1

Repolarization Characteristics in Early Repolarizationand Brugada Syndromes: Insight into an Overlapping Mechanism

of Lethal ArrhythmiasAHMED KARIM TALIB, M.D., Ph.D.,∗ NOBUYUKI SATO, M.D., Ph.D.,∗ NAOKO KAWABATA,

M.D.,∗ EITARO SUGIYAMA, M.D.,∗ NAKA SAKAMOTO, M.D.,∗ YASUKO TANABE, M.D.,Ph.D.,∗ TAKAYUKI FUJINO, M.D., Ph.D.,∗ TOSHIHARU TAKEUCHI, M.D.,∗ YASUAKI SAIJO,M.D., Ph.D.,† KAZUMI AKASAKA, M.D., Ph.D.,‡ YUICHIRO KAWAMURA, M.D., Ph.D.,∗ and

NAOYUKI HASEBE, M.D., Ph.D.∗

From the ∗Department of Cardiology; †Division of Community Medicine and Epidemiology, Department of Health Science; and‡Department of Laboratory Medicine, Asahikawa Medical University, Asahikawa, Japan

Repolarization Indices in ER and Brugada Syndromes. Introduction: We reported impairedQT-rate dependence in early repolarization syndrome (ERS); however, contemporary data have shownpeak incidence of sudden cardiac death (SCD) in ERS and Brugada syndrome (BrS) at mid-night and earlymorning. Taken together, we analyzed the nocturnal QT-rate dependence in both syndromes.

Methods and Results: A total of 172 subjects were enrolled: 11 ERS, 11 BrS patients, 50 subjects with anuneventful ER pattern (ERP), and 100 non-J-wave control subjects. Ambulatory ECG-derived parameters(QT, QTc, and QT/RR slope) and day–night QT difference were analyzed and compared. Among thegroups, there was no significant difference in the average QT or QTc; however, the 24-hour QT/RRslope was significantly smaller in ERS and BrS patients (0.103 ± 0.01 and 0.106 ± 0.01, respectively)than in the control group (0.156 ± 0.03, P < 0.001). Detailed analysis showed a lower day–night QTdifference in ERS and BrS patients (19 ±18.7 and 24 ±14 milliseconds, respectively) than in the controls(40 ± 22 milliseconds, P = 0.007) with the lowest QT/RR slopes seen in the ERS and BrS groups from 0 to3:00 am (QT/RR; 0.076 ± 0.02 vs. 0.092 ± 0.04 vs. 0.117 ± 0.04, for the ERS, BrS, and controls, respectively,P = 0.004) and from 3 to 6 am (QT/RR 0.074 ± 0.03 vs. 0.079 ± 0.02 vs. 0.118 ± 0.04, P < 0.001).

Conclusion: In a large population of age- and gender-matched groups, both ERS and BrS patientsshowed attenuated QT-rate dependence and impaired QT day–night modulation that may provide abaseline reentrant substrate. Importantly, QT/RR maladaptation was most evident at mid-night and earlymorning, which may explain the propensity of such patients to develop SCD during this critical period. (JCardiovasc Electrophysiol, Vol. pp. 1-9)

Brugada syndrome, early repolarization syndrome, QT dynamics, QT interval, sudden death

Introduction

Although a clear potential association between electrocardio-graphic early repolarization (ER) and sudden cardiac death(SCD) was reported more than 5 years ago,1 only a fewelectrocardiographic approaches are available to increase theunderstanding of the arrhythmogenic mechanism of ER, andmost have focused on the J-wave characteristics, such asthe J-wave amplitude and distribution,2,3 and ST-segmentconfiguration.4 Apart from the J-wave characteristics, it wasreported that, among patients with idiopathic ventricular fib-rillation (VF), ER syndrome (ERS) patients have a shortercorrected QT (QTc) interval than non-ER patients.1 Recently,we have found that although the QT and QTc intervals do

No disclosures.

