INTRODUCTION — The vast majority of cases of sudden cardiac arrest (SCA) and sudden cardiac death (SCD), due to ventricular fibrillation (VF), are associated with structural heart disease, particularly coronary heart disease (CHD). SCA in the normal heart is an uncommon occurrence, accounting for 5 percent of cases in an autopsy study of 270 cases [1]. (See "Pathophysiology and etiology of sudden cardiac arrest".)

Some causes of SCA in patients with apparently normal hearts have been identified. These include Brugada syndrome, long QT syndrome, preexcitation syndrome, and commotio cordis. Brugada syndrome (BS) and the related disorder, sudden unexpected nocturnal death syndrome (SUNDS), will be reviewed here. Other causes of SCA in normal hearts are discussed elsewhere. (See "Sudden cardiac arrest in the absence of apparent structural heart disease".)

ECG PATTERNS — The Brugada syndrome is associated with a peculiar pattern on the electrocardiogram (ECG) consisting of a pseudo-RBBB and persistent ST segment elevation in leads V1 to V3 (figure 1) [2-5]. Isolated cases have been described involving the inferior leads; such patients appear to have a unique mutation [6].

Three different patterns of ST elevation have been described (table 1 and figure 2) [7,8].

• In the classic Brugada type 1 ECG, the elevated ST segment (≥2 mm) descends with an upward convexity to an inverted T wave. This is referred to as the "coved type" Brugada pattern.
• The type 2 and type 3 patterns have a "saddle back" ST-T wave configuration, in which the elevated ST segment descends toward the baseline, then rises again to an upright or biphasic T wave. The ST segment is elevated ≥1 mm in type 2 and <1 mm in type 3.

Moving the right precordial chest leads up to the second or third intercostal space may increase the sensitivity of detecting these abnormalities [8,9]. In a study of 98 men with a family member with a type I Brugada ECG, those with a type I Brugada ECG pattern only in high chest leads had a similar rate of cardiac events during >1 year of follow-up as those with type 1 Brugada ECG with standard positioning of chest leads [10]. However, the incidence of false-positive results needs to be better defined in larger studies.

The widened S wave in left lateral leads that is characteristic of RBBB is absent in most patients with BS. This observation suggests that there is a high takeoff of the ST segment in the right precordium, ie, a "J" wave rather than a true RBBB [5].

QT interval prolongation may be seen in the right precordial leads [11]. The degree of prolongation is usually modest but some patients have genetic abnormalities that cause both BS and long QT syndrome [12-14]. (See 'Sodium channel gene' below.)

Transient ECG abnormalities — In some patients, the characteristic ECG changes are transient. This was illustrated in a series of 43 patients with BS, in whom serial ECGs were obtained over a median follow-up of 18 months [15]. The following findings were noted:

• Among 15 patients with a spontaneous type 1 ECG at presentation, 14 had at least one non-diagnostic (type 2, 3 or normal) ECG during follow-up.
• Among 28 patients whose initial ECG was non-diagnostic, eight developed characteristic type 1 ECG abnormalities during follow-up.

Thus, fluctuations in the ECG pattern appear to be common in BS.

Provoking factors — Characteristic ECG abnormalities may be exposed by a sodium channel blocker, such as flecainide, ajmaline, or procainamide, thereby identifying those at risk (table 2) [3,8,16-18]. In addition, flecainide can also induce QT prolongation that is primarily seen in the right precordial leads (V1 and V2) [11]. (See 'Drug challenge' below.)

Pacing, vagal maneuvers, and increased alpha-adrenergic tone also may provoke the typical ECG changes of BS [19,20]. Other factors that can unmask or modulate the ECG pattern of BS are beta blockers, tricyclic or tetracyclic antidepressants, lithium, local anesthetics, fever, hypokalemia, hyperkalemia, hypercalcemia, and alcohol and cocaine toxicity (table 2) [8,21]. The role of genetic predisposition with these factors is not clear, though a case report of bupivacaine-induced ECG changes was associated with an SCN5A mutation [8,22]. (See 'Autonomic tone' below and 'Cocaine abuse' below and 'Psychotropic drugs' below.)

The clinical significance of drug-provoked ECG changes in the absence of symptoms, family history of BS, or family with Brugada ECG is controversial.

Brugada pattern — The characteristic ECG pattern is only one component of the diagnostic criteria for BS. Patients with typical ECG features but no other clinical criteria are said to have the Brugada pattern but not BS [7]. (See 'Diagnostic criteria' below.)

EPIDEMIOLOGY

Prevalence — The prevalence of the typical ECG changes of BS has been evaluated in a number of different populations. Studies in heterogeneous populations suggest that the majority of affected individuals are Asian [5].

The following reports illustrate the range of findings:

• In two reports from Japan, the prevalence was 0.7 and 1.0 percent [23,24]; 0.12 to 0.16 percent of the Japanese population have type 1 (coved type) ST segment elevation [23,25,26].
• In a Finnish cohort, the prevalence of type 2 and 3 ST segment elevation was 0.6 percent; no type 1 patients were found in a screen of over 3000 apparently healthy individuals [27].
• In a large urban teaching hospital in the United States, the prevalence was 0.4 percent [28].

