Introduction

The estimated prevalence of CHD in adulthood is about 1 per 150 people and is growing because of advances in pediatric cardiology, cardiac surgery, and post-operative care. Half of the adult patients with CHD have undergone surgery in the past. Twenty percent of the newborns with CHD have anomalies affecting the right ventricular outflow tract (RVOT), such as tetralogy of Fallot (TOF) with or without pulmonary atresia, truncus arteriosus, transposition of the great vessels, common arterial trunk, and others, requiring surgical reconstruction with the use of trans-annular patch (TAP), bioprosthetic valve, or valved conduits [1]. Homograft implantation between the right ventricle and the pulmonary artery is also a part of the Ross procedure for aortic valve disease. Long-term durability of conduits, bioprosthetic valves, and TAP depends on the patient age, the cardiac defect, the type of tissue or material utilized, and the type of the operation. After RVOT correction with TAP, significant pulmonary regurgitation (PR) is observed in 48% of the patients directly after the operation and in 85% 2 years later. In patients after a correction using a valved homograft, significant RVOT degenerative dysfunction, PR, and/or pulmonary stenosis (PS) is observed in 50–55% of the patients during the first 10 years after the first operation and after 5–6 years after the second operation leading to repeat surgical interventions over lifetime [2, 3]. This can be associated with increased morbidity and mortality [4, 5] due to chest adhesions, bleeding, cardiac ischemia, arrhythmia burden, heart failure, and multi-organ dysfunction [6,7,8,9,10]. Patient management strategies have, therefore, been based on delaying surgical interventions for as long as possible, in order to minimize the number of surgical procedures. However, delaying surgery carries a risk of reaching a point of irreversible right ventricular (RV) dysfunction. A balance between the developing RVOT dysfunction and the need to minimize the total number of lifetime surgeries for a given patient lead to development of minimally invasive valve therapies. Since the introduction of the first balloon expandable valve in the pulmonary valve (PV) position [11], advances in interventional cardiology and tPVR have revolutionized the management of these patients. The availability of these minimally invasive and effective therapies may allow for earlier treatment of right-sided valvular disease before the onset of irreversible ventricular remodeling and dysfunction. Moreover, tPVR options can reduce the need for multiple surgical interventions over a patient’s lifetime, therefore affecting the morbidity of this growing patient population [12].

Pathophysiology and Clinical Presentation

Although PV dysfunction is often better tolerated than left-sided valvular dysfunction in the short and intermediate term, the long-term consequences are numerous and may include progressive exertional limitations and development of ventricular and supraventricular arrhythmias. Decompensated congestive heart failure is a late manifestation that usually occurs well after the development of exertional symptoms. In patients with progressive PS, RV hypertrophy develops as a compensatory mechanism for the increased pressure with resultant RV diastolic dysfunction and eventually systolic dysfunction ensues [13]. Severe chronic PR results in RV volume overload which leads to ventricular dilatation, progressive systolic and diastolic dysfunction, and tricuspid regurgitation (TR) due to annular dilatation. Ventricular and atrial arrhythmias are likely to occur in patients with both PS and PR and result in an increased risk of sudden death in this population [10, 14, 15]. Severe TR is often well tolerated for decades. However, much like PR eventual clinical sequelae will emerge. These include the development of ventricular and supraventricular arrhythmias, elevated central venous pressure with resultant multi-organ congestion (congestive hepatopathy, renal dysfunction, splenomegaly), and progressive RV dysfunction [16, 17].

