Stephanie Mick, MD

In the years since the introduction of robotically assisted mitral valve surgery, surgeons have looked for ways to improve techniques and procedures. A study from Cleveland Clinic presented at the American Association for Thoracic Surgery in 2016 assessed efficacy and safety outcomes associated with 1,000 consecutive robotically assisted mitral valve surgeries at Cleveland Clinic.1 The purpose of the study was to assess the clinical outcomes from these cases and analyze whether the outcomes changed over time as surgeons became more competent with robotic techniques. This analysis was also designed to identify procedural processes that improved outcomes during the trial.

STUDY METHODS

Nearly all cases (96%) were classified as degenerative mitral valve disease (N = 960). Of those, most had posterior leaflet prolapse (68%), about one-third (29%) had bileaflet prolapse, and only 3% had anterior leaflet involvement.

All surgeries were performed through right port incisions and used femoral cannulation for peripheral bypass. The aorta was occluded with either a Chitwood transthoracic clamp or a balloon.

STUDY RESULTS

It is important to remember that with femoral artery perfusion, the blood flow is opposite to the normal direction; thus, it goes up the aorta into the head vessels, which presents its own risks and challenges. Also, during retrograde perfusion, there is a risk of dislodging atherosclerotic plaque leading to brain embolus and stroke.

In these 1,000 cases, 997 were planned mitral valve repairs, 2 were mitral valve replacements, and 1 was resection of a mitral valve fibroelastoma. Results for the mitral valve repairs were excellent, with postoperative mitral regurgitation occurring in less than 1% of patients.

A primary point of interest was to identify procedural improvements that occurred during the course of the study. The areas evaluated in robotically assisted mitral valve surgery were the efficacy of the procedure in time, transfusion rates, stroke risk, how many mitral valve replacements occurred, and how many required conversion to sternotomy. These were assessed to determine whether surgical experience resulted in improvement.

Results showed that those efficiencies improved during the study. Cardiopulmonary bypass time decreased from about 140 minutes to 130 minutes. Cross-clamp time improved more dramatically from about 110 minutes to 90 minutes. And the percentage of cases requiring postoperative or intraoperative blood transfusion improved from about 24% to 10%.

PATIENT SELECTION CRITERIA: ALGORITHM

ALGORITHM IMPACT

What was the effect of this algorithm? In the 500 cases after its implementation, the stroke rate decreased by more than half—from 10 incidents before to 4 incidents after—and mitral replacements dropped from 4 to 0. The rate of conversion from robotic repair to conventional sternotomy in this patient series also improved, although this likely reflects surgical experience more than the algorithm. The conversion rate initially increased as surgeons gained experience with the robotic techniques. It rose to 4% during the first 300 to 400 cases, then dropped to 2% at the 500-case mark. It leveled off for the next 300 cases before dropping to 0 toward the end of the series.

Other metrics improved as well, which were attributed to a combination of surgical experience with robotic assistance and use of the patient-selection algorithm. The stroke risk declined to 0.8%, ischemic and cardiopulmonary bypass times declined, and the transfusion rate declined. No mitral replacements were done in the last 500 cases, and the conversion to conventional sternotomy rate declined to 1%.

In conclusion, this Cleveland Clinic study showed that a combination of a focused preoperative assessment using the patient-selection algorithm and increased surgical experience with robotic techniques enhanced clinical outcomes and improved procedural efficiency associated with robotically assisted mitral valve surgery.

In the years since the introduction of robotically assisted mitral valve surgery, surgeons have looked for ways to improve techniques and procedures. A study from Cleveland Clinic presented at the American Association for Thoracic Surgery in 2016 assessed efficacy and safety outcomes associated with 1,000 consecutive robotically assisted mitral valve surgeries at Cleveland Clinic.1 The purpose of the study was to assess the clinical outcomes from these cases and analyze whether the outcomes changed over time as surgeons became more competent with robotic techniques. This analysis was also designed to identify procedural processes that improved outcomes during the trial.

STUDY METHODS

Nearly all cases (96%) were classified as degenerative mitral valve disease (N = 960). Of those, most had posterior leaflet prolapse (68%), about one-third (29%) had bileaflet prolapse, and only 3% had anterior leaflet involvement.

All surgeries were performed through right port incisions and used femoral cannulation for peripheral bypass. The aorta was occluded with either a Chitwood transthoracic clamp or a balloon.

STUDY RESULTS

It is important to remember that with femoral artery perfusion, the blood flow is opposite to the normal direction; thus, it goes up the aorta into the head vessels, which presents its own risks and challenges. Also, during retrograde perfusion, there is a risk of dislodging atherosclerotic plaque leading to brain embolus and stroke.

In these 1,000 cases, 997 were planned mitral valve repairs, 2 were mitral valve replacements, and 1 was resection of a mitral valve fibroelastoma. Results for the mitral valve repairs were excellent, with postoperative mitral regurgitation occurring in less than 1% of patients.

A primary point of interest was to identify procedural improvements that occurred during the course of the study. The areas evaluated in robotically assisted mitral valve surgery were the efficacy of the procedure in time, transfusion rates, stroke risk, how many mitral valve replacements occurred, and how many required conversion to sternotomy. These were assessed to determine whether surgical experience resulted in improvement.

Results showed that those efficiencies improved during the study. Cardiopulmonary bypass time decreased from about 140 minutes to 130 minutes. Cross-clamp time improved more dramatically from about 110 minutes to 90 minutes. And the percentage of cases requiring postoperative or intraoperative blood transfusion improved from about 24% to 10%.

PATIENT SELECTION CRITERIA: ALGORITHM

ALGORITHM IMPACT

What was the effect of this algorithm? In the 500 cases after its implementation, the stroke rate decreased by more than half—from 10 incidents before to 4 incidents after—and mitral replacements dropped from 4 to 0. The rate of conversion from robotic repair to conventional sternotomy in this patient series also improved, although this likely reflects surgical experience more than the algorithm. The conversion rate initially increased as surgeons gained experience with the robotic techniques. It rose to 4% during the first 300 to 400 cases, then dropped to 2% at the 500-case mark. It leveled off for the next 300 cases before dropping to 0 toward the end of the series.

Other metrics improved as well, which were attributed to a combination of surgical experience with robotic assistance and use of the patient-selection algorithm. The stroke risk declined to 0.8%, ischemic and cardiopulmonary bypass times declined, and the transfusion rate declined. No mitral replacements were done in the last 500 cases, and the conversion to conventional sternotomy rate declined to 1%.

In conclusion, this Cleveland Clinic study showed that a combination of a focused preoperative assessment using the patient-selection algorithm and increased surgical experience with robotic techniques enhanced clinical outcomes and improved procedural efficiency associated with robotically assisted mitral valve surgery.

In the years since the introduction of robotically assisted mitral valve surgery, surgeons have looked for ways to improve techniques and procedures. A study from Cleveland Clinic presented at the American Association for Thoracic Surgery in 2016 assessed efficacy and safety outcomes associated with 1,000 consecutive robotically assisted mitral valve surgeries at Cleveland Clinic.1 The purpose of the study was to assess the clinical outcomes from these cases and analyze whether the outcomes changed over time as surgeons became more competent with robotic techniques. This analysis was also designed to identify procedural processes that improved outcomes during the trial.

STUDY METHODS

Nearly all cases (96%) were classified as degenerative mitral valve disease (N = 960). Of those, most had posterior leaflet prolapse (68%), about one-third (29%) had bileaflet prolapse, and only 3% had anterior leaflet involvement.

All surgeries were performed through right port incisions and used femoral cannulation for peripheral bypass. The aorta was occluded with either a Chitwood transthoracic clamp or a balloon.

STUDY RESULTS

It is important to remember that with femoral artery perfusion, the blood flow is opposite to the normal direction; thus, it goes up the aorta into the head vessels, which presents its own risks and challenges. Also, during retrograde perfusion, there is a risk of dislodging atherosclerotic plaque leading to brain embolus and stroke.

In these 1,000 cases, 997 were planned mitral valve repairs, 2 were mitral valve replacements, and 1 was resection of a mitral valve fibroelastoma. Results for the mitral valve repairs were excellent, with postoperative mitral regurgitation occurring in less than 1% of patients.

A primary point of interest was to identify procedural improvements that occurred during the course of the study. The areas evaluated in robotically assisted mitral valve surgery were the efficacy of the procedure in time, transfusion rates, stroke risk, how many mitral valve replacements occurred, and how many required conversion to sternotomy. These were assessed to determine whether surgical experience resulted in improvement.

Results showed that those efficiencies improved during the study. Cardiopulmonary bypass time decreased from about 140 minutes to 130 minutes. Cross-clamp time improved more dramatically from about 110 minutes to 90 minutes. And the percentage of cases requiring postoperative or intraoperative blood transfusion improved from about 24% to 10%.

PATIENT SELECTION CRITERIA: ALGORITHM

ALGORITHM IMPACT

What was the effect of this algorithm? In the 500 cases after its implementation, the stroke rate decreased by more than half—from 10 incidents before to 4 incidents after—and mitral replacements dropped from 4 to 0. The rate of conversion from robotic repair to conventional sternotomy in this patient series also improved, although this likely reflects surgical experience more than the algorithm. The conversion rate initially increased as surgeons gained experience with the robotic techniques. It rose to 4% during the first 300 to 400 cases, then dropped to 2% at the 500-case mark. It leveled off for the next 300 cases before dropping to 0 toward the end of the series.

Other metrics improved as well, which were attributed to a combination of surgical experience with robotic assistance and use of the patient-selection algorithm. The stroke risk declined to 0.8%, ischemic and cardiopulmonary bypass times declined, and the transfusion rate declined. No mitral replacements were done in the last 500 cases, and the conversion to conventional sternotomy rate declined to 1%.

In conclusion, this Cleveland Clinic study showed that a combination of a focused preoperative assessment using the patient-selection algorithm and increased surgical experience with robotic techniques enhanced clinical outcomes and improved procedural efficiency associated with robotically assisted mitral valve surgery.

In the last 10 years, we have seen a revolution in transcatheter therapies for structural heart disease. The most widely embraced, transcatheter aortic valve replacement (TAVR) was originally intended for patients in whom surgery was considered impossible, but it has now been established as an excellent alternative to surgical aortic valve replacement in patients at high or intermediate risk.^1–3 As TAVR has become established, with well-designed devices and acceptable safety and efficacy, it has inspired operators and inventors to push the envelope of innovation to transcatheter mitral valve replacement (TMVR).

This review summarizes the newest data available for the TMVR devices currently being tested in patients with native mitral regurgitation, bioprosthetic degeneration, and degenerative mitral stenosis.

THE MITRAL VALVE: THE NEW FRONTIER

Whereas the pathologic mechanisms of aortic stenosis generally all result in the same anatomic consequence (ie, calcification of the valve leaflets and commissures resulting in reduced mobility), mitral valve regurgitation is much more heterogeneous. Primary (degenerative) mitral regurgitation is caused by intrinsic valve pathology such as myxomatous degeneration, chordal detachment, fibroelastic deficiency, endocarditis, and other conditions that prevent the leaflets from coapting properly. In contrast, in secondary or functional mitral regurgitation, the leaflets are normal but do not coapt properly because of apical tethering to a dilated left ventricle, reduced closing forces with left ventricular dysfunction, or annular dilation as the result of either left ventricular or left atrial dilation.

Surgical mitral valve repair is safe and effective in patients with degenerative mitral regurgitation caused by leaflet prolapse and flail. However, some patients cannot undergo surgery because they have comorbid conditions that place them at extreme risk.⁴ For example, most patients with functional mitral regurgitation due to ischemic or dilated cardiomyopathy have significant surgical risk and multiple comorbidities, and in this group surgical repair has limited efficacy.⁵ A sizeable proportion of patients with mitral regurgitation may not be offered surgery because their risk is too high.⁶ Therefore, alternatives to the current surgical treatments have the potential to benefit a large number of patients.

