During the past five years, transcatheter (T) aortic valve implantation (AVI) has evolved as one of the most important clinical innovations in cardiovascular disease. The specific merit of TAVI is to allow for minimally invasive treatment of severe symptomatic aortic stenosis (AS) in elderly and high-risk patients. Since 2005 retrograde transfemoral (TF) and antegrade transapical (TA) aortic valve implantations have been performed at an increasing number of centres to treat thousands of patients to date. Alternative approaches include retrograde trans-subclavian (TSc) and retrograde transaortic (TAo) access; both are being performed in some patients. In 2008, Conformité Européenne (CE) Mark approval was granted for the Corevalve™ (CV; Medtronic Inc., St. Paul, MN) and the SAPIEN™ (ES; Edwards Lifesciences Inc., Irvine, CA) prostheses. These first-generation devices are CE approved for TF (CV and ES) and TA (ES) implantations. Due to the fact that transcatheter aortic prostheses are being implanted without surgical suture fixation at the aortic annular level, some oversizing is being used to safely position the devices without major risk of dislocation. Currently available valve diameters are 26 mm and 29 mm for the CV (a 31 mm prosthesis is being introduced into clinical practice shortly) and 23 mm and 26 mm as well as a 29 mm valve (currently for TA implantation only) for the ES device. This allows treatment of the majority of high-risk elderly patients at present. Valve size selection depends on several anatomical factors diagnosed during precise pre-operative screening of the patients.
Indication – Patient Screening for Transcatheter Aortic Valve Implantation
Conventional surgical aortic valve replacement (AVR) has evolved as a standardised procedure with excellent short- and longer-term results;1 therefore, conventional AVR remains the ‘gold standard’ for all routine patients with a regular and not enhanced risk profile. Certainly there are discussions on performing TAVI even in younger and lower-risk patients; however, the excellent outcomes with the current approaches have to be always kept in mind.
Use of conventional AVR and TAVI should not be performed in a competitive, but rather in a complementary manner. Current indications for TAVI are seen in elderly patients with increased risk profile; this has been agreed upon by representatives of the European Society of Cardiology (ESC) and the European Association for Cardio-Thoracic Surgery (EACTS) and documented in a position statement.2
Patient screening before performing TAVI should be standardised. It should include routine assessment as for any patients who require cardiac surgery, including transthoracic echocardiography to exclude significant additional valve diseases and cardiac catheterisation to exclude relevant coronary artery disease. Specific risk assessment is then performed. This includes the calculation of two available scores: the Society of Thoracic Surgeons score and the European System for Cardiac Operative Risk Evaluation (EuroSCORE). There are limitations for both of them, for example, frailty cannot be assessed and the logistic EuroSCORE usually leads to quite some overestimation of the true risk. To overcome these limitations a specific aortic valve risk evaluation score (TAVI-Score) may be established in the future.
In addition, specific risk evaluation should take morphological parameters, especially on the stenosed aortic valve, into account. These include the assessment of the aortic valve annular diameter, the specific cusp anatomy and the amount of cusp calcifications.3 In addition, aortic root diameter, the pattern of coronary artery insertion and its distance to the aortic annulus as well as potential calcifications, angulation and kinking of the aorta should be assessed. Specific imaging required to perform these assessments includes transoesophageal echocardiography (TOE) and specific computed tomography (CT). After all these assessments the interdisciplinary heart team should decide upon the indication and how to treat the individual patient.
TAVI imposes new challenges on all partners, cardiologists and cardiac surgeons, who ‘traditionally’ diagnose and treat aortic stenosis. The treatment of elderly and co-morbid patients leads to an increased risk and the use of new techniques leads to some, at least initial, technical challenges. Therefore, the joint collaborative approach, working together as a heart team, is ideal to combine experience and thus solve all these issues. There may be initial challenges to overcome when working together; however, after some joint procedures all partners will realise that this is certainly beneficial. Teamwork includes shared experiences by one and the other partner, usually cardiologist and cardiac surgeon, so that both perform an even number of implants. Precise planning of the procedures as well as team discussion on who should solve potential complications is essential for success. The team needs to discuss potential adverse events and how to react to them in detail. Clear strategies will thus help to solve unforeseen events for the sake of the patient.
