1 Introduction

In 1992, Brugada and Brugada described eight patients who had the following distinct clinical characteristics: (1) the presence of coved ST-segment elevation followed by a negative T wave in the right precordial leads and (2) life-threatening ventricular arrhythmias that could lead to sudden cardiac death, cardiac arrest, or symptoms caused by spontaneous self-terminating ventricular tachycardia (VT) or ventricular fibrillation (VF) episodes, i.e., syncope, agonal respiration, and seizure (Brugada and Brugada 1992). The entity is later widely recognized as “Brugada syndrome.”

The Brugada syndrome (BrS) has also been linked with sudden unexpected death syndrome (SUDS) that usually occurs at night in young Southeast Asian men with a structurally normal heart (Nademanee et al. 1997; Vatta et al. 2002). BrS patients often have a history of unexplained sudden cardiac death in the family, and the syndrome is a well-known autosomal dominant inherited arrhythmia disorder and associated with gene mutations that are predominantly confined to the SCN5A gene, which encodes for the α-subunit of the cardiac sodium channel, causing loss of sodium current (INa) (Vatta et al. 2002; Wilde et al. 2002; Antzelevitch et al. 2005, 2016; Priori et al. 2013). However, the causal role of these genetic variants remains controversial.

SUDS has the same phenotype as BrS (Nademanee et al. 1997), and shortly thereafter we reported that both syndromes also shared the same genetic and biophysical basis (Vatta et al. 2002). In the past two and a half decades, we have witnessed major progress toward better understanding of the syndrome and gained knowledge in epidemiology, genetic aspects of the syndrome, pathophysiology, and patient management (Wilde et al. 2002, 2010; Antzelevitch et al. 2005, 2016; Priori et al. 2013; Probst et al. 2009; Nademanee et al. 2011; Veerakul and Nademanee 2012). Four consensus reports were published to help in defining diagnostic criteria, risk stratification, and management of BrS patients (Wilde et al. 2002; Antzelevitch et al. 2005, 2016; Priori et al. 2013). However, there are controversies, especially on the role of sodium channel mutations and electrophysiologic mechanisms underlying the syndrome(Wilde et al. 2010), risk stratification (Veerakul and Nademanee 2012; Wilde et al. 2010; Priori et al. 2012; Brugada et al. 2002; Probst et al. 2010; Kamakura et al. 2009; Paul et al. 2007; Eckardt et al. 2005; Takagi et al. 2007; Giustetto et al. 2009), and treatment of asymptomatic patients (Priori et al. 2012; Takagi et al. 2007; Giustetto et al. 2009; Viskin and Rogowski 2007; Veerakul et al. 2008).

2 Clinical Presentation and Diagnostic Criteria

The clinical spectrum of BrS patients ranges from asymptomatic to sudden cardiac death (Antzelevitch et al. 2005; Priori et al. 2013; Veerakul and Nademanee 2012). Patients may have a late onset of VT/VF despite having had an abnormal ECG pattern for decades (Nademanee et al. 1997; Antzelevitch et al. 2005). Syncope or seizures due to self-terminating VT/VF episodes are also common as well as agonal respiration and difficulty to arouse at night time again due to self-terminating VF episodes (Nademanee et al. 1997; Antzelevitch et al. 2005).

Life-threatening ventricular tachyarrhythmias in patients with BrS usually consist of VF or a rapid polymorphic VT although rare cases of monomorphic VT have been reported (Boersma et al. 2001; Shimada et al. 1996; Rodríguez-Mañero et al. 2016). In a study looking at the stored electrogram of the implantable cardioverter-defibrillator (ICD), polymorphic VT was often preceded by ventricular premature beats (VPB) with the same morphology as the one that initiated the VT (Kakishita et al. 2000). The VPB typically occurred at the end of the T wave and only rarely followed a long-short sequence. Site of earliest ventricular activation in patients with VT induced in the electrophysiologic lab almost always involves the right ventricular outflow tract (RVOT), at least for the initial beats. VF and sudden cardiac death mostly occur at rest or during sleep or at night in the early morning hours (Nademanee et al. 1997; Matsuo et al. 1999; Aizawa et al. 2016).

Supraventricular tachycardia, including AV nodal reentry tachycardia, atrial flutter, and atrial fibrillation, is common (Eckardt et al. 2001; Morita et al. 2002a; Rodríguez-Mañero et al. 2013). Depolarization abnormalities including prolonged P wave, PR interval, and QRS duration are also frequently observed (Antzelevitch et al. 2005; Takagi et al. 2007; Maury et al. 2013).

Diagnosis of BrS relies on the signature marker (type I Brugada pattern) on the ECG. The two original Brugada consensus (Wilde et al. 2002; Antzelevitch et al. 2005) reports classified the Brugada ECG pattern into three types (Fig. 8.1):

  1. 1.

    Type 1 pattern has ST elevation >2 mm giving rise to a coved-type ST segment, in electrical continuity with a negative T wave and without a separating isoelectric line.

  2. 2.

    Type 2 has a high take-off ST-segment elevation. In this variant, the J-point elevation (>2 mm) gives rise to a gradually descending elevated ST segment (remaining >1 mm above the baseline) and a positive or biphasic T wave. This ST-T segment morphology is referred to as the saddle-back type.

  3. 3.

