Surgical aortic valve replacement is the reference standard for patients with symptomatic severe aortic valve stenosis.1,2 Despite the fact that the prognosis with medical management is poor, many patients do not undergo surgery due to an increased anticipated operative risk, driven by co-morbidities such as pulmonary hypertension, severe obstructive pulmonary disease, previous cardiac surgery, porcelain aorta, etc.2–5 The results with balloon aortic valvuloplasty, for decades the only interventional option for treatment of aortic valve stenosis, are beneficial in the acute phase, with clinical improvements, but unfortunately only palliative and short-lived;6–8 therefore, this approach is attractive are only as a bail-out strategy in unstable patients with acute valvular heart failure and in symptomatic ‘no-option’ candidates with reduced life expectancy.
Transcatheter aortic valve implantation (TAVI) utilising stent-based prostheses has emerged as a promising new option in recent years and has been used by a number of operators in different centres, with incremental success in line with procedural experience.9–17 This has sparked the evolution of more sophisticated techniques, ranging from the initial anterograde approach to gain access via the venous system with transseptal puncture to the currently used retrograde approach utilising arterial access. The technology has also significantly improved over the years, with the development of delivery catheters with smaller profiles and better prostheses with various size options.
Currently, two techniques are commercially available: the balloon-expandable Edwards SAPIEN prosthesis (ESP; Edwards Lifesciences; Irvine, CA, US) and the self-expanding CoreValve® re-valving prosthesis (CVP) system (Medtronic; Minneapolis, MN, US). Both devices have been described in detail in previous publications.9–17
The ESP (see Figure 1) is based on the prototype Percutaneous Valve Technologies, Inc. (PVT) prosthesis, first described and clinically tested by Cribier et al. in 2002.9 The ESP consists of a tubular slotted stainless steel stent with an attached bovine pericardial trileaflet valve and fabric sealing cuff. Two sizes are available: 23 and 26mm, covering annulus diameters from 18 to 25mm. The most recently introduced Novaflex catheter system houses the prosthesis at the distal tip and has a minimum profile diameter of 18Fr. The CVP (see Figure 2), first described by Grube et al. in 2005,12 consists of a trileaflet bioprosthetic porcine pericardial tissue valve that is mounted and sutured in a self-expanding nitinol stent frame. The prosthetic frame is manufactured by laser cutting of a nitinol metal tube with a length of 50mm. The pericardial valve is located in the transmission zone between the lower and the middle part, resulting in a supra-annular position compared with the native valve. The sizes of three subsequently developed delivery systems have been gradually reduced from 25 to 18Fr over time to facilitate vascular access and deployment of the device. The currently commercially available third-generation device (18Fr) is offered in two different sizes (26 and 29mm) for different annulus dimensions ranging from 20 to 27mm.
Today, TAVI is performed preferably under local anaesthesia and mild sedation or, alternatively, under general anaesthesia, mainly depending on the access site chosen. The transfemoral antegrade approach using a femoral vein with transseptal puncture was primarily used in the early days of transcatheter aortic valve replacement. Because of the complexity and the higher complication rate compared with retrograde access, this approach has been almost completely abandoned; the transapical antegrade approach is now a popular surgical approach for transcatheter aortic valve replacements. This technique was first described for the balloon-expandable valve prosthesis (Cribier–Edwards) in 2006 by Lichtenstein et al.18 The technique requires an anterolateral minithoracotomy, pericardiotomy and then puncture of the apex of the left ventricle using a needle through purse-string sutures. Today, this technique is evolving into a new standard surgical procedure. However, the use of transapical versus truly percutaneous transfemoral approaches varies significantly among heart centres, depending on operator experience and local politics. Being more invasive than a transfemoral approach, transapical access is a valuable route for patients with inappropriate peripheral access. Given the particular deployment characteristics, only the Edwards balloon-expandable valve prosthesis currently has the option of antegrade insertion.
The retrograde approach is currently the most popular method for TAVI. The various arterial access routes are iliac, femoral and subclavian. The first transcatheter retrograde device implantations described by Webb et al. (Cribier–Edwards prosthesis)12 and our group (CVP)14 were performed via iliac access, given the profile size of the early device prototypes. The progressive device size reduction over the past five years has led to truly percutaneous access in the common femoral artery. Subclavian access was first described by our group as an alternative route for patients with diseased iliac/femoral arteries.15 Access can be obtained after surgical cut-down either directly or after placement of a graft trunk onto the subclavian artery to both avoid access dissections and facilitate final closure. Because of the higher technical complexity as well as the need for surgical preparations, this approach should be considered a back-up technique if standard alternative insertion sites are inaccessible.
