Article Text

Ventricular arrhythmias and sudden cardiac death in adults with congenital heart disease
1. Paul Khairy
1. Correspondence to Professor Paul Khairy, Montreal Heart Institute Adult Congenital Center, Montreal Heart Institute, 5000 Belanger St. E., Montreal, Quebec, Canada H1T 1C8; paul.khairy{at}umontreal.ca

## Abstract

Remarkable gains in survival have led to an unprecedented number of adults with congenital heart disease. Arrhythmias collectively comprise the most common complication encountered. Recognising the unique issues and challenges involved in managing arrhythmias in adults with congenital heart disease and the consequential decisions surrounding sudden death prevention, expert societies have proposed evidence-based recommendations. On the whole, acute ventricular arrhythmias are managed according to general cardiology guidelines, while taking into consideration congenital heart disease-specific issues, such as positioning of patches or paddles according to location of the heart. Implantable cardioverter-defibrillators (ICDs) are indicated for secondary prevention in patients with sustained ventricular tachycardia or resuscitated cardiac arrest in the absence of a reversible cause. Pharmacological therapy and catheter ablation can be effective in reducing recurrent ICD shocks. Risk–benefit assessment for primary prevention ICDs is a major challenge. Although a clearer picture has emerged of the high-risk patient with tetralogy of Fallot, ICD indications for those with systemic right ventricles or univentricular hearts remain contentious. Challenges to ICD implantation include obstructed veins, conduits and baffles, atrioventricular valve disease and intracardiac shunts. In selected patients, customised systems with epicardial and/or subcutaneous coils may represent a viable solution. Alternatively, the subcutaneous ICD is an attractive option for patients in whom transvenous access is not feasible or desirable and in whom bradycardia and antitachycardia pacing features are not essential. Continued advances in risk stratification and device technologies carry the potential to further improve efficacy and safety outcomes in this growing population of patients.

## Introduction

The past few decades have witnessed historical shifts in the population demographics of congenital heart disease (CHD).1 While CHD was once considered a predominantly paediatric condition, adults now outnumber children with CHD by a ratio of 2:1.2 Population estimates suggest that there are >3 million adults with CHD in North America and Europe.3 This remarkable triumph is, however, mitigated by long-term complications. In particular, adults with CHD are subject to arrhythmic sequelae related to the original heart defects, displaced or malformed conduction systems, surgical scars, altered haemodynamics, hypoxic tissue injury, myocardial ischaemia and/or genetic influences.4 Resulting arrhythmias span the spectrum of bradyarrhythmias, atrial and junctional tachycardias, and ventricular arrhythmias and sudden cardiac death. Recognising the diverse challenges related to arrhythmias in adults with CHD, the Pediatric and Congenital Electrophysiology Society partnered with the Heart Rhythm Society in establishing evidence-based recommendations.4 This review focuses on ventricular arrhythmias and risk stratification for sudden death in adults with CHD, and includes a discussion of pertinent recommendations. In general, adults with CHD and arrhythmias should be referred to centres specialised in adult CHD care.4

## Overview of ventricular arrhythmias

Ventricular ectopy is common in adults with CHD. The prognostic value of non-sustained ventricular tachycardia varies, in part, according to type of CHD. For example, non-sustained ventricular tachycardia has not been linked to sudden death in a heterogeneous population of patients5 but has been associated with inducible6 and clinical7 ventricular tachyarrhythmias in tetralogy of Fallot. The incidence of sustained ventricular arrhythmias in adults with CHD at large is approximately 0.1%–0.2% per year.4 Sustained monomorphic ventricular tachycardia is most often macroreentrant and dependent on a few so-called critical isthmuses that typically involve either an extensively scarred right ventricular outflow tract (figure 1) or ventriculotomy incision.8 Focal ventricular tachycardia is occasionally encountered, with clinical experience suggesting a predilection towards the native right ventricular outflow tract (figure 2). Polymorphic ventricular tachycardia and ventricular fibrillation are also of concern in a subset of patients, particularly those with extensive myocardial hypertrophy, diffuse fibrosis, myocardial ischaemia and/or severe systemic ventricular dysfunction.

Figure 1

Ventricular tachycardia in tetralogy of Fallot. Shown in (A) are surface electrocardiographic recordings during ventricular tachycardia in a patient with tetralogy of Fallot. The tachycardia had a left bundle branch block morphology and inferior QRS axis (ie, positive in leads II, III and aVF). (B) Shows a posterior view of a voltage map of the right ventricle (RV) using a three-dimensional electroanatomic mapping system. Normal voltages are coloured purple and dense scar grey. The site of the ventricular septal defect (VSD) patch is indicated. The ventricular tachycardia was dependent on a critical corridor of tissue in the RV outflow tract (RVOT), which was transected by radiofrequency catheter ablation between the VSD patch and pulmonary annulus. Ablation lesion sites are indicated by white circles.