    Type 3 is the coved or saddle-back type with <1 mm ST-segment elevation.

Fig. 8.1
figure 1

Three types of Brugada ECG pattern. Type 1 is a coved-type pattern and type 2 is a “saddle-back” type which has the ST elevation >2 mm without T-wave inversion. Type 3 pattern is a J-point elevation without ST elevation >1 mm

One has to be cognizant that the Brugada ECG pattern often is wax and wane. Sodium channel blockers, ajmaline, procainamide, and flecainide, could be used to unmask the ECG pattern—since the details of how to perform a drug challenge test to unmask the Brugada ECG pattern have been nicely reviewed elsewhere, we shall not repeat here(Wilde et al. 2002; Antzelevitch et al. 2005, 2016). In recent years, it has become clear that the right ventricular outflow tract (RVOT) is the likely arrhythmogenic substrate site. And the RVOT is the only cardiac structure lying just beneath the third and second intercostal space. We and others have demonstrated that placement of right precordial lead ECG recordings over the higher intercostal spaces (third and second intercostal space) significantly increases diagnostic yield in bringing the Brugada ECG pattern (Veerakul et al. 2000; Shimizu et al. 2000). In our institution, we always routinely record right precordial lead ECG (V1–V3) from fourth, third, and second intercostal space in every patient suspected of BrS. Figure 8.2 shows an example of the ECG tracings from a patient with BrS: the Brugada ECG pattern is absent in the conventional fourth intercostal space lead placement but become apparent in the higher intercostal spaces (third and second).

Fig. 8.2
figure 2

An example of how placement of right precordial leads in higher intercostal spaces unmasks the Brugada ECG pattern in a BrS patient. The higher intercostal spaces (ICS), third and second ICS of V2 in this patients showed distinct coved-type Brugada ECG pattern (type 1)

The original criteria for the diagnosis of BrS mandate that the patient must have type 1 Brugada ECG pattern in at least two leads with or without a sodium channel blocker challenge test and one of the following clinical manifestations: (1) a history of spontaneous VT/VF episodes or aborted sudden cardiac death, (2) a family history of sudden cardiac death or coved-type ECG, (3) agonal respiration during sleep, or (4) inducibility of VT/VF by programmed electrical stimulation. However, since several potential flaws exist in these criteria, they have since then been refined. While the clinical manifestations mentioned above remain important in recommending treatment and risk stratification, they are no longer listed among the diagnostic criteria, and type I Brugada ECG pattern in at least one lead (recorded from either the second, third, or fourth intercostal space) is now required for the diagnosis (Priori et al. 2013). Recently, another guideline was proposed utilizing a scoring system for the diagnosis of Brugada syndrome in the same fashion as the system for long QT syndrome, the so-called Shanghai score (Antzelevitch et al. 2016). Unfortunately, the benefit and the accuracy of the Shanghai score remain unclear, and one may use the old criteria and merely categorize the patients into two groups: symptomatic or asymptomatic BrS.

A Brugada ECG pattern can also be seen in several other conditions. Thus, the diagnosis can be made in a patient showing the Brugada ECG pattern with demonstration of a grossly normal heart by cardiological tests and in the absence of other conditions that can mimic the Brugada ECG pattern (Table 8.1).

Table 8.1 Conditions which can mimic Brugada ECG pattern (Brugada phenocopy) (Antzelevitch et al. 2016)

3 Epidemiology

The prevalence of type I Brugada ECG pattern was estimated to be 1:2000 based on combined studies encompassing >400,000 patients from North America, Europe, and Asia (Postema 2012). The prevalence, however, varies widely among countries with low prevalence in North America (0.005–0.1%) and Western countries including Europe (0–0.017%) compared to Asia (0.15–0.27% in Japan, 0.18% in the Philippines) (Probst et al. 2007). Interestingly, several epidemiologic studies in Thailand have demonstrated that the prevalence of type I Brugada electrocardiogram (ECG) pattern is higher in Thailand than in other Asian countries and even much higher than that in the European countries and the USA. The incidence is ranging from 0.8% to 1.8% in nonfebrile patients to 4% in febrile patients [fever is a well-known precipitating factor for BrS (Rattanawong et al. 2016)] compared to a much lower incidence in Europe and in the USA.

The majority of BrS patients are relatively young between 20 and 40 years (Antzelevitch et al. 2005; Probst et al. 2007) although cases have been reported in children as young as 2 days old and in elderly up until the age of 84 years (Antzelevitch et al. 2005; Probst et al. 2007). Despite what appears to be an autosomal dominant inheritance pattern, BrS has up to tenfold higher prevalence in males with greater severity (Nademanee et al. 1997; Benito et al. 2008). Worldwide, the syndrome is probably responsible for 4–12% of all sudden deaths and at least 20% of sudden deaths in patients with structurally normal hearts (Wilde et al. 2002; Antzelevitch et al. 2005).