The prevalence of the Brugada pattern is much greater in patients who present with apparent idiopathic VF (3 to 24 percent in one series of 37 patients, depending upon the diagnostic criteria used) [29]. (See "Sudden cardiac arrest in the absence of apparent structural heart disease", section on 'Idiopathic VF'.)

Gender — The Brugada pattern is much more common in men than in women, being as much as nine times higher in one analysis [26], and men had a higher rate of syncope and sudden death in a large prospective registry study [30]. Reasons for the prominent male predominance in an autosomal dominant disorder are unclear. Animal models suggest that the impact of testosterone on ion currents, particularly outward potassium currents, may contribute to the increased incidence of clinical manifestations of BS in males [31,32].

Age at diagnosis — BS is usually diagnosed in adulthood. In two of the larger series of BS patients discussed below, the average patient age was 41 years [33,34]. BS is rarely diagnosed in children.

Brugada syndrome in children — The clinical features of BS in childhood were illustrated in a cohort of 30 children diagnosed at 13 referral centers in Europe [35]. The following findings were noted:

• Patient age ranged from infancy to 16 years (mean 8 years).
• Boys (n=17) and girls (n=13) were similarly affected.
• Twelve patients were diagnosed due to symptoms including unexplained syncope (n=10), aborted sudden death (n=1), and a supraventricular tachycardia. The remaining patients were identified due to family screening (n=17) or a suspicious ECG (n=1).
• Five children were treated with ICDs, and four were treated with hydroquinidine (see 'Quinidine' below.
• During a mean follow up of 37 months, one child died suddenly, and two received an appropriate ICD shock. All three of these patients initially presented with symptoms and had a spontaneous type I ECG (see 'ECG patterns' above.
• Fever was a prominent trigger of events. Five of the 10 patients presenting with syncope had fevers at the time of their initial event and the patient who died during follow up was febrile at the time.

Thus, in children with BS clinical manifestations and arrhythmic events are uncommon, but arrhythmias may be more likely during febrile episodes [35,36]. Potential reasons for delayed presentation include the impact of hormones [31,32] and the progressive structural degradation of myocytes [37].

Because the data are limited, the optimal approach to the management of children with BS is not known.

PATHOGENESIS — A variety of factors may contribute to the clinical manifestations of BS including right ventricular abnormalities, mutations in the cardiac sodium channel gene SCN5A, autonomic tone, and the use of cocaine and psychotropic drugs.

Genetics — The BS demonstrates autosomal dominant inheritance with variable expression. Genetic analysis has led to the identification of genes, encoding subunits of a cardiac sodium channel, that causes Brugada syndrome. Reasons for the variable expression of BS are not completely understood, although compound heterozygosity had been described in one family [38].

A second region on the same chromosome (chromosome 3), has been identified as potentially causative of BS, although a specific gene at this locus has not been characterized.

Sodium channel gene — Mutations in SCN5A, the gene that encodes the alpha subunit of the cardiac sodium channel gene, have been found in 18 to 30 percent of families with BS [8,33,39,40]. The gene locus is on chromosome 3p21-24.

The SCN5A mutations seen in BS are "loss of function" mutations and result in a variety of abnormalities in sodium channel activity including failure of expression, alterations in the voltage and time dependence of activation, and accelerated or prolonged recovery from inactivation [8]. In addition, mutations may explain the ability of sodium channel blockers to expose the ECG changes in some patients with this disorder [3,8,16-18]. (See 'Diagnostic criteria' below.)

The defective myocardial sodium channels reduce sodium inflow currents, thereby reducing the duration of normal action potentials. In addition, a prominent transient outward current, called I(to), in the right ventricular epicardium further shortens the action potential [41].

The relationship between sodium channel abnormalities and ST segment elevation is not fully understood. The ventricular myocardium is composed of at least three electrophysiologically distinct cell types: epicardial, endocardial, and M cells. The ST segment elevation and T wave inversions seen in the right precordial leads in BS are thought to be due to an alteration in the action potential in the epicardial and possibly the M cells, but not the endocardial cells [5,41-43]. The resulting dispersion of repolarization across the ventricular wall, which is most pronounced in the right ventricle, results in a transmural voltage gradient that is manifested in the electrocardiogram as ST segment elevation.

The means by which the SCN5A mutation may predispose to ventricular tachyarrhythmias is discussed below, but like the ST segment elevation, VT and VF may be caused by heterogeneity of the cardiac action potential, both across the three layers of myocardial cells and within the epicardium itself.

• Related disorders — Mutations in the SCN5A gene have also been associated with other electrophysiologic abnormalities:

• Isolated AV conduction defect
• Congenital long QT syndrome type 3 (LQT3)
• Congenital sick sinus syndrome
• Familial dilated cardiomyopathy with conduction defects and susceptibility to atrial fibrillation.

The differences in clinical manifestations are probably due to differences in the electrophysiologic abnormalities induced by the specific mutations [44-48]. Certain SCN5A mutations have been associated with an "overlap" syndrome, with affected patients exhibiting sick sinus syndrome or complete heart block as well as BS [47,48]. (See "Etiology of atrioventricular block", section on Familial.) (See "Genetics of congenital and acquired long QT syndrome", section on LQTS type 3.) (See "Manifestations and causes of the sick sinus syndrome", section on 'Childhood and familial disease' and "Genetics of dilated cardiomyopathy".)

