5 Sudden cardiac death: epidemiology and prevention

Even if sudden cardiac death is considered to be the most frequent cause of death in adults in industrialized countries, its incidence varies widely, depending on the definition and the source and quality of underlying data. It is estimated that about 70–80% of cases are due to coronary heart disease. The remaining 20% are attributable to a wide variety of inborn, genetically determined or acquired diseases, including a small group with hitherto undefined background.

Prevention primarily encompasses the treatment of cardiovascular risk factors to avoid manifestations of coronary heart disease. Furthermore, preventive strategies are targeted to define groups of patients with an increased risk for sudden cardiac death or individuals at risk in specific populations, e.g. competitive athletes. A major target group are patients with impaired left ventricular function, preferentially due to myocardial infarction. These patients, and some less clearly defined patient groups with non-ischaemic cardiomyopathy and heart failure, may benefit from the insertion of an implantable cardioverter–defibrillator. With regard to pharmacological prevention, treatment of the underlying condition is the mainstay, since no antiarrhythmic substance—with the exemption of ß-blockers in some situations—has shown to be of efficacy.

Even if several definitions of the term ‘sudden cardiac death’ (SCD) have been used, in more recent investigations, it is usually defined as a natural death due to cardiac causes, heralded by an abrupt loss of consciousness within 1 hour after the onset of acute symptoms [1]. Regrettably, the existence of different definitions of SCD makes meaningful comparisons between studies difficult, if not impossible [2]. Moreover, the equalization of SCD with sudden cardiac arrest, defined as an unexpected arrest within 1 hour after the onset of symptoms which is reversed by medical intervention, makes interpretations even more complicated [2]. Even the generally accepted definition of SCD as above is hampered by several problems. One is the mostly unproven definitive cardiac cause of death and another the unclear duration of symptoms, specifically in unwitnessed arrests. Indeed, a prolonged time delay from the first signs of change in clinical status to the definitive loss of consciousness and cardiac arrest, as, for example, used in the Maastricht study [3], may increase the number of SCD cases, but at the cost of lowering the proportion of deaths due to definitive cardiac causes. In a recently published study, signs of change or cardiac symptoms preceding cardiac arrest were present for <1 hour in only 116 of 323 patients and were present up to 10 hours in the majority of unexpected cardiac deaths, according to observations of eyewitnesses. [4]. Bypassing the problem of unwitnessed arrest by requiring an observation of the victims alive within 24 hours of their death is therefore not very reliable. Finally, the notoriously low autopsy rates remain to be a major obstacle for the exact definition of the burden of the problem.

The incidence of SCD (see Table 5.1) is varying widely, depending on the data source, e.g. data from rescue services [2, 5] or from death certificates, the definition of SCD, and the population investigated [6, 7]. The geographic zone of the investigation may play a role. In addition, secular changes in the frequency and presentation of SCD have been observed.

The estimations, e.g. for the US population of 300 million, vary between 180 000 and 450 000 SCD cases/year, even in recent publications, corresponding to <60 to about 150 cases of SCD per 100 000 population/year [2, 8]. The highest numbers are derived from death certificates which may lead to a significant overestimation. In prospective studies [6, 9], the SCD rate is pronouncedly <100/100 000 population/year. In an ongoing study in Oregon, which tried to use a multimodality approach (emergency medical system (EMS) data, coroner’s and hospital reports), the incidence was 53/100 000/year [6]; similar findings were reported in Europe [10, 11]. In contrast, in the Maastricht study, which included witnessed and non-witnessed sudden arrests of assumed cardiac origin, the incidence was 100/100 000 in the population aged 20–75 years [3]. In former West-Berlin, with a population of 2.1 million, we observed 93 EMS-initiated resuscitation attempts/100 000/year in witnessed and non-witnessed arrests of assumed cardiac origin in the years 1987–1989 [12]. Calculated for the total of European populations of 730 millions, this incidence would result in a burden of about 500 000–650 000 SCD cases all over Europe per year.

