Unlike the left-sided semi-lunar valve, clinically significant acquired pulmonary valve disease is rare. However, stenotic or regurgitant pulmonary valve lesions are not uncommon following surgery for congenital heart defects, including tetralogy of Fallot, pulmonary atresia and other surgical procedures requiring reconstruction of the right ventricular outflow tract (RVOT).1–4 Even with the use of valved conduits or bioprosthetic valves – such as homografts, porcine valves or Contegra – RVOT stenosis, regurgitation or a combination of both usually occurs five to 15 years following surgery. These valvular abnormalities may result in progressive right ventricular dilatation and dysfunction, leading to arrhythmias and premature death.2,4,5
Pure native pulmonary valve stenosis can usually be successfully treated with percutaneous balloon valvuloplasty.6 However, stenotic lesions occurring in surgically reconstructed RVOTs commonly do not respond well to simple balloon angioplasty; furthermore, they frequently present as a combined stenosis–regurgitant lesion. These combined lesions have been successfully treated with stent placement across the pulmonary valve, overcoming the stenosis but causing free pulmonary regurgitation.7,8 Initially, severe pulmonary regurgitation is tolerated well, but over time this may lead to decreased exercise tolerance, right heart failure, arrhythmias and death.8,9
Until recently, patients with severe pulmonary regurgitation were treated with surgical pulmonary valve replacement using a variety of valves, including homografts, cloth tubes with a valve sewn inside or the more recent Contegra conduits (bovine jugular veins with a valve inside). However, even with successful operations there is no good evidence that survival is better with these valves.10,11 Some authors have suggested that perhaps valve replacement is performed too late, which has raised the question of the appropriate timing for such operations.12,13 Even with good myocardial protection, repeat operations on an already compromised right ventricle may lead to further deterioration in its function.14 Therefore, the timing and indications for pulmonary valve replacement in patients with repaired complex congenital heart disease are controversial.10,12 Commonly accepted indications for pulmonary valve replacement are summarised in Table 1.
For all of the reasons stated above, transcatheter implantation of the pulmonary valve appears to be an attractive option in high-risk patients. Bonhoeffer performed the first transcatheter valve replacement in September 2000.15 Since then, he and others have implanted the Melody® valve (Medtronic, Minneapolis, MN, US) in over 500 patients (personal communication, William E Hellenbrand, 2008). With the recent advances in percutaneous aortic valve implantation, we have tested the Edwards Sapien™ (Edwards Lifesciences, Irvine, CA, US) transcatheter heart valve (THV) for implantation in the pulmonic position as an alternative to the Melody valve.
The Melody Valve
The Melody valve consists of a valve-containing bovine jugular vein sewn inside a balloon-expandable Cheatham platinum–iridium stent (see Figure 1). The valve is delivered via a 22Fr system and can be expanded up to a diameter of 22mm.16 It has been implanted in over 500 patients worldwide, with the largest series published to date reporting outcome data of 155 patients.17 The median age of the cohort was 21 years. Ninety-two per cent of the patients had a previous right-ventricle-to-pulmonary-artery conduit of some sort. In the vast majority of patients the valve was implanted via the femoral venous approach. Following implantation, the mean RVOT gradient fell from 37 to 10mmHg (p<0.001), and no patients had more than mild pulmonary regurgitation on angiography at the conclusion of the procedure. Seven major complications occurred (4.5%), including device dislodgement, homograft rupture and compression of the left main coronary artery. During a median follow-up period of 28 months, freedom from re-operation was 93% at 10 months and 84% at 50 months. Valvular competence was preserved during follow-up. Four patients died (mortality at 83 months: 3.4%), two of whom initially presented with cardiogenic shock and died during the peri-procedural period. The authors concluded that the Melody valve is an effective treatment for RVOT dysfunction and that deployment is highly successful and comparably safe. However, the authors also cautioned that percutaneous valve replacement should generally not be performed in patients with patch reconstruction of the RVOT (risk of rupture) and in very distensible homografts (risk of embolisation). In the same patient population, placement of the Melody valve has been shown to improve right and left ventricular performance, reduce symptoms and improve exercise tolerance.18
The Edwards Sapien Transcatheter Heart Valve
The Edwards Sapien THV is made of three bovine pericardial leaflets sewn inside a 14mm-long stainless steel stent (see Figure 2). The ventricular two-thirds of the stent is covered with a fabric sealing cuff designed to reduce the potential for paravalvular leaks. The valve is currently available in diameters of 23 and 26mm. It is mounted on a 23 or 26mm x 3cm high-pressure balloon and is delivered via a 24–26Fr sheath. The valve was initially designed for aortic valve replacement via an antegrade or retrograde approach. The antegrade approach has been completely abandoned. To facilitate delivery of the valve in a retrograde fashion through the tortuous aortic arch, a specially designed Retroflex catheter is used.19 The Retroflex catheter is also helpful for negotiating the venous system and the right-sided heart chambers. Implanted in the aortic position, no structural valve deterioration has been reported beyond four years following implantation, although mid- or long-term follow-up data are generally limited.20,21 The Edwards Sapien THV has recently received the CE mark in Europe and is currently undergoing a US multicentre randomised clinical trial evaluating its feasibility for replacement of the aortic valve in patients with aortic stenosis at high risk of surgical valve replacement (the PARTNER trial).
