Skip to main content

Percutaneous Mitral Valve Interventions: Overview of New Approaches

Clinical Review

Percutaneous Mitral Valve Interventions: Overview of New Approaches

Author Information:
Mehmet Cilingiroglu, MD, FESC, FACC, FSCAI and Ted Feldman, MD, FESC, FACC, FSCAI


Percutaneous therapy has emerged over the past few years as a rapidly progressing field of development for the treatment of mitral regurgitation. Most of the new percutaneous approaches are modifications of previously-described surgical techniques that have been used for mitral valve repair for several decades. The main surgical approaches to mitral valve repair are annuloplasty and leaflet repair. Catheter-based devices mimic these surgical approaches with the aim of achieving outcomes similar to surgery, with less procedural morbidity and mortality as a consequence of their much less invasive nature. Some of these concepts include mitral valve annulosplasty via either the coronary sinus or directly from retrograde left ventricular access, and leaflet repair using modifications of the surgical edge-to-edge technique. Many of these percutaneous approaches have been accomplished with several devices and have been used with enough acute success to demonstrate proof of concept. Here, we will discuss these novel catheter-based therapies for the treatment of mitral regurgitation.


Key words: mitral regurgitation, percutaneous valve repair, surgical valve repair, mitral annuloplasty

Normal mitral valve closure, which prevents systolic backflow of blood into the left atrium during left ventricular systole, depends on the complex interaction of each of the components of the valve apparatus; the left atrial wall, annulus, mitral valve leaflets, chordea tendineae, papillary muscles and the left ventricle (LV).1,2 Abnormalities in the anatomy and function of any one of these components can lead to incompetence of the valve, resulting in mitral regurgitation (MR). Normal mitral valve closure is thought to be a passive process mediated by flow deceleration though the valve with generation of vortices on the ventricular side of the leaflets in conjunction with an adverse pressure gradient between the LV and left atrium.3–5 During systole, the mitral annulus moves toward the LV apex, whereas contraction of the LV myocardium underlying the posterior annulus results in a decrease in annular area by approximately 25%.6 When MR is caused by pathology specific to the valve leaflets such as myxomatous degeneration, it is referred to as degenerative MR. On the other hand, mitral regurgitation that is a consequence of ventricular or annular abnormalities including cardiomyopathy or left ventricular or papillary muscle dysfunction as a result of ischemia, is referred to as functional MR. MR, a common finding, is clinically significant in part as a result of its detrimental effect on LV function. Patients with mild MR may remain asymptomatic for many years. However, moderate-to-severe MR leads to either gradual or, at times, acute ventricular dilation and contractile dysfunction with ultimate development of symptoms and left ventricular failure.7 In patients with symptomatic moderate-to-severe or severe MR, surgical intervention is recommended. Surgery is also recommended for asymptomatic patients with severe MR who have evidence of left ventricular dysfunction and/or new onset atrial fibrillation or pulmonary hypertension.8 The preferred surgical therapy for MR is mitral valve repair.9,10 Patients undergoing mitral valve repair have demonstrated improved short- and long-term outcomes compared to patients who receive valve replacement with mitral valve prosthesis in prior nonrandomized clinical experience.11 In spite of these more favorable results with repair and the obvious limitations of prosthetic valve replacement, mitral valve repair is still being performed in only about half of patients undergoing mitral valve surgery.12 For functional MR, the most common surgical intervention is annuloplasty.13 Annuloplasty for functional MR is usually performed in association with coronary artery bypass surgery. Degenerative MR is typically treated with leaflet repair, and has better long-term outcomes than isolated annuloplasty for functional MR. These surgical procedures, though effective at reducing MR, are associated with the expected mortality and morbidity of open-heart surgery. Thus, the morbidity and mortality associated with open-heart surgery keeps many patients with clinically significant MR from receiving mitral valve surgery, especially when they are older or at higher risk for surgery. Thus, these basic surgical concepts of annuloplasty and leaflet repair have been adapted for the development of novel catheter-based percutaneous approaches with the goal of expanding the pool of patients who might benefit from a less invasive approach to a reduction in MR. Advances in both technique and development of novel devices have led to a variety of methods to treat MR using a percutaneous route (Table 1). Percutaneous approaches for mitral valve repair have been recently implemented in clinical trials for treatment of MR.

