Mitral valve disease is the most frequent structural heart disease.1 Surgical mitral valve replacement is standard of care in patients with mitral valves beyond repair. This can be due to mitral valve stenosis and complex regurgitation. Surgical repair or replacement is associated with a high surgical risk in symptomatic patients with heart failure, comorbidities, poor LV function, frailty and other comorbidities. Transcatheter therapies have emerged as alternative treatment options. Several surgical procedures have inspired transcatheter devices to treat mitral valve disease, from leaflet repair, to annular remodeling, to valve replacement. Percutaneous mitral valve repair is getting established in standard clinical practice, with transcatheter edge-to-edge repair performed in more than 50 000 patients worldwide. For transcatheter mitral valve replacement (TMVR) about 30 devices are currently in development, 6 of them in clinical use with an overall estimated 200 TMVR performed, and feasibility trials running worldwide. This article summarizes the currently available devices for transcatheter mitral valve replacement highlighting specific challenges and opportunities.
Clinically, there are two forms of mitral regurgitation (MR), functional (FMR) and degenerative (DMR). Functional mitral regurgitation (FMR) involves dilatation of the annulus, leaflets’ retention due to left ventricular dilatation with tethering, and lack of coaptation. Today a high percentage of patients with severe FMR do not undergo surgical treatment due to high surgical risk, although they are highly symptomatic. Degenerative mitral regurgitation is caused by structural alterations of the valve, such as abnormalities on leaflets or chordae. Myxomatous alterations (Morbus Barlow) of the leaflets, deterioration or clefts result in insufficiency of the valve, sometimes combined with stenosis. The chordae can be prolonged or detached, leading to prolapse and flail of the leaflets.
Mitral stenosis occurs almost exclusively after rheumatic fever or post endocarditis and the prevalence in Europe and North America has significantly dropped since 1900,2 but there still is a high prevalence in the rest of the world, e.g. 1% in a current screening study of Samoa.3
Success of transcatheter solutions
15 years ago, in 2002, the first in-human transcatheter aortic valve replacement (TAVR) was performed.4 Today, 45% of isolated aortic valve replacements are performed as transcatheter procedures in Europe.5 This was only the beginning of a shift from open surgery to transcatheter solutions. Transcatheter mitral repair devices have been introduced into clinical practice successfully since 2008, and the development of new devices is continuing. The next milestone for transcatheter therapies is to bring TMVR into clinical practice.
Patient selection and indications
Currently only 1 out of 40 patients with moderate to severe MR receives surgical treatment in the US.6 This is due to a high threshold for structural interventions, especially in patients with a high risk for open heart surgery. For patients with severe FMR, severe non-repairable DMR with comorbidities and frailty, as well as for patients with severe mitral valve stenosis, or previous cardiac surgery, TMVR has the potential to become the new standard of care by reducing the peri-procedural risk while achieving equal results. Possible advantages of TMVR versus transcatheter repair therapies include the elimination of MR and the higher predictability and reproducibility due to ease of use.
FMR is caused by left ventricular dysfunction and dilatation. It is secondary to heart failure caused by conditions like ischemic heart disease. Therefore, surgical therapy of FMR comes with a high risk which is associated with the cause for FMR. Treatment of FMR is merely symptomatic. Open surgery, when feasible, mainly addresses the cause of heart failure, e.g. revascularization, aiming at supporting LV remodeling, and may be combined with annular ring implantation to prevent further dilatation of the annulus (table 1). However, patients undergoing this surgery have mortality rates of up to 20–50% in a 1- and 5-year follow-up,7 as well as a low quality of life with frequent hospitalizations for heart failure decompensation. FMR therefore represents a very interesting target for transcatheter solutions.
Interventional edge-to-edge clip MitraClip (Abbott Vascular, Santa Clara, CA, USA) presents a safe and efficient way to reduce MR, as shown in the EVEREST and ACCESS trials.8, 9 Further transcatheter MR interventional therapies, e.g. targeting the mitral annulus, are in promising clinical and preclinical development. But in patients with severe heart failure, the difference between reducing MR and eliminating MR could be relevant.
For severe DMR, surgical repair is the standard of care (table 1). The indication is given by echocardiography alone. The repair has a curative intention and prevents future heart failure. Whenever DMR progresses to heart failure, is associated with other heart diseases causing heart failure, or in patients with high surgical risk due to other causes, TMVR has a significant therapeutic role.
