Residual right ventricular outflow tract and pulmonary valve disease is common in adults with repaired tetralogy of Fallot. Chronic severe pulmonary regurgitation as a result of surgical repair can lead to myriad complications including right ventricular dysfunction, decreased exercise tolerance, right heart failure and symptomatic arrhythmias. The aim of restoring pulmonary valve integrity is to preserve right ventricular size and function with the intent of mitigating the development of symptoms and poor long-term outcomes. Right ventricular size thresholds by cardiac MRI have emerged beyond which reverse right ventricular remodelling after pulmonary valve replacement is less likely. Though pulmonary valve replacement has been shown to improve right ventricular dimensions and symptoms, no consistent improvement in right ventricular ejection fraction or objective measures of exercise capacity have been demonstrated. Furthermore, there are no long-term studies showing that normalisation of right ventricular size results in improved clinical outcomes. New transcatheter techniques of percutaneous pulmonary valve replacement have emerged with good short-term and mid-term outcomes, further adding to the complexity in determining ‘when’ and ‘how’ right ventricular outflow tract and pulmonary valve intervention should occur. With improved survival of these patients, the trend towards earlier pulmonary valve replacement at smaller right ventricular size and rapidly evolving transcatheter pulmonary valve techniques, the clinician must balance the goals of preserving right ventricular size and function in an attempt to prevent untoward outcomes with risks of multiple interventions in a patient's lifetime.
Statistics from Altmetric.com
Tetralogy of Fallot (TOF) is the most common cyanotic birth defect accounting for 3%–5% of all infants born with congenital heart defects.1 As one of the first congenital heart anomalies to be successfully surgically palliated, there is now a growing adult population living with the long-term sequelae of repair. A continual topic of debate in the field of congenital heart disease is the best management strategy for addressing dysfunction of the right ventricular (RV) outflow tract and pulmonary valve. It has been recognised that long-standing pulmonary regurgitation (PR) is detrimental to RV haemodynamics which is associated with adverse outcomes such as heart failure, arrhythmias and death.2 Over the years, the threshold for optimal timing of pulmonary valve replacement (PVR) has been refined and, along with the advent of transcatheter techniques, the principles of management have likewise evolved. In this paper, we will review current thinking and approach to residual pulmonary valve disease with a focus on PR in repaired TOF.
Pulmonary valve dysfunction as a postoperative phenomenon
Surgical repair of TOF is focused on closure of the ventricular septal defect and relief of RV outflow tract obstruction. Historically, there was little concern about the development of PR and many patients underwent repair with generous transannular patches with or without pulmonary valvectomy. Prolonged chronic volume loading of a postoperative RV with variable degrees of hypertrophy scar and compliance can result in progressive RV dilation and dysfunction (figure 1). Over time, patients develop worsening exercise tolerance, atrial and ventricular arrhythmias as well as sudden cardiac death, all of which become most prominent in the third and fourth decades of life.2
A minority of patients have residual pulmonary stenosis as the dominant lesion. Depending on underlying anatomy, these patients may have undergone valve-sparing techniques as a means to minimise PR.3 In the case of TOF with pulmonary atresia or those with coronary anomalies that cross the RV outflow tract, RV to pulmonary artery conduits can deteriorate over time resulting in conduit stenosis, regurgitation or both. The long-term haemodynamic consequence of pressure overload to the RV results in ventricular hypertrophy, diminished compliance and can lead to increased RV end-diastolic and right atrial pressures. Of concern, RV pressure overload and hypertrophy have recently been implicated as risk factors for poor outcomes including sustained ventricular tachycardia (VT) and sudden death late after TOF repair.4
Rationale and timing for PVR
Due to the recognition of untoward late outcomes associated with RV dilation and dysfunction from surgically induced chronic severe PR, attention has turned towards eliminating PR as the nexus of late complications5 and optimal timing for such elimination (figure 2). There is no debate surrounding the indication of PVR for significant PR in symptomatic TOF patients. However, there is continued controversy in asymptomatic patients.