Similarly, many patients with degenerative mitral stenosis caused by calcification of the mitral annulus also cannot undergo cardiac surgery because of prohibitively high risk. While rheumatic disease is the most common cause of mitral stenosis worldwide, degenerative mitral stenosis may be the cause in up to one-fourth of patients overall and up to 60% of patients older than 80 years.⁷ In the latter group, not only do old age and comorbidities such as diabetes mellitus and chronic kidney disease pose surgical risks, the technical challenge of surgically implanting a prosthetic mitral valve in the setting of a calcified annulus may be significant.⁸

The mitral valve is, therefore, the perfect new frontier for percutaneous valve replacement therapies, and TMVR is emerging as a potential option for patients with mitral regurgitation and degenerative mitral stenosis. The currently available percutaneous treatment options for mitral regurgitation include edge-to-edge leaflet repair, direct and indirect annuloplasty, spacers, and left ventricular remodeling devices (Table 1).^9,10 As surgical mitral valve repair is strongly preferred over mitral valve replacement, the percutaneous procedures and the devices that are used are engineered to approximate the current standard surgical techniques. However, given the complex pathologies involved, surgical repair often requires the use of multiple repair techniques in the same patient. Therefore, percutaneous repair may also require more than one type of device in the same patient and may not be anatomically feasible in many patients. Replacing the entire valve may obviate some of these challenges.

Reprinted with permission from Wolters Kluwer Health, Inc. (Sud K, et al. Degenerated mitral stenosis: unmet need for percutaneous interventions. Circulation 2016; 133:1594–1604).

Compared with the aortic valve, the mitral valve poses a greater challenge to percutaneous treatment due to its structure and dynamic relationship with the left ventricle. Some specific challenges facing the development of TMVR are that the mitral valve is large, it is difficult to access, it is asymmetrical, it lacks an anatomically well-defined annulus to which to anchor the replacement valve, its geometry changes throughout the cardiac cycle, and placing a replacement valve in it entails the risk of left ventricular outflow tract obstruction. Despite these challenges, a number of devices are undergoing preclinical testing, a few are in phase 1 clinical trials, and registries are being kept. Depending on the specific device, an antegrade transseptal approach to the mitral valve (via the femoral vein) or a retrograde transapical approach (via direct left ventricular access) may be used (Figure 1).

NATIVE MITRAL VALVE REGURGITATION

For degenerative mitral regurgitation, the standard of care is cardiac surgery at a hospital experienced with mitral valve repair, and with very low rates of mortality and morbidity. For patients in whom the surgical risk is prohibitive, percutaneous edge-to-edge leaflet repair using the MitraClip (Abbott Vascular, Minneapolis, MN) is the best option if the anatomy permits. If the mitral valve pathology is not amenable to MitraClip repair, the patient may be evaluated for TMVR under a clinical trial protocol.

For functional mitral regurgitation, the decisions are more complex. If the patient has chronic atrial fibrillation, electrical cardioversion and antiarrhythmic drug therapy may restore and maintain sinus rhythm, though if the left atrium is large, sinus rhythm may not be possible. If the patient has left ventricular dysfunction, guideline-directed medical therapy should be optimized; this reduces the risk of exacerbations, hospitalizations, and death and may also reduce the degree of regurgitation. If the patient has severe left ventricular dysfunction and a wide QRS duration, cardiac resynchronization therapy (biventricular pacing) may also be beneficial and reduce functional mitral regurgitation. If symptoms and severe functional mitral regurgitation persist despite these measures and the patient’s surgical risk is deemed to be extreme, options include MitraClip placement as part of the randomized Cardiovascular Outcomes Assessment of the MitraClip Percutaneous Therapy (COAPT) trial, which compares guideline-directed medical therapy with guideline-directed therapy plus MitraClip. Another option is enrollment in a clinical trial or registry of TMVR.

At this writing, six TMVR devices have been implanted in humans:

• Fortis (Edwards Lifesciences, Irvine, CA)
• Tendyne (Tendyne Holding Inc, Roseville, MN)
• NaviGate (NaviGate Cardiac Structures, Inc, Lake Forest, CA)
• Intrepid (Medtronic, Minneapolis, MN)
• CardiAQ (Edwards Lifesciences, Irvine, CA)
• Tiara (Neovasc Inc, Richmond, BC).

Most of the early experience with these valves has not yet been published, but some data have been presented at national and international meetings.

The Fortis valve

Courtesy of Edwards Lifesciences.

The Fortis valve consists of a self-expanding nitinol frame and leaflets made of bovine pericardium and is implanted via a transapical approach.

The device was successfully implanted in three patients in Quebec City, Canada, and at 6 months, all had improved significantly in functional class and none had needed to be hospitalized.¹¹ Echocardiographic assessment demonstrated trace or less mitral regurgitation and a mean transvalvular gradient less than 4 mm Hg in all.

Bapat and colleagues¹² attempted to implant the device in 13 patients in Europe and Canada. The average left ventricular ejection fraction was 34%, and 12 of 13 patients (92%) had functional mitral regurgitation. Procedural success was achieved in 10 patients, but five patients died within 30 days. While the deaths were due to nonvalvular issues (multi­organ failure, septic shock, intestinal ischemia after failed valve implantation and conversion to open surgery, malnutrition leading to respiratory failure, and valve thrombosis), the trial is currently on hold as more data are collected and reviewed. Among the eight patients who survived the first month, all were still alive at 6 months, and echocardiography demonstrated no or trivial mitral regurgitation in six patients (80%) and mild regurgitation in two patients (20%); the average mitral gradient was 4 mm Hg, and there was no change in mean left ventricular ejection fraction.

The Tendyne valve

Reprinted from EuroIntervention (Perpetua EM, et al. The Tendyne transcatheter mitral valve implantation system. EuroIntervention 2015; 11:W78-W79.) © 2015 with permission from Europa Digital Publishing.

The Tendyne valve is a self-expanding prosthesis with porcine pericardial leaflets. It is delivered transapically and is held in place by a tether from the valve to the left ventricular apex.

In the first 12 patients enrolled in an early feasibility trial,¹³ the average left ventricular ejection fraction was 40%, and 11 of the 12 patients had functional mitral regurgitation. The device was successfully implanted in 11 patients, while one patient developed left ventricular outflow tract obstruction and the device was uneventfully removed. All patients were still alive at 30 days, and the 11 patients who still had a prosthetic valve did not have any residual mitral regurgitation.

As of this writing, almost 80 patients have received the device, though the data have not yet been presented. Patients are being enrolled in phase 1 trials.

The NaviGate valve

Courtesy of Jose Navia.

The NaviGate valve consists of a trileaflet subassembly fabricated from bovine pericardium, mounted on a self-expanding nitinol stent, and is only implanted transatrially.

NaviGate valves were successfully implanted in two patients via a transatrial approach (Figure 2). Both patients had excellent valve performance without residual mitral regurgitation or left ventricular outflow tract obstruction. The first patient showed significant improvement in functional class and freedom from hospitalization at 6 months, but the second patient died within a week of the implant due to advanced heart failure.¹⁴ A US clinical trial is expected soon.

The Intrepid valve

Courtesy of Medtronic.

The Intrepid valve consists of an outer stent to provide fixation to the annulus and an inner stent that houses a bovine pericardial valve. The device is a self-expanding system that is delivered transapically.

In a series of 15 patients, 11 had functional mitral regurgitation (with an average left ventricular ejection fraction of 35%) and four had degenerative mitral regurgitation (with an average left ventricular ejection fraction of 57%).¹⁵ The device was successfully implanted in 14 patients, after which the average mitral valve gradient was 4 mm Hg. All patients but one were left with no regurgitation (the other patient had 1+ regurgitation).

A trial is currently under way in Europe.

The CardiAQ valve

Courtesy of Edwards Lifesciences.

The CardiAQ is constructed of bovine pericardium and can be delivered by the transseptal or transapical route.

Of 12 patients treated under compassionate use,¹⁶ two-thirds (eight patients) had functional mitral regurgitation. Two patients died during the procedure, three died of noncardiac complications within 30 days, and one more died of sepsis shortly after 30 days. This early experience demonstrates the importance of careful patient selection and postprocedural management in the feasibility assessment of these new technologies.

Patients are being enrolled in phase 1 trials.

The Tiara valve

Reprinted from EuroIntervention (Cheung A, et al. Transcatheter mitral valve implantation with Tiara bioprosthesis. EuroIntervention 2014; 10:U115-U119.) © 2014 with permission from Europa Digital & Publishing.

The Tiara valve, a self-expanding prosthesis with bovine pericardial leaflets, is delivered by the transapical route.

Eleven patients underwent Tiara implantation as part of either a Canadian special access registry or an international feasibility trial. Their average Society of Thoracic Surgeons score (ie, their calculated risk of major morbidity or operative mortality) was 15.6%, and their average left ventricular ejection fraction was 29%. Only two patients had degenerative mitral regurgitation. Nine patients had uneventful procedures and demonstrated no residual mitral regurgitation and no left ventricular outflow tract obstruction. The procedure was converted to open surgery in two patients owing to valve malpositioning, and both of them died within 30 days. One patient in whom the procedure was successful suffered erosion of the septum and died on day 4.¹⁷

Patients are being enrolled in phase 1 trials.

DEGENERATIVE MITRAL STENOSIS

Reprinted with permission from Wolters Kluwer Health, Inc. (Sud K, et al. Degenerated mitral stenosis: unmet need for percutaneous interventions. Circulation 2016; 133:1594–1604).

In patients with degenerative mitral stenosis, extensive mitral annular calcification may provide an adequate “frame” to hold a transcatheter valve prosthesis (Figure 3). Exploiting this feature, numerous investigators have successfully deployed prosthetic valves designed for TAVR in the calcified mitral annulus via the retrograde transapical and antegrade transseptal routes.

Guerrero and colleagues presented results from the first global registry of TMVR in mitral annular calcification at the 2016 EuroPCR Congress.¹⁸ Of 104 patients analyzed, almost all received an Edwards’ Sapien balloon-expandable valve (first-generation, Sapien XT, or Sapien 3); the others received Boston Scientific’s Lotus or Direct Flow Medical (Direct Flow Medical, Santa Clara, CA) valves. With an average age of 73 years and a high prevalence of comorbidities such as diabetes, chronic obstructive pulmonary disease, atrial fibrillation, chronic kidney disease, and prior cardiac surgery, the group presented extreme surgical risk, with an average Society of Thoracic Surgeons risk score of 14.4%. Slightly more than 40% of the patients underwent transapical implantation, slightly less than 40% underwent transfemoral or transseptal implantation, and just under 20% had a direct atrial approach.

The implantation was technically successful in 78 of 104 patients (75%); 13 patients (12.5%) required a second mitral valve to be placed, 11 patients (10.5%) had left ventricular outflow tract obstruction, four patients (4%) had valve embolization, and two patients (2%) had left ventricular perforation. At 30 days, 11 of 104 patients (10.6%) had died of cardiac causes and 15 patients (14.4%) had died of noncardiac causes. When divided roughly into three equal groups by chronological order, the last third of patients, compared with the first third of patients, enjoyed greater technical success (80%, n = 32/40 vs 62.5%, n = 20/32), better 30-day survival (85%, n = 34/40 vs 62.5%, n = 20/32), and no conversion to open surgery (0 vs 12.5%, n = 4/32), likely demonstrating both improved patient selection and lessons learned from shared experience. At 1 year, almost 90% of patients had New York Heart Association class I or II symptoms. Prior to the procedure, 91.5% had New York Heart Association class III or IV symptoms.

At present, TMVR in mitral annular calcification is not approved in the United States or elsewhere. However, multiple registries are currently enrolling patients or are in formative stages to push the frontier of the currently available technologies until better, dedicated devices are available for this group of patients.

