Author + information
- Received March 29, 2014
- Revision received June 19, 2014
- Accepted July 2, 2014
- Published online January 1, 2015.
- Claire Bouleti, MD, PhD∗∗ (, )
- Amir-Ali Fassa, MD∗,
- Dominique Himbert, MD∗,
- Eric Brochet, MD∗,
- Gregory Ducrocq, MD∗,
- Mohammed Nejjari, MD∗,
- Walid Ghodbane, MD†,
- Jean-Pol Depoix, MD‡,
- Patrick Nataf, MD† and
- Alec Vahanian, MD∗
- ∗Department of Cardiology, Bichat-Claude-Bernard Hospital, Assistance Publique-Hôpitaux de Paris, Université Paris VII, Paris, France
- †Department of Cardiovascular Surgery, Bichat-Claude-Bernard Hospital, Assistance Publique-Hôpitaux de Paris, Université Paris VII, Paris, France
- ‡Department of Anesthesiology, Bichat-Claude-Bernard Hospital, Assistance Publique-Hôpitaux de Paris, Université Paris VII, Paris, France
- ↵∗Reprint requests and correspondence:
Dr. Claire Bouleti, Department of Cardiology, Bichat Hospital, 46 rue Henri-Huchard, 75018 Paris, France.
Objectives This study sought to evaluate the feasibility of transfemoral transcatheter heart valve (THV) implantation in failed mitral bioprostheses and ring annuloplasties.
Background Redo mitral surgery may be high risk or contraindicated due to comorbidity. THV implantation has been recently reported in this setting.
Methods Transfemoral implantation of Edwards Sapien prosthesis was performed in 17 patients for degenerated mitral bioprosthesis or previous ring annuloplasty (6 bioprostheses, 11 ring annuloplasties). The procedure was elective for 14 patients and attempted as a rescue in 3 patients. Mean age was 61 ± 24 years. All patients were in New York Heart Association class ≥III, and the surgical risk was high (EuroSCORE [European System for Cardiac Operative Risk Evaluation]: 37 ± 29%, Society of Thoracic Surgeons score: 18 ± 22%).
Results Procedure was successful in 14 patients (82%). Two complications occurred during rescue procedures: 1 procedural death and 1 THV migration. One patient had moderate paraprosthetic regurgitation following the procedure, whereas residual regurgitation was trace or less in 11 patients (69%) and mild in 4 patients (25%). Mean gradient decreased from 12 ± 6 mm Hg to 8 ± 3 mm Hg. During a mean follow-up of 22 months, 4 patients died, 3 from cardiac cause. The 18-month survival was 68 ± 14% in the overall population and 78 ± 14% for patients with elective procedure. One patient underwent mitral valve replacement due to periprosthetic mitral regurgitation. At last follow-up, 12 patients were in New York Heart Association class ≤II (75%) and 4 in class III (25%).
Conclusions This single-center series suggests that transfemoral THV implantation for deterioration of mitral bioprosthesis or surgical repair is feasible in selected patients and improves early hemodynamic and midterm functional status.
Despite the major progress achieved over the past few decades, a significant proportion of patients who undergo mitral valve surgery require reoperation during follow-up. During the 10 years following mitral valve replacement or repair, reoperation is needed in 20% to 35% of patients (1). Whereas redo surgery is the treatment of choice after bioprosthesis (BP) or ring annuloplasty (RA) failure, it may be associated with significant early mortality (5% to 12%), especially in patients with concurrent comorbidities (2,3).
Following the first description of transcatheter heart valve (THV) implantation for native aortic valve stenosis (4), treatment of failed aortic BP with THV implantation has emerged as a promising alternative to redo surgery in high-risk patients (5,6). Recently, THV implantation has also been reported for treatment of failed mitral BP or RA, mainly through the transapical approach (7–15).