Figure 2

Ventricular tachycardia in a patient with a single ventricle and Damus–Kaye–Stansel (DKS) operation. The DKS operation is a palliative procedure for patients with single ventricle physiology and an obstructed rudimentary outflow tract. The proximal pulmonary artery is divided near its bifurcation and anastomosed to the side of the ascending aortic. In Panel A, contrast angiography shows connections between both right (RVOT) and left (LVOT) ventricular outflow tracts to the reconstructed aortic root. These connections are also depicted in Panel B by means of three-dimensional electroanatomic mapping. A focal ventricular tachycardia (VT) was successfully ablated in the native RVOT at the site of the brown circle with a yellow rim.

## Management of ventricular arrhythmias

### Acute management

Ventricular arrhythmias in adults with CHD should be acutely managed according to general cardiology guidelines, while taking into consideration issues specific to CHD.4 For example, in assessing haemodynamic status, reliability of blood pressure measurements may be compromised by an ipsilateral Blalock–Taussig shunt, subclavian flap repair for aortic coarctation, or other vascular obstructions or anomalies. Oxygen saturation levels should be interpreted within the context of the underlying heart defect and individual baseline value. Bubble traps or air filters should be added to intravenous lines in those with right-to-left shunts. For effective cardioversion or defibrillation, the position of patches or paddles should be adapted according to the location of the heart, particularly in patients with meso- or dextrocardia. In those with pacemakers or implantable cardioverter-defibrillators (ICDs), patches or paddles should ideally be positioned at least 8 cm away from the generator. Anteroposterior and anterolateral positions are acceptable.

Sustained ventricular arrhythmias should be terminated expeditiously. In most instances, electrical cardioversion or defibrillation is the preferred treatment option. However, in those with haemodynamically tolerated monomorphic ventricular tachycardia, as commonly encountered in young adults with CHD and normal ventricular function, it may be reasonable to attempt pharmacological conversion.

### Long-term management

As a Class I recommendation, ICDs are indicated for secondary prevention of sudden death in adults with CHD and spontaneous sustained ventricular tachycardia or resuscitated cardiac arrest in the absence of a reversible cause.4 Antiarrhythmic drugs may be helpful in reducing ICD discharges, although studies specific to the adult CHD population are lacking. The choice of agent should consider factors such as coexisting bradyarrhythmias, systemic or subpulmonary ventricular dysfunction, associated therapies and acquired comorbidities.4 Case series from the era of serial drug testing found that mexiletene9 and phenytoin10 reasonably suppress ventricular arrhythmias. Limited data suggest that sotalol and amiodarone may also be effective for this purpose.11

In the absence of an ICD, amiodarone and dofetilide are the preferred antiarrhythmic drugs in the setting of systemic or subpulmonary ventricular dysfunction.4 However, long-term amiodarone therapy in young adults raises concerns over non-cardiac toxicity. In particular, amiodarone-induced thyrotoxicosis is prevalent in patients with cyanotic heart disease or Fontan surgery,12 and in patients with CHD with a body mass index <21 kg/m2.13 Dofetilide may be a reasonable alternative to amiodarone in the absence of impaired renal function or a prolonged QTc interval.4

Catheter ablation can also be effective in reducing recurrent ICD shocks. In general, ventricular tachycardia ablation is not considered a substitute for ICD therapy in adults with CHD at heightened risk for sudden death. Identification of a select group of low-risk patients in whom catheter ablation may be performed without ICD implantation remains a challenge. For example, the incidence of recurrent ventricular tachycardia or sudden death appears to be low in the subset of adults with CHD who have preserved ventricular function, macroreentrant ventricular tachycardia and successful catheter ablation with demonstrated bidirectional conduction block across the transected isthmus.14 ,15 Recommendations for ablation are summarised in table 1.