Indications and Patient Selection

It is important to review the guidelines relating to surgical pulmonary valve replacement (sPVR) before discussing tPVR indications. It serves as an initial framework for thinking about tPVR. The major cardiovascular professional societies [the American Heart Association (AHA), Canadian Cardiovascular Society (CCS), and European Society of Cardiology (ESC)], have all published guidelines regarding timing of sPVR in asymptomatic patients with repaired TOF and significant PR, as outlined in (Table 1). All these sociaties depends on three major clinical categories of risk factors to write their guidelines: (1) age at initial repair, (2) electrophysiological markers, and (3) hemodynamic consequences related to severe PR [18,19,20,21]. The guidelines focus on significant PR, significant RVOT obstruction, RV volumetric criteria and systolic function, occurrence of arrhythmias, QRS duration, and the presence of hemodynamically significant residual cardiac defects. Symptomatic severe PR and/or PS are class I indications for PVR in all the guidelines. However, once symptoms have developed, the RV deterioration is often irreversible. There are no specific recommendations for tPVR in the abovementioned guidelines, which were all published in 2010 or earlier, prior to availability of data on outcomes of this technology and prior to the approval of the Melody valve in the USA. Therefore, without clear guidelines, it is convenient to adopt the indications for tPVR from the sPVR criteria discussed previously with some refining. However, such a practice is inherently problematic, because the published guidelines are largely focused on TOF patients with severe PR and RV dilation after prior TAP repair and this represent only part of the pathological spectrum of tPVR candidates which may include obstructive lesion or mixed disease, and RV dilation may not be as prominent. The instructions for use (IFU) for the Melody valve (Table 2), which are not strictly followed in practice, specify the existence of a full circumferential RVOT conduit that was ≥ 16 mm in diameter when originally implanted, and RVOT dysfunction with a clinical indication for intervention, with at least moderate PR and/or PS as defined in the US Melody Valve Investigational Device Exemption (IDE) trial. These IFU were derived from the inclusion criteria for the IDE trial, which also differentiated between symptomatic and asymptomatic patients: for New York Heart Association (NYHA) class II–IV, moderate or more PR and/or PS with a mean Doppler gradient 35 mmHg; and for NYHA class I, severe PR and/or PS with a mean Doppler gradient 40 mmHg [22]. Unlike the criteria for sPVR, there is no element of RV volume or function assessment included in these indications, and physicians often use sPVR guidelines for guidance on the timing of tPVR. Some experts argue that all patients with asymptomatic severe PR and/or PS should undergo tPVR prior to the development of RV dilation or dysfunction [23]. This aggressive early intervention approach must be weighed against the procedural risks, the potential for infective endocarditis (IE) and the potential decrease in the internal orifice diameter of surgically placed conduits or bioprosthetic valves with multiple tPVR over a patient’s lifetime [24].

Table 1 Recommendations for surgical pulmonary valve replacement in asymptomatic patients
Full size table
Table 2 Inclusion criteria for clinical trials with both the Melody and SAPIEN valves
Full size table

In our practice, we do a full assessment: history, examination, ECG, 24-h ambulatory ECG (Holter monitor), cardiopulmonary exercise testing (CPET), echocardiography, and cardiac MRI (CMR). Then, we use data generated from these tests to determine the optimal timing for tPVR according to the best available and most recent guidelines (Figs. 1 and 2).

Fig. 1
figure 1

Flow diagram for management of patients with primarily right ventricle outflow obstruction. PPG and MPG—peak and mean pressure gradient across the outflow tract, respectively. PS pulmonary stenosis, RV right ventricle, tPVR transcatheter pulmonary valve replacement

Full size image
Fig. 2
figure 2

Flow diagram for management of patients with primarily pulmonary regurgitation (PR). CMR cardiac magnetic resonance, RV right ventricle, PRF pulmonary regurgitant fraction, RVEF RV ejection fraction, RVEDVi RV end diastolic volume indexed, RVESVi RV end systolic volume indexed, CPET cardiopulmonary exercise testing, LVEF LV ejection fraction, tPVR transcatheter pulmonary valve replacement, RVOT right ventricle outflow tract

Full size image

In addition to clinical indications, several anatomical criteria need to be fulfilled for safe anchoring of the percutaneous valve. The ideal anatomy for tPVR is a uniform diameter from RVOT to pulmonary artery (PA) bifurcation with adequate main PA length to avoid stenting into the PA bifurcation. Current indications for tPVR are limited to dysfunctional surgical RVOT conduits with or without bioprosthetic valves of more than 16 mm in diameter. However, tPVR has been performed off-label in patients with native or post TAP repair, and patients with small sized conduits (< 16 mm) [25]. Absolute clinical contraindications for tPVR include presence of active infection and patients with occluded central veins (a hybrid approach may be performed in such patients). Recurrent IE and active intravenous drug abusers are also considered contraindications for this procedure [19, 21].

Percutaneous Pulmonary Valves

The number of valves developed for tPVR is increasing at a rapidly accelerating pace, with it being virtually impossible to present an overview of all developments in various regions of the world. However, thus far, the only two valves that are being used for tPVR on a large scale include the Melody valve (Medtronic Inc., Minneapolis, MN), as well as the SAPIEN valve (Edwards Lifesciences, Irvine, CA).