Indication for cardiac surgery in mitral stenosis is driven by clinical symptoms. Compared with MR, mitral stenosis in most cases requires the valve to be replaced with a biological or mechanical valve, depending on age and comorbidities (table 1). As we have seen happening with TAVR, this will be performed mainly interventionally. In case of severe annular calcification (MAC) with high risk for annular rupture we are faced with a challenging situation. This needs to be seen as a separate entity and may be considered for TAVR, together with cases of previous surgical ring implantation (MITRAL study10).
The target age group will certainly be influenced by the durability of the TMVR devices.
Sizing can be a limiting factor in FMR cases with extreme annular dilatation. Valve-in-valve procedures are certainly feasible in respect of resulting transvalvular gradients.
Kidney insufficiency of stage III or IV can be a limiting factor due to necessity of preoperative CTA scan for planning and periprocedural guidance by contrast. But further down the road, evaluation and even implantation could become guided by echocardiography only.
The mitral: a complex valve
The mitral valve is more complex than the aortic valve. It is composed of a variety of components: the annulus, two distinct leaflets, different types of ventricular chordae and the papillary muscles. Even nowadays, thanks to new imaging modalities, we discover new details such as the relevance of secondary chordae in the early opening phase.11
Anatomical challenges (table 2)
The annulus is double saddle D-shaped and the annular size changes during the contraction circle by 11%.12 The location of the mitral valve close to the aortic valve and the left ventricular outflow tract (LVOT) presents the challenge of possible LVOT obstruction after TMVR. Due to the proximity of aortic valve and mitral valve, implants in mitral position can lead to obstruction of the left ventricular outflow tract (LVOTO). This problem has already been reported in several cases. Septum hypertrophy and a small mitral-aortic angle are predictive factors for this complication. Imaging based prediction algorithms are available, as the post implant “Neo-LVOT” can be calculated and should be taken into account.13
A low profile and asymmetric shape of the implant reduces the risk and is seen in various device designs (e.g. Caisson device with a D-shaped outer Nitinol stent). Another approach to prevent LVOTO is the intentional laceration of the anterior leaflet as is currently being evaluated in the LAMPOON study.14
Anchoring of the device plays a major role in the stability and long-term outcomes of TMVR. The physiological movement of the whole mitral valve apparatus during a heart cycle, including the change of annulus size, results in high mechanical stress. In contrast to the aortic valve there is no stable anatomical fibrotic ring.
Abroad variation in devices exist that address this issue. CardiAQ (Edwards) uses the native leaflets and grasps them with anchors. There are also an apical plug system (Tendyne) and a bottle-cork shaped double stent available.
Sealing is difficult because of the double saddle, D-shaped asymmetric anatomy of the mitral valve. A mismatch between the preshaped frame of the TMVR and the native mitral annular shape is responsible for the occurrence of paravalvular leakage (PVL). To prevent this from happening, most devices use an atrial sealing skirt (e.g. Intrepid), while CardiAQ (Edwards) uses an intra-annular sealing skirt as well as the native leaflets. Intrepid also uses an outer deformable stent for sealing.
The valves are designed in such a way as to prevent valve thrombosis. The native mitral valve leaflets and chordae partially lose their function after TMVR and the sealing skirts are prone to enhance thrombosis formation. The flow velocity across the mitral orifice is lower than across the aortic due to the larger orifice area. What form of post-procedural anticoagulation is necessary and for what duration has to be defined for each device, due to the uniqueness of their designs.15
Durability of the implanted valve is a determining factor regarding the target age group. Due to the physical stress to the leaflets, durability will become a relevant point of interest in the future. Long-term outcomes will depend on improved leaflet pre-treatment and prosthesis design.
The access by which to perform implantation currently – still – is mainly transapical. However, feasibility of transfemoral solutions has been demonstrated. They will be of benefit for patients at higher surgical risk with progressed heart failure and reduce periprocedural complications. The first transseptally delivered TMVR are expected in 2018.
