Mitral regurgitation (MR) is the second most prevalent valvular heart disease in developed countries and nearly one-third of the patients with isolated, native valve disease are diagnosed with significant MR, according to the Euro Heart Survey on Valvular Heart Disease.1
Mitral valve (MV) complex is a very complicated structure which consists of the mitral annulus (MA), valve leaflets, papillary muscles, chordae tendinae, area of insertion in the left ventricle (LV) and adjacent area in the left atrium. In fact, most of its components are part of the LV itself. Alterations in any of these components may lead to MR and more importantly, some of these mechanisms usually coexist, all contributing to MR generation.
Mitral valve surgery (valve repair or replacement) is the current standard of treatment for symptomatic and severe MR.2 Average operative mortality for these procedures, according to the Society of Thoracic Surgeons (STS) database, ranges between 2–6%.3 Mitral valve repair is more technically demanding than MV replacement (MVR), but in excellence groups it has shown to be superior to mitral replacement in terms of operative mortality, post-surgical preservation of LV function and reduction in prosthesis-related complications (such as thromboembolic phenomena). Therefore, in centres with this expertise available, it should be considered as the first treatment option in suitable MV anatomies.
Nevertheless, surgical results vary widely between surgeons and centres and operative mortality and morbidity are very dependent on patient clinical conditions. Left ventricular ejection fraction (LVEF), comorbidity index and age are all strongly related with outcomes in stable patients. Although published data is limited in patients over 70 years old, several studies show that operative mortality significantly increases with age (odds ratio 1.08 [95% confidence interval, 1.04 to 1.12] per year of increasing age as a continuous variable [p<0.0001]).4 Guyton et al. reported an operative mortality of 16.7% for isolated MVR among octogenarians versus 7% among 60–69 years old, that rose up to 33.3 versus 19.2% for combined MVR and CABG. However, octogenarians more frequently had a lower LVEF, class IV angina, New York Heart Association (NYHA) class IV and more frequently underwent emergent surgery.5 Nevertheless, there is evidence that the elderly obtain similar benefit as younger patients in restoration of life expectancy after MV surgery,4 so age alone should not preclude surgery.
This increased risk in surgical mortality in groups such as the very old or those with severe comorbidities may partially explain why a significant proportion of patients with severe MR never undergo surgery. In fact, the Euro Heart Survey registry showed that 49% of patients with severe (grade III/IV) symptomatic MR were denied surgery and the five characteristics linked to this decision were:
- a lower ejection fraction (EF);
- non-ischaemic aetiology;
- older age;
- higher comorbidity (estimated by Charlson comorbidity index); and
- grade III MR.6
Percutaneous treatment of MR is therefore a very interesting therapeutic option for these patients that would otherwise not be considered as candidates for surgery.
Percutaneous Techniques for Mitral Regurgitation Repair
Several technologies have already demonstrated proof-of-concept (see Table 1) and there are over 2,000 patients treated to date with any of these devices. Nevertheless, despite the initial optimism, it was soon found out that the road to success for percutaneous mitral repair techniques would be harder than anticipated. Clinical results to date have somehow not been as good as expected and several technical challenges are yet to be overcome. This is mostly due to the fact that MR is not just one disease but a disease spectrum, with multiple aetiologies and pathophysiologies that often coexist. In order to simplify the disease we can differentiate three important pathophysiological groups: degenerative MR, heart failure-related functional MR and ischaemic-related functional MR. The first challenge is to really understand the MV disease and to realise how difficult it is to fix this very complicated situation with just one percutaneous device.
Current techniques for percutaneous treatment of MR are based on long-standing surgical techniques. They include Alfieri leaflet plication, mitral annuloplasty, LV remodelling etc. The venues under evaluation are described below.
Percutaneous Leaflet Plication – Edge-to-edge Repair
This is a technique based on the surgical Alfieri repair, which was initially applied to degenerative MR. There is only one device in active development: the MitraClip device (Evalve Inc., Redwood City, California, US, now Abbott Mitraclip).
Through a transeptal puncture and across the MV, a steerable delivery catheter is introduced in the left atrium. A metallic clip (made of a cobalt-chromium alloy and covered with Dacron) is opened and after appropriate alignment, is advanced into the LV. The clip is then repositioned in the open position to grasp the P2-A2 scallops of the MV. Once captured, the clip is closed and the valve leaflets become plicated, generating a double-orifice MV and reducing the MR grade.
