (%)297.2 (97)
       Mean ring size32.7 ± 3.37
      II Pump run, n (%)21.3 (7)
      Conversion to replacement, n (%)9.7 (3.2)17.1 (6.9)
      SoW, Sum of weights; CPB, cardiopulmonary bypass; LAA, left atrial appendage.
      Table 3Postoperative outcomes
      Surgery

      SoW = 305.08
      MitraClip

      SoW = 247.49
      P value
      Residual MR ≥2+8.5 (2.8)67.7 (27)<.001
      LCOS52.8 (17)8.0 (3.3).001
      Inotropes > MD124.7 (41)18.0 (7.3)<.001
      AKI134.2 (45)26.3 (11)<.001
      Ventilation hours15 [11; 19]5 [0; 10]<.001
      Ventilation >24 h37.2 (12)1.1 (0.5)<.001
      NIMV61.9 (24)9.8 (6.2).002
      Sepsis11.3 (3.7)1.2 (0.5).118
      Serious bleeding26.1 (15)6.0 (2.3).003
      Transfusion need111.5 (37)15.1 (6.1)<.001
      Stroke4.2 (1.4)2.5 (1.0).741
      ICU LOS, d1 [1; 2]0 [0; 1]<.001
      ICU LOS >1 d128.4 (42)19 (7.8)<.001
      Postoperative LOS, d7 [5; 9]4 [4; 6]<.001
      In-hospital death2.8 (0.9)2.7 (1.1).853
      Discharge home2.5 (0.8)170.6 (69)<.001
      Numbers in brackets report interquartile range.
      SoW, Sum of weights; MR, mitral regurgitation; LCOS, low cardiac output syndrome; MD, medium dosage, ≥adrenalin 0.05 γ/kg/min or association of 2 inotropes; AKI, acute kidney injury; NIMV, noninvasive mechanical ventilation; ICU, intensive care unit; LOS, length of stay.
      Seven (2.3%) patients were missing at follow-up. Median follow-up time was 1.38 [0.67-2.85] years in MitraClip and 3.2 [1.28-6.33] years in surgery.
      After IPTW, 1-year survival was 97.6% (95% confidence interval [CI], 94.0-99.0) in MitraClip versus 95.3% (95% CI, 92.1-97.2) in surgery, whereas 5-year survival was 34.5% (95% CI, 22.9-46.4) versus 82.2% (95% CI, 75.5-87.2) respectively (Figure 2). The hazard of mortality for patients in the MitraClip group was lower in the first year (HR, 0.09; 95% CI, 0.02-0.37, P = .001) but significantly greater during the following years (HR, 4.12; 95% CI, 2.31-7.34, P < .001). When included into the same Cox extended model both preoperative STS-PROM (HR, 1.18; 95% CI, 1.12-1.24, P < .001) and recurrence of MR ≥3+ (HR, 2.19; 95% CI, 1.07-4.48, P = .033) also predicted increased hazard of mortality (Table 4).
      Figure thumbnail gr2
      Figure 2Average treatment effect–weighted Kaplan–Meier survival estimates up to 5 years (A) and up to 1 year (B) after the procedure. The HR of death with MitraClip compared with surgery is lower within but greater beyond 1 year. HR, Hazard ratio; CI, confidence interval.
      Table 4Weighted Cox model for all-cause death
      HRP value95% CI
      STS-PROM1.18<.0011.12-1.24
      MitraClip (t < 1 y)0.09.0010.02-0.37
      MitraClip (t > 1 y)4.12<.0012.31-7.34
      MR ≥3+ recurrence2.19.0331.07-4.48
      HR, Hazard ratio; CI, confidence interval; STS-PROM, Society of Thoracic Surgeons Predicted Risk Of Mortality; MR, mitral regurgitation.
      The raw number of echocardiograms available along follow-up is reported in Table E1. ATE-weighted longitudinal mixed model with GEE for MR is depicted in Figure 3. Worse MR degrees were significantly more frequent in MitraClip compared with surgical patients: 5-year estimate for MR ≥3+ was 36.9% after MitraClip versus 3.9% after surgery (odds ratio [OR], 11.4; 95% CI, 4.40-29.68, P < .001).
      Figure thumbnail gr3
      Figure 3Average treatment effect–weighted longitudinal mixed model for evolution of MR ≥3+ after the procedure. MR, Mitral regurgitation; OR, odds ratio; CI, confidence interval.
      In the MitraClip group, acute residual 2+ MR was associated with worse follow-up MR (OR, 12.0; 95% CI, 3.22-44.6, P <.001). A difference in follow-up MR was also observed between centers (OR, 0.10; 95% CI, 0.02-0.48, P = .004), Table 5.
      Table 5Longitudinal analysis for predictors of MR recurrence within 3 years in MitraClip patients with initial residual MR ≤2+
      ORP value95% CI
      Valve calcifications1.18.840.29-6.12
      Isolated P2 prolapse0.43.230.1-1.71
      Center (Milan)0.09.0040.02-0.48
      Residual 2+ MR12.0<.0013.22-44.60
      OR, Odds ratio; CI, confidence interval; MR, mitral regurgitation.
      A total of 10 (6 MitraClip and 4 surgical) patients required a new mitral intervention/operation. Median time to reintervention/reoperation was 7.9 (3.4-15.6) months. Four patients (3 MitraClip and 1 surgical) could be retreated with a new MitraClip. The remaining 6 (3 MitraClip, 3 surgical) required valve replacement.