BIOPROSTHETIC VALVE OR VALVE RING FAILURE

Implantation of a TAVR prosthetic inside a degenerated bioprosthetic mitral valve (valve-in-valve) and mitral valve ring (valve-in-ring) is generally limited to case series with short-term results using the Edwards Sapien series, Boston Scientific Lotus, Medtronic Melody (Medtronic, Minneapolis, MN), and Direct Flow Medical valves (Figure 4).^19–23

The largest collective experience was presented in the Valve-in-Valve International Data (VIVID) registry, which included 349 patients who had mitral valve-in-valve placement and 88 patients who had mitral valve-in-ring procedures. Their average age was 74 and the mean Society of Thoracic Surgeons score was 12.9% in both groups.24 Of the 437 patients, 345 patients (78.9%) underwent transapical implantation, and 391 patients (89.5%) received  a Sapien XT or Sapien 3 valve. In the valve-in-valve group, 41% of the patients had regurgitation, 25% had stenosis, and 34% had both. In the valve-in-ring group, 60% of the patients had regurgitation, 17% had stenosis, and 23% had both.

Valve placement was successful in most patients. The rate of stroke was low (2.9% with valve-in-valve placement, 1.1% with valve-in-ring placement), though the rate of moderate or greater residual mitral regurgitation was significantly higher in patients undergoing valve-in-ring procedures (14.8% vs 2.6%, P < .001), as was the rate of left ventricular outflow tract obstruction (8% vs 2.6%, P = .03). There was also a trend toward worse 30-day mortality in the valve-in-ring group (11.4% vs 7.7%, P = .15). As with aortic valve-in-valve procedures, small surgical mitral valves (≤ 25 mm) were associated with higher postprocedural gradients.

Eleid and colleagues²⁵ published their experience with antegrade transseptal TMVR in 48 patients with an average Society of Thoracic Surgeons score of 13.2%, 33 of whom underwent valve-in-valve procedures and nine of whom underwent valve-in-ring procedures. (The other six patients underwent mitral valve implantation for severe mitral annular calcification.) In the valve-in-valve group, 31 patients successfully underwent implant procedures, but two patients died during the procedure from left ventricular perforation. Of the nine valve-in-ring patients, two had acute embolization of the valve and were converted to open surgery. Among the seven patients in whom implantation was successful, two developed significant left ventricular outflow tract obstruction; one was treated with surgical resection of the anterior mitral valve leaflet and the other was medically managed.

CONCLUSION

Transcatheter mitral valve replacement in regurgitant mitral valves, failing mitral valve bioprosthetics and rings, and calcified mitral annuli has been effectively conducted in a number of patients who had no surgical options due to prohibitive surgical risk. International registries and our experience have demonstrated that the valve-in-valve procedure using a TAVR prosthesis carries the greatest likelihood of success, given the rigid frame of the surgical bioprosthetic that allows stable valve deployment. While approved in Europe for this indication, use of these devices for this application in the United States is considered “off label” and is performed only in clinically extenuating circumstances. Implantation of TAVR prosthetics in patients with prior mitral ring repair or for native mitral stenosis also has been performed successfully, although left ventricular outflow tract obstruction is a significant risk in this early experience.

Devices designed specifically for TMVR are in their clinical infancy and have been implanted successfully in only small numbers of patients, most of whom had functional mitral regurgitation. Despite reasonable technical success, most of these trials have been plagued by high mortality rates at 30 days in large part due to the extreme risk of the patients in whom these procedures have been conducted. At present, enrollment in TMVR trials for patients with degenerative or functional mitral regurgitation is limited to those without a surgical option and who conform to very specific anatomic criteria.

In the last 10 years, we have seen a revolution in transcatheter therapies for structural heart disease. The most widely embraced, transcatheter aortic valve replacement (TAVR) was originally intended for patients in whom surgery was considered impossible, but it has now been established as an excellent alternative to surgical aortic valve replacement in patients at high or intermediate risk.^1–3 As TAVR has become established, with well-designed devices and acceptable safety and efficacy, it has inspired operators and inventors to push the envelope of innovation to transcatheter mitral valve replacement (TMVR).

This review summarizes the newest data available for the TMVR devices currently being tested in patients with native mitral regurgitation, bioprosthetic degeneration, and degenerative mitral stenosis.

THE MITRAL VALVE: THE NEW FRONTIER

Whereas the pathologic mechanisms of aortic stenosis generally all result in the same anatomic consequence (ie, calcification of the valve leaflets and commissures resulting in reduced mobility), mitral valve regurgitation is much more heterogeneous. Primary (degenerative) mitral regurgitation is caused by intrinsic valve pathology such as myxomatous degeneration, chordal detachment, fibroelastic deficiency, endocarditis, and other conditions that prevent the leaflets from coapting properly. In contrast, in secondary or functional mitral regurgitation, the leaflets are normal but do not coapt properly because of apical tethering to a dilated left ventricle, reduced closing forces with left ventricular dysfunction, or annular dilation as the result of either left ventricular or left atrial dilation.

Surgical mitral valve repair is safe and effective in patients with degenerative mitral regurgitation caused by leaflet prolapse and flail. However, some patients cannot undergo surgery because they have comorbid conditions that place them at extreme risk.⁴ For example, most patients with functional mitral regurgitation due to ischemic or dilated cardiomyopathy have significant surgical risk and multiple comorbidities, and in this group surgical repair has limited efficacy.⁵ A sizeable proportion of patients with mitral regurgitation may not be offered surgery because their risk is too high.⁶ Therefore, alternatives to the current surgical treatments have the potential to benefit a large number of patients.

Similarly, many patients with degenerative mitral stenosis caused by calcification of the mitral annulus also cannot undergo cardiac surgery because of prohibitively high risk. While rheumatic disease is the most common cause of mitral stenosis worldwide, degenerative mitral stenosis may be the cause in up to one-fourth of patients overall and up to 60% of patients older than 80 years.⁷ In the latter group, not only do old age and comorbidities such as diabetes mellitus and chronic kidney disease pose surgical risks, the technical challenge of surgically implanting a prosthetic mitral valve in the setting of a calcified annulus may be significant.⁸

The mitral valve is, therefore, the perfect new frontier for percutaneous valve replacement therapies, and TMVR is emerging as a potential option for patients with mitral regurgitation and degenerative mitral stenosis. The currently available percutaneous treatment options for mitral regurgitation include edge-to-edge leaflet repair, direct and indirect annuloplasty, spacers, and left ventricular remodeling devices (Table 1).^9,10 As surgical mitral valve repair is strongly preferred over mitral valve replacement, the percutaneous procedures and the devices that are used are engineered to approximate the current standard surgical techniques. However, given the complex pathologies involved, surgical repair often requires the use of multiple repair techniques in the same patient. Therefore, percutaneous repair may also require more than one type of device in the same patient and may not be anatomically feasible in many patients. Replacing the entire valve may obviate some of these challenges.

Reprinted with permission from Wolters Kluwer Health, Inc. (Sud K, et al. Degenerated mitral stenosis: unmet need for percutaneous interventions. Circulation 2016; 133:1594–1604).

Compared with the aortic valve, the mitral valve poses a greater challenge to percutaneous treatment due to its structure and dynamic relationship with the left ventricle. Some specific challenges facing the development of TMVR are that the mitral valve is large, it is difficult to access, it is asymmetrical, it lacks an anatomically well-defined annulus to which to anchor the replacement valve, its geometry changes throughout the cardiac cycle, and placing a replacement valve in it entails the risk of left ventricular outflow tract obstruction. Despite these challenges, a number of devices are undergoing preclinical testing, a few are in phase 1 clinical trials, and registries are being kept. Depending on the specific device, an antegrade transseptal approach to the mitral valve (via the femoral vein) or a retrograde transapical approach (via direct left ventricular access) may be used (Figure 1).

NATIVE MITRAL VALVE REGURGITATION

For degenerative mitral regurgitation, the standard of care is cardiac surgery at a hospital experienced with mitral valve repair, and with very low rates of mortality and morbidity. For patients in whom the surgical risk is prohibitive, percutaneous edge-to-edge leaflet repair using the MitraClip (Abbott Vascular, Minneapolis, MN) is the best option if the anatomy permits. If the mitral valve pathology is not amenable to MitraClip repair, the patient may be evaluated for TMVR under a clinical trial protocol.

For functional mitral regurgitation, the decisions are more complex. If the patient has chronic atrial fibrillation, electrical cardioversion and antiarrhythmic drug therapy may restore and maintain sinus rhythm, though if the left atrium is large, sinus rhythm may not be possible. If the patient has left ventricular dysfunction, guideline-directed medical therapy should be optimized; this reduces the risk of exacerbations, hospitalizations, and death and may also reduce the degree of regurgitation. If the patient has severe left ventricular dysfunction and a wide QRS duration, cardiac resynchronization therapy (biventricular pacing) may also be beneficial and reduce functional mitral regurgitation. If symptoms and severe functional mitral regurgitation persist despite these measures and the patient’s surgical risk is deemed to be extreme, options include MitraClip placement as part of the randomized Cardiovascular Outcomes Assessment of the MitraClip Percutaneous Therapy (COAPT) trial, which compares guideline-directed medical therapy with guideline-directed therapy plus MitraClip. Another option is enrollment in a clinical trial or registry of TMVR.

At this writing, six TMVR devices have been implanted in humans:

• Fortis (Edwards Lifesciences, Irvine, CA)
• Tendyne (Tendyne Holding Inc, Roseville, MN)
• NaviGate (NaviGate Cardiac Structures, Inc, Lake Forest, CA)
• Intrepid (Medtronic, Minneapolis, MN)
• CardiAQ (Edwards Lifesciences, Irvine, CA)
• Tiara (Neovasc Inc, Richmond, BC).

Most of the early experience with these valves has not yet been published, but some data have been presented at national and international meetings.

The Fortis valve

Courtesy of Edwards Lifesciences.

The Fortis valve consists of a self-expanding nitinol frame and leaflets made of bovine pericardium and is implanted via a transapical approach.

The device was successfully implanted in three patients in Quebec City, Canada, and at 6 months, all had improved significantly in functional class and none had needed to be hospitalized.¹¹ Echocardiographic assessment demonstrated trace or less mitral regurgitation and a mean transvalvular gradient less than 4 mm Hg in all.

Bapat and colleagues¹² attempted to implant the device in 13 patients in Europe and Canada. The average left ventricular ejection fraction was 34%, and 12 of 13 patients (92%) had functional mitral regurgitation. Procedural success was achieved in 10 patients, but five patients died within 30 days. While the deaths were due to nonvalvular issues (multi­organ failure, septic shock, intestinal ischemia after failed valve implantation and conversion to open surgery, malnutrition leading to respiratory failure, and valve thrombosis), the trial is currently on hold as more data are collected and reviewed. Among the eight patients who survived the first month, all were still alive at 6 months, and echocardiography demonstrated no or trivial mitral regurgitation in six patients (80%) and mild regurgitation in two patients (20%); the average mitral gradient was 4 mm Hg, and there was no change in mean left ventricular ejection fraction.

The Tendyne valve

Reprinted from EuroIntervention (Perpetua EM, et al. The Tendyne transcatheter mitral valve implantation system. EuroIntervention 2015; 11:W78-W79.) © 2015 with permission from Europa Digital Publishing.

The Tendyne valve is a self-expanding prosthesis with porcine pericardial leaflets. It is delivered transapically and is held in place by a tether from the valve to the left ventricular apex.

In the first 12 patients enrolled in an early feasibility trial,¹³ the average left ventricular ejection fraction was 40%, and 11 of the 12 patients had functional mitral regurgitation. The device was successfully implanted in 11 patients, while one patient developed left ventricular outflow tract obstruction and the device was uneventfully removed. All patients were still alive at 30 days, and the 11 patients who still had a prosthetic valve did not have any residual mitral regurgitation.

As of this writing, almost 80 patients have received the device, though the data have not yet been presented. Patients are being enrolled in phase 1 trials.

The NaviGate valve

Courtesy of Jose Navia.

The NaviGate valve consists of a trileaflet subassembly fabricated from bovine pericardium, mounted on a self-expanding nitinol stent, and is only implanted transatrially.