The US Experience
The Endovascular valve edge-to-edge repair (EVEREST) studies represent the initial experience with this device, with more than 600 patients enrolled to date. Anatomic inclusion criteria in the EVEREST were very stringent, as some leaflet characteristics are deemed to hamper a proper clip positioning. Therefore, inclusion criteria comprehended, among others:
- a regurgitant jet origin associated with the A2 to P2 segments of the MV;
- a coaptation length of at least of 2mm and a coaptation depth of no more than 11mm for patients with functional MR;
- a flail gap <10mm and flail width <15mm for patients with flail leaflet; and
- LVEF 25 to 60% and LV end systolic diameter 40–55mm.
Published results on peer-reviewed journals are only based on the 107 patients included in the EVEREST registry7 (55 patients in the EVEREST Phase I feasibility trial and 52 patients in the EVEREST II pivotal trial). All 107 patients had grade III or IV MR and either symptoms or compromised LV function. Procedural success (successful implantation and residual MR ≤2+) was achieved in 74% of the patients and 64% of the patients were discharged with MR ≤1+. Ten patients (9.1%) had a major adverse event (MAE) at 30 days follow-up and procedural mortality was less than 1% (non procedure-related).
Partial clip detachment occurred in 9% of the patients but there was no embolisation. Fifty of 76 successfully treated patients (66%) were free from death, MV surgery or MR >2+ at 12 months. Kaplan-Meier freedom from death was 95.9, 94 and 90.1% at one, two and three years, respectively. Kaplan-Meier freedom from surgery was 88.5, 83.2 and 76.3% during the same time periods. Thirty per cent of the patients had MV surgery during the three years after the procedure and mitral repair was successful in 84% of the patients in which it was planned, therefore showing that, in most cases, surgical options are preserved after clip implantation.
Although not published yet, results from the randomised, pivotal EVEREST II trial have been orally presented. Two hundred and seventy-nine patients with grade III or IV MR and either symptoms or LV dysfunction were randomised to device versus surgical intervention in a 2:1 fashion (178 patients to device and 95 to surgery). Acute procedural success (including MR ≤2 at discharge) was achieved in 137 of the patients assigned to device. The major finding of the study is that the primary safety end-point significantly favoured device over surgery (MAE rate at 30 days of 15% for the device group and 48% for the surgical group in the intention-to-treat analysis, p<0.05), whilst the primary efficacy end-point (freedom from death, reoperation and MR greater than 2+ at one year) favoured surgery over device (67% efficacy rate for device and 74% for surgery in the intention-to-treat analysis, p<0.05). In fact, approximately 50% of the patients in the device group had MR ≤2 at 12 months compared to about 75% in the surgical group. It is important to remark, nevertheless, that despite a better MR reduction in the surgical group, clinical outcomes were better in the device group, with 97% of patients in NYHA class I/II at 12 months compared to 88% in the surgical group. This preliminary result suggests that it is, in fact, a safety/efficacy trade-off and that we have to choose what to give priority to in each individual patient.
Finally, there is also an ongoing EVEREST high-risk registry with about 100 patients included.8
The European Experience
The device has already obtained the Conformite Europeene mark and more than 800 patients have been treated to date in clinical practice with this technology. The initial experiences in Italy9 and Germany10 have already been published in two different papers, both of which confirm the feasibility and safety of the technique: procedural success was higher than 95% in both studies and the adverse event rate below 8%, with 30 days mortality of 3.2% in the Italian experience and 2% in the German experience.
In the study by Franzen et al., anatomic inclusion criteria were looser than in the US experience. In fact, 69% of the 51 patients enrolled in their series did not fulfil the EVEREST anatomic criteria (named EVEREST - patients) due to adverse MV morphology and/or severe LV dysfunction and would therefore have been excluded from the EVEREST studies. Nevertheless, procedural and clinical outcomes between the EVEREST - patients and the EVEREST + patients (those who did fulfil the EVEREST inclusion criteria) did not statistically differ: implantation was successful in 100% of the EVEREST + patients and in 94% of the EVEREST – patients and no clinical improvement was apparent in only three EVEREST + patients and in two EVEREST – patients.
The feasibility and safety of leaflet plication with the Mitraclip device has therefore been demonstrated. Of note, the spectrum of candidates for this technology has changed with growing experience over time. Whilst initial studies included mainly patients with degenerative MR (up to 72% in the randomised EVEREST cohort), later experiences have moved on to include more functional MR patients (59% of the patients in the EVEREST high registry and nearly two-thirds in the European experiences). We have seen that these functional MR patients do as well as patients with degenerative disease, which opens this technology to a population wider than initially anticipated.