Though physiologically distinct, lessons learnt from the effects of chronic aortic regurgitation on the left ventricle (LV) have suggested that chronic severe PR may also follow a similar course. After a compensatory phase in which semilunar valve regurgitation is well tolerated for many years, there is a maladaptive phase consisting of an initial reversible myopathy followed by an irreversible process characterised by myocardial dysfunction and scar. Current thinking on timing for PVR is focused on eliminating chronic PR before irreversible changes occur structurally, even on the cellular level.
Therrien et al introduced the notion of ‘operating too late’ for recovery or maintenance of RV contractility in patients whose RV function had already begun to deteriorate in the face of chronic PR.6 In addition to the contribution towards right heart failure and relationship with LV dysfunction, RV dysfunction has independent prognostic value for long-term outcomes. The INDICATOR cohort, a large multicentre registry of patients with TOF, examined 873 patients with cardiac MRI (CMR) data to identify risk factors for death.4 RV dysfunction was associated with death or sustained VT in addition to LV dysfunction, atrial tachycardia and RV hypertrophy, as mentioned above. However, there are mixed data as to whether PVR improves RV ejection fraction (EF). A recent meta-analysis found no significant improvement in RV EF and Geva et al found that preoperative RVEF <45% was associated with persistent RV dysfunction postoperatively, suggesting that PVR should occur prior to the development of RV dysfunction.7 ,8
Reverse RV remodelling
Attention has recently been focused on PVR prior to irreversible structural changes in RV size. The advent of CMR for the quantification of RV size and function has enabled and perpetuated the concept of a threshold value for RV size beyond which PVR is less likely to result in reverse RV remodelling. Initial studies suggested an indexed RV end-diastolic volume of 170 mL/m2 or end-systolic volume of 85 mL/m2, but thresholds have evolved to current values with a push towards earlier surgery.9–11 More emphasis has been placed on end-systolic volume as a more sensitive preoperative marker of postoperative RV function and later outcome RV size.12 ,13
Table 1 is a summary of the evidence on size threshold for PVR with a goal towards attainment of normal RV size.
Guidelines for PVR
In addition to parameters of RV size (including RV to LV end-diastolic size ratio greater than 2) and RV function (RVEF <47%), other clinical parameters such as LV function (EF <55%), RV outflow tract aneurysm, QRS>160 msec, sustained tachyarrhythmia and other residual haemodynamic lesions can all be considered in the timing of PVR.14
As mentioned above, some patients have predominant pulmonary stenosis and indications for intervention are similar to those with isolated valvar pulmonary stenosis and PVR is recommended for at least moderate RV outflow tract obstruction or greater than 2/3 systemic RV pressure.15 Catheter-based valvuloplasty may be an alternative to surgery in anatomically appropriate patients.
There are little data guiding clinicians on the management of mixed pulmonary valve disease with conventional recommendations geared towards intervention based on the dominant lesion. However, branch pulmonary stenosis increases afterload to the RV and is associated with worsening PR16 and for this reason.15
Current guidelines for PVR are summarised in box 1.
Current guidelines for PVR in adults with repaired tetralogy of Fallot
American College of Cardiology/American Heart Association Guidelines (2008)
1. PVR should be performed in adults with severe PR and symptoms or decreased exercise tolerance (Class 1, LOE B)
2. PVR should be performed in asymptomatic adults with severe PR and any of the following: (Class IIa, LOE B/C)
Moderate-to-severe RV dysfunction
Moderate-to-severe RV enlargement
Development of symptomatic or sustained atrial and/or ventricular arrhythmias
3. PVR should be performed in those with residual valvular or subvalvular RV outflow tract obstruction and any of the following: (Class IIa, LOE C)
Peak instantaneous echocardiography gradient >50 mm Hg
RV pressure >0.7 left ventricular pressure
Progressive and/or severe dilation of the RV with dysfunction
Canadian Guidelines (2009)
The situations that may warrant PVR are as follows: (Class IIa, LOE C)
Free PR associated with progressive or moderate-to-severe RV enlargement (RV end-diastolic volume >170 mL/m2)
Moderate-to-severe RV dysfunction
Important TR, atrial or ventricular arrhythmias or symptoms such as deteriorating exercise performance
Residual PS with RV pressures at least 2/3 systemic
European Guidelines (2010)
1. PVR should be performed in symptomatic patients with severe PR and/or stenosis (RV systolic pressure >60 mm Hg, TR velocity >3.5 m/s). (Class I, LOE C)
2. PVR should be considered in asymptomatic patients with severe PR and/or PS when at least one of the following criteria is present: (Class IIa, LOE C)
Decrease in objective exercise capacity
Progressive RV dilation
Progressive RV systolic dysfunction
Progressive TR (at least moderate)
RVOT obstruction with RV systolic pressure >80 mm Hg (TR velocity>4.3 m/s)
Sustained atrial/ventricular arrhythmias
LOE, level of evidence; PR, pulmonary regurgitation; PS, pulmonary stenosis; PVR, pulmonary valve replacement; RV, right ventricle; RVOT, right ventricular outflow tract; TR, tricuspid regurgitation.
