Original Contribution

Transcatheter Aortic Valve Implantation in Patients With Pre-Existing Mechanical Mitral Valve Prostheses

Smita Scholtz, MD1;  Cornelia Piper, MD, PhD1;  Dieter Horstkotte, MD, PhD1;  Nobuyuki Furukawa, MD2;  Jochen Börgermann, MD, PhD2;  Jan Gummert, MD, PhD2;  Tanja K. Rudolph, MD, PhD1;  Volker Rudolph, MD, PhD1;  Werner Scholtz, MD1

Smita Scholtz, MD1;  Cornelia Piper, MD, PhD1;  Dieter Horstkotte, MD, PhD1;  Nobuyuki Furukawa, MD2;  Jochen Börgermann, MD, PhD2;  Jan Gummert, MD, PhD2;  Tanja K. Rudolph, MD, PhD1;  Volker Rudolph, MD, PhD1;  Werner Scholtz, MD1

Abstract: Objectives. Transcatheter aortic valve implantation (TAVI) has become standard therapy for aortic stenosis patients with intermediate or high operative risk. Treatment of patients with pre-existing mechanical mitral valve replacement (MVR) is challenging due to possible interference between the TAVI prosthesis and MVR. We present our single-center experience with this special patient cohort. Methods. A total of 1960 patients underwent TAVI at our institution between 2009 and March 2018; of these, 16 patients had pre-existing mechanical MVR. Device success and adverse events were analyzed according to the Valve Academic Research Consortium (VARC)-2 criteria. Patients were followed for at least 12 months. Results. Mean patient age was 81.5 ± 4.4 years. The patients had a mean logistic EuroScore of 37.1 ± 13.5% and STS score of 7.1 ± 3.2%. Successful valve deployment was achieved in all patients, peri-interventional stroke rate was 0.0%, and permanent pacemaker was implanted in 2 patients (12.5%). Two patients experienced major complications, with blockage of the MVR disc in 1 patient and annulus rupture in 1 patient. Hence, 30-day mortality was 12.5% and 1-year mortality was 25.0%. Conclusion. TAVI in patients with mechanical MVR is challenging and requires careful preparation and choice of TAVI device. Repositionable and retrievable devices seem to be a safer option.

J INVASIVE CARDIOL 2019;31(9):260-264.

Key words: aortic stenosis, femoral, transapical, prior cardiovascular surgery, transcatheter aortic valve implantation

Transcatheter aortic valve implantation (TAVI) has become an accepted treatment option for patients with severe aortic stenosis (AS) and expected intermediate to high perioperative morbidity and mortality.1,2 However, patients with previous mechanical mitral valve replacement (MVR) present a challenging anatomy due to the proximity between the aortic annulus and MVR. Major problems may arise during implantation, especially when interference occurs between the TAVI prosthesis and MVR and creates possible interference in the MVR function. Cases of insufficient expansion of the transcatheter heart valve,3 dislocation, or embolization4 have been reported. Therefore, such patients were excluded from the PARTNER trial1 and Medtronic CoreValve U.S. pivotal trial.5 Herein, we report on a real-world, single-center experience including 16 patients with mechanical MVR who underwent TAVI.


Patient cohort. From a total of 1960 patients with symptomatic severe AS treated by TAVI at our center between 2009 and March 2018, a total of 16 had a previous mechanical MVR. Patients were classified to be at high risk for redo conventional surgery due to their age or comorbidities, including a porcelain aorta. All cases were discussed and indications approved by the institutional heart team. All patients had coronary angiography and percutaneous coronary intervention was performed prior to TAVI, if appropriate. Multislice computed tomography (MSCT) was part of the screening in all cases and the CT data set was analyzed with the dedicated 3Mensio Structural Heart software (Pie Medical Imaging BV) in order to determine valve size and access mode. Risk scores, such as the EuroScore I, EuroScore II, and STS score, were calculated as part of the evaluation. All patients gave written informed consent and the study was approved by the institutional ethics committee.