Regarding implant sizing we lack overall experience, however, the completion of current clinical studies will better depict the anatomical needs. Some of the devices at present use a one-size approach. But most companies plan to bring at least 3–5 different sizes to market. The role of pre-procedural sizing is fundamental. Abdelghani M et al. suggest the usage of Dmean as average of intercommissural + septolateral distance measured in CTA.16
Current device concepts
Currently there are over 30 devices undergoing the research and development phases; six devices are being implemented into the clinical phase (table 3). In the following we describe the present status and outcomes of these TMVR solutions, highlighting their main characteristics.
This valve has atrial anchors and hooks that grab the native leaflets on the ventricular side. It has an inflow band, an inflow skirt, and an annulare sealing skirt. The leaflets are bovine, initially they were porcine. The access is transfemoral.
The Tendyne valve anchors on the apex, therefore a transapical access is necessary. Currently it is single size for an effective orifice area (EOA) >3cm2. Sealing is achieved by the atrial outer stent and the apical pad. The outer stent is D-shaped. The first 1-year follow-up data on 30 patients showed no procedural deaths or strokes, hemolysis (1pt) and PVL (1pt). NYHA class was reduced in 60% of the patients. MR was absent at 1 year follow-up in 94,7% (Muller D, TCT 2017, Denver).
Interpid (former Twelve)
Double stent design with an atrial skirt. Access is transapical and transseptal. Sizes are 43, 46, 50mm outer annulus. Bovine pericardium 27mm tricuspid on inner stent. The first results on 49 patients (Bapat V, TCT 2017, Denver), showing mild residual MR in 8 patients and PVL in 3 patients. 30-day mortality was 14%, caused by apical bleeding (3pts) and malposition (1pt). Transfemoral approach is expected in 2018.
The Tiara valve is D-shaped with a Nitinol based stent, with anterior and posterior anchors capturing the native leaflets. The sealing is done with an atrial skirt. The leaflets are of bovine pericardium. Current follow-up shows no PVL, no LVOTO, 1 PM implantation in 34 patients (Cheung A, TCT 2017, Denver).
A two-component device composed of the valve on Nitinol stent with atrial sealing skirt and subannular ring fixation that anchors the device on an annular level (stent indentation). Recent results showed LVOTO in 1/12 patients (Piazza N, TCT 2017, Denver).
Transseptal delivery system 42Fr (29mm valve). For mitral annulus 30– 44mm. Early feasibility study paused because of forming of blood clots. Sealing and anchoring is done by an atrial flange and ventricular paddles. 2-year follow-up showed good results.17
Indications for TMVR should be evaluated and discussed in interdisciplinary heart teams. Clinical indication is instead based on the overall clinical profile, the risk score and the valve pathology assessment from a transthoracic echocardiography.
Pulmonary hypertension of more than 2/3 systolic pressure as well as significant right heart failure are relative contraindications and should be excluded by echocardiography and right heart catheterization.
The pre-procedural evaluation depends on the singular devices. Analogous to TAVR, the preparation includes a cardiac computer tomography of the chest with contrast (CTA) for implant sizing and LVOTO risk evaluation. When performed transfemorally, a CTA is indicated.
Conclusion and outlook
Despite many promising device concepts, development of TMVR is remarkably slower than of TAVR. This is mainly due to the anatomical challenges, including valve anchoring and sealing as well as LVOT obstruction.
First-in-man, early feasibility studies and compassionate use experience in almost 200 patients show a reliable feasibility of TMVR using several devices. Safety concerns as evaluated at midterm follow- up must be further assessed by enlarging the clinical experience. The approach remains mainly transapical but a shift to a transfemoral-transseptal approach is the focus of every development program.
Clinical outcomes include complications due to access, delivery, positioning and fixation, as well as due to impact on surrounding structures or device performance. The important role of a detailed pre-procedural planning by echocardiography and CTA is demonstrated. The peri- procedural imaging needs to be of high quality for successful TMVR. The periand post-procedural haemodynamic management will play an important role as to clinical outcomes. We are at a very early stage of the TMVR experience, and further studies are necessary to provide deeper insight into safety, efficacy and durability of these new devices.
The current indications of severe MR with high surgical risk as well as stenosis will need to be elucidated with evidence and extended by experience. This is even more relevant since DMR and severely calcified stenosis currently are excluded from most ongoing studies.
TMVR most certainly has the potential to become the standard of care for mitral stenosis and mitral regurgitation beyond repair. And at best, it may meet the expectation to be a one-for-all simple and reliable therapy for mitral valve diseases.