Leaflet Repair with the Evalve MitraClip™

Among all of the percutaneous approaches for MR, the greatest amount of experience has been developed with the Evalve MitraClip™ (Abbott Laboratories, Abbott Park, Illinois) (Figure 1). Initially described by the Italian surgeon Ottavio Alfieri for surgical repair of anterior leaflet prolapse in 1991, the edge-to-edge technique involves apposing the middle scallops of the anterior and posterior leaflets with a stitch creating a so-called “dual” or “double-orifice” mitral valve.14 The apposition of the leaflets leads to effective reduction of MR. The tissue bridge that is created when the sutured segment heals may also prevent further annular dilatation (Figure 2). In addition, the anchorage of the leaflets together through the chordae tendinea may exert a supporting or tethering effect on the LV, and help to counteract the progression of ventricular remodeling and recurrence of MR. This approach has been successfully used to treat MR resulting from prolapse of either one or both leaflets involving the mid-segment of the valve, as well as for selected patients with functional regurgitation secondary to ischemia or cardiomyopathy.15,16 In a recent report from Alfieri’s group, central double-orifice repair was performed in 260 patients followed for up to 7 years,17 survival was 94% and freedom from reoperation was 90%. The clinical success and simplicity of this technique has thus prompted interest in the development of a catheter-based technology that would enable the interventional cardiologist to perform a percutaneous endovascular repair in the cardiac catheterization laboratory. A percutaneous method to accomplish the double-orifice repair was subsequently developed (Figure 3). A technique that facilitates valve repair without adjunctive annuloplasty potentially allows for more physiologic ventricular contraction, as annular motion contributes significantly to ventricular dynamics. By fastening the leaflets together, edge-to-edge repair ensures a fixed area of effective coaptation. With this technique, the remainder of the coaptation line closes physiologically without disturbing subvalvular or annular architecture and function. This might significantly impact LV hemodynamics and avoid the compromise in cardiac output frequently seen in patients after mitral valve replacement. The system consists of a 24 Fr steerable guide catheter with a 22 Fr distal end, a separately steerable clip delivery system and an implantable clip. Mounted on the distal end of the delivery system, the MitraClip, is a two-armed, soft tissue-grasping and approximating device that, when closed, has an outside diameter of 15 Fr, and in its grasping position, it spans 20 mm. It is designed to vertically hold up to 8 mm of leaflet height and 4 mm of width to recapitulate the Alfieri’s surgical approach in length and width of tissue apposition. It is constructed of implant-grade cobalt-chromium alloy and the clip is covered with polyester. On the atrial side of the clip are “grippers”, which are small, flexible, multipronged friction elements that appose and stabilize tissue against the clip arms. After tissue is captured between the arm and the gripping components, the clip is closed and deployed in a locked position. The clip is designed to promote leaflet-to-leaflet healing around and into the device in order to allow the development of a tissue-supported permanent leaflet approximation. The clip can be opened, the leaflets released, and the clip can then be repositioned using real-time echocardiographic assessment to attain the best possible result before final deployment. Percutaneous edge-to-edge repair with the MitraClip in humans is performed in the catheterization laboratory with a combination of echocardiographic and fluoroscopic guidance under general anesthesia. Access to the left atrium is obtained via the femoral vein and a 24 Fr guide catheter is placed across the interatrial septum using the transseptal approach. A multidirectional steerable system is introduced through the guide and using a series of iterative steps, positioned above the mitral leaflets at the location of the regurgitant jet. The two arms of the clip are opened in the left atrium once the clip is aligned with the long axis of the heart near the origin of the MR jet. Under echocardiographic guidance, the arms are rotated until they are perpendicular to the line of leaflet coaptation. The open, aligned clip is advanced across the mitral orifice and then retracted to grasp the leaflets during systole. The atrial grippers with frictional elements are lowered onto the atrial side of leaflets, approximating (capturing) and stabilizing leaflet tissue against the arms on the ventricular side of the leaflet. When a double orifice and reduction in MR jet are demonstrated by echocardiography, the clip arms are closed in a locked position. The acute result is then evaluated with two-dimensional color-flow and pulsed Doppler imaging. If inadequate reduction in MR is the case, the leaflets can be released and the clip repositioned. In some cases, a single clip is not sufficient to adequately decrease the magnitude of MR. In that situation a second clip can be placed adjacent to the first clip on the side of the residual MR jet. Repeat hemodynamic and echocardiographic assessments are then