Author + information
- Received September 8, 2010
- Revision received September 27, 2010
- Accepted October 1, 2010
- Published online November 1, 2010.
- Martyn Thomas, MD⁎ ()
- ↵⁎Reprint requests and correspondence:
Dr. Martyn Thomas, Cardiovascular Services, Guys and St Thomas' Hospital, London SE1 7EH, United Kingdom
Transcatheter aortic valve replacement is now a viable option in the treatment of high-risk severe symptomatic aortic stenosis. This review describes the current data with this technology and also the potential for the future role of the technology, including potential ways to yet further improve the short- and longer-term results.
Symptomatic severe aortic stenosis is a lethal condition with high 2-year mortality. Although surgical aortic valve replacement (SAVR) has excellent procedural results in elderly patients, as comorbidities and “risk” of the patients increase, so both the procedural and 1-year outcomes deteriorate (1). In addition, between 30% and 50% of patients with symptomatic aortic stenosis do not receive SAVR, often because they are not referred to a surgeon—generally because of concerns about comorbidities (2,3). Therefore it is highly likely that there might be a hidden population who would benefit from a less invasive way of replacing the aortic valve. Transcatheter aortic valve implantation (TAVI) allows the aortic valve to be replaced without a sternotomy, with a beating heart, and without the need for routine cardiopulmonary support.
The first TAVI with a balloon expandable device was performed in 2002 (4), whereas the self-expanding system was first used in 2005 (5). These 2 devices are now commercially available in Europe. Both the Edwards-SAPIEN Transcatheter Heart Valve (THV) (Edwards Lifesciences, Irvine, California) (balloon expandable) and the Medtronic CoreValve (Medtronic, Minneapolis, Minnesota) (self expanding system) became commercially available in 2007.
The least invasive way of performing a TAVI is the transfemoral (TF) retrograde approach to the aortic valve (5,6). However, in some patients the peripheral vasculature is not sufficiently large to allow the large bore sheaths and catheters to be used. In this circumstance the alternative approach is the transapical (TA) for the Edwards-SAPIEN THV (Edwards Lifesciences) (7,8) and the subclavian for the CoreValve (9). Further novel approaches are being developed, such as the transaortic. This approach is in development for both the Edwards-SAPIEN THV and the CoreValve (10,11). It involves direct access to the aortic arch via an upper mini-sternotomy and might be useful in patients with severe respiratory disease or deformities of the left chest, both of which are relative contraindications to a mini-thoracotomy of the TA approach.
Where Is the Procedure Performed?
It is generally agreed that top-quality imaging is required to safely perform TAVI. Therefore, the procedure should be performed either in a hybrid theater or in the cardiac catheter laboratory. Performing the procedure in the operating theater with a “C arm” is generally not advised. European guidelines have recommended that the procedure should be performed in centers that perform high volumes of SAVR (12). The U.K. recommendations are that TAVI should be performed in cardiothoracic surgical centers that have the ability to provide immediate access to cardiopulmonary bypass and other specialized services such as vascular surgery and renal support (13).
Patient Selection for the TAVI Procedure
Patients are selected for TAVI on the basis of their surgical risk and anatomical suitability. The risk of the patient is generally measured by the logistic European System for Cardiac Operative Risk Evaluation (EuroSCORE) (14) or the Society of Thoracic Surgeons (STS) score (15). Patients are, broadly speaking, considered suitable if they have a logistic EuroSCORE of >20% or an STS score of >10. The anatomical suitability is decided after a number of specialized imaging techniques. These include transthoracic and transoesophageal echocardiography, coronary and peripheral angiography, and computed tomography scanning from the aortic root to the common femoral artery. After the risk assessment and the specialized imaging, the patient should be reviewed by a multidisciplinary team to decide on the best approach to the treatment of the aortic stenosis. Treatment options include open SAVR, TAVI, or medical therapy. Both SAVR and TAVI might be preceded by coronary intervention or balloon aortic valvuloplasty (BAV). The BAV has re-established itself as a bridge to TAVI especially in patients with poor left ventricular function—the aim being to improve function and reduce the risk of the TAVI procedure. In addition, it might be used to treat patients (at least in the short term) who are not considered suitable for SAVR or TAVI. Finally, BAV might be used when there is diagnostic uncertainty as to whether the principal cause of symptoms is the stenotic aortic valve, such as patients with both important aortic stenosis and respiratory disease who are breathless. Response to BAV in addition to the measurement of pro–B-type natriuretic peptide might be useful in these circumstances. Coronary angioplasty is generally limited to those patients presenting with significant angina or, again, those with poor left ventricular function in an attempt to optimize function before TAVI.
