J Interv Card Electrophysiol DOI 10.1007/s10840-013-9870-y
REVIEWS
Early repolarization pattern: its ECG characteristics, arrhythmogeneity and heritability Yuka Mizusawa & Connie R. Bezzina
Received: 9 November 2013 / Accepted: 30 December 2013 # Springer Science+Business Media New York 2014
Abstract Early repolarization (ER) has been accepted as a benign ECG variant for decades. Two seminal studies challenged this notion and have demonstrated that ER pattern is associated with an increased risk of arrhythmic and cardiac mortality in patients with idiopathic ventricular fibrillation (IVF) and in the general population. Recent clinical studies demonstrate its varying impact as an arrhythmogenic substrate on different diseases. For example, in ER syndrome, a primary electrical disease, ER appears as a major arrhythmogenic substrate for development of VF whereas in patients with coronary artery disease, an ER pattern may exist as a silent substrate, increasing the risk of VF during episodes of cardiac ischaemia. Due to the high prevalence of an ER pattern in the general population and a low VF event rate, it remains challenging to differentiate a malignant ER pattern from a benign form. Recent research suggests that a J-wave amplitude of more than 0.1 mV combined with a descending/horizontal ST segment may constitute a malignant ER pattern. Further studies are however necessary to evaluate its prognostic value for cardiac and arrhythmic death in the general population as well as in cases with a malignant ER pattern. While genetic testing has revealed putative causal DNA variants in sporadic cases, the lack of co-segregation with the disease in affected families suggests that ER syndrome is not monogenic but is likely a complex disorder influenced by multiple genetic as well as environmental factors. Y. Mizusawa : C. R. Bezzina AMC Heart Centre, Department of Clinical and Experimental Cardiology, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands C. R. Bezzina (*) Department of Experimental Cardiology, Academic Medical Centre, Room L2-108-1, Meibergdreef 9, 1105AZ Amsterdam, The Netherlands e-mail: [email protected]
Keywords Early repolarization . J-wave . Ventricular fibrillation . Genetics . Cellular mechanism . Therapy
1 Introduction Early repolarization (ER), also known as J-wave, is a common ECG pattern which was for a long time considered to be a benign ECG variation. Recently, case–control studies as well as epidemiological studies conducted in the general population have provided evidence that this ECG pattern is associated with a higher chance of arrhythmic and cardiac death. An ER pattern has been described in patients with different cardiac phenotypes, including patients with idiopathic ventricular fibrillation (VF) and patients with VF in the setting of an acute myocardial infarction. Differentiation of malignant ER patterns, namely ER syndrome, from benign ones has been challenging because while an ER pattern is observed commonly in the general population, most of the individuals with an ER pattern remain asymptomatic for life and fatal arrhythmic events are rarely observed. Early studies focused on the morphology of the J-wave and its localization on the 12-lead ECG to evaluate the prognostic role of an ER pattern, but more recent studies suggest that ST-segment morphology in addition to the J-wave characteristics may help diagnose subjects with a malignant ER pattern. As to the heritability of an ER pattern, some gene variants have been reported in sporadic cases with ER syndrome. However, co-segregation of such variants with an ER pattern in families is lacking. In this review, we summarize the most recent knowledge on the ER pattern including its definition, prevalence, the clinical characteristics as well as its genetic background and the proposed cellular mechanism. To note, the definitions of an ER pattern have been inconsistent in different studies. Furthermore, some reports grouped ER syndrome together with Brugada syndrome under the general term ‘J-wave syndrome’ [1]. In
J Interv Card Electrophysiol
this regard, although the phenotypes of ER syndrome and Brugada syndrome overlap, conflicts remain in the current knowledge in order to group or to differentiate ER and Brugada syndromes from a mechanistic point of view [2, 3]. Our focus in this review is on ER pattern/ER syndrome as a separate entity from Brugada syndrome, which is in accordance with the classification of the primary electrical disorders in the consensus statement by Heart Rhythm Society/European Heart Rhythm Association/Asia Pacific Heart Rhythm Society (HRS/EHRA/APHRS) [4].
2 Historical perspective ER, also known as J-wave, is a common ECG pattern characterized by a QRS-ST junction elevation. After the first report by Shipley and Hallaran in 1936 [5], the ER pattern was deemed a benign ECG variant with the exception of the ‘Osborn wave’, a broad J-wave associated with hypothermia and VF [6, 7]. While case reports of idiopathic VF (IVF) with an ER pattern appeared occasionally in the literature, implying its potential arrhythmogeneity and heritable character [8–12], the case–control study by Haïssaguerre and co-workers in 2008 was the first seminal work to suggest the malignant role of ER pattern in IVF patients [13]. In their study population of 206 IVF cases and 412 controls, ER pattern was more frequently present in patients with IVF compared with the control group (31 % vs 5 %, respectively). The odds ratio (OR) for the presence of ER in the IVF cases compared with the controls was 10.9 (95 % confidence interval (CI) 6.3 to 18.9) after adjustment for age, sex, race and level of physical activity. Other case–control studies reported a similarly higher ratio of an ER pattern in IVF patients (42–60 %) compared with the controls (3.3–13 %, P0.2 mV in the inferior leads was associated with an even more increased risk of death from cardiac causes (adjusted RR 2.98, 95 % CI 1.85 to 4.92) and from arrhythmia (adjusted RR 2.92, 95 % CI 1.45 to 5.89) [16]. In subjects with a J-wave of >0.2 mV, the Kaplan–Meier survival curve for death from cardiac causes started to diverge at 15 years follow-up, which suggests an occurrence of events from around 60 years of age [16]. These data imply that >0.2mV J-wave elevation in this age group is not a sole determinant leading to VF but likely plays a role as a modifier to increase a risk of arrhythmic or cardiac death in patients with an underlying disorder such as coronary artery disease [16, 25]. As a supporting evidence for this hypothesis, the same group demonstrated in a subsequent study that in acute coronary disease, VF victims had a higher prevalence of an ER pattern compared with the controls without VF [30]. In particular, an ER pattern with a horizontal ST segment was an independent predictor of sudden death (OR 2.15, 95 % CI 1.20–3.85) [30]. An increased VF risk in patients with an ER pattern has been shown in other studies on subjects with a first myocardial infarction [31, 32] as well as vasospastic angina cases [33].