made to verify final results, at which point the delivery system is removed. Chronic animal studies in a porcine model showed that at 4 weeks, the entire clip was encapsulated in a layer of tissue. There was evidence of tissue deposition and leaflet-to-leaflet healing. At 12 weeks, tissue encapsulation was further developed with leaflet-to-leaflet bridging between the arms of the clip. At 24 weeks, development of a mature tissue bridge was present (Figure 2).18 The procedure requires a dedicated team of physicians. In addition to an interventional physician, a skilled echocardiographer and an anesthesiologist all work together during the procedure. Clear communication between the interventionalist and the echocardiographer providing the transesophageal echocardiographic guidance is critical to achieve an optimal result.19 While evaluating hemodynamics, changes in blood pressure, afterload and vascular tone need to be kept in mind in the anesthetized patient. The Mitraclip system has been successfully evaluated in a U.S. phase I clinical trial (Endovascular Valve Edge-to-Edge Repair Study; EVEREST I).20,21 The study population consisted of surgical candidates with moderate-to-severe or severe MR and clinical symptoms. Asymptomatic patients were eligible if echocardiographic evidence of LV dysfunction was present. American College of Cardiology/American Heart Association (ACC/AHA) guidelines criteria for surgical intervention were followed and patients were closely screened using American Society of Echocardiography quantitative methods for assessment of MR severity.22,23 All of the echocardiographic exams were reviewed in a core laboratory. Echocardiographic anatomic inclusion criteria included specific leaflet morphologic findings to determine if adequate tissue was available and the location of the MR jet origin (from the central two-thirds of the line of coaptation) (Figure 4). A Phase I trial has been completed in a cohort of 55 patients. Registry data from a nonrandomized group of 107 patients,21 as well as outcomes in a high-risk cohort of 78 patients, have been reported. The primary endpoint of the EVEREST I trial was safety at 30 days. Safety was defined as freedom from death, myocardial infarction, cardiac tamponade and cardiac surgery for failed clip or device, clip detachment, permanent stroke or septicemia. A clip was placed successfully in about 90% of cases. Of those who achieve acute procedural success, defined as adequate reduction in MR without a procedural complication, two-thirds were alive and had no need for repeat procedures after 2-year follow-up. In a high-risk group of 78 patients, in addition to improvement in New York Heart Association functional class, favorable ventricular remodeling with a decrease in LV systolic and diastolic dimensions and reduction in the need for hospitalizations among high-risk patients have been demonstrated. Compared to match-controls who were also considered high risk for surgery, there was improved 1-year survival. A phase II randomized trial (EVEREST II) comparing the clip with surgical therapy in 279 patients has been completed. In this study, eligible patients were prospectively randomized to percutaneous repair versus surgery using 2:1 allocation and are undergoing clinical and echocardiographic follow-up. The trial is prospective, core lab-evaluated and event-monitored. None of the past surgical trials of mitral repair therapy have been prospective, with intention-to-treat methods or core labs. Thus, in most surgical reports, the proportion of patients for whom repair is intended but in whom replacement is ultimately performed is not clearly defined. The MR reduction results of surgical mitral repair have not been assessed using objective criteria through an echocardiography core lab with quantitative MR grading. Thus, EVEREST phase II trial is groundbreaking not only in the development of the percutaneous therapy, but also in defining the results of the mitral valve surgery in a multicenter trial. At the end of 2009, patient enrollment was completed and the 1-year follow-up time point had been reached for the entire group. The results of the trial have not been reported at the time of this writing. The MitraClip has been approved in Europe, and early use shows a pattern weighted towards high-risk patients with about two-thirds with functional and one-third with degenerative MR. Patients treated in the initial European experience have usually been referred by surgeons. Potential limitations of this technique include large device size (a 24 Fr guide catheter), technically demanding procedures and uncertainty about the long-term-durability of the results since results beyond 3 years have yet to be reported. Surgical leaflet repair is almost always done in conjunction with an annuloplasty, and the surgical community sees the lack of annuloplasty as an important limitation of this approach. In addition, the feasibility and efficacy of this technique is limited to specifically suitable anatomy and is not applicable in subsets of patients with extreme pathology of leaflets including rheumatic disease or ruptured papillary muscle. The MitraClip™ appears to serve a clear unmet need in patients who are high risk for surgery. The role of this therapy for surgical candidates will be defined by the randomized EVEREST II trial.