      Discussion

      The main findings of this study were that, in an average daily practice low-to-intermediate risk elderly patient population affected by DMR, compared with surgery, MitraClip was associated with: (1) lower rate of acute complications; (2) improved survival within 1 year but reduced survival beyond 1 year; and (3) increased MR recurrence. Baseline patients' greater-risk profile and MR recurrence predicted reduced survival.
      The present study is the first to try to address a weighted comparison between MitraClip and surgery in the aforementioned patient population. Previously, the failure of the HiRiDe Randomized Trial underlined the difficulties to conduct a proper study in a high-risk elderly population affected by MR. The use of ATE weights allowed us to compare outcomes between MitraClip and surgery as all patients could receive either of the 2, thus simulating a randomized trial, on low-to-intermediate risk all-comers. Indeed, with the expansion of transcatheter technologies, the dilemma we are currently facing in the clinical practice is to understand what is the best treatment option for “gray” elderly patients who typically present a low-to-intermediate risk profile but also a very advanced age (often older than 80 years). In such scenario, the burden of surgery and the patients' perceived short life expectancy often make the new less-invasive transcatheter devices appear particularly appealing.
      As predicted, MitraClip was confirmed to be safer than surgery in the short term. Indeed, all acute postoperative complications, such as low cardiac output syndrome, acute kidney injury, prolonged ventilation, bleedings, as well as postoperative length of stay were significantly reduced, with the vast majority of patients discharged directly home without need of rehabilitation. Quick return to normal life and home environment with early full recovery represents a critical advantage in the delicate elderly setting.
      Moreover, for the first time ever, the hazard of follow-up mortality within 1 year from the procedure was observed to be lower after MitraClip compared with surgery. Although numbers are still relatively small and more data will be needed to confirm this trend, the result is promising and worth to be noted. It is also perfectly reasonable, given MitraClip's reduced invasiveness compared with surgery, resembling the experience with transcatheter versus surgical aortic valve replacement.
      • Reardon M.J.
      • Kleiman N.S.
      • Adams D.H.
      • Yakubov S.J.
      • Coselli J.S.
      • Deeb G.M.
      • et al.
      Outcomes in the randomized corevalve US Pivotal high risk trial in patients with a Society of Thoracic Surgeons risk score of 7% or less.
      Indeed, mitral surgery carries a burden that continues to express over several weeks or even months after the operation.
      • Chikwe J.
      • Goldstone A.B.
      • Passage J.
      • Anyanwu A.C.
      • Seeburger J.
      • Castillo J.G.
      • et al.
      A propensity score-adjusted retrospective comparison of early and mid-term results of mitral valve repair versus replacement in octogenarians.
      Beyond 1 year after the procedure, however, survival after MitraClip fell unfavorably compared with surgery. In this regard, it must be pointed out that possible uncaptured confounding factors (such as frailty) and imperfect statistical balance between the 2 groups due to the limited numbers may partially justify such difference. Certainly, a sicker profile (as described by the STS-PROM) and the recurrence of MR were associated with an increased mortality rate. MR recurrence being frequent after MitraClip and being associated with bad outcomes is no surprise,
      • Feldman T.
      • Kar S.
      • Elmariah S.
      • Smart S.C.
      • Trento A.
      • Siegel R.J.
      • et al.
      Randomized comparison of percutaneous repair and surgery for mitral regurgitation: 5-year results of EVEREST II.
      • Suri R.M.
      • Clavel M.A.
      • Schaff H.V.
      • Michelena H.I.
      • Huebner M.
      • Nishimura R.A.
      • et al.
      Effect of recurrent mitral regurgitation following degenerative mitral valve repair: long-term analysis of competing outcomes.
      but the confirmation of these associations in this specific elderly population warrants an improvement in the efficacy of transcatheter mitral repair, especially in the perspective of expanding indications to lower-risk operable patients.
      Good sustained results are actually achievable after MitraClip (Video 1), as suggested by the fact that the majority of patients undergoing MitraClip in our series were indeed still free from MR recurrence after 5 years. While our analysis for predictors of MR recurrence after MitraClip is limited and should only be considered hypothesis generating, it provides interesting insights.
      Figure thumbnail fx3
      Video 1A typical case of elderly patient treated with the percutaneous MitraClip procedure. Video available at: https://www.jtcvs.org/article/S0022-5223(19)30232-6/fulltext.
      Most important, worse follow-up MR was observed when 2+ residual MR was present immediately at the end of the procedure. This has been already previously described, both after surgical mitral repair
      • De Bonis M.
      • Lapenna E.
      • Lorusso R.
      • Buzzati N.
      • Gelsomino S.
      • Taramasso M.
      • et al.
      Very long-term results (up to 17 years) with the double-orifice mitral valve repair combined with ring annuloplasty for degenerative mitral regurgitation.
      and MitraClip.
      • Buzzatti N.
      • De Bonis M.
      • Denti P.
      • Barili F.
      • Schiavi D.
      • Di Giannuario G.
      • et al.
      What is a “good” result after transcatheter mitral repair? Impact of 2+ residual mitral regurgitation.
      • Toggweiler S.
      • Zuber M.
      • Surder D.
      • Biaggi P.
      • Gstrein C.
      • Moccetti T.
      • et al.
      Two-year outcomes after percutaneous mitral valve repair with the MitraClip system: durability of the procedure and predictors of outcome.
      Residual 2+ MR is common after MitraClip and is still often considered “acceptable” in the elderly people, who have a perceived reduced functional capacity and life expectancy. That concept may be accepted in very old, very sick cases, and we do recognize that ≤1+ residual MR cannot be achieved in each and every single case. Nevertheless, increasing evidence suggests that in the low-to-intermediate risk people, despite advance age, every effort should be made to minimize residual MR, otherwise MR will increase over time and impair survival in the long run.
      Some influence of the institutions on follow-up MR was also observed, underlining the importance of experienced operators and teams in such complex procedures, especially in the very early phase of adoption.
      It should also be underlined that MitraClip's edge-to-edge should not be considered a “procedure” per se but rather just one “tool” to be used within a mitral repair procedure, the same way a 4-0 PROLENE suture is used in surgery. Like in surgery, other tools addressing other mitral targets will be needed to achieve full optimization of transcatheter mitral repair. Most important, it is well documented across the surgical literature that the absence of ring implantation is one of the strongest predictors of MR recurrence after mitral repair, even with the edge-to-edge technique.
      • De Bonis M.
      • Lapenna E.
      • Maisano F.
      • Barili F.
      • La Canna G.
      • Buzzatti N.
      • et al.
      Long-term results (</=18 years) of the edge-to-edge mitral valve repair without annuloplasty in degenerative mitral regurgitation: implications for the percutaneous approach.
      Transcatheter mitral rings have recently started to provide promising results,
      • Maisano F.
      • Taramasso M.
      • Nickenig G.
      • Hammerstingl C.
      • Vahanian A.
      • Messika-Zeitoun D.
      • et al.
      Cardioband, a transcatheter surgical-like direct mitral valve annuloplasty system: early results of the feasibility trial.
      and we are looking forward to combining different transcatheter technologies in the future, targeting both the mitral leaflets and the annulus, as we do in surgery, to provide a tailored treatment and improve the efficacy of transcatheter mitral repair.
      In the complex scenario that is the elderly patient, the choice of the best treatment should be individualized per each single person, being tailored through a Heart Team approach to the patient anatomy, risk, need, and desire.