The incidence of SCD is increasing with age and is more frequent in males, compared to females, as is coronary heart disease (CHD), which is the most common cause of SCD [13]. It is estimated that about 70–80% of cases of SCD are attributed to this condition [14, 15]. Since CHD becomes manifest in females about 10–15 years later than in males, the proportion of females with SCD increases in the old and very old population [16, 17]. SCD may be triggered by acute ischaemia, clinically mostly presenting as an ACS due to a plaque rupture or erosion and partial or complete obstruction of a coronary artery. On the other hand, SCD may be caused by re-entry loops due to post-infarction scares, often initiated by additional ischaemia [15]. The probability for the latter mechanism increases with impaired (left) ventricular function. Although CHD is the cause of SCD, even in the majority of cases in younger adult victims of SCD, several other conditions are exposing predominantly younger people and some specific groups to risk. Incident atrial fibrillation (AF) may increase the SCD risk slightly [18]. Other conditions may be related to drugs, trauma, inflammatory, infiltrative, dilated and structural cardiomyopathies, and a wide variety of genetically determined abnormalities. Also the combination of obesity, sleep apnoea, and a seizure disorder increases the risk for SCD [19]. Finally, a smaller group of SCD cases has a hitherto undefined background.

Several drugs with different indications may increase the risk of sudden arrhythmic death. Diuretics, for example, may lead to critical loss of potassium. Tricyclic antidepressants may prolong the QT-interval, specifically in females and individuals with a genetic disposition; the same may be observed even with cardiac medications, e.g. sotalol. Other cardiac medications may lead to high-degree atrioventricular (AV) block; amphetamines may induce ventricular tachycardia (VT) or VF, and cocaine abuse may provoke coronary spasm and consecutive fatal arrhythmia. Finally, many antiarrhythmic drugs also have proarrhythmic properties [1, 20, 21]. Even antiepileptic medication is under discussion [22]. Specific forms of genetically determined infantile-onset severe epilepsy combined with arrhythmia, the Dravet syndrome, however, may be the actual reason for sudden death in epilepsy [23]. Whereas moderate alcohol consumption may be cardioprotective, massive alcohol abuse goes along with an increased risk of SCD [24, 25].

In rare cases, SCD may be triggered by blunt thorax trauma (so-called ‘commotio cordis’), sometimes even without findings at autopsy [26, 27]. Inflammatory diseases causing sudden death also are preferentially seen in younger patients [28]. Several viruses have been identified to generate myocarditis, which often follows a recent viral syndrome. Enteroviruses are most common, but also adenoviruses. Cytomegalovirus (CMV), herpes simplex, Epstein–Barr virus (EBV), and even human immunodeficiency virus (HIV) infections have been reported to cause myocarditis-associated SCD [29]. Specific subtypes of myocarditis as hypersensitivity and toxic myocarditis or giant cell inflammation of the heart are other seldom observed causes for SCD, as is the cardiac involvement of sarcoidosis [30]. In single cases, SCD may also be provoked by one of the rare storage diseases. Finally, in some patients, coronary abnormalities are causative for SCD [31].

Arrhythmogenic right ventricular dysplasia (ARVD) is a genetic cardiomyopathy, characterized by right ventricular (RV) thinning and fat infiltration, and may be the cause of SCD in younger victims. Interestingly, and in contrast to the findings in North-America, ARVD is accountable for most SCD cases in competitive athletes in northern Italy [32]. In the US, hypertrophic cardiomyopathy is the first-line cause of SCD in competitive athletes [26, 27]. This disease is mostly based on a variety of inherited abnormalities of sarcomeric protein synthesis.

So-called channelopathies (Brugada syndrome, long QT (see Figures 5.1 (A) and (B)) or short QT syndrome, and catecholaminergic polymorphic VT) represent primarily electrical disorders of genetic determination which may be responsible for SCD, preferentially in children and adolescents [28, 30, 33–37]. It is suggested that, on the background of one of these diseases, similar to patients suffering from CHD, additional triggers initiate the final catastrophic event. These triggers may be environmental conditions, physical or emotional stress, or endogenous factors such as the circadian variation of blood pressure, hormonal levels, or endogenous fibrinolytic activity [38–44]. All the listed conditions and an additional small group of unexplained aetiology add up to about 20% of all SCD victims not related to CHD.

VF or pulseless VT is thought to be the most frequent arrhythmia precipitating SCD in up to 80% of patients [15]. The proportion of patients with VF as the documented initial arrhythmia, however, varies widely in the literature [45–50]. In recent years, it has been repeatedly observed that the distribution of arrhythmia found in victims of SCD is changing, with a decreasing proportion of patients in VF and an increase in other arrhythmias [47–50]. Data from the Seattle Fire Department, for example, revealed an absolute decrease in the incidence of SCD, due to a reduction in VF of >50% in blacks and whites (but not in Asians or Pacific islanders) in the years from 1980 to 2000. This observation, however, was offset, in part, by an increase in victims found in pulseless electric activity (PEA) or asystole [47] (see also Figure 5.2). It is suggested that improved prevention of CHD [14, 51] and more widespread use of specific prevention measures in patients at high risk for SCD may at least partially explain this observation. Treatment with antipsychotic drugs seems to be associated with arrest in PEA and also may be involved in that change [52].

