Author + information
- Anelechi C. Anyanwu, MD and
- David H. Adams, MD∗ ()
- ↵∗Reprint requests and correspondence:
Dr. David H. Adams, Department of Cardiovascular Surgery, Mount Sinai Medical Center, 1190 Fifth Avenue, New York, New York 10029.
In this edition of the Journal, Cheung et al. (1) report 2 successful deployments of a catheter-mounted valve for treatment of severe functional mitral valve regurgitation. Both procedures used a novel bovine pericardial valve designed for mitral valve replacement, which was mounted on a self-expanding nitinol frame and advanced through the cardiac apex via surgical cut down. Because of the mitral valve’s intricate pathoanatomy and lack of a rigid landing zone, this transcatheter mitral valve prosthesis has a complex design, including anchoring mechanisms to prevent migration into the atrium during ventricular contraction. The prosthesis is not a symmetrical tube but conforms to the typical D-shape of the mitral valve annulus. In contrast, transcatheter aortic valve replacement (TAVR) prostheses have a simple, symmetrical design, with no need for ventricular anchors or specific orientation. Although prosthesis development and technical execution of transcatheter mitral valve replacement (TMVR) present unique challenges, these 2 patients with successfully deployed catheter mitral valves demonstrate that these challenges are not insurmountable. Cheung et al. are to be congratulated for their pioneering effort, which serves as a proof-of-concept for transcatheter replacement in the noncalcified mitral valve. This makes it probable that routine application of TMVR will be technically possible in the near future, leading the authors to question whether TMVR will revolutionize therapy for mitral valve disease, mirroring the course of TAVR.
Will TMVR follow a similarly rapid, explosive adoption and growth trajectory as TAVR? Rather than valve technology and technical feasibility, the prime determinants of TMVR’s applicability are likely to be some key differences that set catheter replacement of the mitral valve apart from that of the aortic valve (Table 1). Understanding these differences is critical to further development and clinical evaluation of TMVR platforms.
Heterogeneity of Mitral Valve Disease
Unlike aortic valve interventions in adults, the majority of which are conducted for a single etiological diagnosis (degenerative calcific aortic stenosis), a vast array of etiologies cause mitral regurgitation (predominantly) or stenosis, including congenital, degenerative, rheumatic, “functional” (secondary to ventricular or atrial pathology), and post-endocarditis etiologies. Within these categories, there is a wide spectrum of lesion presentation. For example, in the degenerative subset, patients with Barlow’s disease have very different valve morphology than those with simple chordal rupture due to fibroelastic deficiency (2), in terms of the degree of annular dilation, annular deformation, leaflet thickening, leaflet calcification, and annular fibrosis or calcification. Within the functional category, annular characteristics differ depending on various ischemic versus nonischemic etiologies. Thus, a single transcatheter prosthetic solution applicable to all etiologies is more challenging, and prostheses or techniques might need to vary depending on the etiology and pathoanatomy. In contrast, the vast majority of TAVR has been applied to a single disease and a consistent lesion: degenerative calcific stenosis.
Not a Disease of the Elderly
Most TAVR procedures are performed in elderly patients, often in their eighth or ninth decade of life. Thus, evaluation and application of this therapy were easier because there was less concern about durability and long-term implications. It was reasonably expected that a majority of high-risk elderly patients undergoing a TAVR would have a shorter life expectancy than the valve. In contrast, applying a less invasive, untested technology would be more controversial in young mitral regurgitation patients who may have several decades of life expectancy than in elderly aortic stenosis patients with an average life expectancy <10 years. Transcatheter mitral valve disease trials would provide an ethical conundrum because the majority of patients undergoing intervention are not in their eighth and ninth decades (as with TAVR) but are young or middle-aged. Data from the Society of Thoracic Surgeons’ database show that although the median age of U.S. patients undergoing mitral valve surgery was 60 years, with only 8% age ≥80 years (3), the median age of patients undergoing aortic valve replacement was 70 years, with 20% age ≥80 years (4).