Table 1

Indications for catheter ablation of ventricular tachycardia in adults with CHD

## Risk stratification for sudden cardiac death

### Scope of the problem

Heart failure appears to have surpassed sudden death as the leading cause of mortality in adults with CHD.4 A German national study found that heart failure and sudden cardiac deaths accounted for 23% and 29% of all deaths, respectively, between 2001 and 2008, whereas corresponding proportions shifted to 30% and 20% between 2009 and 2015.16 These statistics may reflect successes in risk stratification and prevention, in combination with an increased prevalence and severity of heart failure in ageing survivors. While the overall incidence of sudden death in adults with CHD at large is relatively low (<0.1% per year), a handful of CHD lesions account for the majority of cases, namely tetralogy of Fallot, complete transposition of the great arteries (TGA) with an atrial baffle (ie, Mustard or Senning), congenitally corrected TGA, left ventricular outflow obstruction (ie, aortic stenosis and aortic coarctation) and Eisenmenger syndrome. In contrast, heart failure is the leading cause of mortality in Ebstein anomaly and univentricular hearts.16 ,17

### Challenges to risk stratification

Predicting which patients are likely to have a sudden cardiac arrest before a catastrophic event occurs is an imperfect science. Efforts to accurately estimate probabilities based on risk profiles are hampered by several challenges. To highlight a few, the adult CHD population is in many ways a moving target. Surgical techniques and interventions are in constant flux. They evolve through an iterative process in response to the recognition and cataloguing of long-term sequelae, with the objective of continuously improving outcomes. As a result, risk factors identified in earlier series of patients may become less relevant to younger cohorts. Adding further complexity, risk estimates change as patients age and progress further along their disease course. Moreover, attempts to identify risk factors are limited by the relatively small, although growing, population, and modest event rates. While some risk factors, such as severe systemic ventricular dysfunction, are common to patients with a broad spectrum of disorders, others are more disease-specific. For example, inducible sustained ventricular tachycardia is helpful in risk stratifying selected patients with tetralogy of Fallot,6 ,7 but appears to have no prognostic value in TGA.18 These observations suggest that risk stratification in adults with CHD is best performed on a lesion-by-lesion basis.19

### Risk–benefit considerations

A probabilistic framework for risk stratification incorporates baseline lesion-specific risks of sudden cardiac death together with an individual's profile.20 If the estimated probability exceeds a certain threshold, ICD implantation may be contemplated. Key factors involved in guiding ICD management decisions are listed in box 1. Various methodologies have been developed to weigh risks and benefits in a quantitative fashion to inform objective decision-making. Some approaches, such as estimating numbers needed to treat and harm and their relative value-adjusted versions, rely on subjective schemes in weighing the merit of a life saved versus ICD-related complications. Else, minor complications would counterintuitively be considered at par with deaths. Other methods, such as multicriteria decision analyses, risk–benefit plane and the stated preference method, assess joint distributions of benefit and risk. Since all criteria cannot be considered equally important, weights are ideally derived quantitatively and input into mathematical models to determine the preferred alternative.

Box 1

### Key factors in weighing risks and benefits of implantable cardioverter-defibrillator (ICD) therapy

▸ Estimated risk of sudden cardiac death due to malignant ventricular arrhythmias

▸ Competing risks for mortality

▸ Effectiveness of ICD in preventing sudden cardiac death

▸ Morbidity and mortality associated with ICD therapy

▸ Costs considerations

Importantly, the threshold at which benefits exceed risks depends on economic considerations. Although there is no perfect metric, the WHO proposed that a cost per quality-adjusted life year (QALY) threshold of three times a country's per capita gross domestic product (GDP) is a reasonable ‘rule of thumb’ in guiding healthcare resource allocation (http://www.who.int/choice/en/). A QALY refers to 1 year of life adjusted by an index of functionality or health. For illustrative purposes, the probability of sudden cardiac death in controls randomised to the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) was 3.5% per year.21 In this population, primary prevention ICDs were estimated to cost approximately US$80 000 (or$59 000 €) per QALY.22 At the risk of oversimplification, the WHO rule of thumb suggests that it is reasonable for countries with a per capita GDP >US\$26 700 to allocate healthcare resources for primary prevention ICDs for patients with a risk profile similar to SCD-HeFT. According to 2011–2015 World Bank statistics (http://data.worldbank.org/indicator/NY.GDP.PCAP.CD), this includes North America, Australia, New Zealand, Japan, Israel and Western European countries but not most of Eastern Europe, South America, China, India and Africa.