Current Procedural and Outcome Data for TAVI: First-in-Man/Early Feasibility Studies
The first-in-man and early feasibility studies for TAVI represent the pioneering work and learning curve of centers throughout the world who developed the TAVI technique. The initial report of the Edwards-SAPIEN TF (Edwards Lifesciences) retrograde approach in 18 patients indicated a procedural success rate of only 78% and a 30-day mortality of 2 of 17 (11.1%) (6). A similar report for the CoreValve TF system (Medtronic) reported a procedural success rate of 84% and a 30-day mortality of 5 of 25 (20%) (5). The alternative vascular approach (generally with higher-risk patients) for the 2 devices are the TA approach for the Edwards-SAPIEN Valve and the subclavian approach for the Medtronic CoreValve. The first TA Report indicated a 90% procedural success rate and 13.6% mortality at 30 days (8). The data for the subclavian approach seem particularly encouraging. In 54 patients the procedural success was 100%, procedural mortality 0%, 30-day mortality 0%, and 6-month mortality 9.4% (9).
The Partner EU Registry was a pre-CE mark registry in 130 patients (61 TF and 69 TA), which has now reported 18-month outcomes, respectively (16). The 30-day survival was 92% for the TF approach and 81% for the TA approach. Eighteen-month survival was 71% for the TF approach and a disappointing 43% for the TA approach. These results might at least partly be explained by the higher risk nature of the TA versus TF patients (logistic EuroSCORE 33.8% for TA vs. 25.7% for TF). Importantly, much of the post-procedural mortality was noncardiac in nature and related to the comorbidities of the patients. Similar data from 126 patients in the early CoreValve (Medtronic) experience showed a 30-day survival of 84.8% and a 1-year survival of 71.4% (17).
Most Recent Data on Outcomes After TAVI
The most up-to-date data on the clinical outcomes of TAVI comes from a number of single center reports and multicenter and national registries.
Procedural and 30-day outcomes
The largest registry to report to date is the SOURCE (SAPIEN Aortic Bioprosthesis European Outcome) Registry. This is a post–CE-mark registry reporting on the results of 1,038 patients undergoing TAVI with the Edwards-SAPIEN THV (Edwards Lifesciences) (18,19). The Edwards-SAPIEN valve is currently commercially available in 23- and 26-mm sizes. Most of the published reports and presentations for the TF approach use the Edwards-SAPIEN valve and require a 22-F or 24-F sheath, and for the TA approach, a 26-F TA delivery system is used. The commercially available Medtronic CoreValve system is 18-F–compatible, and much of the reported CoreValve data is with this system.
Procedural results are now highly acceptable, with a success rate of >95% and a very low on-table mortality rate. Valve embolization rates (aortic and ventricular) are low (0.3%), as are coronary obstruction rates (0.6%). If coronary obstruction occurs, it is generally due to a native valve leaflet covering the coronary ostium rather than because of the device itself. This complication can often be predicted by the height and bulkiness of the native leaflets, the “depth” of the sinuses of Valsalva, and the distance between the leaflets and the coronary ostia. Therefore, careful attention should be paid to the pre-procedural morphology of the native valve and to the anatomy of the aortic root and sinuses of Valsalva. This complication seems to be rare and is similar for both the Edwards-SAPIEN THV (Edwards Lifesciences) and the Medtronic CoreValve.
The most important potential complication of both the TF and the TA approach is major vascular access site complications. With the TF approach, this has been reported in approximately 8% to 13% of cases using the Edwards system. Major vascular access site complications include obstructive dissections with limb-threatening ischemia and femoral or iliac rupture. These complications can now be managed with vascular occlusion balloons and stents (both covered and uncovered). Permanent pacemaker requirements for the TF and TA approaches with the Edwards-SAPIEN THV are approximately 5% to 7%, respectively. The incidence of stroke at 30 days is 2.5%, which is the same for both TA and TF. The exact mechanisms for these strokes are unclear but might be different for the 2 approaches. The 30-day mortality for the TAVI population is now in the region of 6% to 10%. In the SOURCE registry, it was 8.5% for the entire population, 6.3% for TF, and 10.8% for TA; this most likely reflects the higher-risk nature of the TA patients rather than an intrinsic difference in risk of the 2 techniques (18).
