Current outcomes and treatment of tetralogy of Fallot

Residual right ventricle outflow tract obstruction

Residual RVOTO is common following repair and results in residual or progressive concentric hypertrophy of the RV. Data obtained from the INDICATOR study suggest that RV hypertrophy, due to increased mass-to-volume ratio, is a more important long-term risk factor for ventricular tachycardia (VT) and death than severity of RV dilation (RV end-diastolic volume index)⁷⁸. Current guidelines provide clear indications for re-intervention for residual RVOTO (Table 1)^75–77. Balloon valvuloplasty or PVR can be performed for valvular pulmonary stenosis (PS). PA branch stenosis can be safely relieved by balloon dilation, stenting, or PA reconstruction⁷⁹. In several large studies, 1% to 7% of patients have undergone PA dilation or stenting at long-term follow-up (median of 5.8 to 36 years)^{61,63,64,80,81}. Surgical relief of the RVOT and PA plasties were performed in 1% to 5% of patients at long-term follow-up^61,63,80,81.

Pulmonary regurgitation

PR is very common at medium- to long-term follow-up. Five to ten years after repair, 40% to 85% of patients have moderate to severe PR^{53,73,82–84}. PR induces RV volume overload of the RV with often progressive RV dilation, which may include the development of tricuspid regurgitation (TR) and RV dysfunction. It is often accompanied by prolongation of the QRS complex, and RV dyssynchrony could contribute to the progression of dysfunction^85–87. There generally is a long period in a compensated state, during which RV function is maintained. In some patients, these compensatory mechanisms fail, leading to progressive RV dysfunction^85,86. The mechanisms of RV adaptation and remodeling, as well as the molecular events contributing to the transition from a compensated to a decompensated state, are still poorly understood. Timely restoration of pulmonary valve competence is considered to halt the progressive adverse RV remodeling resulting in RV dysfunction seen in chronic PR.

Thirty-five years after ToF repair, PVR will have been performed in about 40% of patients^63,65,88. Staged repair and TAP are risk factors for late PVR^12,54,63,80, whereas mild residual PS seems to reduce risk⁸⁹. As more patients with ToF survive into adulthood, PVRs are increasingly being performed⁹⁰.

PVR is effective in decreasing RV volumes, reducing TR, decreasing QRS duration, increasing left ventricle (LV) ejection fraction (EF), and improving functional status^91,92. It should be noted that no improvement in survival following PVR compared with medical management has been demonstrated to date^93,94.

Homograft or bioprosthetic valves are currently the preferred valves for PVR⁹⁵. The current 10-year re-PVR–free survival of ToF patients undergoing homograft PVR ranges from 74% to 89%^95,96.

Tissue-engineered valves with a non-synthetic and non-immunogenic surface have the potential to provide lifelong valve replacement⁹⁷. In situ tissue engineering techniques, in which a decellularized “starter scaffold” of polymers can be used to provide shape and structure to the valve, are of particular interest. This scaffold is infiltrated by endogenous cells to provide a regenerating functional valve. As the scaffold would be non-immunogenic, this could provide a relatively cheap “off the shelf” valve. Current studies evaluating tissue-engineered valves in animals and humans show promising early results⁹⁸.

Several transcatheter PVR strategies have been developed and are increasingly used in a clinical (trial) setting⁹⁹. However, clinical experience compared with (surgical) homograft PVR is limited⁹⁹. Procedural success of transcatheter PVR is generally good (>95%)¹⁰⁰. The hazard rate for re-intervention following transcatheter PVR ranges from 0.4% to 5.9% per patient-year¹⁰⁰. However, high rates of infective endocarditis during follow-up have been described¹⁰¹. Recent results from the MELODY Registry estimate the infective endocarditis risk to be 2.3% per patient-year¹⁰². In comparison, the infective endocarditis risk in surgical PVR has been estimated to be 0.3% per patient-year¹⁰³. Transcatheter PVR has been shown to increase exercise capacity and quality of life 6 months after the procedure^104,105. Direct comparisons with surgical PVR are still lacking.