The Melody Valve

The Melody valve has been available in Canada and Europe since 2006 and was approved by the US FDA in 2010 with a HDE (humanitarian device exemption) [22, 26]. However, in 2015, it was fully approved as a PMA (pre-market approval) valve. The device consists of an 18­mm bovine jugular valve segment that is sutured onto a Cheatham platinum (CP) Stent (NuMED, Inc.; Hopkinton, NY) made of platinum and iridium (Supplement figure 1A). Although the leaflet morphology is highly variable, in general, all available leaflet variations function well. The deep coaptation of the leaflets means that the valve functions well over a wide spectrum of diameters. The initial length of the valve is 28 mm (34–36 mm crimped), but it is shortened in accordance with the final implanted diameter. The valve can be expanded from 16 to 22 mm in diameter and in some instances up to 24 mm. The valve is delivered via a 22F Ensemble® Transcatheter Delivery System (Medtronic) (Supplement figure 1B). This consists of a balloon­in­balloon (BiB) system (available in 16-, 18­, 20­, and 22­mm sizes), which enables the valve to be repositioned, if needed, after the inner balloon has been inflated. The tip of the system is blue in color, to correspond with the outflow suture of the stent. There is a retractable sheath that covers the stented valve during delivery and is pulled back just prior to deployment. Contrast can be injected via the retracted sheath from a side port to confirm the valve positioning. Proximally, there are three distinguished colored ports: the guidewire port (green), inner balloon inflation port (indigo), and outer balloon inflation port (orange).

The Edwards SAPIEN XT Percutaneous Pulmonary Valve

The SAPIEN XT valve consists of three bovine pericardial leaflets sewn to a cobalt chromium balloon expandable stent (Supplement Figure 2A) [27]. The stent has a unique fabric cuff that covers one half of the frame, limiting stent expansion. The valve is available in 20, 23, 26, and 29 mm diameters allowing for implantation in larger RVOT than the Melody valve. Pre­stenting with a baremetal stent to a diameter 2 to 3 mm less than the final valve diameter is performed in most cases, given the relatively short length of this valve (13.5–19.1 mm). Each size of valve is supplied with a matched delivery system, which incorporates a 30-mm-long, appropriately sized, non-compliant high-pressure balloon. The valve is tightly crimped (by specialized manual crimping tool) onto the shaft of the balloon catheter (Supplement Figure 2B) to minimize the valve profile. The NovaFlex catheter (Edwards Lifesciences Inc.) is utilized with the SAPIEN XT valve (Supplement Figure 2). After the introduction through the sheath, the valve is slid over, and aligned with, the deployment balloon. The NovaFlex system requires a 16 French sheath for the 23-mm valve, 18 French sheath for the 26-mm valve, and a 20 French sheath for the 29-mm SAPIEN XT valve. Edwards SAPIEN 3 Valve is the latest generation. It comes in three sizes 23, 26, and 29 mm. It has new enhanced frame geometry and wide strut angles for ultra-low delivery profile and high radial strength. It has low frame height and outer skirt made of polyethylene terephthalate designed to minimize paravalvular leak (for aortic implantation). This valve is currently under a clinical trial in the USA for pulmonary application (COMPASSION S3).