A similar registry report using the CoreValve system is the CoreValve Multicenter, Expanded Evaluation Registry (20). This reported 646 patients and 30-day outcomes. Procedural success was 97%; procedural mortality was 1.5%; 30-day all-cause mortality was 8%; and the combination of death, stroke, and myocardial infarction was 9.3%. With this third-generation 18-F device the vascular complication rate was low at 1.9%. The incidence of stroke at 30 days was 1.9%, and the requirement for a new permanent pacemaker was only 9.3%. A variety of other registries have reported the permanent pacemaker mismatch rate as somewhere between 18.5% and 42.5% (21–23).
The 1-year outcomes for cohort 1 of the SOURCE registry have recently been presented and represent the most up-to-date results with the current Edwards-SAPIEN THV technology (Edwards Lifesciences) (24). The Kaplan-Meier survival curves for the entire cohort, TA patients, and TF patients are shown in Figure 1. Kaplan-Meier survival rates were 76.1% for the cohort as a whole, 72.1% for the TA patients, and 81.1% for the TF patients. Most of the deaths between 30 days and 1 year were noncardiac, as shown in Figure 2, and most likely reflect the comorbidities of the patients.
The 1-year survival rates for the CoreValve (Medtronic) have recently been reported by 4 important European National registries (Fig. 3) with survival rates ranging from 71.9% in the 18-F safety and efficacy study (25), 78.8% in the Italian Registry (21), 79% in the Belgian Registry (22), and 81.6% in the U.K. Registry (23).
Summary of Current Outcomes
The TAVI procedure is now maturing and becoming predictable. One-year survival is more “patient related” rather than “procedure related.” Patients can be reasonably assured to expect the following outcomes: procedural mortality is very low, stroke risk is 2 to 4%, permanent pacemaker requirement is 4% to 8% with the SAPIEN THV and 20% to 40% with the CoreValve, 30-day mortality is 6% to 10%, and 1-year survival is >80% with the TF approach and >70% with the TA approach (the difference mainly being explained by the differences in comorbidities of the patients). The only real difference in outcomes between the 2 devices is the requirement for permanent pacemaker.
Context Versus SAVR
Much of the surgical data on high-risk patients concentrates on inpatient outcomes for patients such as octogenarians. Are there any comparators for the 1-year results of TAVI? The Leipzig group provided data in their paper from 2009 regarding octogenarian patients undergoing SAVR (1) (Fig. 4). The patients were divided into those with a logistic EuroSCORE of <10, 10 to 20, or >20. The 30-day survival rates were similar for all risk groups, ranging from 89% to 93%. The pattern for 1-year survival was very similar to TAVI and seems to reflect the risk of the patient rather than the risk with the procedure. Mortality rates between 30 days and 1 year for patients with a logistic EuroSCORE of <10 was 2.5% (1-year survival, 90%); a score of 10 to 20 was 11.3% (1-year survival, 78%), and for those with a logistic EuroSCORE of >20 (potentially a TAVI population) the mortality rate was 21.3% (1-year survival, 69%). Therefore, the gold standard for 1-year survival of high-risk (TAVI-type) SAVR might be seen as 69%. Thus the 1-year survival rate for TF TAVI of >80% is highly encouraging within this context.
Future Developments of the Devices
The latest iteration of the Edwards valve, the Edwards-SAPIEN XT valve with the Novoflex delivery system (Edwards Lifesciences), involves changes for both the TF and TA system. The valve has changed from stainless steel to cobalt-chromium, and the valve leaflet design has been modified. The Novoflex delivery system is now 18-F for the 23-mm valve and 19-F for the 26-mm valve. This has been achieved by a novel concept of loading/aligning the delivery balloon onto the SAPIEN XT valve in the descending aorta. The new Ascendra TA delivery system (Edwards Lifesciences) is 24-F and has also been made more ergonomically friendly. Improvements are also being made to the Medtronic CoreValve delivery system that should facilitate more accurate placement of the device. Multiple new devices such as the Sadra-Lotus valve and the Direct Flow valve are in development. The Direct Flow Medical aortic valve has reported encouraging early results (26).