Percutaneous Mitral Annuloplasty
Percutaneous techniques mimicking surgical annuloplasty exploit the anatomical proximity of the coronary sinus (CS) to the MA either indirectly, by reshaping the CS (Carillon, Viacor or Viking/Monarc devices) and reducing septo-lateral distance of the MA, or more directly, as an anchor for the Percutaneous Sinus Shortening System (PS3 [Ample Medical Inc., Foster City, California, US]) device, or as a way to get to posterior MA (Mitralign). The CS is a fundamental structure in most of these techniques. Three different options, described below, are currently being explored.
Coronary Sinus Reshaping
This group of percutaneous MR repair techniques deploy a cinching device in the CS-great cardiac vein that can modify CS anatomy and indirectly exert force on the MA to modify its dimensions and improve leaflet coaptation. Three different devices are under evaluation:
The Carillon device (Cardiac Dimension Inc., Kirkland, Washington, US) consists of a catheter based delivery system (introduced through an internal jugular access) and a nitinol device with a distal anchor for the great cardiac vein and a proximal anchor for the proximal CS. These anchors are connected by a nitinol bridge. Tension applied on the bridge results in a reduction of the septum-to-lateral dimension of the MA. Of interest, the device can be retrieved if necessary, for example, in case of coronary flow compromise.
Results of the feasibility Mitral annuloplasty device European Union study (AMADEUS) with the first generation device (Carillon XE), were published in 2009. It enrolled 48 patients with dilated cardiomyopathy, moderate to severe functional MR, ejection fraction <40% and a six-minute walking distance between 150 and 450m. Of them, 18 patients did not receive a device because of inadequate access, insufficient MR reduction or coronary artery compromise. The degree of MR reduction in the 30 patients that did receive a device ranged from 22 to 32% at six months (using five different quantitative echocardiographic measures). Six-minute walking distance improved from 307±87m at baseline to 403±137m at six months (p≤0.001). The major adverse event rate was 13% at 30 days.11
The Tighten the annulus now (TITAN) trial used the second generation (Carillon XE2) device. Combined results from the AMADEUS and TITAN trials were presented by Dr Schofer at Transcatheter Valve Therapies (TVT) (July 2009) and Transcatheter Cardiovascular Therapeutics (TCT) (September 2009, San Francisco, US). Delivery was finally attempted in only 96 of 117 patients enrolled due to access site complications, inadequate venous size, or existence of concomitant coronary artery disease (CAD) that required surgery. Of these, the device had to be recaptured in 30 patients mainly due to insufficient MR reduction or coronary flow compromise and the implant was successful in the remaining 66 patients. Procedural mortality was 2.2% in the AMADEUS trial and 1.9% in the TITAN trial and cardiac perforation rate was 6.5% with the first generation and 0% with the second generation device. Major adverse cardiac event rates were 13 and 1.9% for the AMADEUS and TITAN trials, respectively. Focusing on the effect on coronary artery flow, the authors found that the device crossed a coronary artery in 70% of the patients but originated flow compromise in only 25%.
The Carillon was successfully retrieved in all of these patients and a second device could be successfully implanted in a more proximal location in 11 of these. Globally, the authors estimate that clinical coronary flow compromise precludes the implant in only 15% of the patients.
Percutaneous Transvenous Mitral Annuloplasty Device
The Percutaneous transvenous mitral annuloplasty (PTMA) (Viacor, Wilmington, Massachusetts, US) device consists of a multilumen catheter that is delivered into the CS, into which different nitinol rods of variable stiffness and diameter are inserted. Each rod is formed from a single Nitinol wire that is profiled to provide precise, incremental application of tension to the posterior MA and thus reduce the septal-lateral dimension of the valve, improving valve coaptation. There are two different catheters: the diagnostic catheter allows for studying the effect on MR reduction and is exchanged for a therapeutic catheter, allowing for the definite deployment when the reduction in MR is satisfactory.
Results on the temporary implant on four patients were published in 2007,12 confirming the technical feasibility of this technology. In one patient the device could finally not be implanted due to extreme angulation of the venous system. In the other three, regurgitant volume was substantially reduced (45.5±24.4 to 13.3±7.3ml) via MA anterior-posterior diameter reduction (40.75±4.3 to 35.2±1.6mm).
Safety and efficacy of the percutaneous transvenous mitral annuloplasty device to reduce mitral regurgitation (PTOLEMY I) was a feasibility study that enrolled 27 patients.13 In eight patients the diagnostic procedure could not be completed due to problems with the vascular access and the remaining nineteen patients received a diagnostic PTMA study. The device was successful in reducing MR in 13 patients (MR grade 3.2±0.6 reduced to 2±1) and PTMA implants were finally placed in nine patients (in the other four the device was unstable). There were no major events during the implant procedure. Four devices were removed uneventfully at follow-up and three patients underwent surgery due to device migration and/or loss of efficacy. PTOLEMY II is currently active in Canada and Europe.