The Procedure and Technical Considerations

tPVR is usually performed under general anesthesia with preferably biplane fluoroscopic guidance together with onsite surgical coverage in case it is needed. Covered stents (whether commercial or “self­fabricated”) should be available in the unfortunate rare event of an expanding conduit tear, dissection, or rupture [28, 29]. Femoral approach is the most commonly used approach; alternatively, in some cases, jugular access can provide a more favorable anatomical curvature for valve delivery. Pre-treatment with aspirin and intravenous antibiotics before valve implantation is recommended. Heparin is required to maintain activated clotting time (ACT) longer than 250 s during the procedure. Right heart catheterization is initially performed to assess pressures and saturations with special attention to any relevant PA branch stenosis. It is important to use balloon-tipped catheter when performing these procedures. An inflated balloon ensures that the catheter crosses the largest effective orifice of the TV. This is important because the sheaths and delivery systems used in this procedure are relatively large, and if they are advanced through small chordal spaces, damage to the TV can occur. The balloon-tipped catheter is then replaced with a stiff guidewire (0.035 in [Lunderquist ultra stiff (Cook, Bloomington, IN)] and positioned into a distal branch PA as far as possible. Because of it is vertical orientation, the left lower pulmonary branch provide better support to advance the delivery system. To facilitate advancement, sometimes it is important to pre­shape the guidewire to conform to the shape of the conduit and the branch PA that is the intended target. Great care is needed for maintaining wire stability/position during device advancement to avoid the risk of PA injury. Biplane RVOT angiography is performed in at least two orthogonal projections to assess the proposed site for valve implantation. Measurements are made of the minimal diameter, the largest diameter, and the length of the conduit that needs to be covered by a stent. Implantation site dimension is then measured using a compliant sizing balloon (St Jude Medical, Plymouth, MN). This not only provides another way to judge the diameters within the conduit but, more importantly, gauges the compliance of the conduit. If a significant waist is seen on the balloon, this will alert the operator to an increased risk of conduit tear or dissection. In such circumstances, any stent or valve that needs to be placed in the conduit might need to be implanted at a smaller diameter before the conduit is post-dilated to its final intended diameter. Because the valve is a venous/pericardium graft and therefore “covered,” it is safer to dilate further and at higher pressure. In some patients with borderline or smaller diameter conduits, a significant waist can exclude them from tPVR. During a balloon test dilation, one can perform aortic root injection or selective coronary angiography (or both) in multiple projections to show the relationship of the expanded conduit to the coronary arteries (CA) (Supplement figure 3B). Anatomic variation of CA origin is not uncommon in a patient with CHD [30]. The lateral and LAO/caudal projections usually profile the CA very well in relation to the balloon position. Potential CA compression is not uncommon with 4.4% of the US cohort demonstrating unsuitable anatomy and deaths have been reported as a consequence of coronary compression in this setting [30]. Both the left or right CA may be at risk, although the left main or proximal left anterior descending arteries are the most often affected. Although non-invasive testing may demonstrate significant distance between RVOT and CA, all patients should have simultaneous coronary angiogram with balloon inflation of the RVOT. If there is no evidence of CA compression, tPVR is performed. If there is any question of potential CA compression, either from study of the angiograms or electrocardiographic changes, tPVR is aborted [31,32,33]. Pre-stenting of the conduit with a bare metal or a covered stent has several advantages. It is required in almost all cases: for the Melody valve, to reduce the risk of stent fracture and to maintain a circular configuration of the valve in the long term and to provide a landing zone for the inherently short Edward valve [34]. If significant recoil is seen after implanting the first stent, more than one stent need to be implanted to reinforce the conduit. The stent is usually deployed on a BIB catheter (NuMED Inc., Hopkinton, New York) to a diameter of up to 2 mm less than the original conduit size in stenotic conduits or slightly larger in conduits without stenosis. We typically use one of the following balloon-expandable stent: the IntraStent (ev3, Plymouth, MN), the Palmaz XL stents (Cordis), the CP stent (NuMED Inc.), or the AndraStent (Andramed GmbH, Reutlingen, Germany). Covered stents, when available, can be alternatively used if there are concerns of conduit rupture. In some cases, post-dilatation of the stent using a non-compliant balloon may be required to achieve the intended diameter. Post-stenting pressure measurements should document no to minimal residual gradient across the RVOT prior to proceeding with device implantation. The presence of residual stenosis has potential deleterious effect on exercise capacity and valve degeneration with higher RVOT gradients after tPVR; thus, it is important to abolish any residual gradient before proceeding with tPVR. The appropriate device size is then selected based on the diameter of the balloon at full inflation during pre-stenting. For the Melody valve, the choice of the Ensemble delivery system depends upon the size of the conduit (16, 18, 20, or 22 mm). For the SAPIEN valve, prosthesis size is determined based on the degree of conduit stenosis. After device selection and preparation, the delivery system is advanced over the stiff guidewire, bringing the valve into the desired implantation site. Occasionally, looping the system within the right atrium can facilitate advancing the delivery system when it is at the entrance of the conduit. This generates a forward force helping passage into the conduit. Once the Melody valve is in the appropriate position, it is unsheathed followed by gradual inflation of the inner balloon first then the outer balloons. Slow inflation allows time to adjust the position in case of device slippage. The SAPIEN valve is usually inflated on a single balloon slowly. Repeat angiography and pressure measurements are made to confirm a positive outcome. Occasionally, the implanted valves may require post dilatation with a non-compliant balloon. Transesophageal echocardiography is sometimes used during pulmonary valve implantation and may help with identifying the landing zone, paravalvular leak, and gradient, but the valve leaflets are not easily seen due to scatter from the frame, and if there is a suspicion of valve dysfunction, intracardiac echocardiography is superior.