Two areas where a further improvement might be seen in the near future are the incidence and outcomes of vascular complications and the prevention of stroke. In the SOURCE registry there was an important association between the occurrence of a vascular complication and 1-year survival for both the TF and TA approaches. The presence of a vascular complication reduced the 1-year survival from 83.9% to 72.2% for the TF approach and from 73.2% to 47.4% for the TA approach (22). Reduction in the French size of the Edwards-SAPIEN THV delivery system should result in a reduction in vascular complication rates. In addition techniques and devices are being developed to reduce the incidence of apical bleeding during the TA procedure. Both of these advances, one would anticipate, should lead to a further improvement in 1-year mortality for TAVI. The incidence of stroke during TAVI remains a troublesome 2% to 4%. The exact mechanism of these strokes remains unclear but might include traverse of the aortic arch by large-bore catheters, manipulation of the native diseased aortic leaflets, and hemodynamic insult to the brain during rapid pacing and device deployment. Two recent studies have reported “new” magnetic resonance imaging brain lesions in 70% to 80% of patients after TAVI, although this is rarely associated with clinical sequelae (27,28). The long-term consequences of this phenomenon are unknown. A new device, the Embrella embolic deflector (Embrella Cardiovascular, Inc., Wayne, Pennsylvania), was used for the first time during a TAVI case in January 2010. The device received a CE mark in May 2010, is placed via the radial artery, and is designed to “deflect” any embolic material away from the cerebral circulation (29). Hopefully this type of device will lead to a fall in the silent and clinical incidence of neurological events after TAVI.
Future Clinical Studies
It is widely recognized that the logistic EuroSCORE and the STS score are not ideal tools for measuring the pre-procedural risk of TAVI. Certainly, for the EuroSCORE the predicted mortality is grossly overestimated in the higher-risk patients. Therefore the development of a TAVI risk score is required. We recently attempted to construct such a score with the >1,000 patients in the SOURCE registry and 1-year outcomes (30). The EuroSCORE was not useful in predicting 1-year outcomes, generating a c-statistic of only 0.6 for TA and 0.55 for TF patients. This is hardly surprising, because most of the causes of death between 30 days and 1 year are noncardiac and are not captured in the EuroSCORE. Multivariate analysis was more successful but only moderately useful, generating c-statistics of 0.66 for TA and 0.71 for TF patients. The most likely reason for this failure to develop a risk score is that the numbers of patients required are likely to be much higher (the EuroSCORE and STS scores are based on 20,000 to 80,000 patients) and also perhaps the correct factors are not being captured. It is widely recognized that “frailty” plays a part in outcome and any risk score is likely to require such a measure, along with other clinical parameters. To collect data on such large numbers of patients it will be necessary to harmonize clinical end points in clinical studies. This will be facilitated by the Valvular Academic Research Consortium initiative in which experts in the field have agreed to standardized definitions of clinical end points after TAVI (31).
The most recent clinical study to report was “cohort b” of the Partner US randomized trial (32). This cohort was considered surgically inoperable and was randomized between medical therapy and TAVI. A large proportion of the medically treated patients received BAV. Despite this there is a large difference in 1-year all-comers survival (the primary end point), with a 69.3% survival in the TAVI arm compared with a 50.3% survival in the medical arm (p < 0.001). These results suggest that TAVI should now be considered the standard of care for patients deemed unsuitable for routine SAVR. The results of “cohort a” (a group of high-risk patients randomized between SAVR and TAVI with the TF or TA approach) will be reported in 2011.
The Future of the Technology: Where Are We Going and When Will the Risk Be Lowered?
Unlike new devices in the coronary world (which tend to be introduced into the lower-risk and technically “easy” patients), TAVI has started in the highest-risk and technically most challenging group of patients with aortic stenosis. This is to some extent appropriate, given that SAVR is an excellent gold standard therapy. Selection of the patient who is “ideal” for TAVI will be a major challenge. One important “unknown” is the durability of the transcatheter valves. Given the excellent results of SAVR in low-risk patients, it is unlikely that TAVI will have a major role in these patients for many years given the cost of the devices. Even if the device is highly effective, proving cost-effectiveness under these circumstances will be very difficult. This dilemma is demonstrated in a “theoretical” manner in Figure 5. In low-risk patients (however risk is measured) it is unlikely that TAVI will carry a mortality benefit over SAVR. Equally, in very high-risk patients, it seems unlikely that TAVI will again have any advantage. Somewhere between these 2 opposite ends of the risk spectrum, there should be a middle ground where the TAVI procedure has a mortality advantage over SAVR and therefore will prove to be cost-effective. Finding this patient population is the current challenge of the interventional community.
