Background

Transcatheter aortic valve implantation (TAVI) is a new method enabling off-pump and truly minimally invasive treatment of aortic stenosis by means of retrograde transfemoral, transsubclavian or transaortic aortic valve implantation (AVI) or antegrade transapical AVI. In industrialized communities, aortic stenosis is the most frequently acquired heart valve disease and if untreated is associated with high mortality. Medical therapy is scarce and symptomatic patients have a high attrition rate of up to 50% in the first year . Thus, replacement of the degenerated valve is the only treatment option that improves symptoms and survival . For decades, conventional surgical aortic valve replacement (AVR) via sternotomy, using cardiopulmonary bypass (CPB), was the only therapeutic option and thus evolved as the gold standard. The number of elderly patients presenting with aortic stenosis, however, is steadily increasing in parallel with improved life expectancy, and despite the fact that surgical AVR can be performed in octogenarians with good results , there are many patients with increased risk profiles for surgical AVR due to advanced age and several comorbidities. In the past, these high-risk patients were not even referred to the cardiac surgeon and adequate treatment was denied .

With the introduction of TAVI by Professor Cribier in 2002 and its further development , a truly minimally invasive alternative treatment option for high-risk patients with severe aortic stenosis has evolved. TAVI allows for AVI without cardioplegic arrest and without the use of CPB. Patients are treated on a beating heart and the new prosthesis is implanted within the calcified native valve leaflets that remain in place while being squeezed aside. Since Professor Cribier’s pioneering initial clinical implantation in 2002, the fascinating technical achievements of transcatheter techniques that facilitate off-pump aortic valve implantations have developed further. Besides the mere technical factors that enable standardized transcatheter valve implantations, important additional developments have occurred. Most importantly, cardiologists and cardiac surgeons have come together within heart teams to decide jointly upon the best therapeutic options for each individual patient. Joint collaboration has probably been the most important improvement over the past decade and will lead to even further collaborative efforts for the sake of our patients in the future.

In this manuscript we will focus on some surgical aspects of these exciting, new and truly minimally invasive treatment options for patients with severe aortic stenosis.

Surgical aortic valve replacement (AVR) in 2012

Surgical AVR is a standardized procedure associated with low risks in most referred patients. Use of CPB allows for stable haemodynamics throughout the procedure; after aortic cross-clamping and cardioplegic cardiac arrest an aortotomy is performed and the calcified cusps are excised. Next, any modern xenograft or mechanical prosthesis is implanted using standardized suture techniques, which are associated with a very low risk of paravalvular leakage.

Besides conventional sternotomy, a minimally invasive approach can be routinely applied by means of partial upper sternotomy (J-cut in the third or fourth intercostal space) or parasternal right minithoracotomy in the third intercostal space. Access closure is performed with sternal wires and a standard suturing technique. Current results indicate low mortality rates of between 1% and 3% in an all-comers population.

Transcatheter aortic valve implantation (TAVI) from the beginnings to routine clinical practice

Since the first clinical patient was treated by Professor Cribier 10 years ago, TAVI has evolved as a standard therapy for high-risk elderly patients with aortic stenosis. Initially, TAVI was thought to be a transfemoral procedure, by means of a femoral venous and then a transseptal approach initially and via a retrograde transfemoral arterial approach subsequently. Both of these approaches were pioneered by Professor Cribier.

Cardiac surgeons were ‘non-believers’ in this new technique throughout the early years. Different thoughts, however, led to the initiation of unparalleled joint efforts between cardiac surgeons and cardiologists in order to further develop different therapeutic options for high-risk elderly patients: the awareness that a retrograde transfemoral approach may not be suitable in some patients, especially in presence of severe peripheral vascular disease; the thought that an antegrade approach from the left ventricular apex might be feasible; and the belief that due to their decades of experience of treatment of aortic valve disease, cardiac surgeons should be part of further technological developments. Mike Mack and Fred Mohr started an initiative at Edwards Lifesciences to work on a transapical antegrade access platform for TAVI. By means of the first generation of the Ascendra transapical delivery system (Edwards Lifesciences, Irvine, CA, USA), initial experimental experience was gathered in November and December 2004 at the laboratories in Irvine, USA and Leipzig, Germany. Cardiac surgeons from Dallas (Mike Mack and Todd Dewey), Frankfurt (Gerhard Wimmer-Greinecker and Mirko Doss) and Leipzig (Fred Mohr and Thomas Walther) formed a partnership to develop this procedure further. After ethical approval and based on experiences with transfemoral implantations, initial clinical implantations were done in two patients in Leipzig in December 2004, with the support of Professor Cribier ( Fig. 1 ). Two more patients were treated in 2005 in Frankfurt and Leipzig. However, due to the lack of a 26 mm prosthesis, sufficient oversizing was not applied and patients therefore had severe paravalvular leakage. Further experimental evaluation led to the development of an additional cuff around the SAPIEN prosthesis, to provide a better seal against irregularities at the aortic annulus and thus avoid paravalvular leakage. Towards the end of 2005, a 26 mm SAPIEN prosthesis became available, together with a clear concept of using oversizing. The clinical use of TA AVI was restarted successfully in February 2006 in Leipzig, followed by Vienna and Frankfurt.