We sought to evaluate the immediate and midterm results of transfemoral implantation of balloon-expandable Sapien XT valves (Edwards Lifesciences, Irvine, California) for failed mitral surgical valves in highly selected patients.
The population consisted of symptomatic patients who underwent transfemoral implantation of a THV in mitral position for failing BP or RA at our institution from March 2, 2011 to August 16, 2013. No transapical approaches were performed during the same period. The decision to perform the intervention was based on a consensual agreement between the members of the heart team in patients considered at high risk for redo surgery. Informed consent for the procedure was obtained from all patients.
The choice of the THV size was based on an integrative approach taking into account manufacturer’s inner diameters, and also the mean diameter determined from the long and short diameter measurements as assessed by computed tomography (CT), 3-dimensional transesophageal echocardiography (TEE), and fluoroscopy.
THV implantation procedures
The technique for THV implantation in mitral position through a transfemoral approach has been described previously (8,16,17). All procedures were performed under general anesthesia and TEE guidance, with the exception of 1, which was done under local anesthesia and fluoroscopic guidance due to impossible tracheal intubation. Bilateral femoral venous access was used. A temporary pacemaker lead was placed in the right ventricle as used during transcatheter aortic valve implantation. Transseptal puncture was performed under TEE guidance in a “high “and “posterior” position similar to that used during MitraClip (Abbott Vascular, Santa Clara, California) implantation. After placing a Mullins sheath (Medtronic, Minneapolis, Minnesota) in the left atrium, a bolus of heparin (70 IU/kg) was administered with the aim of achieving an activated clotting time between 250 and 300 s. Then the mitral valve was crossed with a Judkins right 5-F catheter advanced on a standard 0.035-inch guidewire or a 7-F balloon wedge pressure catheter (Arrow International, Reading, Pennsylvania). Then, a 0.035-inch Amplatz Super Stiff or Lunderquist guidewire (Cook Medical, Bloomington, Indiana) with a “J” curve at the end was placed in the apex of the left ventricle (Figure 1A). The atrial septum was dilated using 12- to 14-mm peripheral balloons. Although pre-dilation should generally be avoided, it was performed in 3 cases due to anticipated difficulties in crossing highly stenotic bioprostheses. In addition, 1 procedure was performed immediately after failed balloon valvuloplasty in a case of restenosis after open-heart commissurotomy with annuloplasty. Then the Sapien XT valve, mounted for antegrade implantation (similar to the position in transapical aortic valve procedures) on a NovaFlex catheter (Edwards Lifesciences), was advanced through the atrial septum. The THV was orientated toward the mitral valve and positioned using maximal flexion of the NovaFlex catheter. Implantation was performed by slow balloon inflation under rapid ventricular pacing (160 to 200 beats/min) (Figure 1B). Post-dilation was performed in 3 cases of moderate paravalvular regurgitation. The final result of implantation was assessed by echocardiography and fluoroscopy in the catheterization lab (Figures 1C and 1D) and CT before discharge (Figures 1E and 1F). Contrast medium was not used during mitral implantation, except in 1 case for positioning of a cerebral protection device (Embrella, Edwards Lifesciences) during mitral valve-in-valve implantation (MVIVI) within a severely calcified and stenotic BP.
The femoral veins were closed by manual compression or suture.
Complications were reported according to the VARC-2 (Valve Academic Research Consortium Procedural) criteria (18). Device success was defined as the absence of procedural mortality, the correct positioning of a single THV, and the absence of residual moderate or severe prosthetic regurgitation. Prosthetic function was assessed before discharge according to current guidelines, using the integration of qualitative and quantitative parameters obtained by echocardiography and adapted to prostheses in mitral position (18–20). THV regurgitation was graded as absent, trace, mild, moderate, or severe.
Clinical evaluation and echocardiography (transthoracic and transesophageal) were performed before discharge. All adverse events were prospectively recorded. The 30-day, 6-month, 1-year, and 2-year medical visits were performed in our institution or by the patients’ own cardiologist. Follow-up was complete in all patients. Mean follow-up was 22 ± 3 months, and maximum follow-up reached 28 months.