4 Genetics of BrS

Because of occurrence in siblings and family history of sudden death in affected individuals, genetics was thought to play an important part of BrS since the first description in 1992 (Brugada and Brugada 1992). Transmission in familial cases is consistent with autosomal dominant with incomplete penetrance, especially in females whose incidence is up to 10 times lower than in male. In 1998, Chen et al. reported the first mutation, linked to BrS, in the SCN5A gene which encodes for the α-subunit of the sodium channel (Chen et al. 1998). Since then, more than 300 SCN5A mutations have been reported in BrS and represent currently the most common genotype. Functional studies demonstrate that SCN5A mutations in BrS cause loss of function of the sodium channel due to decreased expression of the sodium channel protein (Nav1.5) on the sarcolemma (Valdivia et al. 2004), expression of nonfunctional channels (Kyndt et al. 2001), or altered gating properties (delayed activation, earlier inactivation, faster inactivation, enhanced slow inactivation, and delayed recovery from inactivation) (Bezzina et al. 1999; Dumaine et al. 1999; Akai et al. 2000; Amin et al. 2005; Nakajima et al. 2015; Kinoshita et al. 2016). The loss of function of the sodium channel results in a decrease in sodium current and in turn impairs the fast upstroke of phase 0 of the action potential causing a slow conduction in the heart.

Even though SCN5A mutations are the most common defect found in 11–28% of BrS probands, the genetics of BrS have become heterogeneous. In addition to SCN5A mutations, more mutations are found in gene encoding protein of potassium and calcium channels. There have been now mutations of at least 18 genes that have been associated with BrS phenotype (Table 8.2). The contribution of the other 17 genes (except SCN5A) as the etiology of BrS, however, is small, and the presence of rare variants in these genes as the sole cause of BrS has been called into question due to similar percentage of normal individual carrying rare variants in these genes compared to BrS patients (Le Scouarnec et al. 2015).

Table 8.2 Gene mutations associated with Brugada syndrome (Antzelevitch et al. 2016)

Despite the fact that many genes have been identified and linked to the syndrome, gene mutation alone cannot explain the phenotype in full. Kapplinger et al. found nearly 300 SCN5A mutations in 211 unrelated probands (Kapplinger et al. 2010). The SCN5A mutations were found in 21% of patients with the Brugada phenotype and 2–5% of healthy controls, respectively. These findings suggest an important role of SCN5A mutations causing loss-of-function sodium channel in the phenotype manifestation. However, 80% of these mutations were only present in a single individual or one family, and a causal role of these mutations in the Brugada syndrome is far from clearly established.

Probst et al. studied 13 large families with SCN5A mutations and revealed the following intriguing findings: Many of the mutation carriers did not have the Brugada signature sign on ECG nor could it be provoked by sodium channel blockers (Probst et al. 2009). Moreover, in 5 of the 13 families with more than five clinically affected individuals, there were one or two affected individuals with the Brugada phenotype who did not have the familial SCN5A mutations. Furthermore, the Brugada ECG pattern was induced in eight mutation-negative patients (Probst et al. 2009). These findings along with a report of a case of identical twins carrying a SCN5A mutation of which only one displayed the phenotype suggest that SCN5A mutations may act as modifiers rather than causative mutations (Sakabe et al. 2002a).

SCN5A mutations may cause not only BrS but other diseases as well. Indeed, SCN5A mutations have also been associated with long QT syndrome (Bezzina et al. 1999; Wang et al. 1995), cardiac conduction disease (Tan et al. 2001), sick sinus syndrome (Benson et al. 2003), atrial fibrillation (Darbar et al. 2008; Olson et al. 2005), and dilated cardiomyopathy with overlap syndromes identified in specific families (Meregalli et al. 2009).

Because of the seemingly complex inheritance pattern in BrS, a genome-wide association (GWAS) study was done to explore the contribution of common single nucleotide polymorphisms (SNP) in this disease. In this study, the investigators could demonstrate association of three common SNP in SCN5A, SCN10A, and HEY2 with BrS. Each “risk allele” was associated with an odd ratio of 1.6–2.8 for having Brugada syndrome (Bezzina et al. 2013). Risk was progressively higher with presence of more risk alleles and was as high as 21.5 if 5 or 6 risk alleles were present in a single individual. This study suggests that the genetic basis of BrS and susceptibility to VF therein in the individual patient is not caused by a single major genetic mutation (classical Mendelian view) but rather by inheritance of multiple susceptibility genetic variants (oligogenic) acting in concert through one or more mechanistic pathways. This finding partially explains why simple genetic testing may have easily missed these rare genetic variants in many of the BrS patients.

Because of the complexity of genetic mechanism in this disease, genetic testing for the diagnosis of BrS is not as helpful as some other monogenic diseases associated with sudden cardiac death like the long QT syndrome or catecholaminergic polymorphic VT (CPVT). Given the background rate of rare variants of unknown significance (VUS) in SCN5A of approximately 2% in normal population and 20% yield in patients with BrS (signal-to-noise ratio of 10:1), the HRS/EHRS Expert Consensus Guidelines written in 2011 listed genetic testing for the diagnosis of Brugada syndrome as “can be useful” (Class IIa) in a proband with clinical diagnosis of BrS (Ackerman et al. 2011). The test is probabilistic rather than deterministic, and interpretation of an abnormal test must consider all the clinical and molecular genetic findings. Due to presence of rare genetic variants in the normal population, a false-positive result is possible. For example, a recent study of 870 whole exome sequencing data and 6161 genotype array data in general population found that of 28 variants associated with BrS, none of whom had Brugada marker on the ECG. Syncope, ventricular arrhythmia, and mortality were also not significantly different between those who had the variants and those who did not (control group) (Ghouse et al. 2017).

