Background

Congenital heart disease is present in 0.9% of living births; currently, 90% of those affected will reach adulthood because of recent progress made in paediatrics, cardiology, surgery and resuscitation . Among what are commonly called “grown-up congenital heart diseases” (GUCHs) or, more recently, “adult congenital heart diseases” (ACHDs), tetralogy of Fallot (TOF) has a preponderant place, because of its relatively high prevalence (7–10% of all congenital heart diseases; 1/3500 to 1/4300 in the adult population) , and because it is possible to have corrective surgery, leading to almost normal anatomy and physiology in adulthood. Indeed, very long-term follow-up has demonstrated that health status is excellent, with a mortality rate that is considered to be low (14% mortality for hospital survivors at 40-year follow-up, after surgery performed in the 1970s), even if it is still higher than in the general population, mainly because of heart failure and ventricular arrhythmias . Thus, the third reason for making TOF one of the main GUCHs of concern is the late risk of malignant ventricular arrhythmias and sudden cardiac death (SCD).

The aim of this review is to highlight the mechanisms of ventricular arrhythmias in TOF, and to present current knowledge of secondary and primary prevention of SCD in this setting.

SCD in GUCH patients

The occurrence of SCD in patients with previous surgical repair of congenital heart defects is a tragic event, as many are usually considered to be “cured” of their congenital heart disease (even if this terminology may be a bit too optimistic), with rather low mortality rates and usually excellent quality of survival. Some of these SCDs are probably linked to paroxysmal high-degree atrioventricular block (e.g. in repaired TOF with relevant intraventricular conduction disturbances, but also in ventricular septal defect, cushion defect or congenitally corrected transposition of the great arteries). Some other SCDs are caused by ventricular fibrillation induced by fast ventricular rates during supraventricular tachycardia (e.g. atrial tachycardia with 1/1 atrioventricular conduction after atrial switch for transposition of the great arteries, or after Fontan procedures for single ventricle). Other causes probably include haemodynamic compromise, embolism, myocardial infarction or aneurism rupture, but it is now clear that most SCDs (a proportion estimated at around 75% ) are secondary to arrhythmias, and among these, malignant ventricular arrhythmias have been documented in 85% of cases at the time of the cardiac arrest .

Even if the culprit GUCH for SCD has changed in recent decades, TOF remains one of the main GUCHs carrying the risk of late SCD. In 1974, congenital aortic stenosis, Eisenmenger’s syndrome, TOF and hypertrophic obstructive cardiomyopathy were responsible for more than half of SCDs in children (non-operated defects in most) ; whereas TOF, systemic to pulmonary anastomosis, pulmonary hypertension caused by left to right shunt and dilated cardiomyopathy were present in half of the SCDs in a report published 10 years later (postoperative in a significant number of cases) . In 1998, Silka et al. found that in 3600 patients with GUCH, most late SCDs were linked to aortic stenosis, aortic coarctation, transposition of the great arteries and TOF, leading to a yearly SCD rate of 0.22% (50 to 200 times higher than in the general population); most were suspected to have an arrhythmic cause . Similar causes of SCD have been found in more recent studies . However, even if it is always a clinically relevant issue, SCD is not the major cause of death in GUCH, representing only one-quarter of all-cause mortalities in adults with GUCH in many series , or even fewer, according to more recent data , while it is known to account for half of cardiac deaths among those with acquired heart disease.

Prospective trials regarding SCD in GUCH are lacking for many reasons:

• •
  

  first, these are very heterogeneous defects;

  
  
• •
  

  second, the number of cases needed to reach statistical power makes any trial almost unachievable (e.g. TOF would require 1700 patients to be followed for 10 years) ;

  
  
• •
  

  third, earlier conclusions may not be relevant to newer generations of patients with GUCH (complete repairs during the neonatal period, improved physiological outcomes, extended survival of patients with more complex forms) .

SCD and ventricular arrhythmias in TOF

The modern management of TOF began before the 1960s, after the first surgical complete repair performed by Dr D. C. Lillehei in 1954. As early as 1975 – around 15 to 20 years later – the first cases of unexpected cardiac arrest in repaired TOF were noticed, and were already suspected to be caused by malignant ventricular arrhythmias .

Since then, SCD has been identified repeatedly as the major cause of mortality in repaired TOF, representing, however, only one-third to half of all-cause mortalities in different series . Over the past 30 years, many studies have evaluated the risk of SCD in TOF to be between 0 and 0.8% yearly . In a meta-analysis including 39 studies and 4583 patients, the late rate of SCD in TOF was evaluated to be 1.8% over more than 8 years of follow-up, i.e. an annual SCD rate of only 0.15% , which is, of course, higher than in the general population, but does not constitute an objectively “high” risk sufficient to warrant systematic prevention. For comparison, a yearly risk of SCD of up to 0.8% in hypertrophic cardiomyopathy does not constitute an indication for an implantable cardioverter defibrillator (ICD) . Interestingly, the risk of late SCD in TOF is not linear but increases over time, especially 20–25 years after the surgery . However, it is still unclear if this increase is simply caused by ageing or is dependent on ongoing progress in surgical techniques.

For the occurrence of sustained ventricular tachycardia (VT), the risk has been evaluated as 4% at 21-year follow-up (i.e. 0.2% yearly) , and as 14% in 556 adults operated on for TOF after a follow-up of around 30 years . The risk also changes with age, with an increase at around 40 years of age . Thus, it is tempting and logical to mainly link SCD in TOF to the occurrence of sustained VT.

Mechanisms of VT and SCD in TOF

Data on unrepaired TOF are scarce, as these usually have a very poor prognosis ; what we know comes from ancient studies, where ventricular arrhythmias were commonly found in older TOF patients before repair .

For repaired TOF, conduction disturbances are probably not a common cause of SCD because long-term evolution from right bundle branch block to bi-/tri-fascicular block is uncommon , and because postoperative bi-fascicular block (without transient high-degree atrioventricular block) has never been shown to lead to late SCD . Only transient complete heart block that persisted beyond the third postoperative day has been correlated with the risk of late SCD . Even if prolonged QRS duration has been correlated with SCD , QRS prolongation relates rather to right ventricular size, and predicts malignant ventricular arrhythmias . Moreover, autopsies in patients with TOF who died from SCD revealed that the conduction tissues were undamaged, but found extensive fibrosis of the right ventricular myocardium at both the ventriculotomy site and the septum . Ventricular arrhythmias such as premature beats and documented ventricular fibrillation were soon suspected to be involved in SCD mechanisms and, finally, repeated documentation of sustained VT during aborted SCD in repaired TOF patients sealed the causal relationship between ventricular arrhythmias and SCD.

After the first cases of surgical or intracardiac mapping and ablation, reentry circuits around the surgical ventriculotomy scars, using fibrotic areas of slow conduction, have been shown to be the cause of sustained VT in repaired TOF . It is not known, however, if ventricular fibrillation is always caused by degenerating monomorphic sustained VT.