Quantitative variables were expressed as mean ± SD. Qualitative variables were expressed as percentages. Comparisons of variables before and after the procedure used the Wilcoxon signed rank test for quantitative variables. In cases in which more than 2 groups were compared, a Friedman rank sum test was performed. Survival curves for time-dependent variables were assessed with Kaplan-Meier estimates. The results were considered significant when 2-sided p values were <0.05. All analyses were performed with the SPSS statistical software package (version 19, SPSS Inc., Chicago, Illinois).
Transfemoral implantation of a THV in the mitral position was attempted in 17 patients. Patients’ characteristics are shown in Table 1.
The type of index surgery was mitral valve replacement in 6 patients and mitral valve repair in 11 patients.
Mean age was 61 ± 24 years (range 18 to 90 years). The majority of patients were female (n = 13, 76.5%). All patients were highly symptomatic; 12 patients (70.6%) were in New York Heart Association (NYHA) class III, and 5 (29.4%) were in class IV. Mean logistic EuroSCORE (European System for Cardiac Operative Risk Evaluation) was 37 ± 29%, mean EuroSCORE II was 20 ± 22%, and mean Society of Thoracic Surgeons score was 18 ± 22%. This population included 3 young women with a desire for pregnancy and at high surgical risk due to major pulmonary hypertension and severe right ventricular dysfunction for 2 of them and a coronary bypass just beneath the sternum for the third.
The mean delay since the last surgery was 8.1 ± 4.7 years. The mode of valve failure was stenosis in the majority of patients (n = 11, 64.7%), whereas 5 patients (29.4%) presented moderate to severe regurgitation and 1 patient (5.9%) had a combination of both mechanisms.
Three MVIVI were performed as rescue interventions in 3 patients with cardiogenic shock, 2 under extracorporeal membrane oxygenation or support: 1 patient with severe stenosis of a BP that had been implanted 10 years previously (Patient #1); 1 patient with a residual severe prosthetic regurgitation 2 days after surgical mitral valve replacement (Patient #2); and 1 patient with severe prosthetic stenosis 23 days following surgical mitral valve replacement (Patient #5). Types of failing BP and RA, along with implanted THV size, are shown in Table 2.
Rings were semirigid in 8 cases (Carpentier Edwards Physio, Edwards Lifesciences), rigid in 1 case (Carpentier Edwards Classic, Edwards Lifesciences), and flexible in 2 cases (Duran AnCore, Medtronic, Minneapolis, Minnesota).
Among the 12 patients with stenotic mitral failure, mean transmitral gradient at baseline was 13.8 ± 4.1 mm Hg.
A Sapien XT valve was implanted in 16 patients (94%). Failure to implant a THV occurred during a rescue procedure complicated by death, which likely resulted from rupture of the inferior vena cava following maneuvers to position the THV within a failing BP that was implanted obliquely in a supra-annular atrial position (Patient #2).
A 26-mm Sapien XT valve was implanted in 12 patients (71%), whereas 2 patients were treated with a 23-mm and 2 patients with a 29-mm Sapien XT valve (Table 2). The size of the valve was chosen by an integrative approach taking into account TEE, CT, and fluoroscopic measurements (Figures 2A and 2D, 2B and 2E, and 2C and 2F, respectively), with fluoroscopy usually providing the largest and TEE the smallest diameters. Mean mitral annulus size was indeed 21.4 ± 2.0 mm for echographic evaluation, 23.7 ± 2.2 mm for CT evaluation, and 25.2 ± 2.1 mm for fluoroscopic evaluation (p < 0.001) (Figure 2G). Valve implantation was performed under rapid ventricular pacing in all but 1 patient in cardiogenic shock, who had no more venous access available, in whom it was made using maximal ventricular unloading from an extracorporeal membrane oxygenator.