EVEREST II (Endovascular Valve Edge-to-Edge Repair Study) randomized patients with degenerative or functional mitral regurgitation to receive either percutaneous clip therapy or mitral valve repair (5). The average age in the clip cohort was 67 years, with 49% in New York Heart Association class I or II. At 12 months, 46% of the clip cohort had moderate or greater mitral regurgitation, and 21% had undergone reoperation. Such results are suboptimal in a low-risk operable patient cohort and may discourage future investigators from conducting randomized trials of transcatheter mitral valve surgery in young, operable patients. Where there is a well-established, effective surgical alternative, placing an untried percutaneous device with unknown short-term and long-term results poses significant ethical dilemma. The hesitation to potentially predispose a patient to an inferior outcome may restrict patient selection for trials or clinical application of TMVR. Patient #2 in the report by Cheung et al. (1) was 61 years old; some may question whether therapies with more predictable midterm outcomes than an experimental device therapy (e.g., surgical repair, ventricular assist device therapy, transplantation) could better serve such a patient.
Success of Mitral Valve Repair
The success of TAVR is partly driven by its design, which replicates surgical treatment for senile aortic valve stenosis (almost invariably aortic valve replacement). It may be oversimplified as essentially the same operation by using a less invasive approach. Provided the end result (i.e., a functioning, durable bioprosthetic valve) is the same, there would be minimal concern that the treatment substantially differs from surgical aortic valve replacement. If safety and efficacy are not substantially inferior to surgical valve replacement, then TAVR will be a viable alternative for most patients, as is happening in practice. This is not true for mitral valve disease, for which bioprosthetic valve replacement is not the default therapy; mitral valve repair is the recommended intervention for treating severe mitral valve regurgitation, and it should generally be undertaken if the valve is repairable (6). More than 60% of U.S. patients undergoing mitral valve surgery (for any etiology) have a valve repair (3); for degenerative disease, the repair rate in specialty centers approaches 100% (7). Surgical mitral valve replacement is avoided when possible because of a significantly higher incidence of early and late complications, as well as a significant reduction in life expectancy, compared with valve repair. Although the long-term results of TMVR have yet to be determined, there seems little reason to believe that, in the long term, TMVR will fare substantially better than surgical valve replacement. Thus, valve repair will remain the treatment of choice for most patients with mitral valve regurgitation. Because repair is desirable and achievable in most patients, it is unlikely that percutaneous mitral valve replacement will have as mainstream a role in mitral valve disease as in aortic stenosis.
However, percutaneous repair approaches are under development and, if they evolve into effective therapies, they may challenge surgical repair’s dominance. Surgical repair tailors a blend of strategies (e.g., annuloplasty, leaflet resections, leaflet repairs, chordal replacement, chordal division), often used in combination, to the individual valve. Equally effective percutaneous repair will likely require multiple devices and techniques for the various combinations of lesions causing the mitral valve dysfunction. Attempts to apply a single therapy (percutaneous clip) to a wide spectrum of mitral disease have yielded unsatisfactory results. Indeed, any repair approach that does not include annuloplasty is likely to have limited long-term durability. Paradoxically, if a percutaneous solution applicable to the majority of mitral valve dysfunction were sought, then, in some ways, TMVR would be the simplest solution because any valve could theoretically be replaced. However, applying TMVR to the spectrum of mitral valve disease means trading reduced invasiveness for limited durability, prosthesis-related complications, and unknown long-term effects. Of note, bioprosthetic valves placed surgically in mitral positions are prone to earlier structural degeneration than aortic prostheses.
Risks to Surrounding Structures
Once successfully deployed, TAVR results are generally stable; it is unusual for patients free of procedural complications to develop late implantation complications. This may not apply to TMVR, particularly with a self-expanding stent. Without a discrete fibrous or calcific rim in many patients, and with further stent expansion after initial implantation, there could be negative effects on surrounding structures such as the circumflex artery, the left ventricular outflow tract (displacement of the native mitral valve could cause systolic anterior motion), the aortic valve, and conduction tissue. In addition, device migration due to high systolic pressures could lead to varying anatomic effects not seen at initial implant.