      Study Limitations

      Several limitations of this study must be acknowledged. First, its retrospective nature may have led to biases like the impossible collection of variables such as quality of life, frailty, or other potential confounding factors. Frailty in particular is an important missing variable, the assessment of which remains not standardized and uncommon in the clinical daily practice. Also, precise scallop-by-scallop anatomy pairing was not considered into the propensity score model because of the too-small number of patients. Second, because of the small numbers of patients and events, the best IPTW model that could be achieved allowed sufficient but not perfect balance between groups, with some ASMD between 0.1 and 0.2; the overall statistical power of the study was also limited. Third, because the present series includes MitraClip procedures performed since the beginning of centers' experiences, a learning curve effect should be taken into account. Fourth, echocardiographic evaluation was performed on site, without a core-lab re-adjudication. The evaluation of MR in the subverted double-orifice valve anatomy remains challenging and may have led to biased estimation of real MR. Greater rigor in the echocardiographic follow-up, technique, and analysis in the MitraClip group could have biased the long-term MR results. For the aforementioned reasons, our findings should only be considered hypothesis-generating and they should be confirmed by larger data.

      Conclusions

      In the average low-to-intermediate risk elderly affected by DMR, MitraClip was associated with lower acute postoperative complications and increased survival within 1 year compared with surgery but also with greater MR recurrence and reduced survival beyond 1 year. Long-term survival was impaired by patients' greater risk profile and MR recurrence. Although early results are promising, in the setting of operable patients with life expectancy beyond 1 year, the quality bar for transcatheter mitral repair needs to be raised.

      Conflicts of Interest

      Dr Denti is a consultant for Abbott Vascular, Valtech, 4Tech Cardio, and InnovHeart. Dr Taramasso is a consultant for Abbott Vascular and 4Tech Cardio. Dr Maisano is a consultant for Abbott Vascular, St Jude Medical, Medtronic, Valtech Cardio, Bioventrix, Direct Flow, Edwards Lifescience, and Xeltis and he is co-founder of 4Tech Cardio, AFfix, and TSP Medical. All other authors have nothing to disclose with regard to commercial support.

      Appendix

      Figure thumbnail fx4
      Figure E1Weighted residual mitral regurgitation at the end of the procedure.
      Table E1Raw un-weighted number of available echocardiograms during follow-up
      Time, yMitraClipSurgery
      0100206
      194178
      259131
      33093
      41972
      51462