The key problem of epidemiology and prevention of SCD, however, is related to the fact that its incidence is high in a very small well-defined, high-risk group of the total population, whereas the vast majority of cases will occur in a population with very low, or nearly no definable, risk. This fact, first described by Myerburg [53], represents the key challenge in SCD prevention (see Figure 5.3).

Preventive strategies are divided in primary prevention for patients at risk which had no event to date and secondary prevention for survivors of a cardiac arrest or life-threatening VT [54–56]. Primary, but also secondary prevention will basically encompass consequent treatment of cardiovascular risk factors, i.e. stopping smoking, appropriate exercise, keeping a prudent diet, and adequate treatment of hypertension, hyperlipidaemia, and diabetes, according to the guidelines. Clearly, cessation of drug and/or alcohol abuse is necessary. For patients with signs of heart failure, optimal treatment of this condition is equally important [56].

Achieving an increased awareness of the problem among the public, and even more among patients at higher risk and their relatives, is also a crucial point for primary prevention. Learning the skills of basic life support (BLS), including the use of automated external defibrillators (AEDs), should be part of the education for schoolchildren and needs broader implementation among the public, as well as the recognition and prompt reaction on warning signs and symptoms of a heart attack by individuals at risk and their relatives [4].

In a more specific sense, primary prevention targets a group of patients with a high risk of SCD by definition [53]. For example, while aiming to find individuals at elevated risk for SCD among young competitive athletes, abnormal findings disclosed in a routine 12-lead ECG may be suggestive of cardiomyopathies [32].

Patients with an LVEF <40 % of ischaemic and non-ischaemic cause represent a typical high-risk group. Electrophysiological testing may be helpful in further defining the risk in those patients and may guide catheter ablation of arrhythmic structures [54, 56, 57]. For ischaemic cardiomyopathy, every attempt should be made for revascularization to prevent ischaemia—the most important trigger for fatal arrhythmias.

Lidocaine has been tested for the prophylaxis of SCD due to VF in AMI in several studies of different design. Even if the incidence of VF could be reduced in some studies, a meta-analysis revealed a trend towards higher mortality with lidocaine prophylaxis, due to an increase of deaths resulting from asystole [58].

Primary prevention with β-blockers—mostly performed in the pre-reperfusion era—resulted in a significant reduction of SCD [59–62]. Whereas, in the ISIS I study [60], IV initiation of β-blocker prophylaxis was an advantage, later studies performed in patients receiving reperfusion treatment revealed no better results with this strategy [57]. In contrast, patients with heart failure at presentation may develop CS if treated with (high-dose) IV β-blockers [63]. Therefore, oral treatment after stabilization of the patient is actually recommended.

In the CAST trials, three class I antiarrhythmics (encainide, flecainide, and moricizine) were tested for the prevention of SCD in post-MI patients [1, 20]. The studies resulted in an increased mortality (SCD and non-sudden death) with active treatment. With D-L-sotalol, a class III antiarrhythmic with some β-blocker properties, a significant reduction in reinfarctions was achieved, but SCD rate was unchanged [64]. Amiodarone, another class III antiarrhythmic, was also tested in several studies for the prevention of SCD. Neither in the large EMIAT [64] nor in the CAMIAT [65] studies was a reduction in total mortality achieved, even if a significant decrease in arrhythmic death was observed. Since amiodarone apparently has no proarrhythmic effect, it has been successfully used in combination with β-blockers to suppress VF/VT episodes in patients treated with an implantable cardioverter–defibrillator (ICD) [66]. In summary, none of the available antiarrhythmic drugs—with the exemption of β-blockers—has proven to be effective in the prevention of SCD [58].

ICD therapy is recommended as a class I, level of evidence A indication for patients with a functional New York Heart Association (NYHA) class II or III, with an LVEF of 30–40% in ischaemic and 30–50% in non-ischaemic cardiomyopathy [55, 67–70]. To select patients for primary prevention, however, is difficult, since no clinical test or electrophysiological study leads to definitive results [7, 17, 54, 56, 71, 72]. Studies performed to select patients effectively, who will benefit from these devices, postulate an intensive electrophysiological and structural investigation of the heart as well as of deoxyribonucleic nucleic acid (DNA) and serum samples [73].