Nonbenign Course of Paravalvular Leaks
Paravalvular leakage remains the Achilles’ heel of transcatheter valve replacement. Such leaks may be more likely in the mitral than in the aortic position. This is the case with surgical valve replacement, partly due to the high systolic pressures to which a mitral prosthesis is exposed, compared with much lower diastolic pressures endured by aortic valve prostheses. As discussed in this initial report of TMVR (1), the transcatheter mitral prosthesis is designed to reduce the likelihood of paravalvular leakage. Although mild aortic paravalvular leaks have been well tolerated and managed and have not impeded the expansion of TAVR, we cannot be certain that the consequences will be similar with TMVR. Surgical experience shows that mitral paravalvular leaks are associated with substantially higher mortality and morbidity than aortic paravalvular leaks (8). Mitral valve paravalvular leaks also tend to be unstable and progressive, becoming more significant over time, with a more aggressive course; this often results in severe symptoms related to heart failure or hemolysis. Therefore, in terms of avoiding paravalvular leakage, the bar is somewhat higher with TMVR than with TAVR. For transcatheter mitral therapy to have widespread application, incidence of paravalvular leakage must be minimal because even mild leaks could be consequential.
The Tricuspid Valve
Any transcatheter solution to advanced mitral valve disease that does not include a solution for the tricuspid valve may be of limited efficacy in many patients. Even if an effective and safe approach to TMVR was developed, surgery may remain strongly favored because of the ability to concurrently tackle functional tricuspid regurgitation. Patients with long-standing mitral valve disease commonly have functional tricuspid regurgitation or significant tricuspid dilation, which may negatively affect long-term survival and quality of life. During mitral valve surgical repair or replacement, concurrent tricuspid valve repair is often undertaken in patients with significant tricuspid regurgitation or annular dilation. This advantage is lost with percutaneous therapy. Correction of mitral valve dysfunction does not always lead to reversal of tricuspid regurgitation. In the experience from percutaneous balloon valvotomy for mitral stenosis in patients with coexisting tricuspid regurgitation, the majority of patients still have tricuspid regurgitation at follow-up, despite successful reversal of severe mitral stenosis (9). Indeed, Patient #2 in the study by Cheung et al. (1) had severe tricuspid regurgitation. This can remain a significant source of morbidity in a heart failure patient, even with successful TMVR. Until percutaneous solutions exist for the tricuspid valve, surgery should remain the procedure of choice for patients with coexisting severe tricuspid regurgitation, even with increased surgical risk.
Who Is an Inoperable Patient?
The success of surgical therapy makes it more likely that TMVR will initially be restricted to patients who are not good surgical candidates. Cheung et al. (1) identified 2 patients deemed not to be surgical candidates as initial candidates for “first-in-human” application of TMVR. Patient #1 was 73 years of age, with advanced ischemic cardiomyopathy and a Society of Thoracic Surgeons' mortality risk score of 47%, clearly not a surgical candidate. However, the inoperability of Patient #2 (as judged by the investigators) is less obvious. Patient #2 was 61 years of age, with cardiomyopathy and a Society of Thoracic Surgeons' predicted mortality risk of 4.5%. Some surgeons would question whether this patient was too high-risk for surgery, as several series show that such patients can undergo surgery with a mortality risk <7% (10). Was this patient’s condition inoperable, or was surgery deemed futile or not worthwhile? Because this patient had been referred for transplantation, most likely an earlier decision had been made that transplantation should be the surgical treatment of choice. In other words, some patients are declined for mitral valve surgery not because their condition is inoperable and they are extreme-risk surgical cases but because it is believed that surgery is either not indicated or not beneficial. Indeed, current American College of Cardiology/American Heart Association guidelines (6) give mitral valve surgery for functional regurgitation a IIb recommendation, so practitioners do not agree on the basis for any valve intervention in such patients. If the benefit of surgery is unclear and unproven in such patients, is experimental percutaneous mitral replacement justified? It must be noted that, as with surgical repair, the benefit of percutaneous edge-to-edge clip repair in heart failure patients with mitral regurgitation is unproven and the subject of an ongoing clinical trial.