Arrhythmia

Ventricular tachycardia. VT is a common arrhythmia in the repaired ToF population. Cuypers et al. reported a 5% cumulative incidence of sustained VT after a median of 35 years after ToF repair⁶³ and these figures are similar to those of most reports^58,106. However, cumulative incidences of up to 15% have been reported in some adult populations¹⁰⁷. Predictors of sustained VT include higher age, number of prior cardiac surgeries, presence of a TAP, LV diastolic dysfunction, and QRS width^63,106–108. Most guidelines recommend implantable cardioverter defibrillators (ICDs) for patients who have had sustained VT or cardiac arrest^76,77. ICDs are also employed for primary prevention, although selecting high-risk patients who would benefit from ICD implantation remains challenging^76,77. Pacemaker and ICD prevalences in adult ToF populations both range from 5% to 10%^63,107,109.

Electrophysiological studies can help to determine the underlying substrate, and radiofrequency ablation can be performed. Ablation of monomorphic VT substrates has excellent short-term outcomes with recurrent VT in 18% of patients after a mean follow-up of 34 months¹¹⁰. Another study found a similar recurrence rate (19%) 10 years after ablation¹¹¹.

Supraventricular tachycardia. The prevalence or cumulative incidence of supraventricular tachycardia (SVT) in adult patients ranges from 4% to 20%^{107–109,112}. In the first 10 to 15 years following ToF repair, SVT is relatively uncommon but the incidence rises steadily after this period¹⁰⁷. Intra-atrial re-entrant tachycardia, typically involving the right atrium, is the most common type of SVT in patients with ToF¹⁰⁷. Two large studies found that SVT was an independent predictor of death or VT^78,108. Few studies have assessed the efficacy of ablation of atrial arrhythmias in corrected ToF, and long-term follow-up is lacking^113–115.

Aortopathy

Dilation of the aorta is seen in 12% to 24% of adult patients with ToF^116–118. In patients with aortic dilation, aortic root size seems to progressively increase over a period of years. Aortic dissection following ToF appears to be a rare complication¹¹⁹. A population-based study in Texas demonstrated no increased risk for thoracic aortic dissection for patients with ToF compared with the general population¹¹⁹. However, progressive aortic root dilation can lead to malcoaptation of the aortic valve and aortic regurgitation. Furthermore, the elasticity of the dilated aortic root was shown to be reduced in patients with ToF, possibly hampering circulatory function¹²⁰. The importance of aortopathy in circulatory function and mortality remains incompletely understood.

Right ventricular adaptation and remodeling

The mechanisms of RV adaptation and remodeling in response to chronic RV volume overload, resulting from PR, are poorly understood¹²¹. In young pig models, chronic PR affects biventricular systolic function, RV myocardial contractility, and LV diastolic performance¹²². Histopathology of several animal models displays early hypertrophy of the chronically volume-loaded RV and, in a later stage, myocardial fibrosis¹²¹. The molecular responses to increased volume or pressure loading of the RV are different from those in the LV^{121,123–125}. In a pig model of repaired ToF with induced PR, PS, and an RVOT scar, RV hypertrophy and dilation were found after 23 weeks. The myocardium was characterized by increased collagen deposition, leading to decreased impulse conduction velocity and dispersion¹²⁶. Similar findings were found in the LV, despite preserved LV function at this stage. This demonstrates biventricular adverse effects are present early in the adverse remodeling process¹²⁷.

Basic research into RV remodeling has focused mainly on the response to increased pressure loading rather than the predominantly volume-loaded RV as seen in PR^124,125. Volume loading and pressure loading increase myocardial metabolic demand. This metabolic stress induces an increased amount of reactive oxygen species. Compensatory anti-oxidant production in the RV is impaired compared with the LV¹²⁵. This might imply that the RV is more vulnerable to oxidative stress, as seen in abnormal loading conditions.

In volume-loaded RV mouse models, a clinical course similar to RV dysfunction with volume-loaded RV in humans is observed. RV function is maintained during a compensated phase, followed by RV dysfunction¹²⁸. Gene expression patterns of the cardiomyocyte in the compensated state differ from those of healthy controls. Several molecular pathways, such as transforming growth factor beta (TGF-β) signaling, p53 signaling, and cytoskeleton-related pathways, are downregulated in the early compensated state but show late upregulation as the RV progressively remodels¹²⁸. However, the exact cellular and molecular mechanisms of transition from a compensated to a decompensated state of the volume-loaded RV have not been fully elucidated^125,129.