Percutaneous Mitral Annuloplasty

Annuloplasty techniques using percutaneous approaches are either indirect or direct. Indirect methods are based on the relationship of the coronary sinus to the mitral annulus. The Coronary sinus parallels the posterior mitral annulus, and thus provides a route for annuloplasty device delivery.

Coronary Sinus Annuloplasty

Functional MR is primarily the result of incomplete coaptation of normal leaflets as a result of progressive mitral annular dilation, alteration in LV geometry and/or papillary muscle dysfunction.24,25 This secondary MR frequently accompanies ischemic or chronic heart failure and triggers a vicious cycle of continuing volume overload, ventricular dilation, progression of annular dilation and increased LV wall tension, and thus further worsening of MR and heart failure.26,27 In addition, this functional disorder can eventually lead to loss of systolic sphinteric contraction of the mitral annulus and retraction of the chordea tendinea with fibrosis. Functional MR secondary to LV dysfunction is a significant clinical problem, representing an independent and strong predictor of mortality in patients with both ischemic and nonischemic heart failure.28 Although medical therapies play a role in alleviating symptoms, they do not arrest or slow the deterioration of LV function caused by chronic MR. In the absence of structural mitral valve abnormalities, the dimension of the mitral valve annulus is the most significant determinant of mitral leaflet coaptation, regurgitant orifice area and subsequent MR. Hence, the dominant modality of the current surgical approach is insertion of a mitral annuloplasty ring that reduces annular circumference and pushes the posterior leaflet forward for better coaptation, thereby decreasing MR.24,29 However, relatively high mortality and morbidity rates may render surgical treatment prohibitive in the majority of heart failure patients.29 Furthermore, clear survival benefits from surgical annuloplasty for functional MR have yet to be demonstrated.30,31 Thus, several less invasive percutaneous annuloplasty modalities via the coronary sinus have become the target of clinical research. The anatomic proximity of the coronary sinus to the posterior mitral annulus, coupled with ease of percutaneous access to this large vein, offers a basis for the development of less invasive catheter-based mitral annuloplasty.32,33 Some of the initial permanent human implants of a coronary sinus annuloplasty device were reported using the Edwards Monarc™ device34 (Edwards Lifesciences, Irvine, California). The prototype device consisted of three elements: two anchor self-expanding stents, one to be deployed distally in the great cardiac vein and the second, larger anchor is deployed in the coronary sinus ostium proximally, and a spring connector bridge (Figure 5). The bridge element is held in the open position by absorbable suture material. The mitral annular circumference decreases as the bridge shortens over a period of 3–6 weeks with dissolution of the absorbable suture. In an animal model, the average reduction in anterior-posterior distance of the mitral annulus was 24% at 9 weeks, resulting in significantly improved MR. Among the first 5 reported patients, the bridge fractured in 3, without any clinical complications, but with loss of efficacy. This has led to redesign of the device, and the trial has been resumed with implants in an additional 80 patients. In this second group, reductions of 1 to 2 grades of MR were seen in 60% of patients. Coronary compression that was clinically evident occurred in 3 with an acute MR resulting in death in 1 patient, and device fractures were noted in a few. Results in a cohort of 70 patients have been presented.35 These initial data confirm feasibility, but also the challenges of percutaneous coronary sinus-based annular remodeling in humans, and demonstrate a potential beneficial effect on functional MR in selected groups of patients. The Carillon™ device (Cardiac Dimensions, Kirkland, Washington) is constructed of a nitinol wire with distal and proximal figures of eight anchors connected by an intervening wire, also placed via jugular venous access (Figure 