NaviGate valves were successfully implanted in two patients via a transatrial approach (Figure 2). Both patients had excellent valve performance without residual mitral regurgitation or left ventricular outflow tract obstruction. The first patient showed significant improvement in functional class and freedom from hospitalization at 6 months, but the second patient died within a week of the implant due to advanced heart failure.¹⁴ A US clinical trial is expected soon.

The Intrepid valve

Courtesy of Medtronic.

The Intrepid valve consists of an outer stent to provide fixation to the annulus and an inner stent that houses a bovine pericardial valve. The device is a self-expanding system that is delivered transapically.

In a series of 15 patients, 11 had functional mitral regurgitation (with an average left ventricular ejection fraction of 35%) and four had degenerative mitral regurgitation (with an average left ventricular ejection fraction of 57%).¹⁵ The device was successfully implanted in 14 patients, after which the average mitral valve gradient was 4 mm Hg. All patients but one were left with no regurgitation (the other patient had 1+ regurgitation).

A trial is currently under way in Europe.

The CardiAQ valve

Courtesy of Edwards Lifesciences.

The CardiAQ is constructed of bovine pericardium and can be delivered by the transseptal or transapical route.

Of 12 patients treated under compassionate use,¹⁶ two-thirds (eight patients) had functional mitral regurgitation. Two patients died during the procedure, three died of noncardiac complications within 30 days, and one more died of sepsis shortly after 30 days. This early experience demonstrates the importance of careful patient selection and postprocedural management in the feasibility assessment of these new technologies.

Patients are being enrolled in phase 1 trials.

The Tiara valve

Reprinted from EuroIntervention (Cheung A, et al. Transcatheter mitral valve implantation with Tiara bioprosthesis. EuroIntervention 2014; 10:U115-U119.) © 2014 with permission from Europa Digital &amp; Publishing.

The Tiara valve, a self-expanding prosthesis with bovine pericardial leaflets, is delivered by the transapical route.

Eleven patients underwent Tiara implantation as part of either a Canadian special access registry or an international feasibility trial. Their average Society of Thoracic Surgeons score (ie, their calculated risk of major morbidity or operative mortality) was 15.6%, and their average left ventricular ejection fraction was 29%. Only two patients had degenerative mitral regurgitation. Nine patients had uneventful procedures and demonstrated no residual mitral regurgitation and no left ventricular outflow tract obstruction. The procedure was converted to open surgery in two patients owing to valve malpositioning, and both of them died within 30 days. One patient in whom the procedure was successful suffered erosion of the septum and died on day 4.¹⁷

Patients are being enrolled in phase 1 trials.

DEGENERATIVE MITRAL STENOSIS

Reprinted with permission from Wolters Kluwer Health, Inc. (Sud K, et al. Degenerated mitral stenosis: unmet need for percutaneous interventions. Circulation 2016; 133:1594–1604).

In patients with degenerative mitral stenosis, extensive mitral annular calcification may provide an adequate “frame” to hold a transcatheter valve prosthesis (Figure 3). Exploiting this feature, numerous investigators have successfully deployed prosthetic valves designed for TAVR in the calcified mitral annulus via the retrograde transapical and antegrade transseptal routes.

Guerrero and colleagues presented results from the first global registry of TMVR in mitral annular calcification at the 2016 EuroPCR Congress.¹⁸ Of 104 patients analyzed, almost all received an Edwards’ Sapien balloon-expandable valve (first-generation, Sapien XT, or Sapien 3); the others received Boston Scientific’s Lotus or Direct Flow Medical (Direct Flow Medical, Santa Clara, CA) valves. With an average age of 73 years and a high prevalence of comorbidities such as diabetes, chronic obstructive pulmonary disease, atrial fibrillation, chronic kidney disease, and prior cardiac surgery, the group presented extreme surgical risk, with an average Society of Thoracic Surgeons risk score of 14.4%. Slightly more than 40% of the patients underwent transapical implantation, slightly less than 40% underwent transfemoral or transseptal implantation, and just under 20% had a direct atrial approach.

The implantation was technically successful in 78 of 104 patients (75%); 13 patients (12.5%) required a second mitral valve to be placed, 11 patients (10.5%) had left ventricular outflow tract obstruction, four patients (4%) had valve embolization, and two patients (2%) had left ventricular perforation. At 30 days, 11 of 104 patients (10.6%) had died of cardiac causes and 15 patients (14.4%) had died of noncardiac causes. When divided roughly into three equal groups by chronological order, the last third of patients, compared with the first third of patients, enjoyed greater technical success (80%, n = 32/40 vs 62.5%, n = 20/32), better 30-day survival (85%, n = 34/40 vs 62.5%, n = 20/32), and no conversion to open surgery (0 vs 12.5%, n = 4/32), likely demonstrating both improved patient selection and lessons learned from shared experience. At 1 year, almost 90% of patients had New York Heart Association class I or II symptoms. Prior to the procedure, 91.5% had New York Heart Association class III or IV symptoms.

At present, TMVR in mitral annular calcification is not approved in the United States or elsewhere. However, multiple registries are currently enrolling patients or are in formative stages to push the frontier of the currently available technologies until better, dedicated devices are available for this group of patients.

BIOPROSTHETIC VALVE OR VALVE RING FAILURE

Implantation of a TAVR prosthetic inside a degenerated bioprosthetic mitral valve (valve-in-valve) and mitral valve ring (valve-in-ring) is generally limited to case series with short-term results using the Edwards Sapien series, Boston Scientific Lotus, Medtronic Melody (Medtronic, Minneapolis, MN), and Direct Flow Medical valves (Figure 4).^19–23

The largest collective experience was presented in the Valve-in-Valve International Data (VIVID) registry, which included 349 patients who had mitral valve-in-valve placement and 88 patients who had mitral valve-in-ring procedures. Their average age was 74 and the mean Society of Thoracic Surgeons score was 12.9% in both groups.24 Of the 437 patients, 345 patients (78.9%) underwent transapical implantation, and 391 patients (89.5%) received  a Sapien XT or Sapien 3 valve. In the valve-in-valve group, 41% of the patients had regurgitation, 25% had stenosis, and 34% had both. In the valve-in-ring group, 60% of the patients had regurgitation, 17% had stenosis, and 23% had both.

Valve placement was successful in most patients. The rate of stroke was low (2.9% with valve-in-valve placement, 1.1% with valve-in-ring placement), though the rate of moderate or greater residual mitral regurgitation was significantly higher in patients undergoing valve-in-ring procedures (14.8% vs 2.6%, P < .001), as was the rate of left ventricular outflow tract obstruction (8% vs 2.6%, P = .03). There was also a trend toward worse 30-day mortality in the valve-in-ring group (11.4% vs 7.7%, P = .15). As with aortic valve-in-valve procedures, small surgical mitral valves (≤ 25 mm) were associated with higher postprocedural gradients.

Eleid and colleagues²⁵ published their experience with antegrade transseptal TMVR in 48 patients with an average Society of Thoracic Surgeons score of 13.2%, 33 of whom underwent valve-in-valve procedures and nine of whom underwent valve-in-ring procedures. (The other six patients underwent mitral valve implantation for severe mitral annular calcification.) In the valve-in-valve group, 31 patients successfully underwent implant procedures, but two patients died during the procedure from left ventricular perforation. Of the nine valve-in-ring patients, two had acute embolization of the valve and were converted to open surgery. Among the seven patients in whom implantation was successful, two developed significant left ventricular outflow tract obstruction; one was treated with surgical resection of the anterior mitral valve leaflet and the other was medically managed.

CONCLUSION

Transcatheter mitral valve replacement in regurgitant mitral valves, failing mitral valve bioprosthetics and rings, and calcified mitral annuli has been effectively conducted in a number of patients who had no surgical options due to prohibitive surgical risk. International registries and our experience have demonstrated that the valve-in-valve procedure using a TAVR prosthesis carries the greatest likelihood of success, given the rigid frame of the surgical bioprosthetic that allows stable valve deployment. While approved in Europe for this indication, use of these devices for this application in the United States is considered “off label” and is performed only in clinically extenuating circumstances. Implantation of TAVR prosthetics in patients with prior mitral ring repair or for native mitral stenosis also has been performed successfully, although left ventricular outflow tract obstruction is a significant risk in this early experience.

Devices designed specifically for TMVR are in their clinical infancy and have been implanted successfully in only small numbers of patients, most of whom had functional mitral regurgitation. Despite reasonable technical success, most of these trials have been plagued by high mortality rates at 30 days in large part due to the extreme risk of the patients in whom these procedures have been conducted. At present, enrollment in TMVR trials for patients with degenerative or functional mitral regurgitation is limited to those without a surgical option and who conform to very specific anatomic criteria.

In the last 10 years, we have seen a revolution in transcatheter therapies for structural heart disease. The most widely embraced, transcatheter aortic valve replacement (TAVR) was originally intended for patients in whom surgery was considered impossible, but it has now been established as an excellent alternative to surgical aortic valve replacement in patients at high or intermediate risk.^1–3 As TAVR has become established, with well-designed devices and acceptable safety and efficacy, it has inspired operators and inventors to push the envelope of innovation to transcatheter mitral valve replacement (TMVR).

This review summarizes the newest data available for the TMVR devices currently being tested in patients with native mitral regurgitation, bioprosthetic degeneration, and degenerative mitral stenosis.

THE MITRAL VALVE: THE NEW FRONTIER

Whereas the pathologic mechanisms of aortic stenosis generally all result in the same anatomic consequence (ie, calcification of the valve leaflets and commissures resulting in reduced mobility), mitral valve regurgitation is much more heterogeneous. Primary (degenerative) mitral regurgitation is caused by intrinsic valve pathology such as myxomatous degeneration, chordal detachment, fibroelastic deficiency, endocarditis, and other conditions that prevent the leaflets from coapting properly. In contrast, in secondary or functional mitral regurgitation, the leaflets are normal but do not coapt properly because of apical tethering to a dilated left ventricle, reduced closing forces with left ventricular dysfunction, or annular dilation as the result of either left ventricular or left atrial dilation.

Surgical mitral valve repair is safe and effective in patients with degenerative mitral regurgitation caused by leaflet prolapse and flail. However, some patients cannot undergo surgery because they have comorbid conditions that place them at extreme risk.⁴ For example, most patients with functional mitral regurgitation due to ischemic or dilated cardiomyopathy have significant surgical risk and multiple comorbidities, and in this group surgical repair has limited efficacy.⁵ A sizeable proportion of patients with mitral regurgitation may not be offered surgery because their risk is too high.⁶ Therefore, alternatives to the current surgical treatments have the potential to benefit a large number of patients.

Similarly, many patients with degenerative mitral stenosis caused by calcification of the mitral annulus also cannot undergo cardiac surgery because of prohibitively high risk. While rheumatic disease is the most common cause of mitral stenosis worldwide, degenerative mitral stenosis may be the cause in up to one-fourth of patients overall and up to 60% of patients older than 80 years.⁷ In the latter group, not only do old age and comorbidities such as diabetes mellitus and chronic kidney disease pose surgical risks, the technical challenge of surgically implanting a prosthetic mitral valve in the setting of a calcified annulus may be significant.⁸

The mitral valve is, therefore, the perfect new frontier for percutaneous valve replacement therapies, and TMVR is emerging as a potential option for patients with mitral regurgitation and degenerative mitral stenosis. The currently available percutaneous treatment options for mitral regurgitation include edge-to-edge leaflet repair, direct and indirect annuloplasty, spacers, and left ventricular remodeling devices (Table 1).^9,10 As surgical mitral valve repair is strongly preferred over mitral valve replacement, the percutaneous procedures and the devices that are used are engineered to approximate the current standard surgical techniques. However, given the complex pathologies involved, surgical repair often requires the use of multiple repair techniques in the same patient. Therefore, percutaneous repair may also require more than one type of device in the same patient and may not be anatomically feasible in many patients. Replacing the entire valve may obviate some of these challenges.

Reprinted with permission from Wolters Kluwer Health, Inc. (Sud K, et al. Degenerated mitral stenosis: unmet need for percutaneous interventions. Circulation 2016; 133:1594–1604).