TAVI procedure. Procedures were performed either by transfemoral or transapical approach in a hybrid operating room equipped with an Artis Zeego imaging system (Siemens). The transfemoral cases were conducted with conscious sedation and local anesthesia under fluoroscopic guidance. Transapical cases were performed in a similar manner under general anesthesia via a left lateral minithoracotomy. The following valve types were implanted: CoreValve (Medtronic); CoreValve Evolut R (Medtronic); Edwards Sapien S3 (Edwards Lifesciences); and Direct Flow Medical valve (DFM). Hemodynamic measurements were taken before, continuously during, and directly after valve implantation. Final angiogram with 30 mL contrast at a flow rate of 15 mL/s documented the final valve position and possible residual paravalvular leakage (PVL). PVL was angiographically classified into four categories:6 none/trace; mild; moderate; and severe. Heparin was administered at the beginning to keep the activated clotting time >250 seconds throughout the procedure and was antagonized by protamine at the end. All patients were on phenprocoumon/warfarin therapy due to their mechanical MVR and received an additional 75 mg of clopidogrel/day for 3 months after the procedure. Procedural success and adverse events were evaluated according to Valve Academic Research Consortium (VARC)-2 criteria.7

Follow-up. Patients were re-evaluated at 3 and 12 months after TAVI in our outpatient department. Clinical and echocardiographic findings were recorded. Thereafter, the research nurse contacted all patients regularly every 12 months to document general health status and cardiovascular events.

Statistical analysis. All data concerning baseline characteristics, procedural outcomes, and follow-up were entered into a dedicated database. Continuous variables are described as mean ± standard deviation or medians (interquartile range), if appropriate. Categorical variables are presented as numbers (percentages). The cumulative incidences of clinical events at follow-up were assessed by the Kaplan-Meier method. The SPSS statistical package was used for all statistical evaluations (SPSS 24).


Patients’ demographic and clinical characteristics are listed in Table 1. Mean patient age was 81.5 ± 4.4 years and female gender was prevalent (75.0%). The majority of patients (81.3%) were highly symptomatic with dyspnea (New York Heart Association class III or IV). Due to typical comorbidities in this age group, the mean logistic EuroScore was 37.1 ± 13.5% and STS-PROM score was 7.1 ± 3.2%. Echocardiographic and MSCT measurements are reported in Table 1; procedural characteristics and postprocedural events are reported in Table 2.

Transfemoral access was used in 13 patients and transapical approach was used in 3 patients. The TAVI prosthesis was deployed in all 16 cases: CoreValve (n = 5); CoreValve Evolut R (n = 2); Direct Flow Medical (n = 3); Sapien XT (n = 2); and Sapien S3 (n = 4). Figure 1 illustrates the three different valve types after implantation and their proximity to the MVR. Table 3 summarizes detailed procedural considerations and outcome for each patient. Nine patients had St. Jude Medical bileaflet mechanical valves and underwent uneventful TAVI. Three of the 4 patients with the Advantage bileaflet mechanical heart valves (Medtronic) underwent uneventful TAVI as well. In 1 patient with Advantage MVR, implantation of an Edwards Sapien 3 valve via transapical approach was performed but resulted in blockage of one MVR disk (Figure 2). In this case, conversion to open-heart surgery was not considered due to extremely high surgical risk, and the patient died a few days after TAVI from multiorgan failure. Retrospective analysis of the MSCT scan of this patient showed a distance from the aortic annulus to the MVR of only 0.5 mm. Three patients had monodisc mitral valve prostheses (Aortech Ultracor, Björk Shiley M, and Edwards Size 4M). In the patient with Aortech Ultracor mitral prostheses, a CoreValve prosthesis was implanted in an aortic position. Postdilation with a 26 mm semicompliant balloon was necessary due to moderate PVL. This unfortunately resulted in annulus rupture, which was most likely due to balloon over-sizing (aortic annulus by CT = 23.6 mm); a conversion to open-heart surgery was not attempted because of poor prognosis and the patient subsequently died.

Thirteen patients had no or trace PVL, 2 patients had mild PVL, and 1 patient had moderate PVL. The invasively measured mean transvalvular gradients were reduced from 35.4 ± 20.3 mm Hg to 4.3 ± 4.0 mm Hg after TAVI. The echocardiographic transvalvular aortic mean gradient was effectively reduced from 39.8 ± 12.5 mm Hg at baseline to 9.6 ± 4.3 mm Hg after TAVI. The rate of pacemaker implantation was 12.5% in this patient group. No coronary obstruction occurred. Stroke rate was 0.0% and minor vascular complication rate was 6.3%; no major vascular complications were observed. Thirty-day mortality rate was 12.5% and overall survival rate was 75.0% at 1 year.