Most of the mutations in BS are found at sites other than those known to contribute to the long QT syndrome [39,44]. One exception is a 1795insD mutation (insertion of aspartic acid) in the C terminus, which can cause both the long QT and BS as a result of its interaction with a heterogeneous physiologic substrate [12,13]. In another family, the mutation was associated with BS, long QT syndrome, and a conduction defect [14].

Further support for an overlap between BS and LQT3 comes from a report of 13 patients with LQT3 who received intravenous flecainide [46]. Although shortening of repolarization and the QT interval occurred in 12 of these patients, six (46 percent) developed ST segment elevation in V1 through V3, compatible with BS. (See "Genetics of congenital and acquired long QT syndrome".)

Additional loci — Since only a minority of affected families have an identified abnormality of SCN5A, it is possible that additional genetic abnormalities may produce the phenotypic characteristics of BS.

• Cardiac calcium channel gene — In a series of 82 probands with a clinical diagnosis of BS, seven individuals (8.5 percent) were found to have mutations in the alpha1 or beta2 subunit of the cardiac L-type calcium channel [49]. Three of these patients had a unique phenotype in that in association with the typical ECG findings of BS, they also had shortened QT intervals (≤360 msec). The relationship of this disorder to the usual form of BS and to short QT syndrome remains to be defined. (See "Sudden cardiac arrest in the absence of apparent structural heart disease", section on 'Congenital short QT syndrome'.)

• A locus on chromosome 3p22-25 has been identified in a large family with an autosomal dominant syndrome similar to BS (RBBB and ventricular arrhythmias) [50]. The mutation is also associated with progressive conduction disease. Compared to patients with SCN5A mutations, affected members of this family had a good prognosis, with a very low incidence of SCD. Also in contrast to mutations in the sodium channel gene, characteristic ECG changes were not exposed by infusions of procainamide.

A specific disease-causing gene at this locus has not yet been identified, which limits full characterization of this disorder and its relationship to the usual form of BS.

• Other sodium channel mutations - Other mutations that decrease the sodium channel current have been identified in isolated kindreds with Brugada syndrome. These include a mutation in SCN1B, an accessory subunit that interacts with the sodium channel, and is also associated with conduction system disease [51], and a mutation in GPD1-L, a gene that affects trafficking of the sodium channel [52].

Structural abnormalities — BS is not usually associated with structural heart disease. Standard cardiac testing, including echocardiography, stress testing, and cardiac magnetic resonance imaging often reveal no abnormalities. However, it is probably more accurate to categorize BS as a disorder that occurs in hearts that are apparently normal, since there is some evidence that subtle structural or microscopic abnormalities occur, including dilation of the right ventricular outflow tract and localized inflammation and fibrosis [37,53-56].

Supporting a pathogenic role of fibrosis in BS, a mouse model of heterozygous SCN5A knockout revealed age-dependent fibrosis and marked slowing of conduction velocity in the right ventricle [53]. Similarly, evaluation of an explanted heart from a transplant recipient with BS revealed microscopic fibrosis and conduction abnormalities [54].

Further evidence of microscopic abnormalities in BS comes from a series of 18 patients who underwent endomyocardial biopsy [37]. Although noninvasive evaluation was normal in each patient, all had evidence of microscopic structural abnormalities, including signs of RV myocarditis in 14. The ECG changes resolved at follow-up in 8 of the 14 patients with signs of RV myocarditis, raising the possibility that the BS ECG seen in these patients at presentation may have been a manifestation of RV myocarditis rather than an intrinsic cardiac ion channel abnormality.

Arrhythmogenesis — The most significant clinical manifestations of BS are ventricular arrhythmias, although there is some evidence that there may also be an increased rate of atrial fibrillation.

Ventricular arrhythmias and phase two reentry — Ventricular arrhythmias may result from the heterogeneity of myocardial refractory periods in the right ventricle. This heterogeneity arises from the presence of both normal and abnormal sodium channels in the same tissue, and from the differential impact of the sodium current in the three layers of the myocardium (see 'Sodium channel gene' above [41,42,57].

Within the epicardium, the juxtaposition of myocytes with different refractory periods can produce the triggers that initiate sustained arrhythmias (eg, closely-coupled premature beats) via a unique type of reentry called phase two reentry. In cardiac myocytes with defective sodium channels, initial depolarization is blunted (phase zero), and the counterbalancing effect of I(to) (phase 1) may be more significant. This phenomenon is more dramatic in the epicardium where I(to) currents are greater. In combination, this results in less initial depolarization, and reduced activation of the calcium channels that maintain the depolarized state during phase two. Thus, phase two of the cardiac action potential can be dramatically shortened.

The cells with impaired sodium channel function may fail to propagate the action potential, resulting in localized conduction block. However, due to the abbreviation of phase two, these same cells have a much shorter refractory period and recover excitability before the surrounding cells. The combination of localized conduction block and a shortened refractory period provides the substrate for localized reentry, which, in this case, is referred to as phase two reentry. The closely-coupled ventricular premature beats that result from phase two reentry may precipitate sustained ventricular arrhythmias. Optical mapping studies have supported this mechanism of initiation of polymorphic VT and VF in a canine model of Brugada syndrome [58]. This mechanism is also seen in flecainide-induced arrhythmias [59] and is similar to the mechanism of arrhythmogenesis during myocardial ischemia.