To evaluate the ease of placement of the Edwards Sapien THV in the pulmonic position, we performed experiments in two animal models. Results of the first 10 animals have been reported.22 It was found that tracking of the valve/stent assembly from either the jugular or femoral approach was easily possible. Secure placement in a previously placed right-ventricle-to-main-pulmonary-artery conduit was also demonstrated. On 13 December 2005 we then performed the first transcatheter placement of a pulmonary valve with the Edwards Sapien THV in a 16- year-old patient.23 The patient is now about 2.5 years post-valve deployment, with no obstruction across the valve and a competent valve. Since this initial experience, five more patients have undergone transcatheter placement of the pulmonary valve with the Edwards Sapien™ THV in the US, and four patients have done so in Canada. The cases in the US are part of an approved Investigational Device Exemption (IDE) clinical trial. Figure 3 demonstrates the steps of valve implantation in another patient with homograft obstruction who underwent recent valve implantation.
Comparison of the Edwards Sapien Transcatheter Heart Valve with the Melody Valve
The Edwards Sapien THV has not been compared with the Melody valve in any clinical setting. However, two important differences between the two valves are apparent. First, the Melody valve can only be expanded up to 22mm in diameter. However, many repaired pulmonary outflow tracts, especially those with a transannular patch repair for tetralogy of Fallot, have larger diameters, sometimes even exceeding 30mm, making placement of the Melody valve impossible.24 To overcome this problem a downsizing technique of the RVOT has been tested in animals using an hourglass-shaped covered stent placed in the RVOT.25 The Melody valve can then be placed within the narrowing of the covered stent. The Edwards Sapien THV is currently available in diameters of 23 and 26mm. Given the manufacturing process, different sizes can be generated, with additional diameters of 20 and 29mm currently planned. This would expand the spectrum of potential THV candidates to those with an aneurysmal RVOT.
Second, stent fractures are a common problem with the Melody valve, occurring in over 20% of cases.26 Some of those fractures may be clinically silent. However, in a recent publication describing 20 cases of early failure of percutaneous pulmonary valve replacement with the Melody valve, almost half of the failures (n=9) were due to stent fractures.27 In the small number of patients treated with the Edwards Sapien THV in the pulmonary position, no stent fractures have been observed to date. Furthermore, to our knowledge no stent fractures have been reported for the same valve in patients following aortic valve replacement, with over 800 implants (personal communication, Edwards Lifesciences, Inc., Irvine, CA, 2008).
Other Transcatheter Pulmonary Valves Under Development
A few other pulmonary THVs have been developed. Although not a percutaneously delivered THV, the development of the Shelhigh valve is probably the most advanced of those ‘other’ valves. It is a porcine pulmonary valve mounted inside a self-expandable stent. Designed for aneurysmal RVOTs, it is ‘injected’ via a small incision of the right ventricle following sternotomy. No cardiopulmonary bypass is necessary for placement of the valve.
Results of 10 patients have been published with favourable short-term results.28,29 Another THV mounted inside a self-expanding stent that can be implanted percutaneously has undergone feasibility testing in animals.30 The conceivable advantage of a self-expanding stent is better placement in ectatic RVOTs. However, the 24Fr delivery system is as large as the ones currently used in clinical practice. The third THV pulmonary valve under development is the small intestinal submucosa (SIS) valve, named after the surface of the valve, which consists of acellular matrix prepared from the jejunum of pigs. The valve has a square stent design and has so far undergone animal testing.31 The theoretical advantage of this valve is its low profile, which allows delivery of the valve through an 8Fr system.
Current Shortcomings and Outlook
More patients need to be enrolled and longer follow-up data have to be gathered in the recently initiated trial evaluating the performance of the Edwards Sapien THV in the pulmonic position in order to judge its performance compared with the Melody valve. The size of the delivery sheath limits the applicability of almost all THVs to patients above a weight of approximately 25–30kg; therefore, a smaller-diameter undeployed stent/valve assembly would be desirable. Another disadvantage of balloon-expandable stents/valves in general is the inability to reposition the device once it is deployed. Furthermore, if the valve cannot be delivered due to excessive tortuosity in the venous system and right-sided cardiac chambers, retrieving the valve once it is outside of its short delivery sheath is a complicated procedure.
Finally, after documenting the safety and deliverability of the THVs, timing of the procedure will remain an issue. When transcatheter valve placement of the pulmonary valve is fully introduced into clinical practice, the timing and indications of the procedure will have to be defined, as the main goal remains preservation of right ventricular function and potentially lethal arrhythmias. Hence, transcatheter valve replacement may be performed at earlier stages than the current practice of surgical pulmonary valve replacement.