Compared with the aortic valve, the mitral valve poses a greater challenge to percutaneous treatment due to its structure and dynamic relationship with the left ventricle. Some specific challenges facing the development of TMVR are that the mitral valve is large, it is difficult to access, it is asymmetrical, it lacks an anatomically well-defined annulus to which to anchor the replacement valve, its geometry changes throughout the cardiac cycle, and placing a replacement valve in it entails the risk of left ventricular outflow tract obstruction. Despite these challenges, a number of devices are undergoing preclinical testing, a few are in phase 1 clinical trials, and registries are being kept. Depending on the specific device, an antegrade transseptal approach to the mitral valve (via the femoral vein) or a retrograde transapical approach (via direct left ventricular access) may be used (Figure 1).

NATIVE MITRAL VALVE REGURGITATION

For degenerative mitral regurgitation, the standard of care is cardiac surgery at a hospital experienced with mitral valve repair, and with very low rates of mortality and morbidity. For patients in whom the surgical risk is prohibitive, percutaneous edge-to-edge leaflet repair using the MitraClip (Abbott Vascular, Minneapolis, MN) is the best option if the anatomy permits. If the mitral valve pathology is not amenable to MitraClip repair, the patient may be evaluated for TMVR under a clinical trial protocol.

For functional mitral regurgitation, the decisions are more complex. If the patient has chronic atrial fibrillation, electrical cardioversion and antiarrhythmic drug therapy may restore and maintain sinus rhythm, though if the left atrium is large, sinus rhythm may not be possible. If the patient has left ventricular dysfunction, guideline-directed medical therapy should be optimized; this reduces the risk of exacerbations, hospitalizations, and death and may also reduce the degree of regurgitation. If the patient has severe left ventricular dysfunction and a wide QRS duration, cardiac resynchronization therapy (biventricular pacing) may also be beneficial and reduce functional mitral regurgitation. If symptoms and severe functional mitral regurgitation persist despite these measures and the patient’s surgical risk is deemed to be extreme, options include MitraClip placement as part of the randomized Cardiovascular Outcomes Assessment of the MitraClip Percutaneous Therapy (COAPT) trial, which compares guideline-directed medical therapy with guideline-directed therapy plus MitraClip. Another option is enrollment in a clinical trial or registry of TMVR.

At this writing, six TMVR devices have been implanted in humans:

• Fortis (Edwards Lifesciences, Irvine, CA)
• Tendyne (Tendyne Holding Inc, Roseville, MN)
• NaviGate (NaviGate Cardiac Structures, Inc, Lake Forest, CA)
• Intrepid (Medtronic, Minneapolis, MN)
• CardiAQ (Edwards Lifesciences, Irvine, CA)
• Tiara (Neovasc Inc, Richmond, BC).

Most of the early experience with these valves has not yet been published, but some data have been presented at national and international meetings.

The Fortis valve

Courtesy of Edwards Lifesciences.

The Fortis valve consists of a self-expanding nitinol frame and leaflets made of bovine pericardium and is implanted via a transapical approach.

The device was successfully implanted in three patients in Quebec City, Canada, and at 6 months, all had improved significantly in functional class and none had needed to be hospitalized.¹¹ Echocardiographic assessment demonstrated trace or less mitral regurgitation and a mean transvalvular gradient less than 4 mm Hg in all.

Bapat and colleagues¹² attempted to implant the device in 13 patients in Europe and Canada. The average left ventricular ejection fraction was 34%, and 12 of 13 patients (92%) had functional mitral regurgitation. Procedural success was achieved in 10 patients, but five patients died within 30 days. While the deaths were due to nonvalvular issues (multi­organ failure, septic shock, intestinal ischemia after failed valve implantation and conversion to open surgery, malnutrition leading to respiratory failure, and valve thrombosis), the trial is currently on hold as more data are collected and reviewed. Among the eight patients who survived the first month, all were still alive at 6 months, and echocardiography demonstrated no or trivial mitral regurgitation in six patients (80%) and mild regurgitation in two patients (20%); the average mitral gradient was 4 mm Hg, and there was no change in mean left ventricular ejection fraction.

The Tendyne valve

Reprinted from EuroIntervention (Perpetua EM, et al. The Tendyne transcatheter mitral valve implantation system. EuroIntervention 2015; 11:W78-W79.) © 2015 with permission from Europa Digital Publishing.

The Tendyne valve is a self-expanding prosthesis with porcine pericardial leaflets. It is delivered transapically and is held in place by a tether from the valve to the left ventricular apex.

In the first 12 patients enrolled in an early feasibility trial,¹³ the average left ventricular ejection fraction was 40%, and 11 of the 12 patients had functional mitral regurgitation. The device was successfully implanted in 11 patients, while one patient developed left ventricular outflow tract obstruction and the device was uneventfully removed. All patients were still alive at 30 days, and the 11 patients who still had a prosthetic valve did not have any residual mitral regurgitation.

As of this writing, almost 80 patients have received the device, though the data have not yet been presented. Patients are being enrolled in phase 1 trials.

The NaviGate valve

Courtesy of Jose Navia.

The NaviGate valve consists of a trileaflet subassembly fabricated from bovine pericardium, mounted on a self-expanding nitinol stent, and is only implanted transatrially.

NaviGate valves were successfully implanted in two patients via a transatrial approach (Figure 2). Both patients had excellent valve performance without residual mitral regurgitation or left ventricular outflow tract obstruction. The first patient showed significant improvement in functional class and freedom from hospitalization at 6 months, but the second patient died within a week of the implant due to advanced heart failure.¹⁴ A US clinical trial is expected soon.

The Intrepid valve

Courtesy of Medtronic.

The Intrepid valve consists of an outer stent to provide fixation to the annulus and an inner stent that houses a bovine pericardial valve. The device is a self-expanding system that is delivered transapically.

In a series of 15 patients, 11 had functional mitral regurgitation (with an average left ventricular ejection fraction of 35%) and four had degenerative mitral regurgitation (with an average left ventricular ejection fraction of 57%).¹⁵ The device was successfully implanted in 14 patients, after which the average mitral valve gradient was 4 mm Hg. All patients but one were left with no regurgitation (the other patient had 1+ regurgitation).

A trial is currently under way in Europe.

The CardiAQ valve

Courtesy of Edwards Lifesciences.

The CardiAQ is constructed of bovine pericardium and can be delivered by the transseptal or transapical route.

Of 12 patients treated under compassionate use,¹⁶ two-thirds (eight patients) had functional mitral regurgitation. Two patients died during the procedure, three died of noncardiac complications within 30 days, and one more died of sepsis shortly after 30 days. This early experience demonstrates the importance of careful patient selection and postprocedural management in the feasibility assessment of these new technologies.

Patients are being enrolled in phase 1 trials.

The Tiara valve

Reprinted from EuroIntervention (Cheung A, et al. Transcatheter mitral valve implantation with Tiara bioprosthesis. EuroIntervention 2014; 10:U115-U119.) © 2014 with permission from Europa Digital &amp; Publishing.

The Tiara valve, a self-expanding prosthesis with bovine pericardial leaflets, is delivered by the transapical route.

Eleven patients underwent Tiara implantation as part of either a Canadian special access registry or an international feasibility trial. Their average Society of Thoracic Surgeons score (ie, their calculated risk of major morbidity or operative mortality) was 15.6%, and their average left ventricular ejection fraction was 29%. Only two patients had degenerative mitral regurgitation. Nine patients had uneventful procedures and demonstrated no residual mitral regurgitation and no left ventricular outflow tract obstruction. The procedure was converted to open surgery in two patients owing to valve malpositioning, and both of them died within 30 days. One patient in whom the procedure was successful suffered erosion of the septum and died on day 4.¹⁷

Patients are being enrolled in phase 1 trials.

DEGENERATIVE MITRAL STENOSIS

Reprinted with permission from Wolters Kluwer Health, Inc. (Sud K, et al. Degenerated mitral stenosis: unmet need for percutaneous interventions. Circulation 2016; 133:1594–1604).

In patients with degenerative mitral stenosis, extensive mitral annular calcification may provide an adequate “frame” to hold a transcatheter valve prosthesis (Figure 3). Exploiting this feature, numerous investigators have successfully deployed prosthetic valves designed for TAVR in the calcified mitral annulus via the retrograde transapical and antegrade transseptal routes.

Guerrero and colleagues presented results from the first global registry of TMVR in mitral annular calcification at the 2016 EuroPCR Congress.¹⁸ Of 104 patients analyzed, almost all received an Edwards’ Sapien balloon-expandable valve (first-generation, Sapien XT, or Sapien 3); the others received Boston Scientific’s Lotus or Direct Flow Medical (Direct Flow Medical, Santa Clara, CA) valves. With an average age of 73 years and a high prevalence of comorbidities such as diabetes, chronic obstructive pulmonary disease, atrial fibrillation, chronic kidney disease, and prior cardiac surgery, the group presented extreme surgical risk, with an average Society of Thoracic Surgeons risk score of 14.4%. Slightly more than 40% of the patients underwent transapical implantation, slightly less than 40% underwent transfemoral or transseptal implantation, and just under 20% had a direct atrial approach.

The implantation was technically successful in 78 of 104 patients (75%); 13 patients (12.5%) required a second mitral valve to be placed, 11 patients (10.5%) had left ventricular outflow tract obstruction, four patients (4%) had valve embolization, and two patients (2%) had left ventricular perforation. At 30 days, 11 of 104 patients (10.6%) had died of cardiac causes and 15 patients (14.4%) had died of noncardiac causes. When divided roughly into three equal groups by chronological order, the last third of patients, compared with the first third of patients, enjoyed greater technical success (80%, n = 32/40 vs 62.5%, n = 20/32), better 30-day survival (85%, n = 34/40 vs 62.5%, n = 20/32), and no conversion to open surgery (0 vs 12.5%, n = 4/32), likely demonstrating both improved patient selection and lessons learned from shared experience. At 1 year, almost 90% of patients had New York Heart Association class I or II symptoms. Prior to the procedure, 91.5% had New York Heart Association class III or IV symptoms.

At present, TMVR in mitral annular calcification is not approved in the United States or elsewhere. However, multiple registries are currently enrolling patients or are in formative stages to push the frontier of the currently available technologies until better, dedicated devices are available for this group of patients.

BIOPROSTHETIC VALVE OR VALVE RING FAILURE

Implantation of a TAVR prosthetic inside a degenerated bioprosthetic mitral valve (valve-in-valve) and mitral valve ring (valve-in-ring) is generally limited to case series with short-term results using the Edwards Sapien series, Boston Scientific Lotus, Medtronic Melody (Medtronic, Minneapolis, MN), and Direct Flow Medical valves (Figure 4).^19–23

The largest collective experience was presented in the Valve-in-Valve International Data (VIVID) registry, which included 349 patients who had mitral valve-in-valve placement and 88 patients who had mitral valve-in-ring procedures. Their average age was 74 and the mean Society of Thoracic Surgeons score was 12.9% in both groups.24 Of the 437 patients, 345 patients (78.9%) underwent transapical implantation, and 391 patients (89.5%) received  a Sapien XT or Sapien 3 valve. In the valve-in-valve group, 41% of the patients had regurgitation, 25% had stenosis, and 34% had both. In the valve-in-ring group, 60% of the patients had regurgitation, 17% had stenosis, and 23% had both.

Valve placement was successful in most patients. The rate of stroke was low (2.9% with valve-in-valve placement, 1.1% with valve-in-ring placement), though the rate of moderate or greater residual mitral regurgitation was significantly higher in patients undergoing valve-in-ring procedures (14.8% vs 2.6%, P < .001), as was the rate of left ventricular outflow tract obstruction (8% vs 2.6%, P = .03). There was also a trend toward worse 30-day mortality in the valve-in-ring group (11.4% vs 7.7%, P = .15). As with aortic valve-in-valve procedures, small surgical mitral valves (≤ 25 mm) were associated with higher postprocedural gradients.