      Supplementary Data

      References

        • Mehta R.H.
        • Eagle K.A.
        • Coombs L.P.
        • Peterson E.D.
        • Edwards F.H.
        • Pagani F.D.
        • et al.
        Influence of age on outcomes in patients undergoing mitral valve replacement.
        Ann Thorac Surg. 2002; 74: 1459-1467
        • Detaint D.
        • Sundt T.M.
        • Nkomo V.T.
        • Scott C.G.
        • Tajik A.J.
        • Schaff H.V.
        • et al.
        Surgical correction of mitral regurgitation in the elderly: outcomes and recent improvements.
        Circulation. 2006; 114: 265-272
        • Chikwe J.
        • Goldstone A.B.
        • Passage J.
        • Anyanwu A.C.
        • Seeburger J.
        • Castillo J.G.
        • et al.
        A propensity score-adjusted retrospective comparison of early and mid-term results of mitral valve repair versus replacement in octogenarians.
        Eur Heart J. 2011; 32: 618-626
        • Badhwar V.
        • Peterson E.D.
        • Jacobs J.P.
        • He X.
        • Brennan J.M.
        • O'Brien S.M.
        • et al.
        Longitudinal outcome of isolated mitral repair in older patients: results from 14,604 procedures performed from 1991 to 2007.
        Ann Thorac Surg. 2012; 94 (discussion 1877-9): 1870-1877
        • Mirabel M.
        • Iung B.
        • Baron G.
        • Messika-Zeitoun D.
        • Detaint D.
        • Vanoverschelde J.L.
        • et al.
        What are the characteristics of patients with severe, symptomatic, mitral regurgitation who are denied surgery?.
        Eur Heart J. 2007; 28: 1358-1365
        • Maisano F.
        • Franzen O.
        • Baldus S.
        • Schafer U.
        • Hausleiter J.
        • Butter C.
        • et al.
        Percutaneous mitral valve interventions in the real world: early and 1-year results from the ACCESS-EU, a prospective, multicenter, nonrandomized post-approval study of the MitraClip therapy in Europe.
        J Am Coll Cardiol. 2013; 62: 1052-1061
        • Buzzatti N.
        • Maisano F.
        • Latib A.
        • Taramasso M.
        • Denti P.
        • La Canna G.
        • et al.
        Comparison of outcomes of percutaneous MitraClip versus surgical repair or replacement for degenerative mitral regurgitation in octogenarians.
        Am J Cardiol. 2015; 115: 487-492
        • Lim D.S.
        • Reynolds M.R.
        • Feldman T.
        • Kar S.
        • Herrmann H.C.
        • Wang A.
        • et al.
        Improved functional status and quality of life in prohibitive surgical risk patients with degenerative mitral regurgitation after transcatheter mitral valve repair.
        J Am Coll Cardiol. 2014; 64: 182-192
        • Baumgartner H.
        • Falk V.
        • Bax J.J.
        • De Bonis M.
        • Hamm C.
        • Holm P.J.
        • et al.
        2017 ESC/EACTS guidelines for the management of valvular heart disease.
        Eur Heart J. 2017; 38: 2739-2791
        • Nishimura R.A.
        • Otto C.M.
        • Bonow R.O.
        • Carabello B.A.
        • Erwin III, J.P.
        • Fleisher L.A.
        • et al.
        2017 AHA/ACC focused update of the 2014 AHA/ACC guideline for the management of patients with valvular heart disease: a report of the American College of Cardiology/American Heart Association task force on clinical practice guidelines.
        J Am Coll Cardiol. 2017; 70: 252-289
        • Feldman T.
        • Kar S.
        • Elmariah S.
        • Smart S.C.
        • Trento A.
        • Siegel R.J.
        • et al.
        Randomized comparison of percutaneous repair and surgery for mitral regurgitation: 5-year results of EVEREST II.
        J Am Coll Cardiol. 2015; 66: 2844-2854
        • Mauri L.
        • Garg P.
        • Massaro J.M.
        • Foster E.
        • Glower D.
        • Mehoudar P.
        • et al.
        The EVEREST II Trial: eesign and rationale for a randomized study of the evalve MitraClip system compared with mitral valve surgery for mitral regurgitation.
        Am Heart J. 2010; 160: 23-29
        • Lancellotti P.
        • Moura L.
        • Pierard L.A.
        • Agricola E.
        • Popescu B.A.
        • Tribouilloy C.
        • et al.
        European Association of Echocardiography recommendations for the assessment of valvular regurgitation. Part 2: mitral and tricuspid regurgitation (native valve disease).
        Eur J Echocardiogr. 2010; 11: 307-332