Procedural success was achieved in 14 interventions (82%). The 3 procedural failures included the previously mentioned death, 1 THV migration, and a moderate paraprosthetic mitral regurgitation. There was no conversion to open heart surgery.
The THV migration occurred during a rescue MVIVI (Patient #1) and was related to undersizing. We used a 26-mm Sapien XT, which was the largest valve available on a NovaFlex catheter at that time and deemed to be the only possible option in this patient in shock and heparin-induced thrombocytopenia (21). Following migration of a first THV, a second 26-mm THV was successfully implanted within the BP with the addition of an extra milliliter of contrast medium in the inflator. Due to the prohibitive risk of surgery, the migrated valve was left in the left atrium and subsequently became entrapped in the left atrial appendage, with an uneventful in-hospital outcome.
The moderate paravalvular regurgitation occurred after mitral valve-in-ring implantation of a 26-mm Sapien XT within a 27-mm Duran AnCore ring (Medtronic). The patient required surgical mitral valve replacement 6 months after discharge for heart failure and severe pulmonary hypertension.
There was 1 minor vascular complication during an emergency procedure requiring an unplanned surgical access closure following laceration of the femoral vein after forced retrieval of a nonimplanted valve.
The other in-hospital complication was a temporary renal replacement therapy after a rescue procedure.
On the echocardiographic evaluation performed at day 7, mean transvalvular gradient decreased from 13.8 ± 4.1 mm Hg to 8.1 ± 2.6 mm Hg (p < 0.003) among the patients with mitral stenotic failure (Figure 3). Three patients had a “significant” residual stenosis (mean gradient >10 mm Hg), according to the American Society of Echocardiography (20). Furthermore, there was no residual regurgitation in 4 cases, whereas it was quantified as trace in 7 cases, mild in 4 cases, and moderate in 1 case (Figure 4). Left ventricular outflow tract obstruction occurred in 3 cases after mitral valve-in-ring implantation, with maximal gradients of 25, 46, and 49 mm Hg at discharge. There was no clinical impact of these left ventricular outflow tract obstructions with no congestive heart failure during hospitalization, the 3 patients being in NYHA class II. No patient needed reintervention due to left ventricular outflow tract obstruction. Furthermore, the maximal gradients decreased to 10, 8, and 15 mm Hg, respectively, on echocardiographic control examinations.
Kaplan-Meier 30-day survival rates were 94.1% in the overall population and 100% in patients who underwent an elective procedure. The events that occurred during this period are detailed in Table 3. One instance of severe bleeding occurred as a possible inferior vena cava rupture in the previously described patient who underwent a rescue procedure. No patients experienced major stroke or severe bleeding when they benefited from an elective procedure.
Mean follow-up was 22.4 ± 2.8 months. There were 4 deaths during follow-up: 1 patient died after 75 days from a cerebral tumor that was unknown at the time of the procedure (Patient #3); 2 patients had sudden death at 158 and 373 days (Patients #1 and #9). The fourth patient underwent successful transcatheter aortic valve replacement 371 days after mitral valve-in-ring implantation for progression of aortic stenosis and died from an unknown cause 253 days after the reintervention (Patient #7). These 4 patients were at extremely high surgical risk (mean Society of Thoracic Surgeons score = 32%), and the patient who suffered sudden death at 158 days had benefited from a rescue procedure. All 4 patients were in NYHA class II, and they had no prosthesis dysfunction at the last follow-up before death occurred. Furthermore, 1 patient underwent mitral valve replacement and tricuspid RA after 195 days because of heart failure, persistent pulmonary hypertension, and moderate mitral regurgitation.
Kaplan-Meier survival rates at 1-year and 18-month follow-up, respectively, were 80 ± 11% and 68 ± 14% in the overall population and 91 ± 9% and 78 ± 14% in the 14 patients who underwent an elective procedure (Figure 5).