Other factors also may be important. Subtle structural abnormalities (eg, interstitial fibrosis or inflammation) in combination with reduced sodium currents could produce localized delay of impulse conduction [53]. The possible significance of conduction delay is less well established than action potential heterogeneity, and the issue is complicated by the possibility of overlap with other arrhythmia syndromes, particularly arrhythmogenic right ventricular cardiomyopathy. (See 'Relation to ARVC' below.)

Atrial fibrillation — Patients with BS may also be at increased risk of atrial arrhythmias, most notably atrial fibrillation (AF) [60-62]. The incidence of AF is higher in BS patients than in control subjects [60], and the presence of AF has been associated with increased disease severity and a higher risk of VF [62].

The relationship between BS and AF is illustrated by the following observations:

• Among 59 patients with BS and 31 matched controls, over three years of follow-up, AF occurred in 12 (20 percent) of the BS patients but none of the control subjects [60]. There was a significant correlation between a history of AF and inducibility of sustained ventricular arrhythmia on EP testing.
• In a series of 73 BS patients, AF occurred in 10 (14 percent) [62]. Patients with AF had a higher incidence of syncope (60 versus 22 percent of patients without AF), and VF (40 versus 14 percent).

These findings are consistent with a diffuse myocardial nature of the sodium channel abnormality.

Relation to ARVC — The Brugada ECG pattern can be an early subclinical manifestation of arrhythmogenic right ventricular cardiomyopathy (ARVC) [42]. ARVC is genetic disorder, usually autosomal dominant, that primarily involves the right ventricle, as the right ventricular myocardium is typically replaced by fat, with scattered residual myocardial cells and fibrous tissue. (See "Genetics and pathogenesis of arrhythmogenic right ventricular cardiomyopathy".)

An association between ARVC and the Brugada ECG pattern is suggested by a report of 96 victims of SCD who were ≤35 years and had a baseline ECG available [63]. Right precordial ST segment elevation with or without RBBB was present in 13 (14 percent); at autopsy all but one had ARVC. However, this study was from southern Italy, where ARVC is an important cause of SCA. Furthermore, mutations in SCN5A have not been described in ARVC.

Patients with ARVC often have abnormalities in the right ventricle that can be seen on echocardiography or cardiac magnetic resonance imaging. In contrast, the vast majority of patients with BS do not have apparent structural heart disease on routine imaging studies [8]. (See "Clinical manifestations and diagnosis of arrhythmogenic right ventricular cardiomyopathy", section on 'Evaluation'.)

Autonomic tone — An imbalance between sympathetic and parasympathetic tone may be important in the pathogenesis of BS, as suggested by the nocturnal occurrence of the associated tachyarrhythmias [64] and the alteration of typical ECG changes by pharmacologic modulation of autonomic tone [19,20].

Further support for the role of autonomic dysfunction comes from a study of 17 patients with BS who underwent scanning with iodine-123-metaiodobenzylguanidine (MIBG), a radiolabeled guanethidine analog that is actively taken up by sympathetic nerve terminals [20]. A segmental reduction in MIBG uptake was seen in eight of the 17 patients, but in none of 10 controls.

Fever — Data from a retrospective review of 111 patients with Brugada syndrome suggest that fever is a trigger for ECG changes and cardiac arrest [65]. In 24 of these patients with ECGs recorded during both fever and normothermia, fever was associated with Brugada syndrome-type ECG changes. Fever was present in four of 22 patients with cardiac arrest.

Cocaine abuse — The ECG findings of BS can be transiently induced by cocaine use [66]. Cocaine acts like a class I antiarrhythmic agent, producing local anesthetic effects via sodium channel blockade in the heart; this could explain the relation to the Brugada pattern. (See "Cardiovascular complications of cocaine abuse".)

Psychotropic drugs — Other drugs that block cardiac sodium channels have been associated with a transient Brugada pattern on the ECG, including overdoses with neuroleptic drugs or cyclic antidepressants [67,68]. In one report, a Brugada pattern was seen in 15 of 98 cases (15.3 percent) of a cyclic antidepressant overdose [68]. One patient with the Brugada pattern and one without died of refractory ventricular fibrillation.

DIAGNOSTIC CRITERIA — Brugada syndrome is not yet well defined and establishing the diagnosis is often difficult. Both ECG and clinical features are important [7,8,69]. The appearance of typical ECG changes alone is considered to represent an idiopathic Brugada ECG pattern but not the BS [7,8]. Genetic testing may also be useful in establishing the diagnosis. (See 'Role of genetic testing' below.)