Sinner et al. reported similar results to the study by Tikkanen et al. [27]. They performed a population-based prospective study on 1,945 patients from the MONICA/ KORA cohort which demonstrated the prevalence of an ER pattern of 13.1 %. ER pattern was associated with cardiac and all-cause mortality, mostly pronounced in those of younger age (35–54 years old) and male sex; hazard ratio (HR) was 2.65 (95 % CI 1.21–5.83) for men between 35 and 54 years old. An inferior ER pattern further increased cardiac mortality (HR of 3.15, 95 % CI 1.58–6.28) for both sexes and particularly in male subjects between 35 and 54 years of age (HR of 4.27, 95 % CI 1.90–9.61). HRs for all-cause mortality were weaker but reached statistical significance. Another study on the incidence of ER appeared from Japan and reported different results compared with the two studies mentioned above. Haruta et al. evaluated ECGs of 5,976 atomic bomb survivors in the Nagasaki area who underwent a biennial health check-up [26]. They reported that the incidence rate of an ER pattern was 715 per 100,000 person-years. Their study cohort showed an association with an increased risk of unexpected death (HR 1.83, 95 % CI 1.12–2.97) but decreased risk of cardiac death (HR 0.75, 95 % CI 0.60– 0.93) and all-cause mortality (HR 0.85, 95 % CI 0.78–0.93). Both slurring and notching patterns of J-wave were related to a higher risk of unexpected death (HR 2.09, 95 % CI 1.29– 4.83) and in both inferior and lateral leads (HR 2.50, 95 % CI 1.29–4.83). The reason for the inconsistent results from the different studies is not clear, but age, ethnicity, and diagnostic methods of sudden death cases (autopsy or death certificate) may play a role in it. A meta-analysis of prospective cohort studies and case–control studies demonstrated that the risk ratios of an ER pattern were 1.70 (95 % CI 1.19–2.42, P=0.003) for arrhythmic death, 0.78 (95 % CI 0.27–2.25, P=0.63) for cardiac death and 1.06 (95 % CI 0.87–1.28, P=0.57) for all-cause mortality [34]. A J-wave amplitude of ≥0.1 and ≥0.2 mV in inferior leads had a higher risk for arrhythmic death (RR 1.58 and 3.02, respectively) and death from cardiac causes (RR 1.48 and 2.98, respectively). Notching J-wave patterns had an increased risk for arrhythmic death (RR 1.48). Importantly, the estimated absolute risk differences of subjects with an ER pattern in the meta-analysis were low (70 cases of arrhythmia death per 100,000 subjects per year) [34]. While a metaanalysis is limited by its heterogeneous population, the reported results seem to reflect the fact that a J-wave elevation is indeed benign in general. Of note, the morphology of an ST segment (horizontal/descending or ascending pattern) was not included in this meta-analysis due to the lack of data. In addition to the localization and morphology of the Jwave, more recent studies have investigated the morphology of ST segments following the J-point to scrutinize an ER pattern carrying a risk of arrhythmic/cardiac death. In the general population, Tikkanen et al. analysed the prognostic
J Interv Card Electrophysiol
significance of horizontal/descending ST segments and successfully showed that a J-wave of ≥0.1 mV with a horizontal/ descending ST segment had an increased HR of arrhythmic death (RR 1.43, 95 % CI 1.05–1.94) while ascending an ST segment did not show an increased RR (0.89), suggesting a benign ER variant [17]. Strikingly, subjects with a higher Jpoint amplitude (>0.2 mV) in the inferior leads with a horizontal/descending ST segment showed an even higher HR of arrhythmic death of 2.14 (95 % CI 1.56–6.30). Rosso et al. investigated IVF patients and age- and gender-matched controls and showed that a J-wave with a horizontal STsegment morphology was strongly associated with a history of VF (OR 13.8, 95 % CI 5.1–37.2) [18]. In spite of the amount of knowledge acquired so far, the utility of the ER pattern as a prognostic indicator of VF risk in asymptomatic individuals remains unclear. For example, in asymptomatic individuals (aged 35–45 years) with a J-wave and horizontal ST segment, the risk of cardiac arrest due to ER syndrome is estimated by Rosso et al. to be as low as 0.03 % [18]. Wu et al. similarly calculated a low absolute incidence rate of arrhythmic death in subjects with an ER pattern (70 cases per 100,000 subjects per year) [34]. Besides, various definitions of an ER pattern in the different studies make it difficult to conclude which ER pattern actually carries a risk for arrhythmic and cardiac mortality. The results of the more recent studies suggest that a J-wave alone may not be sufficient to differentiate a malignant ER pattern, but the definition of an ‘ER pattern’ ought to include both J-wave (amplitude, morphology and localization) and ST-segment morphology to find individuals carrying a risk ER pattern. At present, for suspected IVF cases, a careful assessment of patient history coupled with a watchful follow-up remains a principle to diagnose patients at risk [35, 36]. The ascending pattern of an ST segment may help exclude subjects with an innocuous ER pattern. Further research is certainly necessary to solve such a dilemma in clinical practice.