With respect to outcomes of randomized clinical studies, one of the problems is that, in some ICD trials [74], side effects of antiarrhythmic treatment in the control arm, e.g. amiodarone, may have led to worse outcomes for the controls, compared to the ICD group. The timing of ICD insertion is even more problematic. The MADIT I and MADIT II trials [71, 75] resulted in favour of primary prevention by ICD treatment. The device was implanted 6 months, or later, after the index infarction in more than three-quarters of the patients. In the VALIANT study [57], a secondary prevention study testing an ACE-I against an ARB, a high number of patients died from SCD shortly after the index event. Therefore, an early insertion of the device within 6–40 days after an MI was tested in patients with impaired ventricular function in the DINAMIT and IRIS studies [17, 50, 76, 77]. Indeed, in both trials, the SCD incidence was reduced in the early phase of observation but was later counterbalanced by more deaths of non-arrhythmic causation [78]. For a definitive decision to implant an ICD, often time is needed for ventricular recovery after an MI or for the treatment of another potentially reversible condition, which may be bridged by a wearable defibrillator device [6, 79].

Another specific problem is primary prevention in patients suffering from non-ischaemic cardiomyopathies with low ejection fraction (EF). Until now, studies including this group of patients resulted in encouraging results in NYHA class II patients, but only in trends favouring ICD treatment in more severe heart failure [74, 80, 81]. Better outcomes were seen if an ICD is combined with cardiac resynchronization treatment (CRT) [82].

A further problem is the efficacy of an ICD in the elderly [83]. Most probably due to concomitant diseases, e.g. renal failure or chronic obstructive pulmonary disease (COPD), the efficacy of ICD treatment is of limited efficacy in elderly patients. Concomitant diseases and life expectancy therefore should be a consideration in ICD therapy planning.

In patients, even in children, with genetically determined arrhythmic syndromes like long QT or Brugada syndrome, an ICD may be indicated (together with β-blockers, notably in specific subgroups of the long QT syndrome [84] and in catecholaminergic VT) in high-risk groups, as well as in selected patients with inflammatory, infiltrative cardiomyopathy or groups of exceedingly high risk with hypertrophic cardiomyopathy [35, 55, 85, 86]. Potentially lethal arrhythmias in these patients are often triggered by emotional or physical stress. Competitive sports or other stressful activities therefore should be avoided [19, 34, 85].

Since channelopathies demonstrate ECG abnormalities—spontaneous or provoked—awareness for typical alterations in routine ECG reading is decisive in identifying individuals at risk and in primary preventive measures. Also the observation of these abnormalities offers potential for family screening, including the clarification of genetic mutations in more than two-thirds of cases. Channelopathies do not show any structural or histological abnormality of the heart at autopsy. Therefore, it is of utmost importance to keep these rare conditions in mind in order to identify other affected family members.

Secondary prevention refers to patients who survived a sudden cardiac death or a life-threatening sustained VT. If cardiac arrest occurs in patients beyond 48 hours of onset of an AMI, the risk of a recurrent event may also be high [87]. Early implantation of an ICD in these patients, however, resulted in a trend of reduction of SCD but not a reduction in total cardiovascular mortality [76, 77]. All patients with an ischaemic cause of arrest should have optimal treatment for this condition. In patients suffering from structural heart disease, as well as from heart failure, targeted treatment of the specific condition is necessary to reduce the risk of future events. Implantation of an ICD is indicated for all survivors of a cardiac arrest for secondary prevention [55], regardless of the underlying heart disease.

SCD is the most frequent mode of death in industrialized countries, even if numbers differ widely, depending on the definition and sources of data. CHD and its complications are the commonest causes of SCD, specifically in the presence of an impaired LV function. Inflammatory diseases, trauma, and genetically determined disorders represent a minor group of victims of SCD but are particularly meaningful, since they affect preferentially younger patients, including otherwise healthy athletes.

Primary strategies for the prevention of SCD should be targeted on general measures to prevent CHD and the optimal treatment of the sequelae of this disorder. However, primary prevention also targets a group of patients with manifest heart disease defined by a reduced LVEF of <40%. Since antiarrhythmics are of limited value for this high-risk group and since VF is the most frequent arrhythmia precipitating SCD, implantation of a cardioverter–defibrillator may be indicated for primary prevention. These devices are lifesaving for patients with an EF of 30–40% with ischaemic, and an EF of 30–50% with non-ischaemic, heart disease and for patients who had survived a life-threatening tachyarrhythmia or a cardiac arrest, for secondary prevention. These well-defined target groups, however, comprises only a small percentage of potential victims in the general population. Therefore, awareness of the problem among the general public, even more so in high-risk patients and their relatives, training in BLS, including the use of AEDs, as well as its implementation among the public, and even among schoolchildren, is of outstanding importance to successfully reduce the burden of SCD.