Eleid and colleagues²⁵ published their experience with antegrade transseptal TMVR in 48 patients with an average Society of Thoracic Surgeons score of 13.2%, 33 of whom underwent valve-in-valve procedures and nine of whom underwent valve-in-ring procedures. (The other six patients underwent mitral valve implantation for severe mitral annular calcification.) In the valve-in-valve group, 31 patients successfully underwent implant procedures, but two patients died during the procedure from left ventricular perforation. Of the nine valve-in-ring patients, two had acute embolization of the valve and were converted to open surgery. Among the seven patients in whom implantation was successful, two developed significant left ventricular outflow tract obstruction; one was treated with surgical resection of the anterior mitral valve leaflet and the other was medically managed.

CONCLUSION

Transcatheter mitral valve replacement in regurgitant mitral valves, failing mitral valve bioprosthetics and rings, and calcified mitral annuli has been effectively conducted in a number of patients who had no surgical options due to prohibitive surgical risk. International registries and our experience have demonstrated that the valve-in-valve procedure using a TAVR prosthesis carries the greatest likelihood of success, given the rigid frame of the surgical bioprosthetic that allows stable valve deployment. While approved in Europe for this indication, use of these devices for this application in the United States is considered “off label” and is performed only in clinically extenuating circumstances. Implantation of TAVR prosthetics in patients with prior mitral ring repair or for native mitral stenosis also has been performed successfully, although left ventricular outflow tract obstruction is a significant risk in this early experience.

Devices designed specifically for TMVR are in their clinical infancy and have been implanted successfully in only small numbers of patients, most of whom had functional mitral regurgitation. Despite reasonable technical success, most of these trials have been plagued by high mortality rates at 30 days in large part due to the extreme risk of the patients in whom these procedures have been conducted. At present, enrollment in TMVR trials for patients with degenerative or functional mitral regurgitation is limited to those without a surgical option and who conform to very specific anatomic criteria.

Transcatheter aortic valve replacement (TAVR) has established itself as an effective way of treating high-risk patients with severe aortic valve stenosis. With new generations of existing valves and newer alternative devices, the procedure promises to become increasingly safer. The field is evolving rapidly and it will be important for interventional cardiologists and cardiac surgeons alike to stay abreast of developments. This article reviews the history of this promising procedure and examines its use in current practice.

HISTORICAL PERSPECTIVE

In 1980, Danish researcher H. R. Anderson reported developing and testing a balloon-expandable valve in animals.¹ The technology was eventually acquired and further developed by Edwards Life Sciences (Irvine, California).

Alain Cribier started early work in humans in 2002 in France.² He used a transfemoral arterial access to approach the aortic valve transseptally, but this procedure was associated with high rates of mortality and stroke.³ At the same time, in the United States, animal studies were being carried out by Lars G. Svensson, Todd Dewey, and Michael Mack to develop a transapical method of implantation,^4,5 while John Webb and colleagues were also developing a transapical aortic valve implantation technique,^6,7 and later went on to develop a retrograde transfemoral technique. This latter technique became feasible once Edwards developed a catheter that could be flexed to get around the aortic arch and across the aortic valve.

As the Edwards balloon-expandable valve (Sapien) was being developed, a nitinol-based self-expandable valve system was introduced by Medtronic: the CoreValve. Following feasibility studies,^5,8 the safety and efficacy of these valves were established thorough the Placement of Aortic Transcatheter Valves (PARTNER) trial and the US Core Valve Pivotal Trial. These valves are currently approved by the US Food and Drug Administration (FDA) for patients for whom conventional surgery would pose an extreme or high risk.^9–11

CLINICAL TRIALS OF TAVR

The two landmark prospective randomized trials of TAVR were the PARTNER trial and CoreValve Pivotal Trial.

The PARTNER trial consisted of two parts: PARTNER A, which compared the Sapien balloon-expandable transcatheter valve with surgical aortic valve replacement in patients at high surgical risk (Society of Thoracic Surgeons [STS] score > 10%), and PARTNER B, which compared TAVR with medical therapy in patients who could not undergo surgery (combined risk of serious morbidity or death of 50% or more, and two surgeons agreeing that the patient was inoperable).

Similarly, the CoreValve Pivotal Trial compared the self-expandable transcatheter valve with conventional medical and surgical treatment.

TAVR is comparable to surgery in outcomes, with caveats

In the PARTNER A trial, mortality rates were similar between patients who underwent Sapien TAVR and those who underwent surgical valve replacement at 30 days (3.4% and 6.5%, P = .07), 1 year (24.2% and 26.8%), and 2 years (33.9% and 35.0%). The patients in this group were randomized to either Sapien TAVR or surgery (Table 1).^10,12

The combined rate of stroke and transient ischemic attack was higher in the patients assigned to TAVR at 30 days (5.5% with TAVR vs 2.4% with surgery, P = .04) and at 1 year (8.3% with TAVR vs 4.3% with surgery, P = .04). The difference was of small significance at 2 years (11.2% vs 6.5%, P = .05). At 30 days, the rate of major vascular complications was higher with TAVR (11.0% vs 3.2%), while surgery was associated with more frequent major bleeding episodes (19.5% vs 9.3%) and new-onset atrial fibrillation (16.0% vs 8.6%). The rate of new pacemaker requirement at 30 days was similar between the TAVR and surgical groups (3.8% vs 3.6%). Moderate or severe paravalvular aortic regurgitation was more common after TAVR at 30 days, 1 year, and 2 years. This aortic insufficiency was associated with increased late mortality.^10,12

In the US CoreValve High Risk Study, no difference was found in the 30-day mortality rate in patients at high surgical risk randomized to CoreValve TAVR or surgery (3.3% and 4.5%) (Table 1). Surprisingly, the 1-year mortality rate was lower in the TAVR group than in the surgical group (14.1% vs 18.9%, respectively), a finding sustained at 2 years in data presented at the American College of Cardiology conference in March 2015.^13–16

TAVR is superior to medical management, but the risk of stroke is higher

In the PARTNER B trial, inoperable patients were randomly assigned to undergo TAVR with a Sapien valve or medical management. TAVR resulted in lower mortality rates at 1 year (30.7% vs 50.7%) and 2 years (43.4% vs 68.0%) compared with medical management (Table 1).¹⁷ Of note, medical management included balloon valvuloplasty. The rate of the composite end point of death or repeat hospitalization was also lower with TAVR compared with medical therapy (44.1% vs 71.6%, respectively, at 1 year and 56.7% and 87.9%, respectively, at 2 years).¹⁷ The TAVR group had a higher stroke rate than the medical therapy group at 30 days (11.2% vs 5.5%, respectively) and at 2 years (13.8% vs 5.5%).¹⁷ Survival improved with TAVR in patients with an STS score of less than 15% but not in those with an STS score of 15% or higher.⁹

The very favorable results from the PARTNER trial rendered a randomized trial comparing self-expanding (CoreValve) TAVR and medical therapy unethical. Instead, a prospective single-arm study, the CoreValve Extreme Risk US Pivotal Trial, was used to compare the 12-month rate of death or major stroke with CoreValve TAVR vs a prespecified estimate of this rate with medical therapy.¹⁴ In about 500 patients who had a CoreValve attempt, the rate of all-cause mortality or major stroke at 1 year was significantly lower than the prespecified expected rate (26% vs 43%), reinforcing the results from the PARTNER Trial.¹⁴

Five-year outcomes

The 5-year PARTNER clinical and valve performance outcomes were published recently¹⁸ and continued to demonstrate equivalent outcomes for high-risk patients who underwent surgical aortic valve replacement or TAVR; there were no significant differences in all-cause mortality, cardiovascular mortality, stroke, or need for readmission to the hospital. The functional outcomes were similar as well, and no differences were demonstrated between surgical and TAVR valve performance.

Of note, moderate or severe aortic regurgitation occurred in 14% of patients in the TAVR group compared with 1% in the surgical aortic valve replacement group (P < .0001). This was associated with increased 5-year risk of death in the TAVR group (72.4% in those with moderate or severe aortic regurgitation vs 56.6% in those with mild aortic regurgitation or less; P = .003).

If the available randomized data are combined with observational reports, overall mortality and stroke rates are comparable between surgical aortic valve replacement and balloon-expandable or self-expandable TAVR in high-risk surgical candidates. Vascular complications, aortic regurgitation and permanent pacemaker insertion occur more frequently after TAVR, while major bleeding is more likely to occur after surgery.¹⁹ As newer generations of valves are developed, it is expected that aortic regurgitation and pacemaker rates will decrease over time. Indeed, trial data presented at the American College of Cardiology meeting in March 2015 for the third-generation Sapien valve (Sapien S3) showed only a 3.0% to 4.2% rate of significant paravalvular leak.

Contemporary valve comparison data

The valve used in the original PARTNER data was the first-generation Sapien valve. Since then, the second generation of this valve, the Sapien XT, has been introduced and is the model currently used in the United States (with the third-generation valve mentioned above, the Sapien S3, still available only through clinical trials). Thus, the two contemporary valves available for commercial use in the United States are the Edwards Sapien XT and Medtronic CoreValve. There are limited data comparing these valves head-to-head, but one recent trial attempted to do just that.

The Comparison of Transcatheter Heart Valves in High Risk Patients with Severe Aortic Stenosis: Medtronic CoreValve vs Edwards Sapien XT (CHOICE) trial compared the Edwards Sapien XT and CoreValve devices. Two hundred and forty-one patients were randomized. The primary end point of this trial was “device success” (a composite end point of four components: successful vascular access and deployment of the device with retrieval of the delivery system, correct position of the device, intended performance of the valve without moderate or severe insufficiency, and only one valve implanted in the correct anatomical location).

In this trial, the balloon-expandable Sapien XT valve showed a significantly higher device success rate than the self-expanding CoreValve, due to a significantly lower rate of aortic regurgitation (4.1% vs 18.3%, P < .001) and the less frequent need for implantation of more than one valve (0.8% vs 5.8%, P = .03). Placement of a permanent pacemaker was considerably less frequent in the balloon-expandable valve group (17.3% vs 37.6%, P = .001).²⁰

PREOPERATIVE CONSIDERATIONS AND EVALUATION CRITERIA

Currently, TAVR is indicated for patients with symptomatic severe native aortic valve stenosis who are deemed at high risk or inoperable by a heart team including interventional cardiologists and cardiac surgeons. The CoreValve was also recently approved for valve-in-valve insertion in high-risk or inoperable patients with a prosthetic aortic valve in place.

The STS risk score is a reasonable preliminary risk assessment tool and is applicable to most patients being evaluated for aortic valve replacement. The STS risk score represents the percentage risk of unfavorable outcomes based on certain clinical variables. A calculator is available at riskcalc.sts.org. Patients considered at high risk are those with an STS operative risk score of 8% or higher or a postoperative 30-day risk of death of 15% or higher.

It is important to remember, though, that the STS score does not account for certain severe surgical risk factors. These include the presence of a "porcelain aorta" (heavy circumferential calcification of the ascending aorta precluding cross-clamping), history of mediastinal radiation, “hostile chest” (kyphoscoliosis, other deformities, previous coronary artery bypass grafting with adhesion of internal mammary artery to the back of sternum), severely compromised respiratory function (forced expiratory volume in 1 second < 1 L or < 40% predicted, diffusing capacity for carbon monoxide < 30%), severe pulmonary hypertension, severe liver disease (Model for End-stage Liver Disease score 8–20), severe dementia, severe cerebrovascular disease, and frailty.

With regard to this last risk factor, frailty is not simply old age but rather a measurable characteristic akin to weakness or disability. Several tests exist to measure frailty, including the “eyeball test” (the physician’s subjective assessment), Mini-Mental State Examination, gait speed/15-foot walk test, hand grip strength, serum albumin, and assessment of activities of daily living. Formal frailty testing is recommended during the course of a TAVR workup.

Risk assessment and patient suitability for TAVR is ultimately determined by the combined judgment of the heart valve team using both the STS score and consideration of these other factors.