        • Stone G.W.
        • Adams D.H.
        • Abraham W.T.
        • Kappetein A.P.
        • Genereux P.
        • Vranckx P.
        • et al.
        Clinical trial design principles and endpoint definitions for transcatheter mitral valve repair and replacement: part 2: endpoint definitions: a consensus document from the Mitral Valve Academic Research Consortium.
        J Am Coll Cardiol. 2015; 66: 308-321
        • Zoghbi W.A.
        • Enriquez-Sarano M.
        • Foster E.
        • Grayburn P.A.
        • Kraft C.D.
        • Levine R.A.
        • et al.
        Recommendations for evaluation of the severity of native valvular regurgitation with two-dimensional and Doppler echocardiography.
        J Am Soc Echocardiogr. 2003; 16: 777-802
        • Austin P.C.
        An introduction to propensity score methods for reducing the effects of confounding in observational studies.
        Multivariate Behav Res. 2011; 46: 399-424
        • Lee B.K.
        • Lessler J.
        • Stuart E.A.
        Weight trimming and propensity score weighting.
        PLoS One. 2011; 6: e18174
        • Austin P.C.
        • Stuart E.A.
        Moving towards best practice when using inverse probability of treatment weighting (IPTW) using the propensity score to estimate causal treatment effects in observational studies.
        Stat Med. 2015; 34: 3661-3679
        • Harder V.S.
        • Stuart E.A.
        • Anthony J.C.
        Propensity score techniques and the assessment of measured covariate balance to test causal associations in psychological research.
        Psychol Methods. 2010; 15: 234-249
        • De Bonis M.
        • Lapenna E.
        • Del Forno B.
        • Di Sanzo S.
        • Giacomini A.
        • Schiavi D.
        • et al.
        Minimally invasive or conventional edge-to-edge repair for severe mitral regurgitation due to bileaflet prolapse in Barlow's disease: does the surgical approach have an impact on the long-term results?.
        Eur J Cardiothorac Surg. 2017; 52: 131-136
        • Reardon M.J.
        • Kleiman N.S.
        • Adams D.H.
        • Yakubov S.J.
        • Coselli J.S.
        • Deeb G.M.
        • et al.
        Outcomes in the randomized corevalve US Pivotal high risk trial in patients with a Society of Thoracic Surgeons risk score of 7% or less.
        JAMA Cardiol. 2016; 1: 945-949
        • Suri R.M.
        • Clavel M.A.
        • Schaff H.V.
        • Michelena H.I.
        • Huebner M.
        • Nishimura R.A.
        • et al.
        Effect of recurrent mitral regurgitation following degenerative mitral valve repair: long-term analysis of competing outcomes.
        J Am Coll Cardiol. 2016; 67: 488-498
        • De Bonis M.
        • Lapenna E.
        • Lorusso R.
        • Buzzati N.
        • Gelsomino S.
        • Taramasso M.
        • et al.
        Very long-term results (up to 17 years) with the double-orifice mitral valve repair combined with ring annuloplasty for degenerative mitral regurgitation.
        J Thorac Cardiovasc Surg. 2012; 144: 1019-1026
        • Buzzatti N.
        • De Bonis M.
        • Denti P.
        • Barili F.
        • Schiavi D.
        • Di Giannuario G.
        • et al.
        What is a “good” result after transcatheter mitral repair? Impact of 2+ residual mitral regurgitation.
        J Thorac Cardiovasc Surg. 2016; 151: 88-96
        • Toggweiler S.
        • Zuber M.
        • Surder D.
        • Biaggi P.
        • Gstrein C.
        • Moccetti T.
        • et al.
        Two-year outcomes after percutaneous mitral valve repair with the MitraClip system: durability of the procedure and predictors of outcome.
        Open Heart. 2014; 1: e000056
        • De Bonis M.
        • Lapenna E.
        • Maisano F.
        • Barili F.
        • La Canna G.
        • Buzzatti N.
        • et al.
        Long-term results (</=18 years) of the edge-to-edge mitral valve repair without annuloplasty in degenerative mitral regurgitation: implications for the percutaneous approach.
        Circulation. 2014; 130: S19-S24
        • Maisano F.
        • Taramasso M.
        • Nickenig G.
        • Hammerstingl C.
        • Vahanian A.
        • Messika-Zeitoun D.
        • et al.
        Cardioband, a transcatheter surgical-like direct mitral valve annuloplasty system: early results of the feasibility trial.
        Eur Heart J. 2016; 37: 817-825