The role of genetic testing in risk stratification remains unclear. Crotti et al. reported their findings of a comprehensive mutational analysis of all 12 BrS genes for a single large cohort of BrS patients (Crotti et al. 2012). They found putative pathogenic mutations in 21% of their BrS cohort. Similar to other reports, 78% of the mutations in BrS were still confined to SCN5A. Interestingly, in male patients, the yield of positive testing varied from 11% in those older than 40 years of age to 21% in male patients 20–40 years of age and to 83% in male patients younger than 20 years of age. The BrS patients with prolonged PR interval >200 ms had very high incidence of SCN5A mutations (39%) compared to those with normal PR interval. The yield of identifying the mutations is similar between those who have only the typical type 1 Brugada ECG pattern (asymptomatic cases) and those with symptoms and/or family history of sudden cardiac death, so presence of identifiable mutation did not appear to correlate with the symptoms or prognosis. Based on their findings, the authors recommended genetic testing for all patients who just have the type 1 Brugada ECG pattern (asymptomatic cases) as well as symptomatic cases. Due to their findings of a low prevalence of non-SCN5A mutations, they suggested that it was reasonable to test most patients for SCN5A mutations alone first, with further testing for the other minor BrS genes only in special circumstances. This recommendation concurs with the position paper of the Canadian Cardiovascular Society but not with consensus of the Heart Rhythm Society/European Heart Rhythm Association which states that either comprehensive genetic testing or target testing for SCN5A can be used (Ackerman et al. 2011; Gollob et al. 2011; Kaufman 2012).

5 Pathophysiology

Using arterially perfused wedge preparation of canine right ventricle (RV), Antzelevitch and his colleagues proposed the repolarization theory as the electrophysiologic abnormality underlying BrS (Yan and Antzelevitch 1999). In their experimental studies, they observed the transmembrane voltage gradient between the RV epicardium and endocardium due to the loss of the action potential (AP) dome only in the epicardium but not in the endocardium; RV epicardium is well known to have abundant Ito. Upon exposure to sodium channel blockers in combination with acetylcholine, this area then developed a notch and dome appearance of the epicardium AP leading to a coved-type ST-segment elevation in the right precordial leads. When the loss of the AP dome was further accentuated, it caused marked shortening of the epicardial AP in certain regions causing pronounced heterogeneity of transmembrane voltage potentials and, in turn, causing phase 2 reentry and triggered VF (Yan and Antzelevitch 1999). However, thus far, there have not been clear clinical relevant data in humans to support this theory. Perhaps, the observational study showing that quinidine, a strong Ito blocker, is effective in treating BrS patients could be inferred as weak-indirect evidence that supports the repolarization theory (Belhassen et al. 2004). While the repolarization theory enjoyed its popularity early on, the lack of strong clinical relevant findings to convincingly support the concept and subsequent clinical evidence led to the other theory, that of depolarization disorder.

Using an electrical guidewire to record an epicardial electrogram from a conus branch of the right coronary artery, Negase et al. were the first to show abnormal electrograms characterized by late potentials following the QRS which were recorded from the free wall of RVOT epicardium in BrS patients (Nagase et al. 2002). Their findings suggest conduction delay in the RVOT epicardium. Two studies conducted in an explanted heart in addition to biopsies showing (ultra-) structural changes in the right ventricular outflow tract of BrS patients demonstrated conduction disorder in these patients. The explanted hearts showed no evidence of repolarization abnormality; instead they found evidence of interstitial fibrosis causing conduction delay in one heart (Coronel et al. 2005) and right ventricular excitation failure and activation delay by current-to-load mismatch in the sub-epicardium in the other heart (Hoogendijk et al. 2010). As a result, the delay in the AP of the RVOT causes the electrical gradient from the more positive RV to RVOT, leading to the ST elevation of the right precordial leads—similar to the situation of a myocardial injury at the RVOT—and as the RVOT depolarizes later (during repolarization of the RV), this gradient is reversed, and the net current flows toward the RV, resulting in a negative T wave in the same right precordial leads. The experiment from the same group in this explanted heart also showed that this site is the arrhythmogenic site during programmed stimulation-induced VF.

Perhaps, the most compelling findings to support depolarization disorder came from our study (Nademanee et al. 2011). Our group carried out a study to determine the substrate sites and arrhythmogenic mechanisms in this disease. We found that all our BrS patients had abnormal low-voltage, fractionated late potentials exclusively clustering in the anterior aspect of the RVOT epicardium and not seen anywhere else. Figures 8.3 and 8.4 show an example of low-voltage fractionated electrograms recorded from the anterior RVOT epicardium of a patient who presented with electrical storm. Ablation at this area normalized the Brugada ECG pattern and prevented recurrent VF episodes. As shown in Fig. 8.3, the endocardial site (arrow) displays a single potential of 2.09 mV, with a duration of 58 ms, and did not extend beyond the QRS compared to the epicardial counterpart that showed low-voltage late potential (0.48 mV), with a duration of 236 ms with late potential extended beyond the QRS. Figure 8.4 shows epicardial electrograms recorded from various sites of the epicardium in both the LV and RV epicardium. Note that abnormal fractionated electrograms and double potential electrograms are only localized in the anterior aspect of the RVOT epicardium. Similar observations were found in all our study patients, and these findings clearly provide the strongest clinical evidence that the delayed depolarization at the anterior aspect of the RVOT is the most likely underlying electrophysiologic mechanism underlying BrS.