Implantation approaches

Today, TAVR could be performed by several approaches: transfemoral arterial, transapical, transaortic via partial sternotomy or right anterior thoracotomy,^21,22 transcarotid,^23–25 and transaxillary or subclavian.^26,27 Less commonly, transfemoral-venous routes have been performed utilizing either transseptal²⁸ or caval-aortic puncture.²⁹

The transfemoral approach is used most commonly by most institutions, including Cleveland Clinic. It allows for a completely percutaneous insertion and, in select cases, without endotracheal intubation and general anesthesia (Figure 1).

In patients with difficult femoral access due to severe calcification, extreme tortuosity, or small diameter, alternative access routes become a consideration. In this situation, at our institution, we favor the transaortic approach in patients who have not undergone cardiac surgery in the past, while the transapical approach is used in patients who had previous cardiac surgery. With the transapical approach, we have found the outcomes similar to those of transfemoral TAVR after propensity matching.^30,31 Although there is a learning curve,³² transapical TAVR can be performed with very limited mortality and morbidity. In a recent series at Cleveland Clinic, the mortality rate with the transapical approach was 1.2%, renal failure occurred in 4.7%, and a pacemaker was placed in 5.9% of patients; there were no strokes.³³ This approach can be utilized for simultaneous additional procedures like transcatheter mitral valve reimplantation and percutaneous coronary interventions.^34–36

Transcatheter aortic valve replacement (TAVR) has established itself as an effective way of treating high-risk patients with severe aortic valve stenosis. With new generations of existing valves and newer alternative devices, the procedure promises to become increasingly safer. The field is evolving rapidly and it will be important for interventional cardiologists and cardiac surgeons alike to stay abreast of developments. This article reviews the history of this promising procedure and examines its use in current practice.

HISTORICAL PERSPECTIVE

In 1980, Danish researcher H. R. Anderson reported developing and testing a balloon-expandable valve in animals.¹ The technology was eventually acquired and further developed by Edwards Life Sciences (Irvine, California).

Alain Cribier started early work in humans in 2002 in France.² He used a transfemoral arterial access to approach the aortic valve transseptally, but this procedure was associated with high rates of mortality and stroke.³ At the same time, in the United States, animal studies were being carried out by Lars G. Svensson, Todd Dewey, and Michael Mack to develop a transapical method of implantation,^4,5 while John Webb and colleagues were also developing a transapical aortic valve implantation technique,^6,7 and later went on to develop a retrograde transfemoral technique. This latter technique became feasible once Edwards developed a catheter that could be flexed to get around the aortic arch and across the aortic valve.

As the Edwards balloon-expandable valve (Sapien) was being developed, a nitinol-based self-expandable valve system was introduced by Medtronic: the CoreValve. Following feasibility studies,^5,8 the safety and efficacy of these valves were established thorough the Placement of Aortic Transcatheter Valves (PARTNER) trial and the US Core Valve Pivotal Trial. These valves are currently approved by the US Food and Drug Administration (FDA) for patients for whom conventional surgery would pose an extreme or high risk.^9–11

CLINICAL TRIALS OF TAVR

The two landmark prospective randomized trials of TAVR were the PARTNER trial and CoreValve Pivotal Trial.

The PARTNER trial consisted of two parts: PARTNER A, which compared the Sapien balloon-expandable transcatheter valve with surgical aortic valve replacement in patients at high surgical risk (Society of Thoracic Surgeons [STS] score > 10%), and PARTNER B, which compared TAVR with medical therapy in patients who could not undergo surgery (combined risk of serious morbidity or death of 50% or more, and two surgeons agreeing that the patient was inoperable).

Similarly, the CoreValve Pivotal Trial compared the self-expandable transcatheter valve with conventional medical and surgical treatment.

TAVR is comparable to surgery in outcomes, with caveats

In the PARTNER A trial, mortality rates were similar between patients who underwent Sapien TAVR and those who underwent surgical valve replacement at 30 days (3.4% and 6.5%, P = .07), 1 year (24.2% and 26.8%), and 2 years (33.9% and 35.0%). The patients in this group were randomized to either Sapien TAVR or surgery (Table 1).^10,12

The combined rate of stroke and transient ischemic attack was higher in the patients assigned to TAVR at 30 days (5.5% with TAVR vs 2.4% with surgery, P = .04) and at 1 year (8.3% with TAVR vs 4.3% with surgery, P = .04). The difference was of small significance at 2 years (11.2% vs 6.5%, P = .05). At 30 days, the rate of major vascular complications was higher with TAVR (11.0% vs 3.2%), while surgery was associated with more frequent major bleeding episodes (19.5% vs 9.3%) and new-onset atrial fibrillation (16.0% vs 8.6%). The rate of new pacemaker requirement at 30 days was similar between the TAVR and surgical groups (3.8% vs 3.6%). Moderate or severe paravalvular aortic regurgitation was more common after TAVR at 30 days, 1 year, and 2 years. This aortic insufficiency was associated with increased late mortality.^10,12

In the US CoreValve High Risk Study, no difference was found in the 30-day mortality rate in patients at high surgical risk randomized to CoreValve TAVR or surgery (3.3% and 4.5%) (Table 1). Surprisingly, the 1-year mortality rate was lower in the TAVR group than in the surgical group (14.1% vs 18.9%, respectively), a finding sustained at 2 years in data presented at the American College of Cardiology conference in March 2015.^13–16

TAVR is superior to medical management, but the risk of stroke is higher

In the PARTNER B trial, inoperable patients were randomly assigned to undergo TAVR with a Sapien valve or medical management. TAVR resulted in lower mortality rates at 1 year (30.7% vs 50.7%) and 2 years (43.4% vs 68.0%) compared with medical management (Table 1).¹⁷ Of note, medical management included balloon valvuloplasty. The rate of the composite end point of death or repeat hospitalization was also lower with TAVR compared with medical therapy (44.1% vs 71.6%, respectively, at 1 year and 56.7% and 87.9%, respectively, at 2 years).¹⁷ The TAVR group had a higher stroke rate than the medical therapy group at 30 days (11.2% vs 5.5%, respectively) and at 2 years (13.8% vs 5.5%).¹⁷ Survival improved with TAVR in patients with an STS score of less than 15% but not in those with an STS score of 15% or higher.⁹

The very favorable results from the PARTNER trial rendered a randomized trial comparing self-expanding (CoreValve) TAVR and medical therapy unethical. Instead, a prospective single-arm study, the CoreValve Extreme Risk US Pivotal Trial, was used to compare the 12-month rate of death or major stroke with CoreValve TAVR vs a prespecified estimate of this rate with medical therapy.¹⁴ In about 500 patients who had a CoreValve attempt, the rate of all-cause mortality or major stroke at 1 year was significantly lower than the prespecified expected rate (26% vs 43%), reinforcing the results from the PARTNER Trial.¹⁴

Five-year outcomes

The 5-year PARTNER clinical and valve performance outcomes were published recently¹⁸ and continued to demonstrate equivalent outcomes for high-risk patients who underwent surgical aortic valve replacement or TAVR; there were no significant differences in all-cause mortality, cardiovascular mortality, stroke, or need for readmission to the hospital. The functional outcomes were similar as well, and no differences were demonstrated between surgical and TAVR valve performance.

Of note, moderate or severe aortic regurgitation occurred in 14% of patients in the TAVR group compared with 1% in the surgical aortic valve replacement group (P < .0001). This was associated with increased 5-year risk of death in the TAVR group (72.4% in those with moderate or severe aortic regurgitation vs 56.6% in those with mild aortic regurgitation or less; P = .003).

If the available randomized data are combined with observational reports, overall mortality and stroke rates are comparable between surgical aortic valve replacement and balloon-expandable or self-expandable TAVR in high-risk surgical candidates. Vascular complications, aortic regurgitation and permanent pacemaker insertion occur more frequently after TAVR, while major bleeding is more likely to occur after surgery.¹⁹ As newer generations of valves are developed, it is expected that aortic regurgitation and pacemaker rates will decrease over time. Indeed, trial data presented at the American College of Cardiology meeting in March 2015 for the third-generation Sapien valve (Sapien S3) showed only a 3.0% to 4.2% rate of significant paravalvular leak.

Contemporary valve comparison data

The valve used in the original PARTNER data was the first-generation Sapien valve. Since then, the second generation of this valve, the Sapien XT, has been introduced and is the model currently used in the United States (with the third-generation valve mentioned above, the Sapien S3, still available only through clinical trials). Thus, the two contemporary valves available for commercial use in the United States are the Edwards Sapien XT and Medtronic CoreValve. There are limited data comparing these valves head-to-head, but one recent trial attempted to do just that.

The Comparison of Transcatheter Heart Valves in High Risk Patients with Severe Aortic Stenosis: Medtronic CoreValve vs Edwards Sapien XT (CHOICE) trial compared the Edwards Sapien XT and CoreValve devices. Two hundred and forty-one patients were randomized. The primary end point of this trial was “device success” (a composite end point of four components: successful vascular access and deployment of the device with retrieval of the delivery system, correct position of the device, intended performance of the valve without moderate or severe insufficiency, and only one valve implanted in the correct anatomical location).

In this trial, the balloon-expandable Sapien XT valve showed a significantly higher device success rate than the self-expanding CoreValve, due to a significantly lower rate of aortic regurgitation (4.1% vs 18.3%, P < .001) and the less frequent need for implantation of more than one valve (0.8% vs 5.8%, P = .03). Placement of a permanent pacemaker was considerably less frequent in the balloon-expandable valve group (17.3% vs 37.6%, P = .001).²⁰

PREOPERATIVE CONSIDERATIONS AND EVALUATION CRITERIA

Currently, TAVR is indicated for patients with symptomatic severe native aortic valve stenosis who are deemed at high risk or inoperable by a heart team including interventional cardiologists and cardiac surgeons. The CoreValve was also recently approved for valve-in-valve insertion in high-risk or inoperable patients with a prosthetic aortic valve in place.

The STS risk score is a reasonable preliminary risk assessment tool and is applicable to most patients being evaluated for aortic valve replacement. The STS risk score represents the percentage risk of unfavorable outcomes based on certain clinical variables. A calculator is available at riskcalc.sts.org. Patients considered at high risk are those with an STS operative risk score of 8% or higher or a postoperative 30-day risk of death of 15% or higher.

It is important to remember, though, that the STS score does not account for certain severe surgical risk factors. These include the presence of a "porcelain aorta" (heavy circumferential calcification of the ascending aorta precluding cross-clamping), history of mediastinal radiation, “hostile chest” (kyphoscoliosis, other deformities, previous coronary artery bypass grafting with adhesion of internal mammary artery to the back of sternum), severely compromised respiratory function (forced expiratory volume in 1 second < 1 L or < 40% predicted, diffusing capacity for carbon monoxide < 30%), severe pulmonary hypertension, severe liver disease (Model for End-stage Liver Disease score 8–20), severe dementia, severe cerebrovascular disease, and frailty.

With regard to this last risk factor, frailty is not simply old age but rather a measurable characteristic akin to weakness or disability. Several tests exist to measure frailty, including the “eyeball test” (the physician’s subjective assessment), Mini-Mental State Examination, gait speed/15-foot walk test, hand grip strength, serum albumin, and assessment of activities of daily living. Formal frailty testing is recommended during the course of a TAVR workup.

Risk assessment and patient suitability for TAVR is ultimately determined by the combined judgment of the heart valve team using both the STS score and consideration of these other factors.

Implantation approaches

Today, TAVR could be performed by several approaches: transfemoral arterial, transapical, transaortic via partial sternotomy or right anterior thoracotomy,^21,22 transcarotid,^23–25 and transaxillary or subclavian.^26,27 Less commonly, transfemoral-venous routes have been performed utilizing either transseptal²⁸ or caval-aortic puncture.²⁹

The transfemoral approach is used most commonly by most institutions, including Cleveland Clinic. It allows for a completely percutaneous insertion and, in select cases, without endotracheal intubation and general anesthesia (Figure 1).