Assessing the right ventricle in patients with tetralogy of Fallot

Our limited understanding of the pathophysiology of RV failure hampers our ability to adequately detect failure in the early stages in clinical practice. Imaging techniques are used to assess the RV and follow patients serially, aiming to detect early changes in biventricular size and performance. Cardiovascular magnetic resonance (CMR) imaging is routinely used to reliably quantify RV volumes and function, wall mass, and PR¹³⁰. Adverse clinical events have been related to larger RV volumes, PR severity, biventricular EF, and mass-to-volume ratio^78,131,132. Increased RV volumes, most commonly end-diastolic volume index (EDVi), have been considered a sign of prolonged high PR burden and thus a predictor of RV dysfunction. However, exercise capacity can be preserved even in severely dilated ventricles, demonstrating that compensatory mechanisms can still be adequate to maintain performance of large RVs¹³³. In the INDICATOR cohort, increased RV wall mass-to-volume ratio, among other factors, was found to be an independent predictor of VT and all-cause mortality, whereas RV EDV and end-systolic volume were not predictive of the end-points⁷⁸. RV hypertrophy could be a more sensitive marker of pending dysfunction than EDV, although this might be particularly true for patients with residual PS.

Regional myocardial performance and mechanical synchrony can be assessed by strain imaging studies. Global circumferential or longitudinal strain has been used to assess RV function. Under normal circumstances, the RV ejects mainly by longitudinal shortening while, with increased RV pressure loading, circumferential contraction is increased¹³⁴. The predictive value of global longitudinal or circumferential strain in ToF is still uncertain: Orwat et al. found that RV global longitudinal strain assessed by CMR was a superior independent predictor for death, cardiac arrest, or VT compared with RV volumes¹³⁵. RV global circumferential strain was not predictive of outcome in that study¹³⁵. Diller et al. found a similar relation for LV global longitudinal strain assessed by echocardiography¹³⁶.

Mechanical dyssynchrony has been demonstrated to relate to prolonged or fragmented (QRS complex containing additional spikes without bundle branch block) QRS complexes¹³⁷. The contributions of this mechano-electrical interaction to RV function remain uncertain, as studies assessing mechanical dyssynchrony report conflicting results^{135,136,138–141}. RV circumferential dyssynchrony was shown to negatively predict exercise capacity in one study¹⁴⁰. This association has not been confirmed in other studies^138,141. Cardiac resynchronization therapy is increasingly used in ToF. A recent study found that 12 out of 15 adult patients with ToF had an improved NYHA (New York Heart Association) class or LV function after 2.6 years (median) of cardiac resynchronization therapy¹⁴². Procedural success was high and adverse events were rare.

Right ventricular interactions in tetralogy of Fallot

Atrio-ventricular interactions. Diastolic function after ToF repair is a determinant of the amount of PR. In some patients, end-diastolic forward flow (EDFF) in the main PA during right atrial contraction can be observed¹⁴³. This is considered a sign of “restrictive RV physiology” as the non-compliant RV acts as a conduit during atrial contraction as RV diastolic pressure exceeds PA diastolic pressure^144,145. Restrictive physiology could limit the amount of PR as elevated diastolic RV pressure reduces the amount of PR. A recent study found no relationship between the presence of EDFF and other markers of diastolic dysfunction (that is, RV hypertrophy, atrial dilatation, reduced stroke volume, or reduced PR)¹⁴⁶. Different mechanisms, such as pulmonary arterial capacitance and atrial function¹⁴⁷, may play significant roles in the occurrence of EDFF. Luijnenburg et al. found that bi-atrial function, but not diastolic ventricular function, differed between patients with EDFF and those without it¹⁴⁷. In that study, abnormal atrial function was related to worse exercise capacity and higher N-terminal pro brain natriuretic peptide (NT-proBNP). Kutty et al. found that right atrial longitudinal strain predicted RV performance but not exercise capacity¹⁴⁸.

The effect of EDFF on circulatory function is controversial. Studies found conflicting results regarding the relationship between EDFF and the amount of PR^143,144,146, exercise capacity^144–147, and EDV^144–147. The presence of EDFF might have a different etiology and clinical importance early versus late after repair or in severely dilated versus non-dilated ventricles.