At the latest follow-up, there was an improvement in functional class compared with baseline, with 13 patients in NYHA class ≤II, whereas 3 patients were in class III due to radiation-induced cardiomyopathy (n = 1) or irreversible right ventricular dysfunction (n = 2).
The major finding from the present study is that transfemoral implantation of Sapien XT valves in failed mitral BP and RA is feasible in highly selected patients with an acceptable rate of procedural success and is associated with favorable early and midterm outcomes.
It should be stressed that the procedure was performed in patients at high surgical risk, due to their age (oldest patient was 90 years), their comorbidities, and the advanced stage of their heart valve disease. Indeed, all patients were in NYHA class III or IV before the procedure, and several of them had undergone previous multiple open heart surgeries. In addition, of the 17 patients who underwent the procedure, 3 patients were in cardiogenic shock and underwent rescue interventions (2 under circulatory assistance).
The first 2 cases of MVIVI were performed by Webb et al. (6) via a transseptal and a transatrial approach. However, these routes were abandoned after unfavorable outcomes due to difficulties in achieving a coaxial alignment of the THV and the mitral prosthesis. Thereafter, the team successfully performed all MVIVI via a transapical approach (10), which has been adopted as the default route by most other centers performing THV implantation in the mitral position (9,11–13). Nevertheless, our results, associated with reports from other groups (8,22), suggest that THV implantation in the mitral position may be alternatively performed using the transseptal approach. Although more challenging to obtain than with the transapical route, sufficient coaxiality and stabilization were achieved here using the combination of a TEE-guided transseptal puncture, the flexing properties of the NovaFlex catheter, and slow balloon inflation allowing adjustment during THV implantation.
The transseptal approach, which has the advantage of being less invasive than the transapical route, may therefore have a significant impact on the outcome. In addition, it may allow the procedure to be performed under conscious sedation. In the largest published series of transapical MVIVI, Cheung et al. (10) reported a 26% rate of major bleeding in elective procedures only, whereas in a smaller series of 6 patients treated with the same approach, 1 patient died after the procedure due to acute bleeding from the apical wound, and another required rethoracotomy due to hemothorax (12). In the present series, 1 fatal instance of bleeding occurred during a rescue procedure (the previously described possible rupture of the inferior vena cava). There was no other severe bleeding during the hospital course and thus no bleeding for the elective procedures. When comparing our results to Cheung’s series, the 30-day survival rates were both 100% in elective patients. Thus, our results suggest that transfemoral transseptal approach may be a valid alternative to the transapical approach in centers benefiting from a large experience in transseptal interventions.
Recently, Cullen et al. (7) reported the combination of transseptal and transapical approaches in failed mitral and tricuspid bioprostheses by using the Melody valve (Medtronic). Although such a strategy is likely to improve coaxiality and stability during THV delivery, the use of a left ventricular puncture also increases the complexity and risk of the procedure, as complication rates of 7% to 30% following transapical puncture have been reported (23,24). Accordingly, although all 9 patients successfully underwent implantation of the Melody valve within a failed mitral BP, 2 patients (22%) presented a left hemothorax.
In the present series, the only procedural failure that could be related to the lack of support concerns Patient #2, in whom emergency MVIVI was attempted for a BP implanted obliquely in a supra-annular atrial position due to massive mitral annulus calcifications. Crossing of the mitral BP with the THV was impossible and further maneuvers to advance the THV likely led to rupture of the inferior vena cava and death. It is possible that a transapical approach would have allowed positioning of the THV within the BP, but this treatment option had been ruled out upfront by the heart team due to the patient’s condition.