Fig. 8.3
figure 3

A left lateral view of the right ventricular outflow tract (RVOT) displays the difference in ventricular electrograms between the endocardial and epicardial site of the anterior RVOT of a BrS patient with electrical storm. The left and right insets display bipolar and unipolar electrograms recorded from the epicardium and endocardium from the same site of the RVOT, respectively. Bi-DIST bipolar distal, Bi-PROX bipolar proximal, Uni-DIST unipolar distal, Uni-PROX unipolar proximal. Reproduced with permission from Nademanee K, Veerakul G, Chandanamattha P, et al. Prevention of ventricular fibrillation episodes in Brugada syndrome by catheter ablation over the anterior right ventricular outflow tract epicardium. Circulation. 2011;123:1270–9

Fig. 8.4
figure 4

Comparison of ventricular electrograms recorded from different sites in both the left ventricle (LV) and right ventricle (RV) of the same patient as in Fig. 8.3. Reproduced with permission from Nademanee K, Veerakul G, Chandanamattha P, et al. Prevention of ventricular fibrillation episodes in Brugada syndrome by catheter ablation over the anterior right ventricular outflow tract epicardium. Circulation. 2011;123:1270–9

Recently, Szél and Antzelevitch also observed fractionated electrograms in the RV epicardium, but they suggested that this is due to a heterogeneous epicardial loss of dome and local re-excitation via a concealed phase 2 reentry, rejecting the possibility of abnormal depolarization or structural abnormalities (Szél and Antzelevitch 2014). However, in our patients another mechanism seems in operation. Fractionated signals were clearly associated with diastolic potentials in the BrS patients; we observed that the fractionation was present immediately after activation, arguing against reactivation of calcium channels (which would require a minimum time delay). Fractionated activity and diastolic potentials have been shown to be associated with reentrant arrhythmias and are observed in BrS patients. Nevertheless, while it is quite apparent that depolarization disorder is likely to be the main mechanism underlying the BrS, one must be mindful that repolarization abnormality could contribute to the arrhythmogenesis of BrS patients, along with genetic mutations of ionic channel and other precipitating factors.

Brugada syndrome was classically thought of as primary electrical disease with no structural defect of the heart (Antzelevitch et al. 2005). Despite no structural abnormality detected by the usual cardiologic studies such as echocardiography, angiography, or cardiac MRI, we recently demonstrated minor yet similar pathological findings in six autopsy hearts of SADS victims who had a family history of BrS and six biopsy specimens in patients with BrS (Nademanee et al. 2015). Increased collagen in the RVOT with epicardial and interstitial fibrosis was found with corresponding area of low-amplitude, fractionated electrogram in biopsy cases. Connexin-43 signals were diminished in the RVOT of these patients which raises the possibility of cardiomyocyte electrical uncoupling as one of the pathophysiology in this disease. These findings point to the likelihood of combined structural abnormalities and ion channel defects as the basis of ventricular arrhythmia and sudden death in patients with BrS. A recent MRI study of cases with BrS compared to age- and sex-matched control demonstrated a slight but statistically significant larger RV end-systolic volume (31 vs. 28 mL/m2, p = 0.038) and lower ejection fraction (61% vs. 64%, p = 0.004); these findings attest to these minor structural abnormalities which likely make BrS a cardiomyopathic entity as well as channelopathy (Bastiaenen et al. 2017).

6 Modulating and Precipitating Factors

As mentioned above, the Brugada ECG pattern is often concealed but can be unmasked or modulated by sodium channel blockers, a febrile state, vagotonic agents, autonomic nervous system changes, tricyclic or tetracyclic antidepressants, first-generation antihistamines (dimenhydrinate), a combination of glucose and insulin, hyperkalemia, hypokalemia, hypercalcemia, and alcohol and cocaine toxicity.

6.1 Autonomic Nervous System

The effect of sympathetic stimulation by isoproterenol infusion, resulting in normalization of the BrS pattern, suggests that sympathetic activity could modify the VF substrate (Miyazaki et al. 1996). The presence of the Brugada ECG pattern is probably a prerequisite for the increased risk of SCD, and normalization of the ECG pattern is associated with a decreased risk (Antzelevitch et al. 2005). This concept is strengthened by the fact that some patients with “VF storms” associated with BrS can be effectively treated with isoproterenol infusion (Tanaka et al. 2001). On the other hand, increased vagal tone could be arrhythmogenic in BrS patients. Increased vagal tone, as well as acute β-blockade, was found to promote VF induction in the electrophysiology laboratory (Kasanuki et al. 1997). Abe et al. found that fluctuations in late potentials on signal-averaged ECG (SAECG) occurred predominantly at night, suggesting that conduction delay and, by inference, the arrhythmogenic substrate are autonomically modulated (Abe et al. 2012). Therefore, it is plausible that at night during sleep, when vagal tone is usually increased and associated with the withdrawal of sympathetic activity, the VF substrate is modulated and more susceptible to arrhythmogenesis. Kasanuki et al. also showed a sudden increase in vagal activity, as measured by heart rate variability (HRV), just before episodes of VF in a patient with BS (Kasanuki et al. 1997). However, Krittayaphong et al. studied HRV from 24-h Holter data of SUDS patients with the Brugada ECG marker, aiming to determine the circadian pattern of sympathetic and parasympathetic activity (Krittayaphong et al. 2003). Surprisingly, they found decreased HRV at night in SUDS patients when compared with the control group and suggested that these patients had an abnormal increase in sympathetic activity or decrease in the vagal tone at night. Although the explanation for the different findings of Kasanuki et al. and that of Krittayaphong et al. is unknown, it is clear that the sympathovagal balance in the BrS patients plays a significant role in the circadian variation of VF occurrence. However, further studies are needed to clearly define the complex interplay between the autonomic nervous system and the arrhythmic mechanisms of BrS.