At the end of the procedure, we routinely perform a figure-of-eight suture to achieve hemostasis. Some operators do preclosing the femoral venous access site with the suture-based Proglide device (Abbott Inc., Redwood city, CA) [35]. Patients are typically observed overnight in CCU or telemetry wards. Empiric IV antibiotics are administered. The next day, ECG, PA/lateral chest x-ray, and transthoracic echocardiogram should be performed. We still instruct the patient to take endocarditis prophylaxis before dental procedures. Regarding antiplatelet therapy, there is little data supporting specific antiplatelet regimen, and to date, no definite reports of thromboembolism have been reported. In our institution, we prescribe life-long low-dose aspirin. Patients are typically discharged the following day with arrangement for routine cardiology follow-up care including regular fluoroscopic evaluation. Supplement Figure 3 demonstrates few steps in a patient who had distal conduit narrowing with junction of the right pulmonary artery and pulmonary regurgitation. This patient received a 26-mm Edwards valve after stenting the conduit/RPA junction.

Efficacy and Outcome

Reviewing the data from the largest most recently published studies evaluating tPVR (Melody and SAPIEN valve) in > 650 patients are outlined in Supplement Table S1. Procedural success is generally high, with a mean valve deployment success rate of 95%. The mean procedural major complications rate was greater than 4%. Freedom from re-intervention was 86% over a mean follow-up of 26 months. The most frequent cause for re-intervention with the Melody valve was stent fracture despite pre-stenting (5–16%). Stent fracture has not been reported to date with the SAPIEN valve. The risk of infective endocarditis has been reported between 1 and 4% with the Melody valve [36].

In 2005, Khambadkone and colleagues reported the European results of 59 consecutive Melody valve implants. Their good hemodynamic and clinical results showed the overall safety of the procedure [37]. The initial US Melody Feasibility trial was a multi-center prospective nonrandomized study that focused on safety, procedural success, and short­term outcomes in 30 patients [22]. Significant sequelae were reported in three patients, but no deaths. In the short term, the valve function was good. This led to the FDA approval of use of the Melody valve under HDE status in January 2010. The trial continued, and later in 2010, McElhinney and associates reported midterm results of 136 patients in whom Melody valve implantation had been intended. One hundred twenty­four patients had undergone implantation of the Melody valve, and 12 had been excluded. Importantly, six patients had been excluded for risk of CA compression. Serious sequelae had occurred in 6% of the patients. Freedom from stent fracture of the valve was ~ 78% at 14 months, with a freedom from Melody valve dysfunction/re-intervention of 94% at 1 year. In this trial, not all patients had first undergone implantation of a baremetal stent within the conduit, an important technical advantage that has been shown to decrease the stent fracture rate of the Melody valve and perhaps to improve overall valve function. Another key endpoint of procedural success was the functional and clinical status as evaluated in accordance with NYHA functional class and exercise testing, this has been shown to improve in these patients after tPVR. tPVR resulted in acute reduction of RV pressure to a median of 42 mmHg and a reduction in peak gradient across the RVOT to a median of 12 mmHg. All patients had no or trace PR except one with moderate PR [26]. Cheatham et al. then reported on the hemodynamic outcomes of 148 patients who underwent the Melody valve implantation as part of the original US IDE trial [38•]. The 5-year freedom from re-intervention and explantation was 76 and 92%, with a conduit pre-stenting and a lower discharge RVOT gradient being associated with a longer freedom from re-intervention. Over a median follow-up of 4.5 years, almost all patients were in the NYHA class I or II. All but one of those patients who remained re-intervention free had mild or less PR, and the follow-up gradient remained unchanged from early tPVR. Recently, data from three prospective multi-center studies (300 patients) with Melody valve implantation conducted in Canada and Europe has shown that significant baseline TR, often seen in patients with RVOT dysfunction, was improved in 65% of the patients. Acute reduction of TR persisted over 5 years of follow-up [39]. This finding may extend indications for those patients with RVOT dysfunction suitable for tPVR who were referred for surgery because of concomitant TR.