In patients with difficult femoral access due to severe calcification, extreme tortuosity, or small diameter, alternative access routes become a consideration. In this situation, at our institution, we favor the transaortic approach in patients who have not undergone cardiac surgery in the past, while the transapical approach is used in patients who had previous cardiac surgery. With the transapical approach, we have found the outcomes similar to those of transfemoral TAVR after propensity matching.^30,31 Although there is a learning curve,³² transapical TAVR can be performed with very limited mortality and morbidity. In a recent series at Cleveland Clinic, the mortality rate with the transapical approach was 1.2%, renal failure occurred in 4.7%, and a pacemaker was placed in 5.9% of patients; there were no strokes.³³ This approach can be utilized for simultaneous additional procedures like transcatheter mitral valve reimplantation and percutaneous coronary interventions.^34–36

Transcatheter aortic valve replacement (TAVR) has established itself as an effective way of treating high-risk patients with severe aortic valve stenosis. With new generations of existing valves and newer alternative devices, the procedure promises to become increasingly safer. The field is evolving rapidly and it will be important for interventional cardiologists and cardiac surgeons alike to stay abreast of developments. This article reviews the history of this promising procedure and examines its use in current practice.

HISTORICAL PERSPECTIVE

In 1980, Danish researcher H. R. Anderson reported developing and testing a balloon-expandable valve in animals.¹ The technology was eventually acquired and further developed by Edwards Life Sciences (Irvine, California).

Alain Cribier started early work in humans in 2002 in France.² He used a transfemoral arterial access to approach the aortic valve transseptally, but this procedure was associated with high rates of mortality and stroke.³ At the same time, in the United States, animal studies were being carried out by Lars G. Svensson, Todd Dewey, and Michael Mack to develop a transapical method of implantation,^4,5 while John Webb and colleagues were also developing a transapical aortic valve implantation technique,^6,7 and later went on to develop a retrograde transfemoral technique. This latter technique became feasible once Edwards developed a catheter that could be flexed to get around the aortic arch and across the aortic valve.

As the Edwards balloon-expandable valve (Sapien) was being developed, a nitinol-based self-expandable valve system was introduced by Medtronic: the CoreValve. Following feasibility studies,^5,8 the safety and efficacy of these valves were established thorough the Placement of Aortic Transcatheter Valves (PARTNER) trial and the US Core Valve Pivotal Trial. These valves are currently approved by the US Food and Drug Administration (FDA) for patients for whom conventional surgery would pose an extreme or high risk.^9–11

CLINICAL TRIALS OF TAVR

The two landmark prospective randomized trials of TAVR were the PARTNER trial and CoreValve Pivotal Trial.

The PARTNER trial consisted of two parts: PARTNER A, which compared the Sapien balloon-expandable transcatheter valve with surgical aortic valve replacement in patients at high surgical risk (Society of Thoracic Surgeons [STS] score > 10%), and PARTNER B, which compared TAVR with medical therapy in patients who could not undergo surgery (combined risk of serious morbidity or death of 50% or more, and two surgeons agreeing that the patient was inoperable).

Similarly, the CoreValve Pivotal Trial compared the self-expandable transcatheter valve with conventional medical and surgical treatment.

TAVR is comparable to surgery in outcomes, with caveats

In the PARTNER A trial, mortality rates were similar between patients who underwent Sapien TAVR and those who underwent surgical valve replacement at 30 days (3.4% and 6.5%, P = .07), 1 year (24.2% and 26.8%), and 2 years (33.9% and 35.0%). The patients in this group were randomized to either Sapien TAVR or surgery (Table 1).^10,12

The combined rate of stroke and transient ischemic attack was higher in the patients assigned to TAVR at 30 days (5.5% with TAVR vs 2.4% with surgery, P = .04) and at 1 year (8.3% with TAVR vs 4.3% with surgery, P = .04). The difference was of small significance at 2 years (11.2% vs 6.5%, P = .05). At 30 days, the rate of major vascular complications was higher with TAVR (11.0% vs 3.2%), while surgery was associated with more frequent major bleeding episodes (19.5% vs 9.3%) and new-onset atrial fibrillation (16.0% vs 8.6%). The rate of new pacemaker requirement at 30 days was similar between the TAVR and surgical groups (3.8% vs 3.6%). Moderate or severe paravalvular aortic regurgitation was more common after TAVR at 30 days, 1 year, and 2 years. This aortic insufficiency was associated with increased late mortality.^10,12

In the US CoreValve High Risk Study, no difference was found in the 30-day mortality rate in patients at high surgical risk randomized to CoreValve TAVR or surgery (3.3% and 4.5%) (Table 1). Surprisingly, the 1-year mortality rate was lower in the TAVR group than in the surgical group (14.1% vs 18.9%, respectively), a finding sustained at 2 years in data presented at the American College of Cardiology conference in March 2015.^13–16

TAVR is superior to medical management, but the risk of stroke is higher

In the PARTNER B trial, inoperable patients were randomly assigned to undergo TAVR with a Sapien valve or medical management. TAVR resulted in lower mortality rates at 1 year (30.7% vs 50.7%) and 2 years (43.4% vs 68.0%) compared with medical management (Table 1).¹⁷ Of note, medical management included balloon valvuloplasty. The rate of the composite end point of death or repeat hospitalization was also lower with TAVR compared with medical therapy (44.1% vs 71.6%, respectively, at 1 year and 56.7% and 87.9%, respectively, at 2 years).¹⁷ The TAVR group had a higher stroke rate than the medical therapy group at 30 days (11.2% vs 5.5%, respectively) and at 2 years (13.8% vs 5.5%).¹⁷ Survival improved with TAVR in patients with an STS score of less than 15% but not in those with an STS score of 15% or higher.⁹

The very favorable results from the PARTNER trial rendered a randomized trial comparing self-expanding (CoreValve) TAVR and medical therapy unethical. Instead, a prospective single-arm study, the CoreValve Extreme Risk US Pivotal Trial, was used to compare the 12-month rate of death or major stroke with CoreValve TAVR vs a prespecified estimate of this rate with medical therapy.¹⁴ In about 500 patients who had a CoreValve attempt, the rate of all-cause mortality or major stroke at 1 year was significantly lower than the prespecified expected rate (26% vs 43%), reinforcing the results from the PARTNER Trial.¹⁴

Five-year outcomes

The 5-year PARTNER clinical and valve performance outcomes were published recently¹⁸ and continued to demonstrate equivalent outcomes for high-risk patients who underwent surgical aortic valve replacement or TAVR; there were no significant differences in all-cause mortality, cardiovascular mortality, stroke, or need for readmission to the hospital. The functional outcomes were similar as well, and no differences were demonstrated between surgical and TAVR valve performance.

Of note, moderate or severe aortic regurgitation occurred in 14% of patients in the TAVR group compared with 1% in the surgical aortic valve replacement group (P < .0001). This was associated with increased 5-year risk of death in the TAVR group (72.4% in those with moderate or severe aortic regurgitation vs 56.6% in those with mild aortic regurgitation or less; P = .003).

If the available randomized data are combined with observational reports, overall mortality and stroke rates are comparable between surgical aortic valve replacement and balloon-expandable or self-expandable TAVR in high-risk surgical candidates. Vascular complications, aortic regurgitation and permanent pacemaker insertion occur more frequently after TAVR, while major bleeding is more likely to occur after surgery.¹⁹ As newer generations of valves are developed, it is expected that aortic regurgitation and pacemaker rates will decrease over time. Indeed, trial data presented at the American College of Cardiology meeting in March 2015 for the third-generation Sapien valve (Sapien S3) showed only a 3.0% to 4.2% rate of significant paravalvular leak.

Contemporary valve comparison data

The valve used in the original PARTNER data was the first-generation Sapien valve. Since then, the second generation of this valve, the Sapien XT, has been introduced and is the model currently used in the United States (with the third-generation valve mentioned above, the Sapien S3, still available only through clinical trials). Thus, the two contemporary valves available for commercial use in the United States are the Edwards Sapien XT and Medtronic CoreValve. There are limited data comparing these valves head-to-head, but one recent trial attempted to do just that.

The Comparison of Transcatheter Heart Valves in High Risk Patients with Severe Aortic Stenosis: Medtronic CoreValve vs Edwards Sapien XT (CHOICE) trial compared the Edwards Sapien XT and CoreValve devices. Two hundred and forty-one patients were randomized. The primary end point of this trial was “device success” (a composite end point of four components: successful vascular access and deployment of the device with retrieval of the delivery system, correct position of the device, intended performance of the valve without moderate or severe insufficiency, and only one valve implanted in the correct anatomical location).

In this trial, the balloon-expandable Sapien XT valve showed a significantly higher device success rate than the self-expanding CoreValve, due to a significantly lower rate of aortic regurgitation (4.1% vs 18.3%, P < .001) and the less frequent need for implantation of more than one valve (0.8% vs 5.8%, P = .03). Placement of a permanent pacemaker was considerably less frequent in the balloon-expandable valve group (17.3% vs 37.6%, P = .001).²⁰

PREOPERATIVE CONSIDERATIONS AND EVALUATION CRITERIA

Currently, TAVR is indicated for patients with symptomatic severe native aortic valve stenosis who are deemed at high risk or inoperable by a heart team including interventional cardiologists and cardiac surgeons. The CoreValve was also recently approved for valve-in-valve insertion in high-risk or inoperable patients with a prosthetic aortic valve in place.

The STS risk score is a reasonable preliminary risk assessment tool and is applicable to most patients being evaluated for aortic valve replacement. The STS risk score represents the percentage risk of unfavorable outcomes based on certain clinical variables. A calculator is available at riskcalc.sts.org. Patients considered at high risk are those with an STS operative risk score of 8% or higher or a postoperative 30-day risk of death of 15% or higher.

It is important to remember, though, that the STS score does not account for certain severe surgical risk factors. These include the presence of a "porcelain aorta" (heavy circumferential calcification of the ascending aorta precluding cross-clamping), history of mediastinal radiation, “hostile chest” (kyphoscoliosis, other deformities, previous coronary artery bypass grafting with adhesion of internal mammary artery to the back of sternum), severely compromised respiratory function (forced expiratory volume in 1 second < 1 L or < 40% predicted, diffusing capacity for carbon monoxide < 30%), severe pulmonary hypertension, severe liver disease (Model for End-stage Liver Disease score 8–20), severe dementia, severe cerebrovascular disease, and frailty.

With regard to this last risk factor, frailty is not simply old age but rather a measurable characteristic akin to weakness or disability. Several tests exist to measure frailty, including the “eyeball test” (the physician’s subjective assessment), Mini-Mental State Examination, gait speed/15-foot walk test, hand grip strength, serum albumin, and assessment of activities of daily living. Formal frailty testing is recommended during the course of a TAVR workup.

Risk assessment and patient suitability for TAVR is ultimately determined by the combined judgment of the heart valve team using both the STS score and consideration of these other factors.

Implantation approaches

Today, TAVR could be performed by several approaches: transfemoral arterial, transapical, transaortic via partial sternotomy or right anterior thoracotomy,^21,22 transcarotid,^23–25 and transaxillary or subclavian.^26,27 Less commonly, transfemoral-venous routes have been performed utilizing either transseptal²⁸ or caval-aortic puncture.²⁹

The transfemoral approach is used most commonly by most institutions, including Cleveland Clinic. It allows for a completely percutaneous insertion and, in select cases, without endotracheal intubation and general anesthesia (Figure 1).

In patients with difficult femoral access due to severe calcification, extreme tortuosity, or small diameter, alternative access routes become a consideration. In this situation, at our institution, we favor the transaortic approach in patients who have not undergone cardiac surgery in the past, while the transapical approach is used in patients who had previous cardiac surgery. With the transapical approach, we have found the outcomes similar to those of transfemoral TAVR after propensity matching.^30,31 Although there is a learning curve,³² transapical TAVR can be performed with very limited mortality and morbidity. In a recent series at Cleveland Clinic, the mortality rate with the transapical approach was 1.2%, renal failure occurred in 4.7%, and a pacemaker was placed in 5.9% of patients; there were no strokes.³³ This approach can be utilized for simultaneous additional procedures like transcatheter mitral valve reimplantation and percutaneous coronary interventions.^34–36