Since its initial introduction in 2000, percutaneous PVR has gained widespread acceptance and usage as a non-surgical alternative among patients with repaired TOF who have undergone prior PVR or RV-pulmonary artery conduit placement.17 Currently, valve options are somewhat limited due to small sizes that are inappropriate for native RV outflow tracts. However, there are new options that are expanding the eligibility for patients.
Types of percutaneous pulmonary valves
The Medtronic Melody transcatheter pulmonary valve (Medtronic, Minneapolis, Minnesota, USA) has been commercially available in the USA since 2010 and has received pre-market approval in 2015 by the Food and Drug Administration for this indication. Though initially marked by struggles with stent fracture, altering the approach with systematic pre-stenting of the conduit reduced this risk significantly.18 A limitation of this valve is the relatively small size—it is marketed for conduits larger than 16 mm with two possible diameters which can be expanded to 20 or 22 mm.
The Edwards SAPIEN system is used extensively in the aortic position and was first reported in the pulmonary position in 2006. It can be used for larger conduits having expandable diameters of 23 mm and 26 mm.19 Wilson et al published a recent retrospective review of the Toronto experience of 25 patients with the Edward SAPIEN system in the pulmonary position with good technical success.20 The Edwards SAPIEN system is used extensively in the aortic position and was first reported in the pulmonary position in 2006. It can be used for larger conduits having expandable diameters of 23 mm and 26 mm. in the USA, has also been used successfully in the pulmonary position after prior repair of TOF.21
Two self-expandable valves are currently being tested. The first successful deployment of a self-expanding pulmonary valve was reported in 2010.22 Later called the Native Outflow Tract device (Medtronic, Minneapolis, Minnesota, USA), it has recently completed enrolment in an Investigation Device Exemption trial. The Venus-P valve (Venus Medtech, Shanghai, China) is a novel self-expanding percutaneous stent valve (diameters 20–32 mm) with flailed ends to conform to a dilated RV outflow tract.23 Initial experience of this native RV outflow tract device has shown short-term procedural success and no evidence of stent fracture.24
Though initially approved only for use in prior surgically placed circumferential RV to pulmonary artery conduits, both the Melody valve and Edwards SAPIEN system are being used off-label with increasing regularity in native outflow tracts with acceptable short-term to mid-term results.25 Transcatheter techniques and technology are evolving rapidly, with more than seven trials examining percutaneous PVR currently ongoing (https://clinicaltrials.gov/). Table 2 summarises the specifications of the valves that are available or being investigated.