Type 1 — In a consensus report from the Study Group on the Molecular Basis of Arrhythmias of the European Society of Cardiology, it was proposed that type 1 BS should be strongly considered in patients who meet the following criteria [7]:

• Appearance of type 1 ST segment elevation (coved type) (figure 2) in more than one right precordial lead (V1 - V3) in the presence or absence of a sodium channel blocker, plus at least one of the following:

      a)  Documented ventricular fibrillation
      b)  Self-terminating polymorphic ventricular tachycardia (VT)
      c)  Family history of sudden cardiac death at <45 years
      d)  Type 1 ST segment elevation in family members
      e)  Electrophysiologic inducibility of VT
      f)  Unexplained syncope suggestive of a tachyarrhythmia
      g)  Nocturnal agonal respiration

Type 2 and type 3 — The consensus report proposed that the diagnosis of BS should be strongly considered in patients with a type 2 or type 3 Brugada ECG who meet both of the following criteria [7]:

• Appearance of type 2 or type 3 ST segment elevation (saddle-back type) (figure 2) in more than one right precordial lead under baseline conditions, with conversion to type 1 following challenge with a sodium channel blocker (see 'Drug challenge' below)

• One of the features (a-g) described above

Drug-induced conversion of type 3 to type 2 ST segment elevation is considered inconclusive for the diagnosis of BS [8].

However, not all patients with a type 2 or type 3 Brugada ECG need to undergo drug challenge. In particular, we suggest that drug challenge is not necessary in patients who have documented VF, self-terminating polymorphic VT, unexplained syncope strongly suggestive of a tachyarrhythmia, or nocturnal agonal respiration (clinical features a, b, f, and g). Most such symptomatic patients will receive an implantable cardioverter-defibrillator regardless of the results of drug testing.

Thus, the main indication for drug challenge is in patients with clinical features (c) and (d), which reflect findings in the family. If the drug challenge is positive, EP testing should be considered. (See 'Role of EP testing' below.)

Drug challenge — Among patients with the Brugada type 2 or type 3 ECG pattern, the Brugada type 1 ECG pattern can occasionally be unmasked by sodium channel blockers (eg, flecainide, procainamide, ajmaline, pilsicainide) [8,16-18]. The reported sensitivity of pharmacologic challenge with these drugs has been variable, ranging from 100 percent [17] to as low as 15 percent [40].

The recommended doses by the second consensus conference on Brugada syndrome were [8]:

• Flecainide — 2 mg/kg over 10 minutes intravenously or 400 mg PO
• Procainamide — 10 mg/kg over 10 minutes intravenously
• Ajmaline — 1 mg/kg over five minutes intravenously
• Pilsicainide — 1 mg/kg over 10 minutes intravenously

The test should be performed under continuous ECG monitoring [8]. Indications for termination are development of a diagnostic type 1 Brugada ECG, a ≥2 mm increase in ST segment elevation in patients with a type 2 Brugada ECG, the development of ventricular premature beats or other arrhythmias, or widening of the QRS ≥30 percent above baseline.

A variety of other drugs have been associated with a Brugada-like ECG; it is not clear if a genetic predisposition is involved with these drugs (table 2) [8]. (See 'Provoking factors' above.)

Exclusion of confounders — Confounding factors that could account for either the ECG findings or syncope should be excluded. These include atypical right bundle branch block, left ventricular hypertrophy, early repolarization, acute myocardial infarction, acute pericarditis, and the ECG changes sometimes seen in the right precordial leads in well-trained athletes [8].

Role of genetic testing — Genetic testing for Brugada syndrome (BS) is commercially available and can be useful in confirming the diagnosis. However, the genetic and clinical heterogeneity of BS can limit the test's utility as illustrated by the following observations:

• Genetic testing typically involves sequencing SCN5A, the gene encoding the alpha subunit of the cardiac sodium channel. However, only 15 to 30 percent of patients diagnosed with BS have mutations in SCN5A [70]. This may reflect variations in the testing procedure since mutations in non-coding regions or alterations in splice sites are not always examined even though they can lead to abnormal sodium channel activity. In addition, there is evidence that mutations in other genes can give rise to this disorder [70] (see 'Additional loci' above.
• Not all patients with documented Brugada SCN5A mutations have BS. In one study, the average penetrance among 24 patients in four genotyped families was only 16 percent (one Brugada patient in each family) [40].

In light of these issues, we advise that genetic testing for BS be performed in conjunction with specialists who have expertise in this area.

CLINICAL MANIFESTATIONS AND RISK STRATIFICATION — All clinical manifestations of BS are related to life-threatening ventricular arrhythmias. Sudden cardiac arrest may be the first and only clinical event in BS, occurring in as many as one-third of patients. Arrhythmic events generally occur between ages 22 and 65 and are more common at night than in the day and during sleep than while awake [3,64]. SCA in Brugada patients is usually not related to exercise [63]. Stored electrograms from implantable cardioverter-defibrillators (ICDs) have found that frequent spontaneous premature ventricular beats, which are identical in morphology to those that initiate VF, are often seen before the onset of the arrhythmia [71].

Clinical risk factors for SCD — In one analysis of 200 symptomatic and asymptomatic patients with BS, male sex and a family history of SCD were identified as risk factors for subsequent SCD [33]. However, neither characteristic had a high specificity (26 percent for male sex and 65 percent for family history). The increased risk of male sex has been noted by others [34].

Risk stratification usually begins with determining the presence or absence of associated symptoms.