Besides the ‘core’ team of cardiologist and cardiac surgeon other partners such as dedicated cardiac anaesthetist, pump technician for eventual cardiopulmonary bypass (CPB), catheterisation laboratory assistants and operative theatre nurses should be included in the team.
Procedures such as TAVI should ideally be performed in a hybrid operating theatre. This offers the optimal sterile environment as in all regular operating theatres plus ideal fluoroscopic and angiographic imaging by means of a high-quality X-ray system as in a regular catheterisation laboratory. The second-best solution is to use a regular catheterisation laboratory with special focus on sterility. Optimal imaging is paramount and high-quality X-ray systems have evolved as the common standard for almost all TAVI procedures. Most importantly, the hybrid operative theatre offers sufficient space for all partners to work accordingly and allows for ideal imaging (TOE and fluoroscopic/angiographic imaging) to safely place and implant the prostheses.
Transcatheter Aortic Valve Implantation – Technical Aspects and Complications
TAVI can be performed in a standardised manner once the pre-operative screening has been completed, the patient has been discussed by the heart team and the decision whether to use a TF or TA approach has been taken. Several procedural steps are similar between both approaches: routine monitoring (electrocardiogram [ECG], central venous pressure [CVP], arterial blood pressure), general anaesthesia (TA and some TF patients) or local anaesthesia with conscious sedation, supine positioning, sterile draping and preparation of emergency equipment in case cardiopulmonary bypass (CPB) needs to be initiated. Procedural steps include a femoral venous guidewire that is positioned up to the right atrium and an arterial sheath to provide a ‘safety net’;4 a femoral access sheath or an anterolateral minithoracotomy for a TA access, passing a soft guidewire that is then exchanged to a super-stiff guidewire retrogradely (TF) or antegradely (TA) across the stenosed aortic valve; balloon valvuloplasty under rapid ventricular pacing (RVP); placement of the device; fluoroscopic and eventually TOE control for the ideal valve position; valve implantation; haemodynamic stabilisation; functional control to check the valve; and then retrieval of the application system. Both TF- and TA-AVI have evolved as routine procedures in many centres in the past five years. Technical steps of TA-AVI have been published in detail previously and have been updated recently.5,6
Most importantly, both heart team members, the implanting cardiologist and the cardiac surgeon, should communicate frequently during the procedure, discuss specific steps and decide critical parts together. Clear commands by the leading physician, usually one of the two partners who is leading the case, are helpful for the whole team, especially during critical steps of the procedure such as RVP for example and during periods of haemodynamic compromise. Good collaboration with the anaesthetist is essential as well.
All partners should keep in mind that ‘anything can happen at any time’ at an expected frequency of 10 %. Unfortunately, complications may occur in any patient and at any time; thus occurrence is unpredictable and the whole team needs to be prepared.
There are different potential complications, all of which the heart team should be aware of.
Haemodynamic compromise can occur before valve implantation (elderly high-risk patients usually require some volume loading and eventually a lower dosage of vasopressors, because they usually are on diuretics pre-operatively to preserve respiratory function), during RVP (which in some cases – fewer than 5 % – is not really tolerated) or immediately after valve implantation. The latter may occur most frequently. Besides intravenous vasopressors and volume loading (most importantly the ‘start off’ pressure before RVP should be sufficiently high), some low-dose intra-aortic inotropes (bolus of 10 μg adrenaline – be careful with coronary spasm) may act immediately and thus lead to improved and stable haemodynamics. Some ‘overshoot’ of the blood pressure should be anticipated and eventually treated. Femoral cannulation using the ‘safety wires’ and subsequent CPB support should be the standard approach whenever haemodynamic compromise cannot be treated by means of medical therapy.