Outcomes of percutaneous pulmonary valves
There are good supportive data on the short-term and mid-term outcomes of the Melody valve system with good procedural success, low risk, improved CMR-defined ventricular parameters and improved exercise capacity.26 ,27 Patients undergoing transcatheter valve replacement anticipate a shorter recovery, lower risk and outcomes similar to surgical PVR. No head-to-head studies have directly compared percutaneous with surgically implanted pulmonary valves, though reductions in PR and RV size appear to be comparable.26 ,28
One issue of concern that has received a considerable amount of attention recently is endocarditis after Melody valve placement. This was first noted in a review of 155 cases done between 2000 and 2007 in which five patients developed infective endocarditis.29 ,30 Multiple studies have suggested a rate of endocarditis above what is expected in surgical PVR, estimated to be as high as 10%–15% in the medium term compared with 1%–2% in surgically implanted valves.29–32 Risk factors include RV outflow tract obstruction, incomplete stent apposition and abrupt cessation of aspirin.32
Impact of transcatheter techniques on management of pulmonary valve disease
Historically, timing for surgical referral involved balancing the goal of avoiding RV deterioration (going too late) and the need for repeat surgical interventions (going too soon). In our institution, it is our practice to surgically implant bioprosthetic pulmonary valves at least 27 mm in size with the understanding that the next valve would be placed via transcatheter technique, thereby eliminating the risks and morbidity of a repeat open heart surgery. Currently, the indication for percutaneous PVR is the same as that for surgical PVR.
This paradigm has been challenged by recent data suggesting that earlier percutaneous pulmonary valve implantation in younger patients—especially those with mixed pulmonary valve disease—translates into incremental improvement in RV size and function as well as parameters of exercise stress testing.33 Short-term and mid-term freedom from reintervention for the Melody valve is good18 but long-term data on transcatheter pulmonary valve durability are not known. Depending on anatomy and rate of prosthetic valve dysfunction, the implication of multiple valve-in-valve procedures in a patient's lifetime akin to Russian ‘nesting dolls’ may translate into the need for surgical conduit revision later in life. On the other hand, there may be significant benefits to ‘going early’ with increased chances of RV preservation and avoidance of the untoward outcomes of heart failure, arrhythmia and death that accompany it.
Refinement of PVR timing
While echocardiography is a simple and readily available test for assessing RV size and function, RV geometry and its retrosternal location often make echocardiography challenging. Traditionally used quantitative measures of RV function have been found to be insensitive in patients with repaired TOF and therefore echocardiography is considered an important first-line modality for qualitative assessment of the RV.34 CMR currently serves as the gold standard for evaluation of RV size and function in patients with repaired TOF.34 Newer CMR techniques are being used to better understand and predict outcomes after PVR in patients with repaired TOF.
Irreversible RV dysfunction resulting from chronic volume overload may reflect ventricular fibrosis. Late gadolinium enhancement for the detection of focal scar has been shown to correlate with adverse outcomes in TOF35 but T1 mapping to quantify extracellular volume (ECV) has been increasingly used to assess diffuse fibrosis. In repaired TOF, ECV diffuse fibrosis has been associated with RV volume overload but not with pressure overload. Furthermore, there is a positive correlation between LV ECV and RV ECV supporting the concept of LV dysfunction in TOF as a result of ventricular interactions.36 Elevated LV ECV in TOF correlates with QRS duration, B-type brain natriuretic peptide and 6 min walk distance among other clinical variables.37 Together, these studies suggest that ECV may serve as a biomarker to predict subclinical deterioration and could be used to refine timing for referral for PVR.
The wide variability of RVEF after PVR and the inability to accurately predict postoperative RV systolic function were addressed by another study using CMR. Tang et al38 demonstrated that out of six possible mechanical and morphological factors, RV wall stress by CMR-based computational modelling is a highly specific predictor of RV response to PVR. Though this was a small study, it may enable providers to tailor the timing for PVR on an individual basis.
There are little data on longitudinal changes in RV dimensions over time in the face of severe PR. A recent study by Wald et al showed that RV dimensions remain relatively stable but 15% exhibited ‘rapid disease progression’ which is defined as an increase in RV end-diastolic volume of ≥30 mL/m2, decrease in RVEF of ≥10% or decrease in LVEF of ≥10%.39 They determined that 3 years was the optimal time interval required to detect significant changes in size and function between studies. Most of these patients fulfilled generally accepted criteria for PVR but of interest, one-third of these ‘rapid disease progressors’ did not, thus highlighting the heterogeneity in clinical behaviour of this patient population.