Ventriculo-arterial interactions. Adequate atrio-ventricular coupling and ventriculo-arterial (VA) coupling are required for an energetically efficient transfer of blood through the right heart. VA coupling has not been studied extensively in ToF. Latus et al. assessed VA coupling as the relationship between pulmonary arterial elastance and ventricular end-systolic elastance in adult patients with ToF by using CMR and catheter-derived measurements both in resting conditions and during dobutamine stress¹⁴⁹. VA coupling was impaired during resting conditions. EF and load-independent parameters of RV contractility increased during dobutamine stress. Pulmonary arterial elastance increased accordingly and the impaired VA coupling that resulted during dobutamine stress was similar to that under resting conditions.

Interventricular interactions. Interactions between the RV and LV have been extensively described. The LV and RV have common myocardial fibers, the interventricular septum, the anatomic space confined by the pericardium, and a common neurohumoral system¹⁵⁰. Not unexpectedly, the effects of chronic PR are not limited to the RV, although the mechanisms of this ventriculo–ventriculo interaction in chronic PR remain poorly understood. A linear correlation between LV and RV EF has been described^150,151. Severe RV dilation causes abnormal diastolic septal positioning, influencing LV filling¹⁵². The role of the LV in outcomes in ToF is increasingly appreciated, as LV function has been associated with increased mortality and increased risk of VT^136,153. In the INDICATOR registry, LV EF was one of three independent predictors of mortality and VT¹⁵⁴. Geva et al. found that LV EF, independent of RV parameters, predicted poor functional status¹⁵¹. Remarkably, parameters of LV function are not considered in current guidelines for the timing of PVR (Table 1).

Drug therapy for right ventricular failure

Pharmacotherapy is important in the treatment of LV failure and improves outcomes. However, the effects of the use of heart failure medication for RV failure have been disappointing^155–157. In patients after ToF repair, RAAS (renin–angiotensin–aldosterone system) inhibitors do not appear to influence RV EF or exercise capacity¹⁵⁸. In a randomized controlled trial of 33 patients with ToF, beta blockers showed no beneficial effects after 6 months of treatment and an increase in NT-proBNP was noted¹⁵⁹. Increasing our understanding of the pathophysiology of RV failure might elucidate new targets for medical treatment unique to the RV.

Current guidelines on the timing of pulmonary valve replacement

Restoring pulmonary valve function before irreversible RV dysfunction occurs could be important to prevent RV failure. However, the durability of currently used pulmonary prosthetic valves is limited. Therefore, the timing of PVR always is a compromise: It should be timed early enough to prevent irreversible adverse remodeling but late enough to limit the number of re-interventions. Because of the difficulties in assessing RV function, predicting decline in RV function is difficult, and the optimal timing of PVR is controversial. Guidelines by the European Society of Cardiology, the Canadian Cardiovascular Society (CCS), and the American College of Cardiology/American Heart Association provide some recommendations on indications for performing PVR^75–77. These indications are summarized in Table 1.

Indications differ between guidelines and have several limitations. Most guidelines do not provide specific cutoff points since these are statistical constructs that do not work for individual patients. The 2009 CCS guideline provides an absolute cutoff value for EDVi but does not take into account the considerable differences in normal (indexed) RV volumes between genders and age¹⁶⁰. End-systolic volume index and RV mass-to-volume ratio have been proposed as superior predictors compared with EDV^78,161. Progressive RV dilation is considered an indication for PVR, but there is no consensus on what too much progression is^162–165. Longitudinal changes in RV size and function following ToF repair have been reported in several studies^166–173. RV volumes increase non-linearly and seem to stabilize in adolescence. These factors need to be taken into account when assessing progressive RV dilation.

Furthermore, the recommendations in current guidelines are often based on long-term outcomes of studies in patients who have been operated at a much older age than has been the practice in the past 20 years. This warrants caution when extrapolating these results to current adolescent or younger patients.

Careful interpretation of current guidelines seems to be justified. Individual patient parameters and views should always be taken into consideration. In clinical practice, an approach using information from different sources, including history, physical examination, electrocardiogram, imaging techniques, exercise testing, and blood biomarkers, may be most useful¹⁷⁴.