## Primary prevention ICDs in selected lesions

### Biventricular physiology with a systemic left ventricle

Table 2 summarises indications for primary and secondary prevention ICDs.4 In a recent systematic review that included 412 adults with CHD and ICDs, 41 (10%) had simple forms of CHD with a systemic left ventricle.23 In general, ICDs are indicated in adults with CHD who meet standard primary prevention criteria, namely biventricular physiology with a systemic left ventricular ejection fraction ≤35% and New York Heart Association (NYHA) Class II or III symptoms.24 These patients would have qualified for inclusion in randomised primary prevention trials such as SCD-HeFT, which did not exclude subjects on the basis of CHD.21 In patients in whom programmed ventricular stimulation is likely to be of prognostic value, such as those with ventriculotomy incisions or scars, it may also be reasonable to consider a primary prevention ICD in the setting of syncope of unknown origin with inducible sustained ventricular tachycardia.4

Table 2

ICD indications in adults with CHD

### Tetralogy of Fallot

Tetralogy of Fallot is the most common congenital cardiac diagnosis in ICD recipients, outnumbering all other forms of CHD combined.23 Sudden death due to ventricular arrhythmias is the leading cause of mortality, with an estimated incidence of 0.2% per year.4 The event rate is non-linear, with most sudden deaths transpiring decades after corrective surgery.25 A multitude of observational studies have painted a reasonably consistent picture of the pathophysiological substrate for ventricular tachyarrhythmias and predisposing factors. Nevertheless, risk stratification remains challenging. Beyond severe left ventricular systolic dysfunction, no single risk factor has sufficient discriminative power to reliably guide ICD management decisions. As such, combinations of risk factors should be considered.

Multivariable approaches to risk stratification must address the complexities of correlated factors. Identification of independent predictors is highly dependent on the mix of variables included in statistical models. For example, a transannular right ventricular outflow tract patch, severe pulmonary regurgitation, right ventricular dilation, tricuspid regurgitation and QRS ≥180 ms are highly correlated factors associated with ventricular arrhythmias and sudden death. When considered together, a transannular patch and QRS ≥180 ms emerge as independent predictors.25 If programmed ventricular stimulation is incorporated into the mix of candidate variables, inducible sustained ventricular tachycardia carries prognostic value independent of such non-invasive markers.6 Programmed ventricular stimulation is particularly helpful in stratifying patients deemed at moderate risk for sudden death and is not indicated in low-risk patients.20 Table 3 summarises a risk score derived in a population with tetralogy of Fallot and primary prevention ICDs. Freedom from appropriate ICD shocks according to risk category is plotted in figure 3.7

Table 3

Risk score for appropriate implantable cardioverter-defibrillator shocks in patients with tetralogy of Fallot

Figure 3

Freedom from appropriate implantable cardioverter-defibrillator (ICD) shocks in primary prevention patients with tetralogy of Fallot according to their risk category. In patients with tetralogy of Fallot and primary prevention ICDs, Kaplan–Meier survival curves for freedom from first appropriate ICD shock are plotted and compared according to risk score classification. Risk score, corresponding risk category, number of patients and annualised rate of appropriate shocks are summarised below. Reproduced with permission from Khairy et al.7

Importantly, risk scores have yet to be prospectively validated. Moreover, it remains unknown whether factors associated with sudden death not considered in generating this risk score, such as extent of myocardial fibrosis,26 carry independent prognostic value and improve performance measures. Metrics of left ventricular dysfunction appear to outperform right ventricular size and function in predicting sudden deaths.7 Considering the totality of evidence, a Class IIA recommendation based on Level B evidence states that ICD therapy is reasonable in selected adults with tetralogy of Fallot and multiple risk factors for sudden death, such as left ventricular systolic or diastolic dysfunction, non-sustained ventricular tachycardia, QRS duration ≥180 ms, extensive right ventricular scarring or inducible sustained ventricular tachycardia.4

### Systemic right ventricle

Despite a threefold higher incidence of sudden death in patients with TGA and atrial switch surgery compared with tetralogy of Fallot,27 the profile of the high-risk patient is less clear. Risk stratification is even murkier for patients with congenitally corrected TGA in whom the incidence of sudden cardiac death is poorly characterised. The notion that bradyarrhythmias are important contributors to sudden death was later refuted by evidence suggesting that pacemakers did not alter outcomes.

Factors most consistently associated with sudden cardiac death in TGA with Mustard or Senning baffles include systemic ventricular dysfunction, severe tricuspid regurgitation, prolonged QRS duration, and atrial tachyarrhythmias.4 ,28 While there is little doubt that poor systemic ventricular function is a marker of adverse outcomes, questions abound as to the ideal cut-off value for risk stratification. It cannot be assumed that the 35% threshold value for a systemic left ventricle is applicable to a systemic right ventricle, considering that normal values for a subpulmonary right ventricle are approximately 20% lower than for a systemic left ventricle. Moreover, the few studies on primary prevention ICDs in TGA have uniformly reported incongruously low rates of subsequent ventricular arrhythmias. For example, in a multicentre cohort in whom 35% of patients had a systemic right ventricular ejection fraction <35%, the appropriate ICD shock rate was 0.5% per year.18 These studies highlight limitations in reliably identifying suitable ICD candidates.