Previous SCA or syncope — Patients with a previous history of SCA and those with a history of syncope (unexplained syncope suggestive of a tachyarrhythmia) are at increased risk for subsequent arrhythmic events compared to asymptomatic individuals [33,34,72,73]. This was demonstrated in a review of 334 patients with the Brugada pattern [72]. Of these patients, 71 had presented after cardiac arrest (group A), 73 after a syncopal episode (group B), and 190 with ECG findings alone (group C). During a mean follow-up of 33 months, a new arrhythmic event (SCD or VF) occurred in 62 and 19 percent of group A and B patients, while only 8 percent of group C patients had a first arrhythmic event (graph 1).

A later study from the same group identified risk factors for SCD or VF among patients with syncope [34]. Among 547 patients with an ECG diagnostic of type 1 BS and no prior cardiac arrest; 124 had at least one prior syncopal episode and 302 had a family history of SCD. The ECG was diagnostic spontaneously in 391 and only after administration of an antiarrhythmic drug in 156. EP testing was performed at the discretion of the physician in 408 patients, 163 of whom had an inducible sustained ventricular arrhythmia. The following findings were noted:

• On multivariate analysis, inducible sustained ventricular arrhythmia and a history of syncope were adverse predictors.
• At a mean follow-up of 24 months (range 1 to 160 months), 45 patients (8.2 percent) had SCD or VF. Patients with a previous history of syncope, inducible sustained ventricular arrhythmia, and a spontaneously abnormal ECG had a 27.2 percent probability of SCD or documented VF. The lowest probability (0.5 percent) was seen in patients lacking all of these features, while patients with one or two features had an intermediate risk (eg, 4.1 percent in patients with syncope and a spontaneously abnormal ECG who were noninducible) (table 3).

Syncope was also identified as a risk factor for subsequent arrhythmic events in the report of 200 patients cited above, with a sensitivity of 36 percent and a specificity of 85 percent [33]. The combination of syncope and spontaneous ST segment elevation in leads V1 to V3 identified a subgroup with a particularly high risk for subsequent SCD (hazard ratio 6.4).

Asymptomatic patients — The risk of cardiac arrest is much lower in asymptomatic patients, although subgroups of asymptomatic patients with increased risk can be identified [33,34,72,73]. The largest experience comes from a review of 547 patients with the type 1 Brugada ECG pattern, 422 of whom were asymptomatic [34]. Among the asymptomatic patients, two features were important determinants of arrhythmic risk:

• The presence of a type 1 ECG abnormality spontaneously versus only after drug challenge (see 'Drug challenge' above)

• Inducible ventricular tachyarrhythmia on EP testing

At a mean follow-up of 24 months, the following probabilities of arrhythmic events (SCD or documented VF) were reported (table 3):

• Among patients with a spontaneous type 1 ECG, event rates for those with positive and negative EP testing results were 14 and 1.8 percent, respectively.
• Among patients with the type 1 ECG abnormality only after drug challenge, event rates for those with positive and negative EP testing results were 4.5 and 0.5 percent, respectively.

A low risk of an arrhythmic event (5 percent by age 41) among asymptomatic patients who had ST elevation only with sodium channel blocker administration was also noted in the study of 200 patients [33].

Role of EP testing — The role of EP testing in patients with known or suspected BS depends largely upon the presence or absence of associated symptoms. Patients with a Brugada ECG pattern and certain high risk clinical features (ie, a history of SCA, sustained ventricular tachyarrhythmias, or unexplained syncope) are known to have an increased risk of SCD [3,5,34,42,72,74]. EP testing may provide additional information, but because such patients already have an indication for ICD therapy, test results are unlikely to impact management decisions. (See 'Use of an ICD' below.)

The role of EP testing in asymptomatic patients remains an area of investigation and debate [75,76]. The report of 547 patients described above is the largest study of the predictive value of EP testing in Brugada syndrome [34]. These results demonstrate that abnormal EP test results (inducible VT or VF) predict an increased risk of life-threatening arrhythmias in asymptomatic patients with type 1 Brugada ECG abnormalities, particularly if of the type 1 ECG is spontaneous rather than drug-induced. However, although patients with inducible VT or VF had a several-fold increase in arrhythmic events compared to those who were not inducible, the positive predictive value of EPS was low.

In the report of 200 symptomatic and asymptomatic patients cited above, among the 86 patients who underwent EP testing, a positive study had a sensitivity and specificity of 66 and 34, respectively [33]. Several factors may have contributed to the failure to detect a benefit from EP testing in this cohort: the number of patients who underwent EP testing was small; the population was not restricted to higher risk patients with type 1 ECG abnormalities, and 49 percent of the patients who did have type 1 abnormalities had this finding only after drug challenge. Thus, the overall population was at lower risk than in the study of 547 patients in which all patients had type 1 ECG abnormalities, 71 percent of which occurred spontaneously [34].

Risk stratification summary — Recommendations for risk stratification for SCD in BS have been addressed in reports from several major societies.

• The 2006 American College of Cardiology/American Heart Association/European Society of Cardiology (ACC/AHA/ESC) guidelines for the management of ventricular arrhythmias and the prevention of SCD included the following comments and recommendations with regard to risk stratification [77]:

    - A history of syncope together with spontaneous ST segment elevation is included as a predictor of SCA. This risk factor is also included as an indication in the 2008 American College of Cardiology/American Heart Association/Heart Rhythm Society (ACC/AHA/HRS) guidelines for device-based therapy (see 'ICD therapy summary' below.