5 ER syndrome ‘ER syndrome’ refers to the patients with IVF having an ER pattern. ER syndrome is predominantly observed in male subjects and is more likely to develop sudden cardiac death during sleep [13]. The reported mean age of the first VF episode is 35–43 years old [13–15, 22, 37]. Before ER syndrome is definitely diagnosed, differential diagnosis is necessary to exclude other primary electrical disorders (long/short QT syndrome, Brugada syndrome and catecholaminergic polymorphic ventricular tachycardia) and structural heart diseases. At this time, there is no established method to differentiate ER syndrome from the benign ER pattern. The electrocardiographic and electrophysiological characteristics mentioned below are described in observational studies with a
small number of cases and may help diagnose ER syndrome in clinical practice. 5.1 ECG: modulation of a J-wave in ER syndrome Modulation of the J-wave amplitude is affected by heart rate, heart rhythm, autonomic tone and drugs. J-wave amplitude increases during slow heart rates as well as a short-long-short sequence [23] and a pause [37]. Circadian variation of the Jwave amplitude is also known to occur, and its augmentation is in concordance with a vagal tone [22]. Accordingly, sympathetic stimulation by isoproterenol decreases the J-wave amplitude [37, 38]. Sodium channel blockers also attenuate the J-wave amplitude [3, 38]. J-wave in ER syndrome increases prominently just before arrhythmic events [12, 13, 23], an occurrence which is now recognized as a hallmark of the disease. By performing serial ECG recordings during an electrical storm in 18 subjects with ER syndrome, Haïssaguerre et al. nicely showed that the amplitude of the J-wave increased to a mean of 4.1 mm during an arrhythmic period compared with 2.6 mm at baseline [13]. As another ECG feature of ER syndrome, Nam et al. demonstrated a global appearance of a J-wave on 12-lead ECG within 30 min of VF storms in four out of five cases [23]. Such J-waves may completely disappear within weeks after VF events [37]. Their dynamic manifestation can therefore hinder the diagnosis of malignant cases with ER syndrome. In contrast to these dynamic J-wave changes in ER syndrome, the ER pattern in healthy individuals does not vary over time [16, 39, 40]. 5.2 Non-invasive electrophysiological study Signal-averaged ECG, T wave alternans (TWA), QT dispersion (QTD) and heart rate variability (HRV) are frequently used tests in clinical practice to evaluate arrhythmic risk and the electrophysiological mechanism of primary electrical disease as well as structural heart disease. Abe et al. used these non-invasive methods to investigate the pathophysiology of an ER pattern and its circadian changes in IVF cases with/without a J-wave and age- and gender-matched controls [22]. The results showed that repolarization parameters (TWA and QTD) were not different between IVF cases with and without J-wave whereas abnormal depolarization parameters (late potentials (LPs)) were more frequently observed in IVF with a J-wave (ER syndrome) compared with those without a J-wave or controls. Interestingly, in ER syndrome, the time at which LPs showed abnormal values were concordant with the time of the VF events (six of seven patients at midnight and one during the day), implying the potential role of LPs in the arrhythmogeneity of ER. However, the result of this study is inconsistent with other clinical as well as experimental studies which suggest the involvement of repolarization abnormality
J Interv Card Electrophysiol
in the pathogenesis of an ER pattern [3, 41]. For the mechanistic comprehension of an ER pattern, further studies are required. 5.3 Invasive electrophysiological study in ER syndrome Clinical electrophysiological study is not helpful in the risk stratification of ER syndrome. Haïssaguerre et al. performed such studies in 132 IVF cases to test the inducibility of VF [13]. Stimulation was performed from two ventricular sites with up to three extrastimuli. VF inducibility was low in patients with a history of VF (34 %), and besides, the VF inducibility was not different in IVF patients with or without an ER pattern. The same investigators also tested the relationship between the lead displaying the ER pattern and the location of VF-initiating premature beats. In patients with inferior lead ER, VFinitiating ectopies originated from the inferior wall of the left ventricle whereas in patients with inferolateral ER, ectopies originated from multiple regions [42, 43]. 5.4 Therapy for ER syndrome In patient management, there is no doubt that cases resuscitated from VF require an implantable cardioverter defibrillator (ICD) [4]. In acute instances of recurrent VF in ER syndrome, an isoproterenol infusion effectively suppresses VF. Here, isoproterenol counterbalances the high vagal tone (in the setting of which arrhythmic events often occur) to terminate VF. Deep sedation is also useful to inhibit an electrical storm [13, 38]. For long-term therapy, quinidine is effective [13, 38]. β-blockers, lidocaine, mexiletine and verapamil were reported to be ineffective in both the acute and chronic phase. Amiodarone was reported to be effective only in the acute phase [13, 38]. Yet, the exact mechanism of the antiarrhythmic drugs to suppress VF in ER syndrome is not fully understood.