Mitral prosthesis size evaluation
An important point of discussion relates to sizing issues, particularly for annuloplasty rings, which have an oval shape and various degrees of rigidity. In the absence of definite recommendations, the choice relies on an integrative approach taking into account the 3 available measurement methods (TEE, CT, fluoroscopy). Despite the small sample size, differences in mitral annulus diameters according to the method used highlight the need not to rely on a single technique. In particular for stenosed rings, transcatheter prosthesis sizing should take into account not only the inner ring diameter but also the adjacent valvular tissue, which is not detected by CT scanner or fluoroscopy. The measurement of the diameter of the inflated balloon during pre-dilation may help to assess the amount of valvular tissue and avoid implanting a too large valve.
The midterm results of elective procedures are good with 91% and 78% survival at 1 year and 18 months, respectively, despite the high-risk profile of this selected population. Deaths were cardiac-related in 2 of 3 patients in this population and consisted of sudden death. The results of the procedure were good whatever the mode of valve failure (stenosis or regurgitation). Moreover, a large majority of patients experienced functional improvement with 12 out of 14 patients (86%) in NYHA class ≤2 at the latest follow-up. As well as the safety of the elective procedures, with no severe complication or death in the present series, we report encouraging midterm results.
This is a single-center, observational study in a small cohort of highly selected patients, without comparison to surgical management, which is considered the treatment of choice in this setting. However, a single-center series ensures homogeneity in patient selection and the performance of the technique, which is of particular importance in innovative procedures.
This single-center series suggests that transfemoral THV implantation for degenerated mitral surgical bioprosthesis, or repair using annuloplasty rings, is feasible and may improve early hemodynamic and functional status in selected patients, in particular those at high risk for surgery. The various reported approaches imply the absence of a consensus on the best strategy for these complex procedures at the present time. Larger series with long-term follow-up are needed to assess the potential role of this technique. Moreover, reporting of all the experience with such procedures is essential to improve results and outcomes and will certainly be useful for future mitral transcatheter valve therapies with dedicated devices.
Dr. Himbert is a proctor for Edwards Lifesciences and Medtronic, Inc. Dr. Ducrocq has received consulting/speaker fees from AstraZeneca and Eli Lilly. Dr. Nataf is a consultant for Medtronic, Inc.; and a proctor for Edwards Lifesciences. Dr. Vahanian is on the advisory board for Medtronic, Inc.; receives speaker fees from Edwards Lifesciences; and receives consulting fees from Abbott and Valtech. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- computed tomography
- mitral valve-in-valve implantation
- New York Heart Association
- ring annuloplasty
- transesophageal echocardiography
- transcatheter heart valve
- Received March 29, 2014.
- Revision received June 19, 2014.
- Accepted July 2, 2014.
- American College of Cardiology Foundation
- Thourani V.H.,
- Weintraub W.S.,
- Guyton R.A.,
- Jones E.L.,
- Williams W.H.,
- Elkabbani S.,
- Craver J.M.
- Vohra H.A.,
- Whistance R.N.,
- Roubelakis A.,
- et al.
- Cribier A.,
- Eltchaninoff H.,
- Bash A.,
- et al.
- Dvir D.,
- Webb J.,
- Brecker S.,
- et al.
- Webb J.G.,
- Wood D.A.,
- Ye J.,
- et al.
- Cullen M.W.,
- Cabalka A.K.,
- Alli O.O.,
- et al.
- Descoutures F.,
- Himbert D.,
- Maisano F.,
- et al.
- Cheung A.,
- Webb J.G.,
- Barbanti M.,
- et al.
- Seiffert M.,
- Conradi L.,
- Baldus S.,
- et al.
- Himbert D.,
- Brochet E.,
- Radu C.,
- et al.
- Himbert D.,
- Descoutures F.,
- Brochet E.,
- et al.
- Kappetein A.P.,
- Head S.J.,
- Genereux P.,
- et al.
- Zoghbi W.A.,
- Chambers J.B.,
- Dumesnil J.G.,
- et al.
- Hasan B.S.,
- McElhinney D.B.,
- Brown D.W.,
- et al.