Similar results to the Melody valve trial(s) were seen with the Edward SAPIEN valve. The COMPASSION trial (COngenital Multicenter trial of Pulmonic vAlve regurgitation Studying the SAPIEN interventIONal THV) was the first prospective multi-center study to assess the safety and efficacy of the SAPIEN valve for the treatment of dysfunctional RVOT conduits with moderate to severe PR with or without PS [27]. Among 36 patients, successful valve deployment was achieved in 33 of 34 attempts. On pullback peak-to-peak systolic gradient across the conduit decreased from 27 to 12 mmHg (p < 0.001). At 6 months follow-up, the number of patients in NYHA functional class I increased from 5 at baseline to 27. PR was moderate or less in 97% of the patients. Haas et al. reported their experience with 22 patients, the procedural success rate was 95.5%, with significant improvement in the hemodynamic measures [40]. Wilson et al. reported the Toronto experience with good technical success [41]. Faza et al. compared the results of tPVR of the SAPIEN and the Melody valves, with no significant difference in immediate results pertaining to residual gradients and no difference in PR on follow-up [42]. When looking after early results of SAPIEN valve implantation (up to 6 months follow-up), Chowdhury et al. found improvements in PV gradient, PR, and RV size [43]. The valve has received the CE Mark in Europe and recently the FDA approval in USA. This was supported by data from the multi-center COMPASSION clinical trial and additional clinical data from Europe.

Randomized data comparing the outcome of the surgical and tPVR approaches on the RV are lacking. Cost­analysis models have been proposed to compare the two methods of therapy. Although the actual tPVR is more expensive than surgically placed conduits, the total initial procedural costs are similar. Despite the less invasive nature of the procedure, tPVR resulted in slightly higher costs in the intermediate and long terms than did surgery [44, 45]. These estimates are, however, based on models, because long-term data for tPVR and its impact over a patient’s lifetime still remain to be seen.

Complications

Reported procedural complications after tPVR are outlined in Supplement Table S2. Acute hemodynamic complications during the procedure can result from (1) obstruction of pulmonary blood flow caused by valve dislodgement into the PA, (2) coronary ischemia resulting from coronary compression, and (3) major hemorrhage resulting from conduit rupture. The rate of serious complications in the US Melody trial was reported at 6%, including death from coronary dissection (n = 1), conduit rupture (n = 1), unstable arrhythmia (n = 1), wire perforation in distal PA (n = 2), and femoral vein thrombosis (n = 1). In the COMPASSION trial, the rate of serious complications was 21% (seven patients) with no deaths reported. Valve or stent migration occurred in four patients (three requiring surgical retrieval and one was deployed in the inferior vena cava), unstable arrhythmias in one patient, and self-limited wire perforation in the distal PA in two patients.

Stent Fracture

Stent fracture remains an important event with the Melody valve (5–16%) and remains the most common reason for re-intervention despite routine pre-stenting. Risk factors associated with stent fracture include: younger age; higher pre- and post-procedural RVOT gradient; smaller angiographic conduit diameter; stent recoil or compression after deployment; and valve position directly under the sternum [34, 46]. Nordmeyer et al. proposed a Stent fracture classification system based on the initial European experience [47]. Type I fracture involves disruption of one strut without loss of stent integrity. These can be seen in up to 40% of the patients but usually not associated with adverse effect. Type II involves fracture with loss of stent integrity, and type III describes fractures associated with separation of fragments. The extent of stent fracture is relevant to clinical outcomes. Type I stent fracture is likely to occur initially, and only 56% of the patients are free from type I stent fracture at 2 years from initial diagnosis of stent fracture and therefore require careful monitoring. Type II and III stent fracture may require re-intervention either by surgical replacement or repeat tPVR, as they are associated with early conduit restenosis and valve failure. Follow-up fluoroscopy at regular intervals is recommended for early detection of stent fracture. A second tPVR, if needed, can be performed similar to the initial implant. It is advisable, however, to implant another stent in all cases of stent prior to valve implantation [48]. Up to date, no stent fracture has not been reported with the SAPIEN valve.