6 Cellular mechanism of an ER pattern The cellular mechanism of an ER pattern was proposed by Yan et al. in 1996 on the basis of experiments using the arterially perfused wedge preparation of canine left ventricle [41]. The proposed mechanism rests on the intrinsic transmural differences in action potential morphology across the ventricular myocardium where the more prominent transient outward potassium current (Ito) in the epicardium (compared with endocardium) confers a spike-and-dome action potential morphology, which is absent in the endocardium. It is postulated that the different action potential morphologies across the ventricular wall result in a transmural voltage gradient, generating the J-wave. Yan and Antzelevitch showed that the amplitude of the J-wave may be modulated by hypothermia (increases J-wave), by premature stimulation
(decreases J-wave), or by the application of the Ito blocker 4aminopyridine (decreases J-wave) [41]. Because the Ito current is rich in the epicardium but not in the endocardium, enhancement of Ito deepens the phase 1 notch of the epicardial action potential and shortens the epicardial action potential, thereby accentuating transmural heterogeneity and consequently the J-wave. An increase in other cardiac potassium currents (IKr, IKs, IKAch, and IKATP) or a decrease in the cardiac sodium (INa) or calcium currents (ICaL) can also theoretically lead to a transmural gradient and a manifestation of an ER pattern [1]. This model thus suggests the possible involvement of ion channel genes in ER syndrome, the involvement of which has been tested in a number of genetic studies on individuals with this disorder. There are however some clinical as well as experimental aspects which cannot be fully explained by this hypothesis. First, suppose that the transmural heterogeneity of repolarization caused by Ito is the fundamental mechanism for developing VF in individuals with an ER pattern, it is as yet not explained why the risk of arrhythmic death is so low (70 cases per 1,000,000 subjects per year) while an ER pattern in the general population is common (1–13 %) [16, 26, 27, 34]. Studies in the general population as well as ER syndrome suggest that the ER pattern exists as an arrhythmogenic substrate but needs triggers to develop VF [44]. For example, patients with acute coronary disease and VF showed a higher prevalence of an ER pattern compared with those without VF, implying that an ischemic attack may be a trigger for VF in individuals with an ER pattern [30]. In ER syndrome, a Jwave elevation is observed just before arrhythmic events [13, 15]. These observations suggest that multiple factors are involved in the augmentation of the J-wave and subsequent development of VF in subjects carrying an ER pattern. However, a different impact of the malignant or the benign ER pattern on the prognosis remains unexplained. Second, in experimental models, equilibration of the action potential duration is observed in adjacent cells unless intercellular coupling is decreased [44, 45]. Accordingly, in the intact heart, the duration of the action potential in different layers of the ventricular wall is expected to differ gradually and consequently the action potential cannot be extremely short in one region, e.g. in the epicardium to cause an extrasystole leading to VF. On the other hand, myocardial structural changes that may be even clinically undetectable could allow the development of transmural heterogeneity of repolarization. Again, this implicates that transmural heterogeneity of repolarization alone does not seem to explain the arrhythmogeneity of the ER pattern. Finally, the malignant ER pattern may be caused by a different mechanism, and the ECG pattern only morphologically mimic the benign ER pattern in the general population [44]. Historically, repolarization abnormality was proposed to explain the mechanism of both ER pattern and Brugada syndrome [1]. Indeed, abnormal ECG patterns are observed
J Interv Card Electrophysiol
during the same phase of the action potential (the transition phase of depolarization to repolarization) in both conditions [1]. Notwithstanding, there are phenotypical differences between ER and Brugada syndromes. For example, the abnormal ECG patterns affect different regions of the heart: the inferolateral leads in ER syndrome and the right pre-cordial leads in Brugada syndrome. Further, ajmaline attenuates the Jwave amplitude in subjects with an ER pattern while it causes ST-segment elevation in patients with Brugada syndrome [3]. There is clearly a lack of knowledge to explain all these differences, which requires further research.