Direct surgical aortic access with thoracotomy and retrograde catheter-based implantation of a valve prosthesis is a new technique that has only recently been tested by cardiac surgeons.19 However, this technique is still in its infancy.
Clinical and Echocardiographic Outcome
TAVI is now a standardised interventional procedure with fairly predictable safety and efficacy. The procedural success rate has increased remarkably over the past five years, with rates between 90 and 99% in experienced hands. The mortality rate at 30-day follow-up is between 8 and 12% and the procedural stroke rate between 2.3 and 3.6% with current devices (see Table 1).15,17,20–23 Direct coronary obstruction is rare (<1%) and is reported only for the Edwards valve. Permanent pacemakers due to significant post-procedural atrioventricular conduction disturbances are needed in up to 20% of patients. It is currently unknown whether one device is superior to another due to the lack of randomised head-to-head comparisons. Available evidence suggests that the CVP is associated with a larger need for permanent pacemaker implantation, but this needs to be confirmed in properly conducted trials.
Patients with successful implantation procedures and uneventful 30-day follow-up improve clinically and benefit durably from the procedure. Immediate or late embolisation of a valve prosthesis is also rare, reported only for the Edwards device.24 As seen in all published series, haemodynamic status improves immediately after successful device implantation, with a reduction of the mean transvalvular gradient to around 10mmHg with a mean neo-orifice area of 1.5–1.8cm2.
Paravalvular aortic regurgitation, mostly mild or mild to moderate, is observed in around 70% of patients and is partially related to focal excessive load of calcifications, which prohibits optimal device expansion with complete paravalvular sealing.15,25 However, severe regurgitations are rare given the improved screening process (matching dimensions of annulus and prosthesis) as well as various procedural techniques and strategies to deal with high-degree regurgitations occurring immediately after device deployment. In long-term follow-up, serial echocardiographic studies have consistently shown good prosthetic valve function with no structural deterioration of valve tissue.15,16,26,27 Reports on transapical aortic valve implantation are available only for the ESP, mainly from two experienced centres.18,28 The implantation success rate was about 90%. The majority of patients are being treated off-pump, and the rate of peri-operative conversion is 9–12%. Mortality rates range from 9 to 18%. Strokes occur in about 0–6%. Non-randomised comparison of transapical and transfemoral approaches from the SAPIEN Aortic Bioprosthesis European Outcome (SOURCE) registry and Placement of Aortic Transcatheter Valve (PARTNER EU) trial indicate a higher mortality rate for the transapical approach in selective patients with higher estimated risk than patients undergoing retrograde TAVI. Randomised studies are needed to definitively assess the superiority of one technique over the other.
Optimal patient selection and pre-procedural screening play a crucial role in this context and should consist of multidisciplinary consultation between cardiologists, cardiac surgeons, anaesthesiologists and other specialists previously involved in the treatment of the patient. In previously published studies, the indication for TAVI was patients with severe (aortic valve area <1cm2 or <0.6cm2/m2) symptomatic aortic stenosis who are refused for conventional surgery. Correspondingly, the initial studies with the ESP and CVP included patients with contraindications conventional open-heart surgery, such as porcelain aorta, but also high-risk patients for surgery with a logistic EuroScore >20% or an STS score >10%.
After publication of the first clinical results showing procedural success rates of approximately 90% and following CE approval of the two devices, the European Association of Cardio-Thoracic Surgery (EACTS) and the European Society of Cardiology (ESC), in collaboration with the European Association of Percutaneous Cardiovascular Interventions (EAPCI), published a position statement29 that TAVI should be restricted to patients at high risk or with contraindications for surgery. They underscored the point that it is too early to establish TAVI in patients who are good surgical candidates.32 Therefore, the estimation of risk for surgery is based not only on quantitative assessment of risk scores but also on clinical judgement.
TAVI success is influenced by numerous factors including operator experience and patient selection; in particular, the operator learning curve affects the outcome in the first cases of each centre. However, even beyond this learning period, procedural outcome varies, with procedural failure rates of approximately 5–25%. Therefore, it is of importance to identify factors that predict the outcome of TAVI in clinical practice.