This is a rare complication (overall incidence should be between 1 % and 2 %); however, it is one of the most dreadful ones. Coronary obstruction can be complete or partial, with or without immediate haemodynamic consequences that should be treated immediately (see Figure 1). It can be due to the frame of the implanted device directly or by some extensive native aortic valve cusp calcifications that are being pushed upwards and lateral towards one of the coronary ostia.
Patients with a narrow aortic root without prominent sinuses of valsalva are at an increased risk. First-line therapy is to place a guidewire and to solve the obstruction by coronary angioplasty and eventually stent implantation. In patients where such a procedure is not possible coronary artery bypass grafting (CABG) will be the therapy of choice.
Non-perfect Valve Position
This can be either too high with potential coronary artery obstruction or distal valve embolisation (in case the stent is not safely ‘anchored’ within the aortic annulus) or too low. A too-low position may lead to an instable position of the valve at the level of the left ventricular outflow tract. The native and stenosed aortic valve cusps may impose on top of the transcatheter valve during TOE imaging. Placement of a second prosthesis slightly higher than the first will stabilise the overall situation.7
This probably is the most common complication. Some minimal or mild paravalvular (PV) leakage occurs in almost every second patient. About 10 % of patients present with severe (equal or greater than 2°) PV leakage (see Figure 2). PV leakage most certainly is due to eccentric calcifications of the native aortic valve cusps, leading to imperfect apposition of the stent at the annular level. It may be treated by postdilatation using a slightly larger balloon (usually filled with one extra millilitre of fluid). In case of PV and central leakage postdilatation may not lead to any improvement; eventually overdilatation of the first valve and subsequent placement of a second valve (technical concept of a ‘washer’) should be considered.
This is rare, occurring in approximately 1 % of patients, and cannot be really predicted. Too much oversizing may be the major risk factor; however, the pattern of native cusp calcification will have an impact as well. Conversion to conventional surgery, usually by placement of a pericardial patch, is the best therapeutic concept (see Figure 3).
Partial Annular Perforation
This can occur in case the perforation appears to be at the membranous septum and causes a ventricular septal defect. This complication is rare as well and can be treated by placing a second prosthesis to cover the perforated parts. Surgical therapy can be considered as well.
New-onset Atrioventricular Block
This occurs in variable percentages between 10 % and 40 % in elderly high-risk patients. Standard therapy will be pacemaker implantation.
Transcatheter Aortic Valve Implantation – Results:
Before reflecting on the current TAVI results we should keep in mind that conventional aortic valve surgery is associated with an overall low 30-day mortality of 3 % in Germany.1
Of course, in most cases different patients are treated by TAVI, usually higher-risk profiles with frequent co-morbidities. However, as soon as we glimpse in the direction of treating lower-risk and younger patients with AS by means of TAVI we need to keep the very good conventional results in mind. Any new technique should keep up with the conventional standard. A prospectively unfortunately is not available, should be performed in an all-comers situation first, before the current indication for TAVI, which is restricted to higher-risk and elderly patients, can be changed.
There was quite some initial enthusiasm with TAVI that may reach a plateau after five years. Studies from the initial years 2006–2008 indicate some mortalities in high-risk patients of course; they usually were between 10 % and 15 % in patients at an estimated logistic EuroSCORE between 15 % and 30 %. Stroke rates were persistent between zero (exclusively TA series) and 5 %.8–14
Despite all optimism these results were not yet perfect. Explanations may be the initial learning and the (partially) high-risk profiles of the patients treated. Further distribution of TAVI to many sites, usually with proctored initial cases, led to a steady increase in patients treated in the years 2008–2010. More-recently published results, however, show similar outcomes. Thirty-day mortalities still are double digit in many studies. Besides, recommended inclusion criteria are not commonly applied, and as such patients with lower than 20 % logistic EuroSCORE risk profile account for almost 50 % of patients in a recent multicentre publication. Other potential inclusion criteria such as frailty were present in 17 % of patients only.15
Outcomes of TAVI heavily depend on proper risk assessment and on the selection process. An all-comers series usually carries higher risks than a selected series of patients where some critically ill patients may have been excluded. Patients on a waiting list usually have better outcomes in comparison with patients included regularly whenever coming to the hospital. Due to the fact that sicker patients may die on a waiting list such a situation may almost be ‘prohibitive’ to obtaining improved outcomes.