For patients with degenerative or rheumatic disease, no matter how advanced, scenarios in which surgery would not be recommended are infrequent. Severe comorbidity would be the overriding objective reason why patients with mitral disease are inoperable. Because most mitral valve patients are young or middle-aged, they generally do not have the degree of comorbidity or advanced age that would preclude surgery. Even elderly patients with mitral regurgitation are often surgical candidates (11). Notable exclusions are when technical factors preclude surgery in an otherwise low-risk patient; examples would include those with porcelain aorta, hostile mediastinum, or very calcified mitral annulus, and these cases would be particularly suited to TMVR. In the PARTNER (Placement of Aortic Transcatheter Valves) trial (12), with the TAVR example, there was a large cohort of truly inoperable extreme-risk patients with severe aortic stenosis, who would definitely not have been offered surgery before the TAVR era. These patients were randomized to undergo TAVR or receive medical therapy in the PARTNER B cohort; one-half of such patients in the medical arm had died within 1 year. There are no similar mitral disease patient cohorts with such poor survival if treated medically. Indeed, TAVR trials had no difficulty enrolling hundreds of extreme- or high-risk patients for randomization (12–14). In contrast, recruitment in transcatheter mitral trials that include inoperability as an inclusion criterion has been slow, which may partly reflect the difficulty in identifying truly inoperable patients with mitral disease.
Mitral Valve Disease Is Not Associated With High Short-Term Mortality
Mitral valve disease is rarely associated with sudden or rapid progression to death. Although it does shorten life expectancy, this effect is more indolent, occurring over years, as opposed to months. The urgency and short-term benefits of mitral valve intervention (over medical therapy) are therefore minimal and swing the bias away from using percutaneous device therapy as a life-extending intervention. Deriving survival benefit from mitral valve intervention generally requires few or several years of survival. If catheter therapy is reserved for extreme- or high-risk patients (as will probably be the case initially), the survival benefit will then likely be limited because many patients will not survive long enough to experience survival benefit. If patients are subjected to a procedural risk for no survival benefit (especially with typically better short-term survival without intervention), physicians may be more cautious in experimentally or clinically extending this therapy. This is in contradistinction to TAVR, in which the procedure is indeed lifesaving. Medically treated patients in the TAVR PARTNER trial (12) had a 50% 1-year mortality rate compared with 30% in patients treated with TAVR. This high short-term mortality rate likely accelerated TAVR adoption because patients otherwise facing imminent death due to aortic stenosis suddenly had a lifesaving option. This situation will not exist for TMVR because mitral valve disease is not generally life threatening in the short term, and the likelihood of 1-year survival on medical treatment for most patients is high. The main driver of the decision to undergo TMVR will therefore be quality of life. It is unlikely that asymptomatic or minimally symptomatic patients would be subjected to “prophylactic” TMVR to improve life expectancy (indeed, only valve repair is recommended in asymptomatic patients ; surgical valve replacement is not recommended because valve-related mortality may reduce long-term survival).
In addition to overcoming the technical hurdles of mitral valve replacement via transatrial, transapical, or transfemoral approaches, valve specialists should further consider prerequisites for application in clinical practice and the target patient subgroups. Although transcatheter valve replacement for de novo valve disease would be a true game changer for subsets of patients, current surgical data and experience suggest that surgical valve repair and replacement will remain dominant therapies in the near term.
↵∗ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
The Icahn School of Medicine at Mount Sinai receives royalties for intellectual property related to mitral and tricuspid repair products from Edwards Lifesciences and Medtronic, Inc. Dr. Anyanwu has reported that he has no relationships relevant to the contents of this paper to disclose. Dr. Adams is the national co-principal investigator of the US Medtronic CoreValve Trial. The Icahn School of Medicine at Mount Sinai receives royalties for intellectual property related to mitral and tricuspid repair products from Edwards Lifesciences and Medtronic, Inc.
- American College of Cardiology Foundation
- Cheung A.,
- Webb J.,
- Verheye S.,
- et al.
- Nishimura R.A.,
- Otto C.M.,
- Bonow R.O.,
- et al.
- Di Salvo T.G.,
- Acker M.A.,
- Dec G.W.,
- et al.
- Heterogeneity of Mitral Valve Disease
- Not a Disease of the Elderly
- Success of Mitral Valve Repair
- Risks to Surrounding Structures
- Nonbenign Course of Paravalvular Leaks
- The Tricuspid Valve
- Who Is an Inoperable Patient?
- Mitral Valve Disease Is Not Associated With High Short-Term Mortality
- What’s Next?