Due to the ageing population and the consequent increased prevalence of aortic stenosis,1 the management of this disease in elderly patients, especially when associated with complex co-morbidities, is becoming an important issue for physicians. While surgical valve replacement remains the ‘gold standard’ treatment, operative mortality and morbidity increase with age and become significantly higher when pre-existing co-morbidities are present.2,3 As demonstrated by the EuroHeart Survey on Valvular Disease,4 as many as one-third of patients with symptomatic aortic stenosis are not undergoing surgery because of age, high-risk cardiac disease, other co-morbidities and/or simply lack of referral to surgery.5 However, with the development of transcatheter heart valve implantation the treatment of aortic stenosis for this subset of patients has entered a new era: with this technique, valves may be placed either through the femoral vessels, vein or artery (transfemoral), or through the apex of the left ventricle (transapical).
To date, two models of percutaneous aortic valve prosthesis have been used in investigational studies in elderly patients with excessive co-morbidities who have been deemed ineligible or very high-risk for surgical valve replacement: the balloon-expandable prosthesis from Edwards Lifesciences (first-generation: Cribier-Edwards valve; second-generation: Edwards-Sapien valve; Edwards Lifesciences Inc., Irvine, CA), first implanted in a human in 2002 by our group;6 and the self-expandable CoreValve prosthesis (CoreValve Inc., Irvine, CA), which was first described as an alternative technique in 2005.
The balloon-expandable prosthesis consists of a tubular slotted stainless-steel stent available in two sizes (23 and 26mm in diameter at full expansion) with an attached equine (first-generation) then bovine (second-generation) pericardial trileaflet valve. A sewn fabric cuff covers the left ventricular half of the prosthesis. A mechanical crimping device is utilised to compress the prosthesis onto a specially constructed valvuloplasty balloon catheter in order to reduce the balloon/valve assembly profile to make it compatible with a 22F (8mm external diameter) or 24F (9mm external diameter) sheath, depending on the diameter of the valve used. This prosthesis is delivered in the subcoronary position, anchored within the native calcified valve.
The CoreValve revalving system is a self-expanding prosthesis that uses a porcine bioprosthesis within a nitinol frame. There have been several device modifications since 2005, leading to a reduced sheath size from 24F to the current 18F device (third-generation). In contrast to the balloon-expandable prosthesis, anchoring with the self-expandable valve is achieved within the ascending aorta as well as within the native valve. Both prostheses have now been demonstrated to be effective, resulting in marked haemodynamic and clinical improvements and overcoming the numerous technical and clinical challenges encountered during the first clinical steps.
The balloon-expandable prosthesis was initially implanted on a compassionate basis in a population of elderly patients who had debilitating, life-threatening aortic stenosis with numerous co-morbidities; the challenging antegrade transeptal approach was used. After transeptal catheterisation, a stiff guidewire was advanced through the left atrium and the left ventricle and across the native aortic valve to the descending aorta, and externalised through the contra-lateral femoral artery. This guidewire offered an optimal support for advancing the prosthesis from the femoral vein to the aortic valve, but could exert traction on the anterior leaflet of the mitral valve with subsequent transient massive mitral regurgitation and haemodynamic collapse during the procedure. Even under optimal conditions, the management of this population of patients with severe aortic stenosis treated with a transcatheter heart valve represented a big challenge.
The results of these series – the Registry of Endovascular Implantation of Valves in Europe (I-REVIVE) and the Registry of Endovascular Critical Aortic Stenosis Treatment (RECAST) trials – have now been reported.7–9 Early mortality and morbidity rates were high, but there were also stunning successes: two patients are now reaching four years of follow-up having returned to normal life. These feasibility studies had a considerable impact on the future development of transcatheter valve implantation, validating the concept of non-surgical valve implantation. A homogeneous increase in valve area to 1.70cm2 (with the 23mm valve size, the only model then available) was demonstrated, with a negligible residual transvalvular gradient and an early increase in left ventricular ejection fraction. A long-standing clinical improvement was observed in all patients who survived their co-morbidities. There has been no coronary occlusion, no mitral regurgitation, no valve migration and no prosthesis dysfunction during follow-up.