Monarc Percutaneous Transvenous Mitral Annuloplasty Device
The development of the Monarc PTMA device (Edwards Lifesciences, Inc., evolution of the Viking device) has been discontinued. An angiographic substudy of 50 patients treated confirmed the possibility of coronary flow compromise with these devices: coronary compression was demonstrated in 15 patients, three of whom experienced clinical events at follow-up.
Despite initially high expectations, the CS approach is developing more slowly than expected, with only over 200 patients treated to date. Several factors may explain this issue:
- The devices have had to undergo redesign due to problems with the nitinol (fractures).
- There is a risk of potentially compressing the coronary arteries. Coronary sinus and great cardiac vein may cross over adjacent arteries, compromising flow and originating ischaemia. Different studies have shown that the CS crosses over the left circumflex artery (LCX) in up to 63–85% of patients.14–16
- Cardiac venous anatomy is highly variable and this may preclude device implantation in some patients. Coronary sinus, great cardiac vein and middle cardiac vein are relatively constant structures, but there is evident anatomical variability in the distribution and number of the posterior and marginal veins of the LV,14,17,18 particularly in patients with a prior history of acute myocardial infarction (AMI). Therefore, knowledge of the individual venous anatomy seems essential for planning a percutaneous approach.
- The reduction in MR has been variable and somehow less than expected. This may be in part due to the complex anatomical relation between the CS and the MA, which are three-dimensional structures that do not lie in the same plane.14 In fact, in most cases the CS does not lie directly over the MA but is located in a higher position on the LA wall and therefore, in most cases, these devices will not directly cinch the MA but rather, they will shrink the inferior portion of the LA close to it.14,15,19
Consequently, evaluation of CS and cardiac vein anatomy, as well as their relation to MA and other adjacent structures (basically coronary arteries) is of great relevance prior to planning a percutaneous approach. Multidetector computed tomography (MDCT) seems a very useful tool in patient evaluation prior to the procedure. It has shown excellent correlation with invasive angiography for the identification of cardiac venous structures in small series17 and it not only allows for evaluation of the cardiac venous system, but also for the anatomical relation between the CS and the coronary arteries and the CS and the MA.14
These techniques reduce the dimensions of the MA acting directly on it. Therefore, they should theoretically overcome the limitations inherent to annuloplasty through the CS. There are two different venues ongoing.
The first one implants several anchors in the posterior MA and connects them with a wire/suture that applies tension in order to bring them closer and reduce posterior MA dimensions. This includes two different technologies, with about half a dozen patients treated to date:
In the case of the Mitralign device (Mitralign, Tewksbury, Massachusetts, US) the LV is accessed through a retrograde approach and a fork-like device is placed on the ventricular portion of the posterior MA. A catheter is also placed in the CS through a venous access. There are magnets in the extremes of the fork-like structure and on the CS catheter, which once coupled, stabilise the whole system. Next, three anchors are delivered and sutured to the ventricular side of the MA with stitches applied from the ventricular to the atrial side of the MA. These anchors are connected with a suture and when tension is applied to this suture, the MA is cinched. There are ongoing feasibility studies in humans in Paraguay and Germany.
Accucinch Annuloplasty System
In the Accucinch Annuloplasty system (Guided Delivery Systems [GDS, Santa Clara, California, US]), anchors, connected by a wire, are similarly placed in the subannular region of the LV through a retrograde approach. Technical feasibility studies in animals have demonstrated a reduction in the MA perimeter when tension is applied to this wire and a feasibility study in humans was started in Europe.
The second approach is in pre-clinical development and is based on the application of radiofrequency energy to the atrial region of the posterior MA (QuantumCor device, San Clemente, California, US) in order to originate scarring and constriction of the tissue treated. Transeptal access to the LA is necessary to introduce a catheter with a loop tip on which the radiofrequency electrodes are placed. No device is left in place in the atrium.
In addition, ReCor Medical (Ronkonkoma, New York, US) is working on a new technology that utilises therapeutic ultrasound to heat and shrink the collagen-rich MA. Through a transeptal access, a non-occlusive ‘floating’ balloon with the capability to beam ultrasound is positioned in the mitral orifice. The heat generated by the ultrasound acts on the collagen present on the MA, inducing its shrinkage and reducing MA dimensions.
The PS3 was based on the concept of atrial remodelling in an attempt to reduce septal-to-lateral distance of the MA. Nevertheless, this research programme has been discontinued due to economical issues.