The importance of exact preprocedural planning (especially the measurement of the distance between the aortic annulus and the MVR) was realized late. Hence, the measurement of this parameter was performed retrospectively in the majority of cases (Figure 3). Mean distance was 6.9 ± 4.1 mm, and 5 patients had a distance <4 mm. However, only 2 patients had a fatal outcome, as described above. The other 3 patients (2 treated with CoreValve and 1 treated with a Direct Flow Medical valve) had good outcomes.


Data on TAVI in the setting of a pre-existing MVR are limited. Randomized TAVI studies usually have excluded patients with mechanical MVR. The literature contains only a few case reports and case series describing variable outcomes ranging from excellent procedural success to fatal outcomes. Device embolization and interference with the MVR leaflets are potential complications. The largest cohort is presented in a multicenter registry with 91 patients including mechanical and biological MVR.8 Device embolization was the major problem and occurred in 6.7% of patients. Therefore, special care is necessary for the preprocedural planning, which should consider the following points: (1) distance between MVR and aortic annulus; (2) type of MVR; (3) choice of access mode; and (4) choice of TAVI prosthesis.

Distance between MVR and aortic annulus. Preprocedural imaging and planning are crucial in this patient subgroup. Transesophageal echocardiography allows the measurement of annular dimensions and even distance between aortic annulus and MVR, but MSCT remains the gold standard for exact measurement of the relevant aortic root structures. Correct sizing of the aortic annulus is essential because under-sizing or over-sizing of the TAVI prosthesis may lead to complications such as embolization or under-expansion.9 A distance of at least 4 mm between the MVR and aortic annulus is recommended to allow secure deployment of the TAVI prosthesis.10 Our data support this recommendation, since 2 patients in the present study who had a fatal outcome due to interference by the prosthesis had a distance <4 mm. However, our data suggest that it would be safer to treat these patients with a self-expanding device. Intraoperative transesophageal echocardiography could be helpful for careful monitoring of MVR function during self-expanding valve implantation.

Type of MVR. Mechanical mitral valves have a rigid housing cage with or without protruding pivot guards. The St. Jude Medical mitral valve seems to have the lowest risk for complications in our cohort, as 9 cases with this valve type had successful TAVI even though the distance between aortic annulus and MVR was <4 mm in 2 cases. Whether other MVR types have higher risk remains unclear, as the cohort is too small to allow such interpretation. Balloon predilation of the native aortic valve may help to anticipate the degree of balloon displacement and possible interference. Kahlert et al9 suggested balloon valvuloplasty at a relatively slow ventricular rate of about 160 beats/min in order to maintain flow through the mitral valve. However, balloon postdilation should be avoided due to the risk of annulus rupture (as seen in 1 patient in the present study) and risk of secondary valve displacement.11

Choice of access mode. Both transfemoral and transapical access modes have been described with successful outcomes. Soon et al10 favored the transapical route because the operator can stabilize both the delivery sheath and catheter and anticipate displacement. In our opinion, the access route should be chosen according to the preprocedural screening and TAVI center’s experience.

Choice of TAVI prosthesis. According to our experience, the Evolut R system, which is fully repositionable, recapturable, and resheathable, seems to be the safest option in this setting. The repositioning and even retrieval of the prosthesis is possible if any complication/interference with the MVR is observed during implantation. Asil et al12 present 6 MVR cases successfully treated with the CoreValve, which supports this point.

The Direct Flow Medical valve system is a second-generation transcatheter valve with a non-metallic design, allowing several repositioning maneuvers and complete retrieval if necessary. Three patients in the present study were successfully treated with this valve type; however, this system is no longer available. The Sapien valve allows only one-shot positioning and seems acceptable if the distance between MVR and aortic annulus is large enough. The JenaValve system (JenaValve Technology) might be a viable alternative, as the lower frame does not reach more than 2 mm into the left ventricular outflow tract and its fixation is on the aortic valve leaflets.13 A successful transfemoral case with implantation of the Symetis Acurate neo has also been described14 in which the distance between aortic annulus and MVR was only 2.4 mm.

Study limitations. This is an observational single-center study data analysis with a relatively small sample size.


Treatment of severe AS by TAVI with the pre-existence of a mechanical MVR is demonstrated to be safe and effective, but should be regarded as a challenging procedure. Further studies in larger cohorts will be necessary to confirm these results.


1. Smith CR, Leon MB, Mack MJ, et al. Transcatheter versus surgical aortic-valve replacement in high-risk patients. N Engl J Med. 2011;364:2187-2198.