It can be hypothesised that difficulties in risk stratification may be due, in part, to unconventional triggers for ventricular arrhythmias. Indeed, analyses of ICD tracings reveal that supraventricular arrhythmias are a common trigger for malignant ventricular arrhythmias in this population.18 This may be due to a confluence of factors that result in myocardial ischaemia, including a reduction in stroke volume with faster heart rates combined with an inefficient coronary circulation. Supportive evidence includes a strong association between exercise and sudden cardiac death28 ,29 and a seemingly protective role of beta-blockers.18 Recommendations reflect current ambiguities. A Class IIB indication based on Level C evidence states that ICD therapy may be reasonable in adults with a systemic right ventricular ejection fraction <35%, particularly in the presence of additional risk factors such as complex ventricular arrhythmias, unexplained syncope, NYHA functional Class II or III symptoms, QRS ≥140 ms, or severe systemic atrioventricular valve regurgitation.4

### Single ventricle

The subgroup with Fontan surgery has one of the highest mortality rates among patients with CHD.17 However, in a large single-centre study, 52% of deaths were heart failure-related, in comparison to 13% due to sudden cardiac death.30 To date, patients with univentricular hearts represent a minute fraction (<2%) of ICD recipients with CHD.23 Identification of risk factors for sudden death is hindered by the low event rate, with no clear independent predictor.17 Atrial tachyarrhythmias may be poorly tolerated haemodynamically and have anecdotally been associated with cardiac arrest.17 Ventricular tachyarrhythmias are not well characterised in this patient population. Intuitively, severely depressed systemic ventricular function and syncope raise concern over risk for sudden death,19 although they have not been formally established as risk factors. Class IIB recommendations based on Level C evidence state that ICD therapy may be considered in the presence of a single ventricular ejection fraction <35% or syncope with a high clinical suspicion of ventricular arrhythmia and no cause identified by thorough invasive and non-invasive investigations.4

## Special ICD considerations

Careful preprocedural planning is critical prior to device implantation. Common ICD-related challenges and complications, including obstacles to transvenous lead placement, are listed in table 4. Transvenous leads carry risks of dislodgement or failure, venous occlusion, endocarditis and thromboembolic events in the presence of an intracardiac shunt.31 In selected patients, customised systems that include epicardial and/or subcutaneous coils may be suitable, but are subject to a higher rate of dysfunction, particularly from increased defibrillation thresholds and shock coil failure.32 The subcutaneous ICD has emerged as an attractive solution for patients in whom transvenous access is not feasible or desirable and in whom bradycardia and antitachycardia pacing features are not essential. It may be compatible with a permanent pacemaker if pacing leads are bipolar and preimplant screening reveals the absence of T-wave oversensing during intrinsic and paced rhythms. Regardless of the type of system implanted, inappropriate shocks are common, occurring in 25% over 3–4 years of follow-up.23 Tailored programming may substantially reduce inappropriate and potentially avoidable shocks.33 While the impact of ICDs on quality of life remains uncertain, a high level of shock-related anxiety has been reported, in association with depression and sexual dysfunction.34

Table 4

Common challenges and complications associated with implantable cardioverter-defibrillator therapy in adults with congenital heart disease

## Conclusion

Management of ventricular arrhythmias and risk stratification for sudden death in adults with CHD involves numerous challenges and decisions of great consequence. Predictions and judgements are rendered on the basis of less than definitive observational data. Nevertheless, major strides have been achieved in our understanding of pathophysiological mechanisms and associated factors. A clearer picture has emerged of the high-risk patient with tetralogy of Fallot, in contrast to the adult with a systemic right ventricle or univentricular heart in whom ICD indications remain more contentious. Drafted by an international panel and endorsed by leading US, Canadian and European societies and associations, the first evidence-based consensus statement on arrhythmias in adults with CHD includes recommendations on pharmacological therapy for ventricular arrhythmias, catheter ablation and ICD indications. These recommendations are not the final word on the topic. Rather, they mark the beginning of a continuous process of refinement and improvement, as further research deepens our knowledge and propels the field forward.

## Footnotes

• Competing interests PK has received research grants from Medtronic and St. Jude Medical.

• Provenance and peer review Commissioned; externally peer reviewed.

## Request Permissions

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.