Conduit Rupture

This complication may occur with either the Melody or SAPIEN valves. Pre- and post-deployment balloon dilations have the potential to cause a tear or rupture in the homografts/Contegra grafts and conduits. Risk factors include heavy calcification and a homograft substrate [49]. Although the incidence of such complications has been reported to be as high as 9%, most cases are not associated with hemodynamic compromise and can be successfully managed with a covered stent [25]. Rarely, surgical conduit replacement may be required after a rupture [34]. Despite the fact that Melody valve is a covered stent, the stent may not completely oppose against the inner surface of irregularly calcified conduit leading to an ineffective seal and leaking.

Valve Migration/Embolization

Valve embolization or migration may require surgical explanation. Retrieval of the valve to inferior vena cava (IVC) followed by stenting to flatten the valve leaflets has been reported; however, this bears the risk of injury to the RV, TV, and IVC. Specific to the SAPIEN valve, if wire position is lost prior to valve deployment, the valve cannot be removed out of the body percutaneously and surgical removal from the venous system may be required. Alternatively, the valve can be deployed in the IVC and stented as described above. On the other hand, since the Melody valve is entirely covered until deployed at the target lesion, the valve can be removed from the femoral vein before deployment if needed [27, 50, 51].

Infective Endocarditis

The risk of infective endocarditis following tPVR has been estimated at 2.4% per patient-year. More than one half of the cases do not directly involve the implanted PV, and most respond to antibiotics without the need for re-intervention However, infective endocarditis can also lead to valve explantation, need for a second procedure, or even sepsis-related mortality. A high residual RVOT gradient, the resulting turbulence, and in situ thrombosis have been implicated in the pathophysiology of post tPVR endocarditis [36]. Noncompliance with recommended antibiotic prophylaxis is also commonly reported in post tPVR endocarditis [52]. Most common organisms involved are Streptococcus viridans and Staphylococcus aureus. Rare organisms also have been isolated (Coxiella burnetii).

Mortality

Mortality associated with tPVR is rare and is most often related to comorbidities, rather than the procedure itself. Khanna et al. reported in-hospital mortality and rate of major adverse events of 0.9 and 2.2% after sPVR using the STS congenital heart surgery database respectively and 4.1 and 20.9% using the STS adult cardiac surgery database, respectively [53].

Future Development

Indications for tPVR are restricted to patients with RVOT diameter up to 22 mm for the Melody valve and up to 27 mm for the SAPIEN valve [54]. The majority of patients [> 80%] who are potential candidates for tPVR do not fulfill these criteria [55, 56]. There is worldwide interest in developing new devices (valves or RVOT reducers) and/or techniques to make these patients suitable for tPVR.

RVOT Reducing Techniques

One of the techniques reported involving double or triple metal stent implantation (Russian doll technique) preceding valve insertion performed in selected cases [57]. Another technique with anchoring multiple overlapping stents in one of the branch PA to then allow implantation of the valve into the meshwork protruding into the MPA (thereby jailing the opposite PA) [57]. Similar results have been documented for the Edwards SAPIEN transcatheter valve [58]. RVOT reducers were proposed by Boudjemline et al., who designed and developed in experimental studies several versions of a self-expandable stent, forming a covered double cylinder with external diameters 30–40 mm and the internal diameter which enables implantation of the valve [59].

Self-Expanding Valve Systems

In 2010, a new self-expandable percutaneous pulmonary valve (the Native Outflow Tract device, Medtronic Inc., Minneapolis, MN, USA) was successfully implanted in a 42-year-old patient with PR. The valve (Harmony™) is made off porcine pericardium and a self-expanding nitinol stent characterized by hourglass geometry (i.e., larger diameters at the proximal and distal end, smaller diameters in the central portion holding the valve) which should help the stability of the device in a large RVOT [60]. In April 2016, Medtronic announced the first clinical data of the Harmony tPVR in 20 patients. The early feasibility study is a non-randomized prospective study in three sites in the USA and Canada. Six-month follow-up data showed positive early outcomes [61].

The Venus P Valve (Venus Medtech, Shanghai, China) is another novel self-expanding percutaneous pulmonary device. It is composed of a tri-leaflet porcine pericardial valve mounted on a covered nitinol stent frame (Supplement Figure 4). It is manually crimped onto a delivery system that ranges from 14- to 22-F, depending on valvular size. The diameters and the lengths of the straight part of the stent range from 18 to 34 mm (in 2 mm increments) and from 20 to 35 mm (in 5 mm increments), respectively. Small case series have demonstrated short-term safety and efficacy of the Venus P Valve [62,63,64,65].

