The evolution of transcatheter pulmonary valve replacement

Keywords::

• bioresorbable
• melody
• pulmonary
• SAPIEN
• stent
• transcatheter
• valve

It is just over 10 years since Bonhoeffer et al. reported successful transcatheter delivery of a valved stent into a 12-year-old boy with stenosis and insufficiency of a prosthetic conduit from the right ventricle to the pulmonary artery [1]. This set the platform for nonsurgical replacement or repair of other heart valves and this is proving to be the most exciting advancement in the treatment of valve disease in many years. In less than a decade, transcatheter pulmonary valve replacement (PVR) has evolved rapidly through the inevitable technical learning curve and design modifications to establish itself as an acceptable therapy for right-ventricular (RV)–pulmonary artery (PA) conduit and bioprosthetic valve dysfunction with over 4000 valves implanted in this timeframe. Limitations remain in relation to valve size for patients with native outflow tracts. Although transcatheter PVR has been described with both established and novel approaches in this setting [2–5], surgery remains the preferred approach in the majority of cases. Extensive data collection through early clinical experience and clinical trials were necessary to prove safety and efficacy of the approach. Evolving data from these studies have demonstrated beneficial effects of transcatheter PVR in RV volume reduction [6], left-ventricular filling properties [7], exercise capacity [8] and electrical remodeling [9].

Initial human experience with Bonhoeffer’s design (consisting of valved bovine jugular vein sewn into a bare metal stent) consisted of 59 patients with 58 successful valve implants [6]. Three patients required acute surgical intervention due to stent dislodgment or conduit rupture. During a mean follow-up of just under 10 months, there was no mortality; however, device related complications were seen in 14 out of the 56 patients (25%). These included in-stent stenosis, referred to as the ‘hammock effect’ in seven patients due to lack of apposition of the valve to the stent. This observation lead to a change in device design with suturing of the whole length of the bioprosthetic valve tissue to the stent. Stent fracture, which has continued to be clinically relevant for this valve (Melody^® Valve [Medtronic Inc, MN, USA]) was noted in seven patients, with one patient undergoing a second ‘valve-in-valve’. More recent studies with the Melody valve have demonstrated improved outcomes with reduction in adverse events [10,11]. Further follow-up data from Bonhoeffer’s group following the initial report demonstrated reduction in procedural complications from 6 to 2.9% [10]. Recently a multicenter US clinical trial evaluating the Melody valve demonstrated impressive medium-term outcomes in 124 patients with dysfunctional RV–PA conduits. Freedom from Melody valve dysfunction or reintervention was almost 94% at 1 year [11]. An alternative transcatheter pulmonary valve has also become available, achieving CE approval in the EU and undergoing trials in the USA. The SAPIEN^™ transcatheter heart valve (Edwards Lifesciences LLC, CA, USA) has achieved widespread acceptance in the aortic position. A multicenter international clinical trial assessing short-term safety and efficacy in the pulmonary position demonstrated effective reduction of right-ventricular outflow tract (RVOT) gradient (27–12 mmHg; p < 0.0001) with improvement in clinical symptoms and maintenance of pulmonary valvular competence at 6 months follow-up [12].

As clinical experience has evolved so has the appreciation of the technical challenges and potential complications. Potential for coronary artery compression by the rigid stent in the RVOT was reported at 4.4% in the Melody US clinical trial [11]. The risk for conduit rupture has been reported also at 4% with many operators now choosing elective placement of a covered stent, if available, in patients with heavily calcified conduits. Stent fracture also remains an important event with the Melody Valve despite prestenting (5–16%) and is the most common reason for reintervention. The extent of stent fracture is also relevant to clinical outcomes, with higher grades of stent fracture more likely to need repeat intervention. Attempts to understand the impact of the hostile environment of the stenotic RVOT conduit on valve function are ongoing with loss of stent circularity and apposition to the anterior chest wall also associated with increased likelihood for reintervention [13]. Further attempts to use advanced imaging and computational models to categorize the impact of anatomical and pathophysiological variants on successful transcatheter PVR are ongoing [14].