Professor Alain Cribier supporting the first transapical aortic valve implantations at the Heart Centre in Leipzig in December 2004, immediately after the procedure. This image shows him between Professors Mike Mack (Dallas, USA; left) and Friedrich W. Mohr (Leipzig, Germany; right).

Current prostheses available for transcatheter aortic valve implantation (TAVI)

Over the years, different prostheses have become available for performing TAVI. Initially there were two prostheses, both of which received Conformité Européenne (CE) approval in 2008 and then became commercially available.

The CoreValve ^® (Medtronic, Minneapolis, MN, USA) is a porcine pericardial valve mounted on a self-expandable nitinol stent, which is available in three sizes (26, 29 and 31 mm); the delivery system allows for retrograde implantation only. The CoreValve ^® prosthesis has not yet obtained Food and Drug Administration approval.

The SAPIEN XT™ (Edwards Lifesciences, Irvine, CA, USA) prosthesis consists of bovine pericardial leaflets mounted on a balloon-expandable cobalt-chromium stent. The prosthesis is available in three sizes (23, 26 and 29 mm) and can be implanted using either the retrograde transfemoral approach or the antegrade transapical approach; the 29 mm valve is for the transapical approach only. A large randomized trial for FDA approval was completed recently using the previous SAPIEN™ valve . FDA approval has been granted for ‘inoperable’ patients according to the PARTNER cohort B study and will probably be granted for high-risk but ‘operable’ patients according to the PARTNER cohort A study in spring 2012.

In October 2011, two second-generation transcatheter valves for the transapical approach obtained CE approval: the JenaValve™ (JenaValve, Munich, Germany) and the ACURATE TA™ valve (Symetis, Ecublens, Switzerland). Both are porcine valves mounted on a self-expandable nitinol stent.

The main steps of transcatheter aortic valve implantation (TAVI)

The transfemoral and TA techniques allow for minimally invasive off-pump AVI. At the beginning of both procedures a ‘safety net’, including a venous back-up wire and an arterial sheath, should be placed in the femoral vessels to facilitate immediate conversion (femoral-femoral percutaneous cannulation) to CPB should any complications occur during the procedures .

For transfemoral AVI, puncture of both femoral arteries is necessary: one to deliver the device and one to insert the pigtail catheter. For balloon valvuloplasty, the native aortic valve is crossed retrogradely. Balloon valvuloplasty is performed under rapid pacing using a transvenous pacing wire. The prepared valve on its delivery system is introduced and positioned under fluoroscopic control after haemodynamic recovery from rapid pacing. The valve is deployed under a second brief episode of rapid ventricular pacing, once a good position is confirmed. After implantation, the valve function is assessed by either fluoroscopy and/or transoesophageal echocardiography (TEE). If any significant paravalvular leak is noticed, it is possible to reballoon the valve with a slightly larger balloon. For femoral artery closure after removing the delivery sheath, a closure device such as Prostar XL™ (Abbott Vascular Devices, Redwood City, CA, USA) can be used. Surgical closure of the femoral artery is an alternative option.

The transfemoral approach allows for a percutaneous procedure, which can be performed under local anaesthesia and then without TEE control. The drawbacks are the retrograde access with relatively long access wires and the necessity to cross the aortic arch retrogradely, thus risking stroke and leading to quite difficult valve positioning in some patients.

The transapical AVI approach is performed through a 5 cm short left anterolateral minithoracotomy at about the midclavicular line in the fifth or sixth intercostal space. After the pericardium is opened, the apex is secured with two teflon-pledgeted purse-string sutures. The apex can then be punctured, allowing for straightforward antegrade valve implantation. The native aortic valve is crossed antegradely with a soft guidewire, which is then changed to a super-stiff wire positioned down into the descending aorta. A pigtail catheter is positioned just above the aortic valve through the arterial sheath of the ‘safety net’.

The apical guidewire is used for the valvuloplasty balloon and the valvuloplasty is performed under rapid ventricular pacing. After haemodynamic recovery, the prepared valve with its delivery system is inserted through the apex and deployed under fluoroscopic guidance. Valve function and position are again assessed by TEE and fluoroscopy. Also, for TA-delivered valves, reballooning might be indicated in case of significant paravalvular leaks. After removal of the sheath and the guidewire, the apex is closed with the prepared purse-string sutures. Detailed step-by-step descriptions for both TAVI approaches have been published earlier .

The transapical approach offers some potential advantages, such as the short distance between the apex and the aortic valve, which allows for very precise implantation. As opposed to the transfemoral approach, where the sheath diameter is limited to the diameter of the femoral arteries, with the transapical approach there is no limitation to sheath diameter. The avoidance of the retrograde crossing of the aortic arch may be advantageous and may translate into a lower stroke rate with the transapical approach.

Transaortic and transsubclavian AVI have been introduced as two alternative approaches for implanting the CoreValve ^® prosthesis in patients where a transfemoral approach is not possible due to pronounced peripheral vascular disease. After performing either an upper partial sternotomy or a parasternal minithoracotomy, the ascending aorta is punctured for the transaortic approach, while for the transsubclavian approach the subclavian artery is exposed by surgical cut-down. For both approaches, the valve is implanted retrogradely in an analogous manner to the transfemoral approach.