Recently, Kim et al. reported first-in-human implantation of a new self-expandable pulmonary valve in a patient with a native RVOT lesion using a newly made knitted nitinol-wire stent mounted with a tri-leaflet porcine pericardial valve developed in South Korea. It was feasible and showed short-term effectiveness. At present, a clinical trial is ongoing in South Korea to evaluate the safety and short-term effectiveness of the implantation of this newly made valved stent for the treatment of native RVOT lesions [66].

Valve in Valve

Gillespie et al. reported 104 cases of tPVR in bioprosthetic valves. At 1-year median follow-up, restenosis was observed in four patients, whereas none had more than mild PR. Two stent fractures were identified during follow-up, neither of which required RVOT re-intervention. Overall, freedom from re-intervention was > 90% at 2 years, and there was no procedure-related mortality [67].

The Small Patient’s Population

Berman et al. reported on 25 patients with a weight below 30 kg (13.8–29 kg) who underwent tPVR, with results very similar to those seen in larger patients [25].

Hybrid Procedure

Perventricular hybrid implantation through a subxiphoid approach is an option that should be considered in very small patients, or patients with a difficult percutaneous approach. This hybrid approach has been successfully used in both the Melody and the SAPIEN valves. Recently, Polish surgical-interventional teams reported on surgical reduction of the RVOT and trans-apical pulmonary valve implantation [68, 69].

Bifurcation Stenosis

Wilhelm et al. described a technique of implanting the Melody valve over two balloons using a Flower-Blossom technique into both branch PA, which is suitable in patients with a very distal obstruction of the MPA/conduit junction [70]. Another technical consideration for patients with bifurcation stenosis reported by Violini et al. is the implantation of a larger stent into a branch PA (thereby jailing the contralateral branch) followed by stent strut breakage with a high-pressure balloon with subsequent valve implantation in the proximal part of the stent that extend to the main pulmonary artery [71]. A further technical variation is bilateral Melody valve implantation into branch pulmonary arteries in patients that are not suitable candidates for implantation in usual pulmonary position [72]. For short landing zone, Jalal et al. reported a technique of folding both ends of the Melody valve stent, which was successfully performed in 10 patients (with use of the Ensemble delivery system), with excellent procedural and hemodynamic results [73].

SAPIEN XT and SAPIEN 3

For large RVOT (> 26 mm), there are currently the 29-mm SAPIEN XT and SAPIEN 3 valves available. Both have been recently used successfully in large native or patched RVOT with excellent results [74, 75].

Recently, Edwards Lifesciences introduced the Altera stent, one size (40 × 45 mm) nitinol self-expanding covered stent (hourglass) providing a rigid landing zone for SAPIEN S3 29 mm, suitable for RVOT diameters up to 38 mm.

Conclusion

tPVR has been consolidated as a safe and effective nonsurgical therapeutic alternative for RVOT dysfunction. In patients with dysfunctional surgical RVOT conduits, tPVR has a high procedural success rate, with immediate and durable resolution of RV-to-PA gradient as well as a very low incidence of post-procedure PR. Less than 25% of the patients who undergo tPVR will require re-interventions in a 5-year period, most commonly secondary to stent fracture in the percutaneous valve stent frame. Future developments in the field aim to reduce the incidence of complications, improve freedom from re-intervention rates, and, most importantly, expand the population eligible for this elegant procedure. The features of the new valves/devices should include a lower introducer profile, low inflammatory response, and long durability, low opening resistance with maximal valve area, as well as fast and reliable closure and as for all cardiovascular implants non-thrombogenicity. Clinical studies in off-label populations, innovative devices, and new techniques will likely expand the indications to native RVOTs, small-diameter conduits, and oversized patched RVOTs.

Parallel to that, advancements in imaging techniques will help better understanding the disease mechanism, it should be recognized that not all RV dysfunction arises from PS and/or PR, but other contributing factors such as intrinsic contractile abnormalities and mechanical dysynchrony should be considered, as tPVR may not treat the actual cause of dysfunction. In addition to the standard evaluation of biventricular function and size by echocardiography and CMR, the assessment of diastolic function, deformation imaging, and objective exercise capacity, may be able to better define and refine the appropriate timing for tPVR [76••].