2. Baumgartner H, Falk V, Bax JJ, et al. 2017 ESC/EACTS guidelines for the management of valvular heart disease. Eur Heart J. 2017;38:2739-2791.

3. Sarkar K, Speciale G, Ussia GP. Core valve implant failure in the presence of mechanical mitral prosthesis: importance of assessing left ventricular outflow tract. Catheter Cardiovasc Interv. 2015;85:920-924.

4. Maroto LC, Rodríguez JE, Cobiella J, Silva J. Delayed dislocation of a transapically implanted aortid bioprosthesis. Eur J Cardiothorac Surg. 2009;36:935-937.

5. Arnold SV, Reynolds MR, Wang K, et al. Health status after transcatheter or surgical aortic valve replacement in patients with severe aortic stenosis at increased surgical risk: results from the CoreValve US pivotal trial. JACC Cardiovasc Interv. 2015;8:1207-1217.

6. Sinning JM, Vasa-Nicotera M, Chin D, et al. Evaluation and management of paravalvular aortic regurgitation after transcatheter aortic valve replacement. J Am Coll Cardiol. 2013;62:11-20.

7. Kappetein AP, Head SJ, Genereux P, et al. Updated standardized endpoint definitions for transcatheter aortic valve implantation: the Valve Academic Research Consortium-2 consensus document. Eur Heart J. 2012;33:2403-2418.

8. Amat-Santos IJ, Cortés C, Nombela Franco L, et al. Prosthetic mitral surgical valve in transcatheter aortic valve replacement recipients: a multicenter analysis. JACC Cardiovasc Interv. 2017;10:1973-1981.

9. Kahlert P, Eggebrecht H, Thielmann M, et al. Transfemoral aortic valve implantation in a patient with prior mechanical mitral valve replacement. Herz. 2009;34:645-647.

10. Soon JL, Ye J, Lichtenstein SV, Wood D, Webb JG, Cheung A. Transapical transcatheter aortic valve implantation in the presence of a mitral prosthesis. J Am Coll Cardiol. 2011;58:715-721.

11. Vavuranakis M, Vrachatis DA, Kariori MG, et al. TAVI in the case of pre-existing mitral prosthesis: tips & tricks and literature review. J Invasive Cardiol. 2014;26:609-613.

12. Asil S, Şahiner L, Özer N, et al. Transcatheter aortic valve implantation in patients with a mitral prosthesis; single center experience and review of literature. Int J Cardiol. 2016;221:390-395.

13. Wachter K, Ahad S, Rustenbach CJ, Franke UF, Baumbach H. Transapical aortic valve implantation in patients with pre-existing mitral valve prostheses: a case report. J Cardiothorac Surg. 2016;11:133.

14. Bagur R, Pestrichella V, Montesanti R, Alemanni R, Cassese M. Transfemoral transcatheter ACURATE-neo aortic valve replacement in a patient with a previous mechanical mitral valve. J Card Surg. 2017;32:358-360.

From the 1Clinic for General and Interventional Cardiology/Angiology; and 2Clinic of Thoracic and Cardiovascular Surgery, Herz- und Diabeteszentrum NRW, Ruhr-Universität Bochum, Bad Oeynhausen, Germany.

Disclosure: The authors have completed and returned the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Börgermann reports speaker honoraria from Edwards Lifesciences, Medtronic, Maquet, and Symetis; proctoring income from Medtronic. Dr Piper reports travel compensation from Edwards Lifesciences and Medtronic. Dr T. Rudolph reports speaker honoraria from Boston Scientific, Edwards Lifesciences, Medtronic, and JenaValve; proctoring income from Boston Scientific and Edwards Lifesciences. Dr V. Rudolph reports speaker honoraria from Medtronic. Dr S. Scholtz  and Dr W. Scholtz report proctoring income, speaker honoraria, and travel compensation from Direct Flow Medical; travel compensation from Edwards Lifesciences and Medtronic. The remaining authors report no conflicts of interest regarding the content herein.

Manuscript submitted January 24, 2019, provisional acceptance given March 11, 2019, final version accepted on April 4, 2019.

Address for correspondence: Smita Scholtz, MD, Herz- und Diabeteszentrum NRW, Clinic for General and Interventional Cardiology/Angiology, Georgstr. 11, D-32545 Bad Oeynhausen, Germany. Email: akleemeyer@hdz-nrw.de