In a previous publication,21 we analysed various clinical, quantitative morphological and procedural parameters potentially affecting procedural outcome based on the combined TAVI experience of two institutions. In this analysis, the pre-procedural functional performance status of the patient was identified as an important outcome predictor for patients undergoing TAVI. This observation is of importance for future TAVI studies, as it might shift the clinical relevance of scoring systems from the historically used, more co-morbidity-based EuroScore or STS score towards functional assessment scales (frailty indices). It is not the sum of various individual co-morbidities that predicts the procedural risk, but rather their consequences for the functional and clinical status of the patient.
This finding of a correlation between pre-procedural frailty and post-procedural outcome may open the door towards inclusion of a healthier patient population. To date, it has been uncertain whether a complication after TAVI is procedure-related – suggesting that the surgical gold standard might have been a better alternative – or mainly related to the patient’s functional status and co-morbidities. The current data support the notion that the healthier the patient is before the procedure, the better the outcome afterwards. Accordingly, the complication and success rates of TAVI, as currently reported in patients at high surgical risk, cannot be translated to lower-risk patients with a good functional performance status. TAVI in patients at low surgical risk will almost certainly yield better feasibility and safety results than those currently reported.
Degenerative, formerly senile, aortic stenosis represents the only considerable indication for TAVI in a selective, high-risk patient population. Aortic insufficiency embodies another potential target for the use of transcatheter bioprostheses. However, an only mildly calcified native valve may hinder sufficient anchoring of the prosthesis and, moreover, the often-associated dilatation of the ascending aorta might contraindicate the sole treatment of the valve problem. Nevertheless, in patients with prior surgical valve replacement (biological valve), TAVI appears to emerge as a valuable treatment option.
Vascular access limits the use of transcatheter-based devices in a substantial number of patients. Further technical improvements are required to provide smaller sheaths and to reduce the associated risk of peripheral vascular complications. The latter have been reported with an incidence ranging from 2 to 15% and remain a substantial cause of morbidity and mortality. However, recent data reveal that an operator learning curve and the use of smaller sheaths as well as more detailed screening help to reduce this type of complication.16 Major cerebrovascular events occur in 2–4%17,21–23 and may be related to balloon dilatation, passing of the aortic arch with the device or a haemodynamic compromise during valve placement. Indirect comparison with surgical data has shown similar event rates,30 but further technical developments need to address this procedurerelated limitation of TAVI.
Mild to moderate aortic regurgitation is observed in up to 70% of the treated patient population, but severe, clinically relevant insufficiency remains rare in the case of correct matching of the prosthesis with the size of the aortic annulus as well as in the case of correct positioning.
The need for a permanent pacemaker implantation after TAVI is reported in up to 20% of cases. CVP seems to induce atrioventricular disturbances more frequently, probably because of the sustained radial force of the nitinol frame and an interference with the subvalvular, closely situated conduction system owing to a deeper extension into the left ventricular outflow tract.
Several of these limitations and issues with regard to the percutaneous treatment of aortic stenosis need to be addressed in future studies, as do future device development processes.
There are multiple new concepts under scientific as well as clinical evaluation designed to provide new features overcoming some of the shortcomings of the first-generation techniques. For example, the SADRA Lotus valve is a self-expanding device providing the ability to correct and reposition the valve prosthesis before releasing and permanently seating the device. The user is able to verify functionality and anchoring properties as well as the position of the device in relationship to the coronary arteries, the native aortic valve and the mitral apparatus before final implantation.31 As opposed to the ‘one-shot’ release of the balloon-expandable techniques, this technology of repositioning aims to offer a more controlled method of device deployment. Once the device is in the correct position, a locking mechanism stabilises the system. In addition, to avoid a paravalvular leak, a sealing membrane is attached to the outside of the prosthesis, filling paravalvular gaps. The first-in-man study using the SADRA Lotus valve is ongoing. The DirectFlow valve is another new concept that is repositionable and retrievable at all time-points during the procedure.32 The device consists of a stentless tissue valve with bovine pericardial leaflets that are located within an expandable non-metallic framework. Results of the ongoing first-in-man study will be available shortly.
The indication for TAVI is currently limited to a selective, high-risk patient population with severe, symptomatic aortic stenosis, but further possible indications have already been tested clinically or by compassionate use. As outlined above, this will further expand the spectrum of indications towards a lower-risk population, finally allowing TAVI to compete with the current surgical gold standard therapy even in this patient population.