6.2 Hypokalemia

Hypokalemia has been implicated as a contributing cause for the prevalence of SUDS in the northeastern region of Thailand where potassium deficiency is endemic (Nimmannit et al. 1991). Serum potassium levels in the northeastern population are significantly lower than those of the population in Bangkok, which lies in the central part of Thailand where potassium is abundant in food.

Hypokalemia is a well-known predisposing factor to ventricular arrhythmias. Furthermore, it has been shown that there is commonly a shift of serum potassium into the muscular compartment between midnight and 7 am, decreasing the amount of serum potassium (Andres et al. 1957). If this phenomenon indeed occurs in BS/SUDS patients in Thailand, then it is likely that low serum potassium is a key factor that precipitates VF at night in these patients.

6.3 Sleep, Heavy Meals, and Alcohol

Because the majority of VF episodes occur at night, the question is whether a sleep disorder is a trigger of VF. Thus far, none of our sleep studies in BrS patients has found any evidence of a sleep disorder, including sleep apnea. One theory that many SUDS researchers have informally discussed as a possible precipitating factor is eating a heavy meal at dinnertime before retiring to bed. A Thai Ministry of Public Health Report (1990) suggested that a large meal of glutinous rice (“sticky rice”) or carbohydrates ingested on the night of death precipitated SUDS attacks. Both carbohydrates and glutinous rice have been shown to shift potassium into cells and thus lower the serum potassium level. Postprandial increased ST segment elevation in lead V2 has been seen in patients with symptomatic BrS on a Holter study (Mizumaki et al. 2007). A study by Nogami et al. showed that glucose and insulin could unmask the Brugada ECG marker or accentuate the J-junction elevation of the ST segment (Nogami et al. 2003). They observed a slight decrease in the serum potassium levels of their study patients, but it did not reach statistical significance. Nevertheless, these findings bode well for a heavy carbohydrate meal being a precipitating factor for sudden death in SUDS patients. Alcohol has also been implicated in triggering recurrent VF in a patient with Brugada syndrome (Ohkubo et al. 2013).

6.4 Body Temperature and Febrile Illness

Dumaine et al. discovered that the T1620M missense mutation causes accelerated inactivation of the sodium channel at physiologic body temperature but not at room temperature (Dumaine et al. 1999). Identification of this temperature-sensitive mutation that precipitates the net loss of sodium current prompted investigators to recognize that a hot climate and body temperature may be important modulating factors. Indeed, several case reports have emerged recently, demonstrating that febrile illness or external heat could unmask BrS and/or precipitate VF occurrence (Antzelevitch and Brugada 2002; Morita et al. 2002b; Saura et al. 2002; Porres et al. 2002; Kum et al. 2002; Canpolat et al. 2017; Chung et al. 2017; De Marco et al. 2012; Skinner et al. 2007). In children, ventricular arrhythmia triggered by fever can be mistaken for episodes of febrile seizure (Skinner et al. 2007). We have encountered a case of a young male patient who died suddenly after a spiking fever of 40 °C after abdominal surgery. Upon review of the ECG, the patient had the typical BrS pattern, but had not had a prior medical problem and had been asymptomatic. The northeastern part of Thailand where SUDS is prevalent is well known for its hot climate, with temperatures reaching as high as 41 °C. It is again unclear how much climate influences the occurrence of SUDS in Thailand, but a study is underway. It is entirely possible that high climatic temperatures or a febrile state could modulate the functional expression of mutant channels in other genes responsible for BrS. In the meantime, physicians should factor in temperature as a cause for arrhythmogenesis in BrS. They should be cognizant of the association between temperature and BrS during diagnosis and treatment, advising patients to promptly treat fevers.

Figure 8.5 proposes pathophysiology of BrS and demonstrates how these modulating and predisposing factors can affect the arrhythmia and the clinical outcomes in many ways: (1) modifying the VF substrates, (2) affecting the gene expression of the ion channel defects, (3) affecting the triggering PVCs and the initiating process of VF, and (4) influencing the sustaining process of the VF episodes.

Fig. 8.5
figure 5

Proposed pathophysiologic mechanisms of BrS with respect to predisposing factors

7 Combined Syndromes

In addition to the above precipitating factors, BrS patients often have other concomitant arrhythmias or arrhythmic syndromes. Atrial fibrillation is one of the common arrhythmias in BrS. Incidence of combined early repolarization syndrome in the BrS patients occurred in 15%, and in this subset (Letsas et al. 2008), the incidence of recurrent VF episodes is significantly higher than that of the BrS alone. Similarly, combined syndromes of a progressive conduction defect and BrS and long QT syndrome and BrS are occasionally observed (Bezzina et al. 1999; Shirai et al. 2002).