    - There is agreement that patients with spontaneous ST-segment elevation have a worse prognosis than individuals with ST-segment elevation only after pharmacologic drug challenge. Clinical monitoring for the development of a spontaneous ST-segment elevation pattern is reasonable for the management of patients with ST-segment elevation induced only with provocative pharmacological challenge with or without symptoms.

    - Family history is not endorsed as a risk factor. It should NOT be assumed that asymptomatic individuals with the characteristic ECG but without family history are at low risk or that family members of an individual with SCD at at increased risk.

    - The evidence is less well established for the use of EP study to risk-stratify patients with a spontaneous ST elevation with or without mutation in the SCN5A gene.

• In contrast, the second consensus conference on Brugada syndrome recommended an approach with greater reliance on EP testing; this approach was endorsed by the Heart Rhythm Society and the European Heart Rhythm Association (algorithm 1A-B) [8]. EP testing was recommended in most asymptomatic patients and, in symptomatic patients, only for assessment of supraventricular arrhythmias.

TREATMENT — Although pharmacologic therapy has been tried, the only therapy with proven efficacy in preventing sudden death is an implantable cardioverter-defibrillator (ICD).

Pharmacologic therapy — There are no proven pharmacologic treatments for preventing SCD in BS, although there are data suggesting benefit from quinidine. In a report of 63 patients, an ICD was implanted in 35, pharmacologic therapy with beta blockers and/or amiodarone was administered in 15, and 13 were not treated [78]. During a 34 month follow-up, the incidence of arrhythmic events was similar in the three groups, but there were no deaths in the ICD group compared to mortality rates of 26 and 31 percent in the other two groups. All deaths were due to SCD. It was concluded that amiodarone and/or beta-blockers do NOT protect patients with BS against SCD.

Other antiarrhythmic drugs may be deleterious, particularly sodium channel blockers. As noted above, a sodium channel blocker, such as flecainide, ajmaline, or procainamide, can transiently induce the characteristic type 1 ECG changes [3,8,16-18]. In addition, sodium channel blockade can induce ventricular premature beats or ventricular tachycardia in patients with BS, particularly in symptomatic patients (six of 10 in one report), and T wave alternans [79]. (See 'Drug challenge' above.)

Quinidine — In contrast to the above observations, quinidine and hydroquinidine (which are also sodium channel blockers) may be beneficial in patients with BS [80,81]. This effect may be mediated by blockade of I(to), the transient outward current, that increases heterogeneity and may promote ventricular premature beats that act as the trigger for VT/VF [41].

The potential efficacy of quinidine was illustrated in a report in which EP testing was performed on 25 patients with BS (15 symptomatic and 10 asymptomatic) before and after treatment with quinidine bisulfate (mean dose 1483±240 mg/day) [80]. Ventricular fibrillation was inducible in all patients at baseline but in only three patients after three to seven days of quinidine therapy. Quinidine treatment was continued in 19 patients for a mean of 56 months. None of these patients had an arrhythmic event, although two had non-arrhythmia-related syncope. Side effects occurred in 36 percent and resolved after cessation of therapy. Quinidine can also attenuate ST segment elevation in some patients [8,41].

Similar benefits have been noted with hydroquinidine, including a reduction in defibrillator discharges in four patients who had received an ICD [82].

Pharmacologic therapy summary — Despite these encouraging findings, there is not enough clinical experience with the use of quinidine or related agents to recommend their use in preference to an ICD. (See 'ICD therapy summary' below.)

However, antiarrhythmic drugs may have a role in patients with an ICD who continue to have frequent discharges. Amiodarone is usually the drug of choice but quinidine or hydroquinidine may be an alternative in patients with BS [80,82,83]. (See "Pharmacologic therapy in survivors of sudden cardiac arrest", section on Adjunctive therapy with an ICD.)

The 2006 ACC/AHA/ESC guidelines include the following recommendations regarding treatment of electrical storm in Brugada syndrome [77]:

• Isoproterenol is reasonable for treatment of electrical storm in Brugada syndrome.
• Quinidine might be reasonable treatment of electrical storm in Brugada syndrome.

Use of an ICD — ICD implantation can prevent SCD in patients with BS. However, selecting which patients should receive an ICD is challenging. The arrhythmic risk varies among BS patients, and ICD implantation carries some risk, including a relatively high rate of inappropriate shocks (6 to 9 percent per year) [8,78,84,85].

In the study of 63 patients cited above, there were no deaths in the 35 patients receiving an ICD compared to mortality rates of 26 and 31 percent among those treated pharmacologically or not treated, respectively [78]. However, these patients represent a high-risk cohort, as all had a spontaneous type 1 ECG and 41 had a history of syncope or SCA. (See 'Clinical risk factors for SCD' above.)

Benefit from an ICD was also documented in a review of 690 patients with BS in a multicenter registry; 258 received an ICD because of perceived high risk of SCD due, for example, to a history of symptoms or a family history of SCD and/or a Brugada ECG pattern [8]. A sustained ventricular arrhythmia was induced during EP testing in 77 percent. At a mean follow-up of 2.5 years, 69 patients (27 percent) had at least one appropriate ICD discharge.

