Description of the condition

From a clinical point of view, any unexpected death can be considered a 'sudden death', brought on by conditions as diverse as arrhythmias, aortic dissection, subarachnoid haemorrhage, acute myocardial infarction or massive pulmonary embolism. Traumatic death is usually excluded from this category. As there may be prognostic and therapeutic differences (i.e. subarachnoid haemorrhage or pulmonary embolism), researchers and clinicians recognise a distinct category known as 'sudden cardiac death' (SCD). A widely accepted definition is "natural death due to cardiac causes, heralded by abrupt loss of consciousness within an hour of the onset of acute symptoms; pre‐existing heart disease may have been known to be present, but the time and mode of death are unexpected" (Myerburg 2004).

SCD is one of the leading causes of cardiac death. Incidence increases with age and is three to four times more frequent in men than women at all ages (Merghani 2013; MMWR 2002). Accurately estimating its real incidence is difficult, but according to data obtained from death certificates, SCD may cause 63.4% of total cardiac mortality in the United States (MMWR 2002). This data probably overestimates SCD prevalence, as it is based only on clinical presentation (MMWR 2002; Zheng 2001). Incidence rates, varying from 0.36 to 1.28/100,000 participants per year, have been reported by some emergency services, but these tend to underestimate the real incidence, as it only refers to participants who survive to the hospital (Sara 2014). Incidence increases from 1/100,000 for those aged < 35 years to 100/100,000 in individuals aged ≥ 35 years old (John 2012). A prospective observational study reported that 7% to 18% of overall mortality in the general population (of all ages) in the USA was due to SCD (Stecker 2014). Globally, estimated incidence of SCD would then be approximately 4 to 5 million cases every year. However, this number may be inaccurate, as, on the one hand, SCD incidence rates in low‐ and middle‐income countries may not be equivalent to those in high‐income countries (Vedanthan 2012), and on the other hand, incidence has declined over the past two decades, from 4.7 per 1000 person‐years in 1990–2000 to 2.1 per 1000 person‐years in 2001–2010 (Niemeijer 2015).

The main pre‐existing heart disease leading to SCD in high‐income countries is coronary heart disease (CHD); there is a general acceptance that SCD accounts for around 50% of all CHD‐related death and that the proportion of all SCDs resulting from CHD is around 80% (Myerburg 2012). Other types of cardiopathy (e.g. hypertrophic cardiomyopathy, non‐ischaemic cardiomyopathy, arrhythmogenic right ventricular dysplasia) can also lead to SCD, while there is no structural abnormality in only 5% of cases (Consensus 1997). The SCD event is most commonly caused by the sudden onset of monomorphic ventricular tachycardia (VT) that degenerates into ventricular fibrillation (VF), and less frequently by the abrupt onset of polymorphic VT/VF, bradyarrhythmias or heart blocks (Priori 2001; Zipes 2006). However, the proportion of participants with pulseless electrical activity or asystole has increased over the past two decades (Teodorescu 2010).

More recently, research has identified diabetes mellitus as an independent risk factor for SCD (Jouven 2005).

The downward trend in SCD incidence might be due to better diagnosis and treatment of heart disease and most importantly primary prevention of SCD and cardiovascular disease in general through improved management of behaviours and other risk factors (Niemeijer 2015).

In this context, clinicians use electrophysiological (EP) testing with intracardiac recording and electrical stimulation at baseline, followed by administration of antiarrhythmic drugs for arrhythmia assessment and risk stratification. EP testing has been used to document the inducibility of VT, evaluate drug effects, assess the risks of recurrent VT or SCD, and assess the indications for implantable cardiac defibrillator (ICD) therapy. For example, in participants with CHD, asymptomatic non‐sustained VT and a left ventricular ejection fraction (LVEF) less than 40%, the inducibility of sustained VT ranges from 20% to 40% and confers worse prognosis, with an increased risk of SCD or death from other causes (Buxton 2000). However, in participants with CHD and a lower LVEF (less than 30%), non‐inducibility doesn't necessarily portend a good prognosis (Buxton 2002), and persistent inducibility while receiving antiarrhythmic drugs confers an even worse prognosis (Wilber 1990).

Description of the intervention

Amiodarone, one of the main class III antiarrhythmics, is a benzofuran derivative approved by the US Food and Drug Administration (FDA) for the treatment of patients with life‐threatening ventricular tachyarrhythmias when other drugs are ineffective or not tolerated (FDA 2013). Researchers have proposed the drug as an alternative to ICD, categorising it as having a 2a level of evidence (weight of evidence is in favour of usefulness/efficacy) for prophylaxis of SCD in participants with CHD and left ventricular (LV) dysfunction (Zipes 2006).

The onset of action after intravenous administration is generally within one to two hours, but after oral administration, the onset of action may require anywhere from two to three days and often from one to three weeks. On occasion, it may take even longer, to the point that achieving a steady state without a loading dose takes about 265 days (Braunwald 2001).

However, research has also associated the use of amiodarone with toxicity involving the lungs, thyroid gland, liver, eyes, skin and nerves (Connolly 1997). Pulmonary toxicity is the drug's most serious potential adverse effect, and some series have described its frequency as high as 17%, although the incidence when compared with placebo is less than 1% (Pollak 1999). Thyroid toxicity is the most common complication requiring intervention, occurring in up to 10% of participants receiving long‐term amiodarone therapy. Minor adverse effects are nausea, anorexia, photosensitivity, and a blue discolouration of the skin (Siddoway 2003).