8 Risk Stratification

There is little to debate that BrS patients who survived an out-of-hospital cardiac arrest are at high risk of recurrent VF episodes and need ICD treatment. Likewise, symptomatic patients with recurrent syncope, agonal respiration at night during sleep, or unknown seizure are at risk of dying suddenly without protection and have Class I indication for ICD treatment. One large study involving 1029 BrS patients in 11 European centers found annual cardiac events rate of 7.7% in patients who presented with sudden cardiac arrest, 1.9% in patients with syncope, and 0.5% in asymptomatic patients (Probst et al. 2010). The heated debate is more on how one best identifies high-risk patients for sudden death and the need of ICD treatment in asymptomatic BrS patients. At the beginning, the Brugada registry reported a significantly high risk of asymptomatic patients with positive VT inducibility by programmed electrical stimulation (Antzelevitch et al. 2005; Brugada et al. 2011). However, more recent studies found a much lower incidence of sudden death or VF in this group and questioned the specificity of programmed electrical stimulation (PES) in risk stratifying asymptomatic BrS patients (Veerakul and Nademanee 2012; Priori et al. 2012). Our own experience of asymptomatic patients shows that the annual cardiac event rate (VF or death) is so low (0.25% per year) that it will be very unlikely for any risk stratification strategy to be able to identify high-risk patients for ICD treatment as a primary prevention (Veerakul et al. 2008).

Electrocardiographic risk factors include presence of spontaneous type I Brugada pattern which carries a higher risk of arrhythmic events compared to drug-induced Brugada pattern (Adler 2016). QRS fragmentation (Priori et al. 2012; Morita et al. 2008; Tokioka et al. 2014), early repolarization pattern (Tokioka et al. 2014; Kawata et al. 2013; Kaneko et al. 2014), a significant S wave in lead I (≥0.1 mV and/or ≥40 ms) (Calò et al. 2016), and prolonged Tpeak-Tend or Tpeak-Tend/QTc (Zumhagen et al. 2016; Maury et al. 2015) were reported to be associated with high risk for ventricular arrhythmia. Other studies found exercise testing (Makimoto et al. 2010; Amin et al. 2009), signal-averaged ECG (Huang et al. 2009), and shortening of the ventricular refractory period (<200 ms) (Priori et al. 2012) to be valuable tools for identifying high-risk patients. However, it is unclear how useful any of these parameters would be in identifying asymptomatic BrS for ICD treatment. In our own asymptomatic cohorts of 115 patients, we found that only two patients after 10 years of follow-up had either VF or sudden death (one EPS positive and the other EPS negative) (Veerakul et al. 2008). With this low event rate and after a decade of follow-up, it is very clear to us that any risk stratification strategy would be very unlikely to be of any value in selecting patients for an ICD treatment in our population.

9 Treatment

BrS patients should be informed of various modulating and precipitating factors—as discussed above—that could bring about malignant arrhythmias: fever, electrolyte abnormalities, alcohol consumption, and a whole host of drugs as listed in www.brugadadrugs.org (Postema et al. 2009). VF and sudden death in BrS usually occur at rest and at night. Therefore, one has to be cognizant of the circadian variation of sympathovagal balance, hormones, and other metabolic factors which are likely to contribute to this circadian pattern (Wilde et al. 2002; Antzelevitch et al. 2005).

9.1 Anti-arrhythmic Drugs

Quinidine has been the only drug that consistently shows benefits in preventing recurrent VF episodes. The drug has been shown to suppress inducibility of VF on PES and reduce the number of appropriate ICD shocks (Belhassen et al. 2004, 2015; Hermida et al. 2004; Mizusawa et al. 2006). Blockade of Ito is thought to be the mechanism by which quinidine is effective. Unfortunately, there are two major problems with quinidine: (1) Only two-third of patients could tolerate the drug, and serious side effects such as thrombocytopenia could be very serious. (2) Quinidine is not available in many countries (Viskin et al. 2007). In Thailand, there is no supply of the drug, and the only drug we could use is amiodarone with variable success. Bepridil—another Ito blocker—which is only available in Japan has been used in BrS patients to suppress VF (Kaneko et al. 2014; Ohgo et al. 2007; Murakami et al. 2010). Cilostazol—an oral phosphodiesterase inhibitor—has also been shown to be of benefit in preventing recurrent VF in Brugada syndrome (Ohgo et al. 2007; Tsuchiya et al. 2002).

However, there is very little data to objectively assess the safety and efficacy of both Bepridil and Cilostazol in BrS patients at this time.

9.2 Implantable Cardioverter-Defibrillator

Symptomatic BrS patients (with a past history of VT/VF or syncope) have a Class I indication for ICD treatment (Wilde et al. 2002; Antzelevitch et al. 2005, 2016; Priori et al. 2013). In our DEBUT study (Tsuchiya et al. 2002)—Defibrillator versus β-blocker in Unexplained Death in Thailand: A Randomized Clinical Trial—we found that ICDs provided full protection from death related to primary VF in the study population which included 59% of patients with BrS. However, we also found that unwanted effects of the ICD were also frequent (30%). Most of the complications were minor; they included defibrillation discharges caused by supraventricular tachycardia or sinus tachycardia and T-wave over-sensing. All of the complications were corrected by reprogramming the devices without major intervention. However, one patient had pocket erosion with infection that required removal of the ICD, and one patient needed to have his ICD lead replaced because of an insulation break. Other studies also show similar results that long-term follow-up of Brugada ICD patients has a high complication rate up to a third of patients (Nademanee et al. 2003; Sacher et al. 2006; Sarkozy et al. 2007; Steven et al. 2011). The majority of the complications, similar to our study, were mostly inappropriate shocks occurring in 36% of patients at follow-up; however, one registry recorded an 18% rate of serious vents including pericardial effusion, lead fracture, infection, and subclavian vein thrombosis (Steven et al. 2011).