Cardiopulmonary exercise testing
Exercise testing has long been used as an objective way to monitor functional capacity in patients with TOF prior to PVR and can support referral for intervention. Babu-Narayan et al reviewed 220 patients at a single centre who underwent PVR and found that peak VO2, VE/VCO2 slope and heart rate reserve on preoperative cardiopulmonary exercise stress testing predicted early mortality after PVR.40 These data show that referral for PVR when patients still have reasonable functional capacity is associated with improved postoperative outcomes and support routine use of cardiopulmonary exercise testing as a means of risk stratification.
Outcomes of PVR
PVR—both surgical and transcatheter—eliminates PR and improves RV dimensions but does not consistently improve RVEF. Clinically, patients feel better despite lack of improvement in objective measures of exercise capacity. LV diastolic filling can be impaired from severe PR, resulting in decreased LV systolic function, but studies show variable effects on LVEF after PVR. Table 3 summarises the clinical and haemodynamic response to PVR.
It is not currently known whether PVR timing based on the goal of normalising RV size postoperatively translates into a hard clinical end point. Furthermore, no study has demonstrated that PVR alone has been associated with a reduction in death or VT postoperatively,41 though patients who undergo cryoablation at the time of PVR have a decrease in both atrial and ventricular arrhythmias postoperatively.42 Patients seem to remain at risk for sudden cardiac death and ventricular arrhythmias despite PVR.41
Patients with primary residual pulmonary stenosis represent the minority of adult patients undergoing PVR and their outcomes may be different. In contrast to patients with TOF having PR, those with pulmonary stenosis have improvement in RVEF secondary to improvement in afterload as well as a significant improvement exercise capacity and RVEF on exercise stress testing.43 ,44
In the long term, patients with TOF who undergo PVR generally do well, though often there are residual lesions that will require intervention. Prosthetic pulmonary valve dysfunction valve durability is limited with the average lifespan estimated to be 10–15 years.8 ,14
Though the RV initially improves in size, there is evidence of progressive prosthetic pulmonary valve dysfunction and RV deterioration with a return to preoperative RV size over the ensuing 10 years.45 Given this and the variability of RV remodelling after restoration of RV outflow tract and pulmonary valve competence, it is difficult to know what criteria clinicians should use to refer for repeat PVR in the absence of symptoms.
Approach to residual pulmonary valve disease
In our practice, we use a combination of imaging studies, ECG and exercise testing to determine optimal timing for PVR. On annual visits, we perform routine history and physical examination along with ECG and echocardiogram. Cardiopulmonary exercise stress testing and CMR are performed at least every 3 years depending on the patient.
Our approach to the dysfunctional RV outflow tract and pulmonary valve in repaired TOF is outlined in figure 3.
The current approach towards the management of PR in patients with repaired TOF is predicated on mitigating the downstream effects of PR, namely heart failure, life-threatening arrhythmia and death as a result of progressive RV dilation and dysfunction. In addition to treating symptoms, recommendations are based on referral to surgery before irreversible damage to RV size or function ensues. The use of CMR has enabled us to refine volume thresholds past which PVR has a lower likelihood of restoring RV dimensions but normalisation of RV size has yet to be associated with decrease in mortality or incidence of life-threatening arrhythmia. That being said, determining predictors of these outcomes are limited by their relative infrequency and follow-up time required to assess for such events. As of yet, subclinical markers of myocardial compromise have not been successfully identified.
With the trend towards PVR with smaller RV size, important questions arise regarding valve longevity, the fate of the RV after PVR, criteria for repeat PVR and feasibility and outcomes of valve-in-valve procedures. Adding to the decision tree complexity is the rapidly evolving technology of transcatheter PVR, which at the current time shares the same indications for valve replacement as the surgical approach. With continued advances in these techniques and improvements in surgical valve durability, as well as with better understanding of the natural history of palliated TOF, the approach to residual pulmonary valve disease in TOF will continue to evolve.
Contributors Both authors have contributed equally to the writing of this review.
Competing interests None declared.
Provenance and peer review Commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.