Left Ventricular Remodelling
The goal is to compress the LV to geometrically remodel the chamber, treating the target problem in functional MR that is ventricular distortion. Subsequently, the MA is also distortioned and compressed.
The percutaneous iCoapsys technology is based on the off-pump Coapsys surgical system (Myocor, Maple Grove, Minnesota, US), which places pads on both sides of the LV connected by a cord that traverses the LV cavity in a submitral position, approximately 2cm below the MA, therefore applying tension to the MA and basal LV wall. Initial surgical data indicated that the device can be implanted safely with acute reduction in MR and positive LV remodelling.21 Results of the Randomized Evaluation of a Surgical Treatment for Off-Pump Repair of the Mitral Valve (RESTOR-MV) study, comparing surgical annuloplasty versus Coapsys LV remodelling in 165 patients undergoing CABG, were presented at the American College of Cardiology (ACC) (2010, Atlanta, Georgia, US) by Dr Grossi. Remarkably, the Coapsys group showed less mortality and a better event-free survival than surgical annuloplasty (13 versus 23% and 85 versus 71%, respectively, p<0.05 for both). Unfortunately, this study was terminated early because of funding issues and the research programme has been abandoned.
The percutaneous device is implanted transpericardially via a sub-xiphoid approach (similar to a pericardial drainage procedure), with insertion of a sheath that allows for the deployment of two pads that are clipped to the LV wall. The wire does not traverse the LV but runs over the lateral surface of the heart, connecting both pads. Technical feasibility was tested in animals and there was a study in humans planned, however this programme has also been stopped due to the economic crisis.
Hybrid Treatment of Mitral Regurgitation
These techniques combine surgical implantation of an annuloplasty ring and later adjustment of the size of the ring to optimise MR reduction via radiofrequency (RF) application (Dynamic Annuloplasty Ring System, Micardia Corp., Irving, California, US) or mechanical traction (Mitral Solutions Inc. annuloplasty ring, Fort Lauderdale, Florida, US). Adjustment can be performed in the operating room after ring implantation or at a later stage in the catheterisation laboratory.
The Dynamic Annuloplasty Ring is available in different sizes (28 to 36mm) and shapes (‘C’ form or ‘D’ form). It has a basal state similar to other conventional annuloplasty rings, which may be modified to an active shape through the application of radiofrequency. Application of RF is able to reduce anterior-to-posterior MA distance in 0.5 to 3mm and inter-commissural distance in 1 to 3.5mm.
The Mitral Solutions Annuloplasty ring is available in just one size. Adjustment is currently achieved off-pump through a superior pulmonary vein with a catheter that matches specific anchors in the ring. Catheter rotation reduces the perimeter of the annuloplasty ring and therefore MR. A similar catheter to be deployed percutaneously via transeptal puncture is also being developed.
Transcatheter Mitral Valve Implantation
Finally, three companies are working in the development of a mitral prosthesis to be delivered percutaneously (Endovalve, CardiAQ and Lutter valve), although they are in the very early stages of the research. In fact, the Endovalve programme has been refocused and the company is now working on a direct atrial access instead of the initially planned transeptal access. Designing a prosthesis that can be percutaneously deployed in a stable position within the MA without interfering with the LV outflow is a real challenge.
Many patients with symptomatic, severe MR never undergo surgery due to a deemed excessive surgical risk and percutaneous approaches to MR repair are therefore an exciting alternative for these patients. Several technologies are being investigated, most of them based on long-standing surgical techniques. Over 1,000 patients have been treated worldwide with the Mitraclip system for leaflet plication. This device has gained the CE mark recently and there has been a shift in the spectrum of candidates suitable for this technique, as recent experiences have shown that adverse mitral anatomies and functional MR patients may also benefit. Devices for indirect annuloplasty through the CS raised great expectations because they seemed so user-friendly, but the anatomic relation between the CS, the MA and the coronary arteries is very complex and procedural and clinical outcomes have not been as good as expected. Anatomic evaluation seems crucial in order to select the most suitable patient for the procedure and cardiac computed tomography angiography is probably the best tool for patient selection. Several other projects are under development and evaluation, but the recent economic crisis has affected many of the research programmes that have been temporarily or permanently halted.
Mitral regurgitation is a complex disease with many aetiologies and pathophysiologies and therefore it seems unlikely that a single device can fix the problem. The first challenge is in fact to really understand MV disease, then, a combination of techniques will be necessary for a satisfactory MR repair. Close collaboration between interventional cardiologists, cardiac surgeons, imaging cardiologists, basic scientists and the industry is essential for these technologies to progress and become a real alternative for this group of patients.