7 Heritability of an ER pattern 7.1 Heritability of an ER pattern in IVF cases and in the general population The familial nature of the ER pattern has been suggested by case–control studies as well as by studies in the general population [13, 46–48]. Haïssaguerre et al. were the first to report that a family history of unexplained sudden death was more frequently observed in sudden cardiac arrest cases with an ER pattern compared with those without (16 % vs 9 %) [13]. Similar results were reported by Watanabe et al. who studied 50 IVF cases with an ER pattern, in which 16 % had a family history of sudden death [46]. In another study, firstdegree relatives of unexplained sudden arrhythmic death syndrome (SADS) probands were studied. The authors reported that 23 % of SADS relatives showed a J-wave in the inferolateral leads compared with 11 % of the control subjects (OR 2.54, 95 % CI 1.66 to 3.90, P
REVIEWS
Early repolarization pattern: its ECG characteristics, arrhythmogeneity and heritability Yuka Mizusawa & Connie R. Bezzina
Received: 9 November 2013 / Accepted: 30 December 2013 # Springer Science+Business Media New York 2014
Abstract Early repolarization (ER) has been accepted as a benign ECG variant for decades. Two seminal studies challenged this notion and have demonstrated that ER pattern is associated with an increased risk of arrhythmic and cardiac mortality in patients with idiopathic ventricular fibrillation (IVF) and in the general population. Recent clinical studies demonstrate its varying impact as an arrhythmogenic substrate on different diseases. For example, in ER syndrome, a primary electrical disease, ER appears as a major arrhythmogenic substrate for development of VF whereas in patients with coronary artery disease, an ER pattern may exist as a silent substrate, increasing the risk of VF during episodes of cardiac ischaemia. Due to the high prevalence of an ER pattern in the general population and a low VF event rate, it remains challenging to differentiate a malignant ER pattern from a benign form. Recent research suggests that a J-wave amplitude of more than 0.1 mV combined with a descending/horizontal ST segment may constitute a malignant ER pattern. Further studies are however necessary to evaluate its prognostic value for cardiac and arrhythmic death in the general population as well as in cases with a malignant ER pattern. While genetic testing has revealed putative causal DNA variants in sporadic cases, the lack of co-segregation with the disease in affected families suggests that ER syndrome is not monogenic but is likely a complex disorder influenced by multiple genetic as well as environmental factors. Y. Mizusawa : C. R. Bezzina AMC Heart Centre, Department of Clinical and Experimental Cardiology, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands C. R. Bezzina (*) Department of Experimental Cardiology, Academic Medical Centre, Room L2-108-1, Meibergdreef 9, 1105AZ Amsterdam, The Netherlands e-mail: [email protected]
Keywords Early repolarization . J-wave . Ventricular fibrillation . Genetics . Cellular mechanism . Therapy
1 Introduction Early repolarization (ER), also known as J-wave, is a common ECG pattern which was for a long time considered to be a benign ECG variation. Recently, case–control studies as well as epidemiological studies conducted in the general population have provided evidence that this ECG pattern is associated with a higher chance of arrhythmic and cardiac death. An ER pattern has been described in patients with different cardiac phenotypes, including patients with idiopathic ventricular fibrillation (VF) and patients with VF in the setting of an acute myocardial infarction. Differentiation of malignant ER patterns, namely ER syndrome, from benign ones has been challenging because while an ER pattern is observed commonly in the general population, most of the individuals with an ER pattern remain asymptomatic for life and fatal arrhythmic events are rarely observed. Early studies focused on the morphology of the J-wave and its localization on the 12-lead ECG to evaluate the prognostic role of an ER pattern, but more recent studies suggest that ST-segment morphology in addition to the J-wave characteristics may help diagnose subjects with a malignant ER pattern. As to the heritability of an ER pattern, some gene variants have been reported in sporadic cases with ER syndrome. However, co-segregation of such variants with an ER pattern in families is lacking. In this review, we summarize the most recent knowledge on the ER pattern including its definition, prevalence, the clinical characteristics as well as its genetic background and the proposed cellular mechanism. To note, the definitions of an ER pattern have been inconsistent in different studies. Furthermore, some reports grouped ER syndrome together with Brugada syndrome under the general term ‘J-wave syndrome’ [1]. In
J Interv Card Electrophysiol
this regard, although the phenotypes of ER syndrome and Brugada syndrome overlap, conflicts remain in the current knowledge in order to group or to differentiate ER and Brugada syndromes from a mechanistic point of view [2, 3]. Our focus in this review is on ER pattern/ER syndrome as a separate entity from Brugada syndrome, which is in accordance with the classification of the primary electrical disorders in the consensus statement by Heart Rhythm Society/European Heart Rhythm Association/Asia Pacific Heart Rhythm Society (HRS/EHRA/APHRS) [4].
2 Historical perspective ER, also known as J-wave, is a common ECG pattern characterized by a QRS-ST junction elevation. After the first report by Shipley and Hallaran in 1936 [5], the ER pattern was deemed a benign ECG variant with the exception of the ‘Osborn wave’, a broad J-wave associated with hypothermia and VF [6, 7]. While case reports of idiopathic VF (IVF) with an ER pattern appeared occasionally in the literature, implying its potential arrhythmogeneity and heritable character [8–12], the case–control study by Haïssaguerre and co-workers in 2008 was the first seminal work to suggest the malignant role of ER pattern in IVF patients [13]. In their study population of 206 IVF cases and 412 controls, ER pattern was more frequently present in patients with IVF compared with the control group (31 % vs 5 %, respectively). The odds ratio (OR) for the presence of ER in the IVF cases compared with the controls was 10.9 (95 % confidence interval (CI) 6.3 to 18.9) after adjustment for age, sex, race and level of physical activity. Other case–control studies reported a similarly higher ratio of an ER pattern in IVF patients (42–60 %) compared with the controls (3.3–13 %, P0.2 mV in the inferior leads was associated with an even more increased risk of death from cardiac causes (adjusted RR 2.98, 95 % CI 1.85 to 4.92) and from arrhythmia (adjusted RR 2.92, 95 % CI 1.45 to 5.89) [16]. In subjects with a J-wave of >0.2 mV, the Kaplan–Meier survival curve for death from cardiac causes started to diverge at 15 years follow-up, which suggests an occurrence of events from around 60 years of age [16]. These data imply that >0.2mV J-wave elevation in this age group is not a sole determinant leading to VF but likely plays a role as a modifier to increase a risk of arrhythmic or cardiac death in patients with an underlying disorder such as coronary artery disease [16, 25]. As a supporting evidence for this hypothesis, the same group demonstrated in a subsequent study that in acute coronary disease, VF victims had a higher prevalence of an ER pattern compared with the controls without VF [30]. In particular, an ER pattern with a horizontal ST segment was an independent predictor of sudden death (OR 2.15, 95 % CI 1.20–3.85) [30]. An increased VF risk in patients with an ER pattern has been shown in other studies on subjects with a first myocardial infarction [31, 32] as well as vasospastic angina cases [33].