6). The distal anchor is released from the 9 Fr guide catheter and placed in the great cardiac vein and the guide catheter is mechanically pulled upward, resulting in tension on the coronary sinus with resultant shortening of the circumference of the coronary sinus. In animal models of heart failure, after device insertion, the mitral annular septal-lateral dimension was reduced by 25%, with substantial improvement in heart failure-induced MR and significant improvement in pulmonary capillary wedge pressure.34,36 Initial experience involving 50 patients clearly demonstrates the ability to reduce MR by 1 or 2 grades. The reduction in MR is immediate and can be modulated during the procedure in the catheterization laboratory. If the distal anchor crosses over the circumflex artery or an obtuse marginal branch, it can be pulled back to relieve coronary compression. Of course, if the device is pulled back, there is less efficacy for MR reduction. Although investigation continues in Europe, the device has not yet had a trial in the United States. Another coronary sinus device, the Viacor™ (Viacor, Inc., Wilmington, Massachusetts) uses a 7 Fr triple-lumen delivery catheter from a subclavian venous access into the coronary sinus and tracked into the great cardiac vein. The implant consists of a composite nitinol and stainless-steel construct coated with medical-grade Teflon and polyethylene plastic. Nitinol rods of varying stiffness are placed into the three lumens within the delivery catheter so that pressure is placed in the mid-coronary sinus compressing the septal-lateral dimension. Rod stiffness can be changed until some diminution in MR is achieved. The rod can be swapped out at a future time if additional stiffness is needed. Preliminary human experience with this device also shows its ability to reduce MR.37,38 While the major tension is placed in the center of the posterior leaflet, it is still possible for circumflex coronary artery compression to occur with this device. Further studies will define its ultimate utility. A heat energy approach could be a unique alternative to the implantable coronary sinus annuloplasty devices (QuantumCor™, Lake Forest, California). The principle mechanism is to reduce the circumference of the mitral annulus by shrinkage of collagen via application of the noncoagulating heat generated with a coronary sinus radiofrequency probe, thus improving mitral competence without implanted materials. The effects of the collagen shrinkage are immediate and the shrinkage does not regress over time because the collagen, heated to a selected level, retains its intrinsic strength. Additionally, the fibrotic healing response to the annular heating can add strength and possibly further enhance the initial shrinkage of the annulus. The prototype catheter has eight electrodes and the tip of the probe is malleable to conform to the annulus shape. Delivery of radio-frequency energy to the electrodes is computer-controlled by maximum temperatures sensed by adjacent thermocouples. In a preliminary sheep study, variable degrees of acute annular contraction was achieved; the reduction in A-P distance of the mitral annulus was to 26%.39 At necropsy, there was no evidence of thrombosis or damage to the coronary arteries or cardiac vein. The system has been used in a preclinical model, and significant development is necessary for a percutaneous system. Coronary sinus-based mitral valve repair is still in its early development. Enough preliminary work has been done to provide proof of principle. Several limitations remain. The relative position of coronary sinus is about 1 cm away from the true mitral annulus, which may detract from the effectiveness of this technique. Preliminary work in autopsy specimens and early implants have shown that the postulated position of the coronary sinus in the left posterior coronary sulcus was found in only 12% of the cases. The branches of the left circumflex coronary artery frequently cross under the coronary sinus, and have been compressed with resultant myocardial infarction in early human trials. Thus, compression of a coronary artery by the device may limit the degree to which the coronary sinus may be encircled.