With the above relatively high complication rate of ICD in BrS patients, one has to be extremely cautious to use ICD in asymptomatic BrS patients. The fact that the event rates in the asymptomatic BrS population are quite low in most series makes ICD treatment in this subset questionable with respect to whether ICD benefits would outweigh the risk. As mentioned in the risk stratification section, thus far there have not been any convincing methods of risk stratification to identify high-risk asymptomatic patients for ICD therapy. One could follow the guideline established by the HRS/EHRA/APHRS expert consensus statement (Priori et al. 2013) recommendations for ICD implantation, as summarized in Table 8.3.

Table 8.3 Indication for ICD implantation in patients with Brugada syndrome (Priori et al. 2013)

The recent approval of leadless ICD, which has been shown to be quite effective in terminating VT/VF episode, is a welcome addition to therapeutic modalities for BrS patients 116. Early users reported issues with sensing and higher sensing screening failure rate compared to other inherited arrhythmia syndrome due to the dynamic nature of QRS and T-wave morphologies in patients with BrS (Kamakura et al. 2017; Conte et al. 2017; Olde Nordkamp et al. 2016). Further studies and clinical trials to determine efficacy and safety are warranted in the high-risk BrS population.

9.3 Catheter Ablation

The early attempt of catheter ablation in treating BrS syndrome patients was limited to a few report cases of patients with electrical storms. The initial approach was designed to target initiating PVCs that trigger VF, which were found to come from the RVOT (Haïssaguerre et al. 2003; Nakagawa et al. 2008; Darmon et al. 2004). The ablation was performed on the endocardial site of the RVOT. However, this approach has not been widely successful largely because patients with BrS rarely had frequent PVCs to be mapped, and therefore it was quite difficult to identify precise targets for ablation and clearly assess the acute outcomes of the ablation. We have reported our epicardial approach for substrate ablations that was indeed safe and effective (Nademanee et al. 2011). We identified and proved that anterior RVOT epicardium is the most common arrhythmogenic substrate sites for our BrS patients. However, subsequently, in a significant number of patients, the RV body and the inferolateral aspect and the area near the tricuspid valve are also infrequently involved. These sites consistently have abnormal late potentials and low-voltage fractionate ventricular electrograms; these abnormal electrograms tended to cluster exclusively in this area but not anywhere else. After ablations at this RVOT epicardial site, the Brugada ECG pattern normalizes, and VT/VF episode subsides. We have now performed 54 BrS patients with frequent ICD discharges. Long-term outcomes (median 30 months) have been excellent with no recurrent VT/VF in all patients off medication. More recent studies ranging from individual case reports as well as collective collaborative studies have confirmed our findings that BrS arrhythmogenic substrates are ubiquitous in the RVOT epicardium and catheter ablations are beneficial in treating symptomatic BrS (Brugada et al. 2015; Zhang et al. 2016). Whether ablation would substitute for an ICD in high-risk BrS patients remains unknown. Further studies clearly need to be done to assess values and limitations of catheter ablation in patients with BrS.

10 Conclusions and Future Perspective

The two decades of BrS research have witnessed an impressive progress of our understanding of several aspects of the syndrome with respect to role of genetics, electrophysiological mechanisms, and clinical characteristics. However, many questions and controversies remain regarding the role of genetic background including polymorphisms as well as rare variants and gene-gene or gene-environmental interaction and other confounding factors such as fever and gender in this population. The cause and the role of fibrosis and anatomical abnormality in the RVOT in arrhythmogenesis will likely be the subject of intense investigation in the coming years. It’s certain that the debate will continue regarding the role of repolarization in BrS and whether our findings of abnormal delayed depolarization in the anterior RVOT epicardium is seen in other centers and more specifically other population besides ours. How does one best treat asymptomatic BrS patients? Will there be any better risk stratification strategy to identify high-risk groups? Why is the anterior RVOT epicardium the arrhythmogenic sites for these patients? Why is there such a male preponderance and why do most of the VF episodes usually occur at night time? Whole genome sequencing may identify rare and common variants in genes modulating ion channels that may combine as multiple hits to cause abnormal conduction in the RVOT areas and in turn increase susceptibility to VF. Further research will continue to answer these questions. Meanwhile, refinement of treatment is needed. We will need to know the efficacy and safety of quinidine in the randomized trial studies which currently are being conducted. In the near future, more study of the subcutaneous leadless ICD will be forthcoming. The expanding role of catheter ablation of the epicardial substrate beyond the population of BrS with frequent VF needs to be also evaluated by assessing the risks and benefits of the procedures, especially with respect to complications related to the epicardial ablation approach. We indeed anticipate with excitement that the next decades will have these answers and in turn advance our ability to care for our BrS patients.