Sinner et al. reported similar results to the study by Tikkanen et al. [27]. They performed a population-based prospective study on 1,945 patients from the MONICA/ KORA cohort which demonstrated the prevalence of an ER pattern of 13.1 %. ER pattern was associated with cardiac and all-cause mortality, mostly pronounced in those of younger age (35–54 years old) and male sex; hazard ratio (HR) was 2.65 (95 % CI 1.21–5.83) for men between 35 and 54 years old. An inferior ER pattern further increased cardiac mortality (HR of 3.15, 95 % CI 1.58–6.28) for both sexes and particularly in male subjects between 35 and 54 years of age (HR of 4.27, 95 % CI 1.90–9.61). HRs for all-cause mortality were weaker but reached statistical significance. Another study on the incidence of ER appeared from Japan and reported different results compared with the two studies mentioned above. Haruta et al. evaluated ECGs of 5,976 atomic bomb survivors in the Nagasaki area who underwent a biennial health check-up [26]. They reported that the incidence rate of an ER pattern was 715 per 100,000 person-years. Their study cohort showed an association with an increased risk of unexpected death (HR 1.83, 95 % CI 1.12–2.97) but decreased risk of cardiac death (HR 0.75, 95 % CI 0.60– 0.93) and all-cause mortality (HR 0.85, 95 % CI 0.78–0.93). Both slurring and notching patterns of J-wave were related to a higher risk of unexpected death (HR 2.09, 95 % CI 1.29– 4.83) and in both inferior and lateral leads (HR 2.50, 95 % CI 1.29–4.83). The reason for the inconsistent results from the different studies is not clear, but age, ethnicity, and diagnostic methods of sudden death cases (autopsy or death certificate) may play a role in it. A meta-analysis of prospective cohort studies and case–control studies demonstrated that the risk ratios of an ER pattern were 1.70 (95 % CI 1.19–2.42, P=0.003) for arrhythmic death, 0.78 (95 % CI 0.27–2.25, P=0.63) for cardiac death and 1.06 (95 % CI 0.87–1.28, P=0.57) for all-cause mortality [34]. A J-wave amplitude of ≥0.1 and ≥0.2 mV in inferior leads had a higher risk for arrhythmic death (RR 1.58 and 3.02, respectively) and death from cardiac causes (RR 1.48 and 2.98, respectively). Notching J-wave patterns had an increased risk for arrhythmic death (RR 1.48). Importantly, the estimated absolute risk differences of subjects with an ER pattern in the meta-analysis were low (70 cases of arrhythmia death per 100,000 subjects per year) [34]. While a metaanalysis is limited by its heterogeneous population, the reported results seem to reflect the fact that a J-wave elevation is indeed benign in general. Of note, the morphology of an ST segment (horizontal/descending or ascending pattern) was not included in this meta-analysis due to the lack of data. In addition to the localization and morphology of the Jwave, more recent studies have investigated the morphology of ST segments following the J-point to scrutinize an ER pattern carrying a risk of arrhythmic/cardiac death. In the general population, Tikkanen et al. analysed the prognostic
J Interv Card Electrophysiol
significance of horizontal/descending ST segments and successfully showed that a J-wave of ≥0.1 mV with a horizontal/ descending ST segment had an increased HR of arrhythmic death (RR 1.43, 95 % CI 1.05–1.94) while ascending an ST segment did not show an increased RR (0.89), suggesting a benign ER variant [17]. Strikingly, subjects with a higher Jpoint amplitude (>0.2 mV) in the inferior leads with a horizontal/descending ST segment showed an even higher HR of arrhythmic death of 2.14 (95 % CI 1.56–6.30). Rosso et al. investigated IVF patients and age- and gender-matched controls and showed that a J-wave with a horizontal STsegment morphology was strongly associated with a history of VF (OR 13.8, 95 % CI 5.1–37.2) [18]. In spite of the amount of knowledge acquired so far, the utility of the ER pattern as a prognostic indicator of VF risk in asymptomatic individuals remains unclear. For example, in asymptomatic individuals (aged 35–45 years) with a J-wave and horizontal ST segment, the risk of cardiac arrest due to ER syndrome is estimated by Rosso et al. to be as low as 0.03 % [18]. Wu et al. similarly calculated a low absolute incidence rate of arrhythmic death in subjects with an ER pattern (70 cases per 100,000 subjects per year) [34]. Besides, various definitions of an ER pattern in the different studies make it difficult to conclude which ER pattern actually carries a risk for arrhythmic and cardiac mortality. The results of the more recent studies suggest that a J-wave alone may not be sufficient to differentiate a malignant ER pattern, but the definition of an ‘ER pattern’ ought to include both J-wave (amplitude, morphology and localization) and ST-segment morphology to find individuals carrying a risk ER pattern. At present, for suspected IVF cases, a careful assessment of patient history coupled with a watchful follow-up remains a principle to diagnose patients at risk [35, 36]. The ascending pattern of an ST segment may help exclude subjects with an innocuous ER pattern. Further research is certainly necessary to solve such a dilemma in clinical practice.