In addition, the choice of procedures, especially whether to perform a ‘TF first’ or even a ‘TF and TA’ approach will lead to differing results. Series with a TF-first approach, like the US Placement of aortic transcatheter valves (PARTNER) trial,16,17 will necessarily lead to better outcomes in the TF cohort, as only patients that have been ‘selected out’ will reach the TA cohort. Comparison of results will not be possible on a scientific basis.
Functional outcome of patients may be acceptable for most implants in higher-risk elderly patients. However, relatively high incidence of new-onset pacemaker implantation as well as the frequent occurrence of PV regurgitation is of concern. Especially younger patients with AS require a perfect result, therefore improved outcomes need to be guaranteed before treating lower-risk patients. Results of TA-AVI are similar to TF, especially when considering the usually higher-risk profile of patients receiving TA-AVI. Randomised comparative studies are not available; however, the Canadian comparative study proved similar outcomes at one and two years.18 Based on the current evidence therefore there is no indication for a ‘TF-first’ approach; patients should receive the antegrade and direct TA approach in a similar number. The heart teams should distribute the patients evenly.
The US PARTNER trial is an important randomised trial comparing TAVI with conventional surgery. Cohort A of the PARTNER trial consists of high-risk patients who are considered operable, while cohort B consists of patients who are considered inoperable.16,17 Overall results of this trial are excellent, especially the very low 30-day mortality. However, patients obviously were selected; this clearly is no all-comers trial, therefore many patients who would require therapy could not be included. From a European perspective it remains relatively unclear how patients with severe symptomatic AS are considered ‘inoperable’. Interestingly, cohort B was completed much earlier than cohort A. Despite all the enthusiasm about the good outcomes of the PARTNER trial, the incidence of stroke, which was higher than expected, is of some concern. Further efforts need to be undertaken to reduce the respective risk of stroke during TAVI.
For the future a prospectively randomised all-comers and three-arm study, conventional surgery versus TF-AVI versus TA-AVI, would be ideal. While such a study is neither in view nor can be realistically performed with sufficient patient numbers we will have to rely on national registries to obtain further outcome data from the ‘new’ versus the ‘old’ therapies. The German national registry (www.AKE.de), set up to include all patients treated for AS in the whole country for a total of five years, will give some answers in due course.
Perspectives for Transcatheter Aortic Valve Implantation
Despite all the achievements with TAVI, further improvements are required for the next generation of devices. Aspects are standard anatomic orientation of the prosthesis, improved sealing for PV leaks (the TA approach would even allow for improved devices that require larger implantation sheaths) and full repositioning. Several devices, Engager™ (Medtronic Inc., Minneapolis, MN, US) (see Figure 4), JenaValve™ (JenaValve Inc., Munich, Germany) (see Figure 5) and Accurate™ (Symetis Inc., Geneva, Switzerland) (see Figure 6), all of them based on Nitinol frames, have undergone first-in-man trials and are currently undergoing pivotal multicentre trials to obtain CE approval in Europe.18–21
Development of a TA closure device together with improved 3D online imaging may allow for a truly percutaneous TA approach and thus lead to even less invasive procedures for the sake of future patients. In addition, improved imaging modalities will lead to further simplified implantation procedures.22
In summary, over recent years TAVI has evolved to a routine, standardised, truly minimally invasive procedure with a sufficient safety profile and acceptable outcomes to treat high-risk elderly patients with symptomatic AS. TF and TA approaches show similar outcomes whenever similar patients are included. Newer, next-generation devices will allow for improved functionality, including repositioning and better sealing to prevent PV leakage.
In parallel to the development of TAVI a fruitful collaboration between cardiologists and cardiac surgeons has evolved. Within this collaboration we as heart teams are at the beginning of promising further technological developments for the sake of our patients.