Transcatheter aortic valve implantation is maturing as an interventional technique for the treatment of severe symptomatic aortic stenosis. The procedural results are now predictable, and the “learning curve” can be shortened by proper training of the multidisciplinary teams. Our current challenge is finding the right patient population who should undergo the technique. This will be facilitated by the development of a “TAVI risk score,” which will replace the logistic EuroSCORE and the STS score as a measure of risk for the procedure. This technology is undoubtedly the biggest advance for patients and the interventional community since the onset of drug-eluting stent technology.
Dr. Thomas is an advisory board member for Edwards Lifesciences and has received research support from Edwards Lifesciences >$10,000 in the last year
- Abbreviations and Acronyms
- balloon aortic valvuloplasty
- European System for Cardiac Operative Risk Evaluation
- surgical aortic valve replacement
- Society of Thoracic Surgeons
- transcatheter aortic valve implantation
- Received September 8, 2010.
- Revision received September 27, 2010.
- Accepted October 1, 2010.
- American College of Cardiology Foundation
- Lung B.,
- Baron G.,
- Butchart E.,
- et al.
- Lung B.,
- Cachier A.,
- Baron G.,
- et al.
- Cribier A.,
- Eltchaninoff H.,
- Bash A.,
- et al.
- Grube E.,
- Laborde J.C.,
- Gerckens U.,
- et al.
- Webb J.G.,
- Chandavimol M.,
- Thompson C.R.,
- et al.
- Walther T.,
- Falk V.,
- Kempfert J.,
- et al.
- Petronio A.S.,
- De Carlo M.,
- Bedogni F.,
- et al.
- Bapat V.,
- Thomas M.,
- Hancock J.,
- Wilson K.
- Vahanian A.,
- Alfieri O.,
- Al-Attar N.,
- et al.
- British Cardiovascular Intervention Society
- Roques F.,
- Nashef S.A.,
- Michel P.,
- et al.
- ↵De Bruyne B. Medium term survival and functional status in patients with severe aortic stenosis treated by transcatheter aortic valvular implantation in the partner EU trial. Presented at: TCT September 21–25, 2009; San Francisco, California.
- ↵CoreValve Investigators. 12 months safety and performance results of transcatheter aortic valve implantation using the 18F CoreValve Revalving prosthesis. Presented at: TCT September 21–25, 2009; San Francisco, California.
- Thomas M.,
- Schymik G.,
- Walther T.,
- et al.
- Wendler O.,
- Walther T.,
- Nataf P.,
- et al.
- Piazza N.,
- Grube E.,
- Gerckens U.,
- et al.
- ↵Petronio AS. Italian Registry: TAVI facts, figures and national registries. Presented at: Euro PCR May 25–28, 2010; Paris, France.
- ↵Bosmans J. Belgian Registry: TAVI facts, figures and national registries. Presented at: Euro PCR May 25–28, 2010; Paris, France.
- ↵Ludman P. UK Registry: TAVI facts, figures and national registries. Presented at: Euro PCR May 25–28, 2010; Paris, France.
- ↵Thomas M. The SOURCE family of registries: 1-year results from cohort 1 of the SOURCE registry. Presented at: Euro PCR May 25–28, 2010; Paris, France.
- ↵Gerckens U, Safety MD. Durability and effectiveness at two years with the 18Fr CoreValve transcatheter aortic valve. Presented at: Euro PCR May 25–28, 2010; Paris, France.
- Ghanem A.,
- Müller A.,
- Nähle C.P.,
- et al.
- Kahlert P.,
- Knipp S.C.,
- Schlamann M.,
- et al.
- Nietlispach F.,
- Wijesinghe N.,
- Gurvitch R.,
- et al.
- ↵Wendler O. Development of a Risk Score for Transcatheter Aortic Valve Implantation: 1-year Outcomes from over 1,000 patients in the SOURCE Registry. Presented at: ESC Congress 2010; August 28–September 1, 2010; Stockholm, Sweden.
- ↵Serruys PW. Valvular Academic Research Consortium-consensus of event definition. Presented at: Euro PCR May 25–28, 2010; Paris, France.
- Leon M.B.,
- Smith C.R.,
- Mack M.,
- et al.
- Vascular Access
- Where Is the Procedure Performed?
- Patient Selection for the TAVI Procedure
- Current Procedural and Outcome Data for TAVI: First-in-Man/Early Feasibility Studies
- Most Recent Data on Outcomes After TAVI
- Summary of Current Outcomes
- Context Versus SAVR
- Future Developments of the Devices
- Future Clinical Studies
- The Future of the Technology: Where Are We Going and When Will the Risk Be Lowered?