The role of the ICD in BS patients with a lower risk profile is less clear. The issue was addressed in a heterogeneous series of 220 patients [84]. Although all patients had a type 1 ECG, this abnormality was spontaneous in 62 percent and induced with sodium channel blockade in the remaining 38 percent. In addition, more than half of these patients (114) were asymptomatic at presentation, and the asymptomatic patients received ICDs on the basis of an abnormal EP study (99 patients), or a family history of SCD or nonsustained VT (15 patients). At a mean follow-up of 38 months, the following findings were noted:

• Appropriate ICD shocks occurred in 8 percent of the overall population (2.6 percent per year). Total arrhythmic events over the three years were significantly more common in patients who had a history of SCA or syncope prior to ICD implantation (22 and 10 percent respectively, compared to 4 percent among those who were asymptomatic).
• 45 patients (20 percent) experienced inappropriate shocks.
• The overall complication rate was 28 percent, which included the inappropriate shocks.

Thus, in this heterogeneous cohort of BS patients, including a significant percentage of asymptomatic patients and patients with a provocable (as opposed to spontaneous) type 1 ECG, overall event rates were relatively low, but were significantly higher in patients with a history of syncope or SCA.

ICD therapy summary — Based upon the high incidence of SCD in selected patients with BS, major societies have reached the following conclusions regarding the indications for an ICD in patients with BS.

The 2008 American College of Cardiology/American Heart Association/Heart Rhythm Society (ACC/AHA/HRS) guidelines for device-based therapy of cardiac rhythm abnormalities included the following statements regarding ICD therapy in BS [86]:

• There is evidence and/or general agreement supporting ICD implantation in all BS patients with a prior cardiac arrest.
• The weight of evidence/opinion supports ICD implantation in BS patients who have had a history of syncope.
• The weight of evidence/opinion supports ICD implantation in BS patients with a history of VT that did not result in cardiac arrest.

In all of the above settings, the guidelines stated that ICD implantation should only be considered in patients who have a reasonable expectation of survival with a good functional capacity for more than one year.

The implications of a positive family history of SCD are uncertain, and the 2008 guidelines did not address the prognostic significance of SCD in a family member with BS. As noted above, the 2006 ACC/AHA/ESC guidelines noted that family history is not reliable risk factor. (See 'Risk stratification summary' above.)

We generally agree with this approach, with one caveat regarding syncope: other causes of syncope, such as typical vasovagal events, bradycardia, or neurologic causes, must be excluded before proceeding to ICD implantation.

• The second consensus conference on Brugada syndrome recommended a somewhat more aggressive approach; this approach was endorsed by the Heart Rhythm Society and the European Heart Rhythm Association [8].

However, ICD use may be too expensive in some countries. In such patients, the consensus conference concluded that relatively high doses of quinidine (1200 to 1500 mg/day) may be beneficial [8,80]. (See 'Quinidine' above.)

Investigative approaches — Other, still investigative approaches have been evaluated in patients with BS who are at risk for SCD. These include:

• Focal radiofrequency ablation to prevent the ventricular premature beats that trigger VT/VF. In an initial report that included three patients with BS who had episodes of VF or polymorphic VT, there was no recurrence of symptomatic ventricular arrhythmia at a mean follow-up of 17 months [57].
• The administration of cilostazol, which is a phosphodiesterase inhibitor that impairs platelet aggregation and is approved for the treatment of intermittent claudication. Possible benefit in BS may be mediated by an increase in calcium current and reduction in I(to) due to an increase in heart rate. In a case report of a 67 year-old man with BS who had daily early morning episodes of VF, the episodes were completely prevented by cilostazol, recurred when the drug was discontinued, and were again prevented when the drug was restarted [87].

SUDDEN UNEXPECTED NOCTURNAL DEATH SYNDROME — A sudden unexpected nocturnal death syndrome (SUNDS, also called sudden unexpected death syndrome or SUDS) has been described in young, apparently healthy males from Southeast Asia; this syndrome has several names including lai tai (death during sleep) in Thailand, bangungut (to rise and moan in sleep followed by death) in the Philippines, and pokkuri (unexpected SCD at night) in Japan [88-90].

A low serum potassium may contribute to SCA in these patients [8]. It has been suggested that a high carbohydrate meal may precipitate SCA, perhaps by increasing the secretion of insulin which drives extracellular potassium into cells.

A relationship to BS was initially suggested by the observation that a majority of patients with SUNDS have the ECG manifestations of BS [90]. This association was confirmed by the finding that these patients have mutations in the same cardiac sodium channel gene, SCN5A, that is abnormal in BS [91].

Based upon these observations, it has been concluded that SUNDS and BS are phenotypically, genetically, and functionally the same disorder [8,91]. Thus, the management of these patients should be the same as that for classic BS (algorithm 1A-B).

The DEBUT trial evaluated the role of an ICD compared with beta blockers in 66 patients who were considered definite or probable SUNDS survivors [92]. The trial was prematurely terminated after a mean follow-up of 24 months because of four deaths in the beta blocker arm, compared with none in the ICD arm; seven patients in the latter group had recurrent VF that was appropriately terminated by the ICD. Similar benefits had been noted in a pilot study of 20 patients. (See "Role of implantable cardioverter-defibrillators for the secondary prevention of sudden cardiac death".)