      Linked Article

      • Commentary: Looking for support for the MitraClip
        The Journal of Thoracic and Cardiovascular SurgeryVol. 158Issue 1
        • Preview
          Treatment options for degenerative mitral regurgitation have evolved from medical management to valve replacement to mitral valve repair, with each step of the way associated with improvement in patient outcomes.1 In 1991, Alfieri's group2 began repairing the mitral valve with a novel edge-to-edge technique. Their description of 35 (out of 299) repairs in which they used this technique had early positive results, especially when the prolapsing leaflet was the anterior leaflet and an annuloplasty ring was used.
        • Full-Text
        • PDF
        Open Archive
      • Commentary: Moving the guidelines for mitral repair
        The Journal of Thoracic and Cardiovascular SurgeryVol. 158Issue 1
        • Preview
          Both European1 and American2 guidelines currently recommend percutaneous mitral repair as a “weak” class IIb recommendation that “may be considered” for patients with symptomatic, severe primary mitral regurgitation who are judged by a Heart Team not to be surgical candidates. However, given increasing facility with percutaneous techniques, likely technological advances, and compelling clinical need in an aging population, indications are likely to expand. Careful analysis of clinical experience is an essential complement to the criterion standard of the randomized clinical trial and is frequently the source helpful findings that can drive future research.
        • Full-Text
        • PDF
        Open Archive