5 ER syndrome ‘ER syndrome’ refers to the patients with IVF having an ER pattern. ER syndrome is predominantly observed in male subjects and is more likely to develop sudden cardiac death during sleep [13]. The reported mean age of the first VF episode is 35–43 years old [13–15, 22, 37]. Before ER syndrome is definitely diagnosed, differential diagnosis is necessary to exclude other primary electrical disorders (long/short QT syndrome, Brugada syndrome and catecholaminergic polymorphic ventricular tachycardia) and structural heart diseases. At this time, there is no established method to differentiate ER syndrome from the benign ER pattern. The electrocardiographic and electrophysiological characteristics mentioned below are described in observational studies with a
small number of cases and may help diagnose ER syndrome in clinical practice. 5.1 ECG: modulation of a J-wave in ER syndrome Modulation of the J-wave amplitude is affected by heart rate, heart rhythm, autonomic tone and drugs. J-wave amplitude increases during slow heart rates as well as a short-long-short sequence [23] and a pause [37]. Circadian variation of the Jwave amplitude is also known to occur, and its augmentation is in concordance with a vagal tone [22]. Accordingly, sympathetic stimulation by isoproterenol decreases the J-wave amplitude [37, 38]. Sodium channel blockers also attenuate the J-wave amplitude [3, 38]. J-wave in ER syndrome increases prominently just before arrhythmic events [12, 13, 23], an occurrence which is now recognized as a hallmark of the disease. By performing serial ECG recordings during an electrical storm in 18 subjects with ER syndrome, Haïssaguerre et al. nicely showed that the amplitude of the J-wave increased to a mean of 4.1 mm during an arrhythmic period compared with 2.6 mm at baseline [13]. As another ECG feature of ER syndrome, Nam et al. demonstrated a global appearance of a J-wave on 12-lead ECG within 30 min of VF storms in four out of five cases [23]. Such J-waves may completely disappear within weeks after VF events [37]. Their dynamic manifestation can therefore hinder the diagnosis of malignant cases with ER syndrome. In contrast to these dynamic J-wave changes in ER syndrome, the ER pattern in healthy individuals does not vary over time [16, 39, 40]. 5.2 Non-invasive electrophysiological study Signal-averaged ECG, T wave alternans (TWA), QT dispersion (QTD) and heart rate variability (HRV) are frequently used tests in clinical practice to evaluate arrhythmic risk and the electrophysiological mechanism of primary electrical disease as well as structural heart disease. Abe et al. used these non-invasive methods to investigate the pathophysiology of an ER pattern and its circadian changes in IVF cases with/without a J-wave and age- and gender-matched controls [22]. The results showed that repolarization parameters (TWA and QTD) were not different between IVF cases with and without J-wave whereas abnormal depolarization parameters (late potentials (LPs)) were more frequently observed in IVF with a J-wave (ER syndrome) compared with those without a J-wave or controls. Interestingly, in ER syndrome, the time at which LPs showed abnormal values were concordant with the time of the VF events (six of seven patients at midnight and one during the day), implying the potential role of LPs in the arrhythmogeneity of ER. However, the result of this study is inconsistent with other clinical as well as experimental studies which suggest the involvement of repolarization abnormality
J Interv Card Electrophysiol
in the pathogenesis of an ER pattern [3, 41]. For the mechanistic comprehension of an ER pattern, further studies are required. 5.3 Invasive electrophysiological study in ER syndrome Clinical electrophysiological study is not helpful in the risk stratification of ER syndrome. Haïssaguerre et al. performed such studies in 132 IVF cases to test the inducibility of VF [13]. Stimulation was performed from two ventricular sites with up to three extrastimuli. VF inducibility was low in patients with a history of VF (34 %), and besides, the VF inducibility was not different in IVF patients with or without an ER pattern. The same investigators also tested the relationship between the lead displaying the ER pattern and the location of VF-initiating premature beats. In patients with inferior lead ER, VFinitiating ectopies originated from the inferior wall of the left ventricle whereas in patients with inferolateral ER, ectopies originated from multiple regions [42, 43]. 5.4 Therapy for ER syndrome In patient management, there is no doubt that cases resuscitated from VF require an implantable cardioverter defibrillator (ICD) [4]. In acute instances of recurrent VF in ER syndrome, an isoproterenol infusion effectively suppresses VF. Here, isoproterenol counterbalances the high vagal tone (in the setting of which arrhythmic events often occur) to terminate VF. Deep sedation is also useful to inhibit an electrical storm [13, 38]. For long-term therapy, quinidine is effective [13, 38]. β-blockers, lidocaine, mexiletine and verapamil were reported to be ineffective in both the acute and chronic phase. Amiodarone was reported to be effective only in the acute phase [13, 38]. Yet, the exact mechanism of the antiarrhythmic drugs to suppress VF in ER syndrome is not fully understood.