Address for correspondence: Ahmed Talib, M.D., Ph.D., Departmentof Cardiology, Asahikawa Medical University, Midorigaoka Higashi2-1-1-1, Asahikawa 078–8510, Japan. Fax: 81-166-68-2449; E-mail:ahmed@asahikawa-med.ac.jp

Manuscript received 3 June 2014; Revised manuscript received 21September 2014; Accepted for publication 14 October 2014.

doi: 10.1111/jce.12566

not differ between uneventful ER pattern (ERP) subjects andERS patients with a history of VF, the latter have a depressedadaptation of the QT interval to the heart rate, which may actas an abnormal substrate.5

On the other hand, epidemiological data have shown thatthe incidence of ventricular arrhythmias in ERS and Brugadasyndrome (BrS) patients exhibits marked circadian varia-tions with significant peaks of cardiac arrest and appropriateshocks observed from mid-night to 6:00 am.6,7

Taken together, we hypothesized that the QT-rate mal-adaptation is most prominent from mid-night to 6:00 am inERS and BrS patients and it may explain the propensity ofsuch patients to develop SCD during such a critical period. Inaddition, we analyzed the QT day–night modulation in bothsyndromes.

Methods

Study Populations

A total of 172 subjects were studied and divided into4 groups; the first 2 groups were ERS and BrS patients, mostof whom had a history of an aborted SCD. The third groupincluded ERP subjects without any cardiac events, and thefourth group included control subjects without ERP.

2 Journal of Cardiovascular Electrophysiology Vol. No.

TABLE 1

Patient Characteristics

ERS BrS P

Number of patients (male %) 11 (81%) 11 (100%) NSAge (years) 51.2 ± 12 48.7 ± 10 NSPR interval (milliseconds) 172 ± 21 169 ± 19 NSVF 8 8 NSSyncope 3 2 NSElectrical storm 3 1 NSFamily history of SCD 1 2 NSPolymorphic VT/VF induction during EPS 0/3 1/3 NS

NS = not significant; ERS = early repolarization syndrome; BrS = Brugadasyndrome; EPS = electrophysiology study; SCD = sudden cardiac death.

ERS and BrS Patients

We analyzed the 24-hour ECG recordings of our patientswho were referred to our hospital for an implantable car-dioverter defibrillator (ICD) implantation from September1999 to January 2014. During admission, a standard inves-tigation protocol was performed in each patient includinga 12-lead surface ECG, chest radiographs, 2-dimensionalechocardiography, 24-hour Holter recordings, myocardialperfusion imaging, cardiac magnetic resonance imaging,cardiac catheterization including acetylcholine provocationwhen indicated, endomyocardial biopsy, and an electrophys-iology study (EPS) when appropriate.

After excluding those with structural heart diseases, weidentified 24 patients who underwent an ICD implantationdue to BrS and ERS. One patient was excluded from thestudy because he was taking an antiarrhythmic medicationfor the treatment of paroxysmal atrial fibrillation; anotherpatient was excluded due to low-amplitude T waves re-sulting in inappropriate QT measurements. Eleven patients(9 males, mean age 51.2 ± 12 years) had a J-wave elevationof �0.1 mV in at least 2 leads within the inferior and/orlateral chest leads; namely, 5 patients had inferolateral, 4had inferior and 2 had lateral ER. Eleven patients (11 males,mean age 48.7 ± 10 years) had a spontaneous ST-segmentelevation in leads V1–V2 including 9 patients with a coved-type ST-segment elevation and 2 patients with a saddle-backtype. In the ERS patients, the presence of BrS was ruled outby a comprehensive work-up including a pilsicainide test. AnEPS was performed in 6 patients. From the right ventricu-lar apex and right ventricular outflow tract, burst pacing anda maximum of up to triple programmed extrastimuli weredelivered to induce ventricular tachycardia (VT)/VF.

Table 1 shows the patient characteristics. In the ERSgroup, 2 patients presented with recurrent syncopal episodesand 9 patients presented with aborted SCD. In the BrS group,8 patients presented with VF, 2 patients presented with syn-cope, and 1 presented with palpitations. Among the 6 (3 ERSand 3 BrS) patients who underwent an EPS, polymorphic VTdeteriorating to VF could be induced in only 1 BrS patient.