Although rapid advances have taken place over the past 10 years with transcatheter PVR, longer-term outcomes are still lacking. The impressive rate of RV remodeling seen within the first 6 months following transcatheter PVR slows considerably with limited further changes in right-ventricular end-diastolic volume or ejection fraction as measured by MRI at 1 year [15]. However, it is likely that a similar effect occurs following surgical valve replacement, and it may be more prudent to focus on valve and stent durability rather than comparative effects on RV remodeling. Surgical valve replacement has had variable success and reports have demonstrated pulmonary regurgitant fractions ≥30% at 1 year in 7% [16]. This may be in part due to distortion or angulation of the configuration of the valve within the RV–PA conduit once it is placed within the chest, an issue somewhat circumvented by the valve being placed within a rigid stent. A prospective randomized clinical trial to compare surgical and transcatheter PVR to answer some of these questions may be unrealistic due to varying anatomical subtypes and potential patient preference for transcatheter PVR. The largest most contemporary dataset evaluating valve dysfunction and reintervention in adolescent patients undergoing surgical PVR revealed mean freedom from valvar dysfunction of 72% and mean freedom from reintervention of 90% at 5 years and mirroring these figures is a shorter-term goal [17]. Although risk factors for reintervention are being identified, the exact pathological mechanisms of valve degeneration in a host of different conduits have only been postulated on through case reports [18]. The impact of residual gradients, stent distortion and innate immunological responses on valve longevity need to be studied to ensure optimal results are achieved. The potential for valve-in-valve replacement potentially extending the number of repeat percutaneous valve replacements to an as yet undefined number [19] should not lead to complacency with regards to optimizing conditions to protect the initial valve delivered.

Therefore, the gauntlet for transcatheter PVR for the next 10 years has been well and truly laid down. Strategies should focus on four main issues: the first should be consolidating and improving upon current techniques to minimize procedural risk and simplify follow-up protocols, reducing cost, (which is not inconsiderable [20]) and inconvenience to the patient. Postprocedural cardiac MRI and cardiopulmonary exercise testing should only be implemented if clinical trial data suggests benefit. The use of a transthoracic echocardiogram has been demonstrated to provide a good estimate of the RV and RVOT indices in the setting of transcatheter PVR and is considerably more accessible and less costly than cardiac MRI. Complete follow-up data through postmarket approval registries is essential to ensure potential associations with patient substrate and valve dysfunction are identified early and approaches modified.

The second strategy should be aimed at valve development to extend transcatheter PVR to patients with severe PR and native RVOTs. Size restriction on the currently available valves (Melody, 22 mm [has been dilated up to 24 mm] and SAPIEN, 26 mm [29 mm SAPIEN valve available in Europe for the pulmonic position]) prevents deployment in the majority of patients (>70%) with significant pulmonary regurgitation. Clinical reports of a new valve sewn into a self-expanding nitinol frame have been described [5], however whether this valve will function appropriately over the anatomical and dynamic variability that exists within the native RVOT remains questionable and further modifications may be necessary. A pilot clinical trial assessing this valve is due to start in North America before the end of 2013. Other self-expanding valve systems are in development; however, clinical data are required before longer-term applicability is assessed.

The third endeavor should be to miniaturize the delivery system so that smaller children with dysfunctional conduits/bioprosthetic valves may be treated percutaneously. The newer generation of the SAPIEN valve (the SAPIEN XT) requires a smaller delivery system of 18–19 Fr (NovaFlex). However, this system has not been tested yet for deployment in the pulmonary position. In addition, newer low-profile pulmonary valves are being evaluated in animal models. One such a new valve is the Colibri Heart Valve (Colibri Heart Valve, LLC, CO, USA). This valve has been tested in a swine model and it requires 12–16 Fr delivery systems for valves ranging in size from 20 to 30 mm.

Ultimately, future aspirations must include efforts to merge these approaches with tissue engineering technologies to provide living autologous valve replacements with growth potential. The concept of a ‘living valve’ delivered on a bioresorbable scaffold with the potential to grow with the patient is one of the final destinations of this exciting journey although whether this is achievable in a 10-year window is uncertain. Animal models assessing this approach have been published [21] and if the rate of progress seen over the past decade continues into the next then perhaps we should not bet against it!

Financial & competing interests disclosure

ZM Hijazi is a non-paid consultant for Edwards Lifesciences, the company that manufactures the Edwards SAPIEN THV. ZM Hijazi is also a consultant for Colibri and has stock options. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript, apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.