The frequency of most adverse effects is related to total amiodarone exposure (Siddoway 2003), but amiodarone is slowly, variably and incompletely absorbed, which makes adverse events unpredictable. Extensive hepatic metabolism occurs with desethylamiodarone as a major metabolite, both extensively accumulating in the liver, lung, fat, 'blue' skin, and other tissues.

How the intervention might work

Generally speaking, class III antiarrhythmic drugs act by prolonging the action potential´s duration of the myocardial cell by lengthening the repolarisation phase and thus the effective refractory period. This prolongation is believed to facilitate termination and prevention of both ventricular re‐entry arrhythmias by producing block within re‐entry circuits, and thus providing both an elevation of the ventricular fibrillation threshold and a reduction of the ventricular defibrillation threshold (Brendorp 2002).

As a class III antiarrhythmic drug, amiodarone prolongs the QT interval, slows the heart rate and atrioventricular nodal conduction (via calcium channel and beta‐receptor blockade), prolongs refractoriness (via potassium and sodium channel blockade), and slows intracardiac conduction (via sodium channel blockade) (Siddoway 2003). By blocking the potassium repolarisation currents, it can inhibit or terminate ventricular arrhythmias by increasing the wavelength for reentry (Zipes 2006).

Why it is important to do this review

According to current evidence, ICD therapy, when compared with antiarrhythmic drugs, reduces mortality in high risk participants with reduced LVEF in both CHD and non‐ischaemic cardiopathy, for both primary and secondary prevention (AVID 1997; Connolly 2000; Desai 2004; Kuck 2000).

While ICD therapy may improve survival in selected patient populations, it may diminish patients' quality of life (Gehi 2006). In a study comparing ICD versus no ICD in participants who underwent coronary artery bypass graft surgery, the use of ICD was associated with lower levels of psychological well‐being and reduced physical and emotional role functioning (Namerow 1999). On the other hand, a recent analysis by Marks et al. from the Sudden Cardiac Death ‐ Heart Failure Trial (SCD‐HeFT) showed that subjective measures of physical function did not differ significantly between the ICD and placebo groups at any time point, but there was a short‐term increase in psychological well‐being among participants with ICD therapy throughout the first year after implantation, a benefit that did not persist at 30 months (Bardy 2005–SCD‐HeFT). The occurrence of ICD shocks reduced the quality of life, but only if quality of life was measured within one to two months after the shock (Bardy 2005–SCD‐HeFT).

However, the elevated up‐front costs of ICD therapy (between EUR 11,000 and 19,000) impede its ready availability in the health systems of low‐ and middle‐income countries, even though costs tend to diminish along patients' longevity (Biffi 2011).

Regarding amiodarone, evidence from trials has been inconsistent, with some studies showing a moderate effect and others no effect at all (Bardy 2005–SCD‐HeFT; Heidenreich 2002; Strickberger 2003). Indirect evidence of effect comes from a recent systematic review concluding that ICD discharges were reduced in participants with ICD plus amiodarone compared to participants with ICD alone. Assuming ICD discharges follow ventricular arrhythmias, one could infer that amiodarone reduces the number of arrhythmic episodes (Ferreira 2007), provided the ICDs were programmed with similar arrhythmia detection times, as different detection times could mean different thresholds for considering any disturbance to be an arrhythmic episode (Scott 2014).

On the other hand, data collected since the 1980s has convincingly proven that beta‐blocking treatment is associated with an improved clinical outcome in several patient groups. The efficacy of this treatment in people with post‐myocardial infarction (MI) relates to a drug‐associated reduction in all‐cause mortality and is not necessarily related to the time after the acute event, when therapy starts (Yusuf 1985). People with a history of congestive heart failure (CHF) or depressed left ventricular function tend to experience the greatest benefits in mortality reduction.

Data suggests that in participants post‐MI, potasium channel blockers such as dofetilide or d‐sotalol have neutral or even harmful effects regarding all‐cause mortality (Køber 2000; Torp‐Pedersen 1999; Waldo 1996), and, for example, both of these drugs result in a higher rate of Torsade de Pointes than amiodarone (Brendorp 2002). However, calcium‐channel blockers such as verapamil have shown favourable effects, if only in people without heart failure (DAVIT II 1990).

It is also important to note that chronic treatment with antiarrhythmic drugs is associated with severe adverse effects, including the potential induction of life‐threatening arrhythmias (e.g. increased mortality is associated with the long‐term use of quinidine; Coplen 1990).

A previous systematic review of randomised controlled trials, which evaluated amiodarone versus other antiarrhythmics or placebo for the prevention of SCD in participants with or without ICD, concluded that it significantly reduced SCD and cardiac mortality, but not all‐cause mortality. However, the authors did not carry out a separate analysis for participants with or without ICD, and the review only evaluated amiodarone for primary prevention (Piccini 2009).

If amiodarone proves to be beneficial in SCD prevention, it would constitute a valid alternative in situations where economic constraints limit the widespread use of ICD.