With the first-generation delivery system, the retrograde approach – which was attempted in seven patients – was transiently abandoned. Advancing the large-profile device through diseased and tortuous femoro-iliac arteries and crossing the native calcified valve could be achieved in only 57% of cases. Technical improvements were needed to revive this simpler and more familiar approach to the aortic valve. This was obtained with the development of a deflectable guiding catheter (RetroFlex catheter). Simultaneously, a larger valve (26mm) was provided to decrease the high incidence (19%) of severe (grade 3) paravalvular leaks, which were observed particularly when the 23mm prosthesis was implanted in a large aortic annulus. At this time, the retrograde prosthesis implantation technique is the only transfemoral approach used that represents a significant enhancement with respect to delivery and procedural simplicity. The steerable RetroFlex catheter facilitates advancement of the prosthesis around the aortic arch (avoiding friction of the device over the aortic wall) and through the aortic valve without it getting caught in the commissures. The prosthesis is positioned accurately using aortic valvular calcification as a marker, and is then delivered using rapid ventricular pacing to decrease cardiac output and subsequently stabilise the device during balloon inflation. The initial experience reported by John Webb10 showed satisfactory short-term results with no intra-procedural deaths. The overall experience (Special Access in Canada, REVIVAL 2 in the US and REVIVE in Europe) is now based on the results of 200 implantations. The success rate is over 85%, with a 30-day mortality rate significantly lower than the EuroScore-predicted surgical mortality (7.8–12.1 versus 30.3–34.3%) in each series and with lasting functional improvement at one year (unpublished data, communications at Euro PCR 2007).
The transapical approach for prosthesis implantation has recently been described. In this approach, the same prosthesis and its delivery catheter are introduced under direct vision into the left ventricle via a mini-thoracotomy without cardiopulmonary bypass, through a ventricular puncture, and directed across the aortic valve with fluoroscopic guidance. This obviates the concerns of access sites in the presence of small-calibre femoral vessels and/or vascular occlusive disease. The feasibility of this technique was shown in animal experiments,11 and its application reported in humans.12,13 To date, the method has been applied in 143 patients with promising results in Canada, Europe and the US. The overall clinical experience with the balloon-expandable valve has reached 400 patients.
Several reports on the results obtained with the self-expanding CoreValve have been published since its first description in 2005.14,16 Within two years, numerous device modifications were made, leading to a reduction of the sheath size from 24F to 21F and to 18F for the current third-generation device. Using this smallest size, the technique can today be performed percutaneously using the retrograde approach. Transoesophageal echocardiographic guidance and haemodynamic support are no longer mandatory, representing a remarkable procedural improvement for the CoreValve. In the last report on 86 patients16 implanted with the 21F (n=50) and 18F (n=36) devices, acute device success was 88%; 30-day outcome showed a mortality rate of 12%, a myocardial infarction rate of 1% and a stroke rate of 10%. In the 63 patients with procedural success (acute device success and no peri-procedural major cardiovascular or cerebral events in the first 48 hours), mortality rate was 5%, and there was no stroke nor myocardial infarction. The results of longer-term follow-up of this series are expected.
Transcutaneous aortic valve implantation is a promising new technology with the potential to benefit many patients who remain untreated today. The feasibility of this technology has been demonstrated with both the balloon-expandable valve and the self-expanding system in a population of high-surgical-risk elderly patients. The retrograde approach is the simplest implantation technique, but remains limited by vascular access difficulties in frail older patients. A solution to this issue has been provided by the introduction of the transapical approach, which provides a rapid, sternal-sparing, beating-heart solution. The short- and mid-term haemodynamic results and clinical outcomes after transcatheter valve implantation have been encouraging, with no out-of-hospital valve failures so far. More data regarding late valve function and late clinical outcomes are still needed. A number of new, innovative devices are under investigation that may further decrease the delivery system profile and provide the potential for prosthesis recovery and repositioning to optimise implantation and placement.
Finally, a randomised clinical trial comparing transcatheter balloon-expandable valve implantation with surgical valve replacement in highsurgical- risk patients and with medical treatment in patients with contraindications to surgery (the PARTNER-US trial) has started in the US. The outcomes of this study will be crucial to determine the future role of the procedure in a broader spectrum of patients with aortic stenosis.