Direct Annuloplasty

Direct annulopasty eliminates some of the limitations of the coronary sinus approach, but it requires access to the LV. Two devices that can be implanted directly into the mitral annulus are in the early phases of development. The Mitralign™ percutaneous annuloplasty system (Mitralign, Inc., Tewksbury, Massachusetts) uses a transaortic approach for direct cannulation of the LV to deliver a guide catheter into the space just beneath the posterior mitral leaflet on the LV side. Four wires are driven through the mitral annulus and anchored to the annulus using pledgets. The anchors are connected by a drawstring, which can be used to apply tension to the mitral annulus and shorten its circumference by 20%. The current version of the device uses two pairs of anchors to plicate the mitral annulus. The technique has been used in only a few patients, and phase I clinical trials have yet to be undertaken. The Guided Delivery System™ (Guided Delivery System, Santa Clara, California) follows a similar approach, placing up to 12 nitinol anchors in the mitral annulus with a tether. This device has been implanted during conventional cardiac surgery, with some patients having results at least beyond 1 year. The percutaneous method is in the early phases of development, and favorable acute results from first-in-man procedures have been presented.

Chamber and Annular Remodeling

A novel approach to treating functional MR by remodeling the LV chamber has been evaluated surgically using the Coapsys™ device (Myocor, Inc., Minneapolis, Minnesota). It is composed of a pair of epicardial pads that are anchored on the LV surface with a tensioning cable passed through the LV cavity to pull the pads together, thereby reducing the septal-to-lateral dimension of the mitral annulus and diminishing the LV chamber diameter.40,41 Experience with the surgical device has shown sustained reductions in MR and LV chamber dimensions for more than 1 year. The iCoapsys™ device is a percutaneous transpericardial version of the surgical system.42,43 Successful percutaneous placement has been completed in 2 patients. Another chamber remodeling approach uses a cord passed through the left atrial cavity with anchors in the coronary sinus and fossa ovalis. The PS3™ percutaneous septal sinus shortening system (Ample Medical, Foster City, California) has been shown to acutely and chronically reduce functional MR in an ovine model tachycardia model, and has been applied in temporary human implants before planned conventional mitral valve surgery.44 The companies that were developing both of these devices are no longer in operation, and it remains to been seen if other devices or systems for chamber remodeling will emerge.


With the advent of novel catheter-based techniques, an exciting new era for percutaneous management of patients with MR is evolving. The evaluation of these new devices for MR therapy poses new challenges. Comparisons with surgical results are inevitable. Annuloplasty devices may develop as heart failure therapies, with trial comparisons to medical therapy. On the other hand, there are almost no prospective data on the results of mitral repair surgery. Intention-to-treat analyses of mitral repair surgery have not been performed in the past. Multicenter trials, core echocardiographic laboratories and events committees have not been utilized for surgical trials. Thus, comparisons with surgery will require randomized trials such as EVEREST II. The obvious lesser invasiveness of these devices, by avoidance of an open chest incision, cardiac arrest and cardiopulmonary bypass, makes the treatment of MR in higher-risk patients attractive, and at the other end of the spectrum offers the chance to evaluate MR reduction earlier in the course of heart failure. None of these devices are likely to result in a treatment for all patients with MR, and also it is likely that not all will be ultimately successful.45 Continued development will require strong collaboration between surgeons and interventional cardiologists.46–48 Ongoing trials will ultimately define how these devices will be used clinically, alone or in combination. COMPLETE FIGURE LEGENDS Figure 2. Edge-to-edge repair with suture apposing anterior and posterior leaflets. Mitral valve is viewed from the left atrial side. The middle scallops of the anterior and posterior leaflets have been approximated with the MitrClip™, which creates a double orifice, edge-to-edge or bow-tie repair. Figure 3. (a) MitraClip™ advanced out of the guide catheter into the left atrium. (b) Open clip is advanced into the left ventricle. (c) Clip is pulled back to grasp the leaflets. (d) Clip is closed on the leaflets. (e) Clip is detached. (f) Left ventricular angiogram showing absence of residual mitral regurgitation. Figure 4. Specific morphologic features of the mitral valve are required for the Evalve MitraClip™ to have a good probability of success. There must be enough coaptation length for the clip to be grasped. A flail gap larger than 10 mm makes success of adequate reduction in MR unlikely, as does a flail width of more than 15 mm. The need for some coaptation length effectively excludes patients with an extremely dilated mitral annulus. In these cases, left ventricular failure with chamber dilatation and annular dilatation causes the leaflet edges to be pulled apart. Annuloplasty is more likely necessary in this anatomic setting.


From NorthShore University HealthSystem, Evanston, Illinois. Disclosure: Dr. Feldman has received grant support from Abbott and Edwards Lifesciences and has been a consultant to Abbott and Edwards Lifesciences. Address for correspondence: Ted Feldman, MD, FESC, FACC, FSCAI, Evanston Hospital, Cardiology Division-Walgreen Building 3rd Floor, 2650 Ridge Avenue, Evanston, IL 60201. E-mail:


Back to Top