6 Cellular mechanism of an ER pattern The cellular mechanism of an ER pattern was proposed by Yan et al. in 1996 on the basis of experiments using the arterially perfused wedge preparation of canine left ventricle [41]. The proposed mechanism rests on the intrinsic transmural differences in action potential morphology across the ventricular myocardium where the more prominent transient outward potassium current (Ito) in the epicardium (compared with endocardium) confers a spike-and-dome action potential morphology, which is absent in the endocardium. It is postulated that the different action potential morphologies across the ventricular wall result in a transmural voltage gradient, generating the J-wave. Yan and Antzelevitch showed that the amplitude of the J-wave may be modulated by hypothermia (increases J-wave), by premature stimulation
(decreases J-wave), or by the application of the Ito blocker 4aminopyridine (decreases J-wave) [41]. Because the Ito current is rich in the epicardium but not in the endocardium, enhancement of Ito deepens the phase 1 notch of the epicardial action potential and shortens the epicardial action potential, thereby accentuating transmural heterogeneity and consequently the J-wave. An increase in other cardiac potassium currents (IKr, IKs, IKAch, and IKATP) or a decrease in the cardiac sodium (INa) or calcium currents (ICaL) can also theoretically lead to a transmural gradient and a manifestation of an ER pattern [1]. This model thus suggests the possible involvement of ion channel genes in ER syndrome, the involvement of which has been tested in a number of genetic studies on individuals with this disorder. There are however some clinical as well as experimental aspects which cannot be fully explained by this hypothesis. First, suppose that the transmural heterogeneity of repolarization caused by Ito is the fundamental mechanism for developing VF in individuals with an ER pattern, it is as yet not explained why the risk of arrhythmic death is so low (70 cases per 1,000,000 subjects per year) while an ER pattern in the general population is common (1–13 %) [16, 26, 27, 34]. Studies in the general population as well as ER syndrome suggest that the ER pattern exists as an arrhythmogenic substrate but needs triggers to develop VF [44]. For example, patients with acute coronary disease and VF showed a higher prevalence of an ER pattern compared with those without VF, implying that an ischemic attack may be a trigger for VF in individuals with an ER pattern [30]. In ER syndrome, a Jwave elevation is observed just before arrhythmic events [13, 15]. These observations suggest that multiple factors are involved in the augmentation of the J-wave and subsequent development of VF in subjects carrying an ER pattern. However, a different impact of the malignant or the benign ER pattern on the prognosis remains unexplained. Second, in experimental models, equilibration of the action potential duration is observed in adjacent cells unless intercellular coupling is decreased [44, 45]. Accordingly, in the intact heart, the duration of the action potential in different layers of the ventricular wall is expected to differ gradually and consequently the action potential cannot be extremely short in one region, e.g. in the epicardium to cause an extrasystole leading to VF. On the other hand, myocardial structural changes that may be even clinically undetectable could allow the development of transmural heterogeneity of repolarization. Again, this implicates that transmural heterogeneity of repolarization alone does not seem to explain the arrhythmogeneity of the ER pattern. Finally, the malignant ER pattern may be caused by a different mechanism, and the ECG pattern only morphologically mimic the benign ER pattern in the general population [44]. Historically, repolarization abnormality was proposed to explain the mechanism of both ER pattern and Brugada syndrome [1]. Indeed, abnormal ECG patterns are observed
J Interv Card Electrophysiol
during the same phase of the action potential (the transition phase of depolarization to repolarization) in both conditions [1]. Notwithstanding, there are phenotypical differences between ER and Brugada syndromes. For example, the abnormal ECG patterns affect different regions of the heart: the inferolateral leads in ER syndrome and the right pre-cordial leads in Brugada syndrome. Further, ajmaline attenuates the Jwave amplitude in subjects with an ER pattern while it causes ST-segment elevation in patients with Brugada syndrome [3]. There is clearly a lack of knowledge to explain all these differences, which requires further research.
7 Heritability of an ER pattern 7.1 Heritability of an ER pattern in IVF cases and in the general population The familial nature of the ER pattern has been suggested by case–control studies as well as by studies in the general population [13, 46–48]. Haïssaguerre et al. were the first to report that a family history of unexplained sudden death was more frequently observed in sudden cardiac arrest cases with an ER pattern compared with those without (16 % vs 9 %) [13]. Similar results were reported by Watanabe et al. who studied 50 IVF cases with an ER pattern, in which 16 % had a family history of sudden death [46]. In another study, firstdegree relatives of unexplained sudden arrhythmic death syndrome (SADS) probands were studied. The authors reported that 23 % of SADS relatives showed a J-wave in the inferolateral leads compared with 11 % of the control subjects (OR 2.54, 95 % CI 1.66 to 3.90, P