ERP and Control Groups

From December 2008 until January 2014, 150 healthysubjects were recruited for this study and subdivided into thefollowing 2 subgroups:

(1) The ERP group included 50 subjects (40 males, meanage 50.9 ± 19 years) in whom medical disorders were

excluded by a detailed medical history and examina-tion that included a 12-lead ECG and chest X-ray. All50 subjects exhibited a notched or slurred J-wave ele-vation of �0.1 mV in at least 2 of the inferior and/orlateral leads. In this group, 23 subjects had inferolateral,20 inferior, and 7 lateral ER.

(2) The control group included 100 healthy volunteers(76 males, mean age 49.6 ± 18 years). All of the con-trol subjects clearly had neither J-wave nor ST-segmentelevation. None of the subjects had any history or ra-diological or ECG evidence of cardiovascular diseasenor were they taking any medication known to affectcardiac repolarization. Subjects with T wave inversion,wide QRS complexes, and ventricular preexcitation wereexcluded from the study. All of the subjects and patientswere allowed to carry out their usual daily activities.The local ethical committee approved the study and allparticipants gave their informed consent.

Holter Recordings and Analyses

All of the study groups underwent 24-hour ECG recordingusing a 2-channel Holter recording system (with NASA andCM5 leads) analyzed by Synetec Holter analysis software(ELA Medical, Sorin Group, Paris, France). The QT inter-vals were automatically measured using the CM5 lead, anda visual review was always performed to confirm the correctmeasurement of the QT intervals. We described the methodpreviously;5 briefly, after conversion of the 24-hour record-ings into 2,880 templates sampled at 30-second intervals, theintervals with �80% eligible QRS complexes were consid-ered as ectopy or noise and therefore eliminated. The averageQT intervals were then measured automatically at the apex(QTa) of the T wave, where the peak of the parabola wasfitted to the highest amplitude change after the QRS and, atthe end of the T wave (QTe), defined as the intersection of thetangent to the downslope of the T-wave and the isoelectricline. The average RR interval in each template was mea-sured. For each template, the mean QT values were plottedagainst the mean R–R of a 30-second interval and the linearregression (QT/RR) and slopes of the linear regressions wereautomatically measured. The Tpeak–Tend (Tp–Te) interval,which is the interval from the peak to the end of the T wave,was calculated as described previously.8

Regarding the 24-hour recording, in order to focus on theintranocturnal QT/RR, the night recording time was dividedinto the following 3 periods: (i) early nighttime (from 18:00to 00:00), (ii) mid-night (from 00:00 to 03:00), and (iii) earlymorning (from 03:00 to 06:00). To improve the reliabilityof the T-wave offset determination, we excluded any poorquality and/or markedly low-amplitude T waves (<0.1 mV)from the study.

Statistical Analysis

The values are presented as the mean ± standard devia-tion. Statistical comparisons among the groups were testedwith a t-test or one-way analysis of variance (ANOVA) as ap-propriate. The ANOVA was followed by a post hoc Tukey’stest for significant intergroup differences. The significant dif-ferences in proportions were assessed by a chi-square test.All statistical analyses were performed on a personal com-puter with the Sigma Plot for Windows statistical package

Talib et al. Repolarization Indices in ER and Brugada Syndromes 3

TABLE 2

QT Indices in the ERS, BrS, and Control Groups

ERS (n = 11) BrS (n = 11) Control (n = 100) P

Age (years) 51.2 ± 12 48.7 ± 10 50.8 ± 18 0.365HR (beat/min) 57.4 ± 7.9 68.9 ± 5.5 67.5 ± 10.2 0.111QT (milliseconds) 413 ± 29 388.8 ± 21 399.8 ± 29 0.209QTc (milliseconds) 405.5 ± 26.5 411.7 ± 18 420.5 ± 23 0.609Tp–Te (milliseconds) 81.2 ± 9† 90.3 ± 9.7‡± 70.9 ± 7.4 <0.00124-hour QT/RR 0.103 ± 0.01¶ 0.105 ± 0.01‡ 0.156 ± 0.03 <0.001Day-time QT (milliseconds) 392 ± 22 377 ± 27 381 ± 29 0.301Night-time QT (milliseconds) 415 ± 28 399 ± 21 418 ± 26 0.154QT day–night difference (milliseconds) 19 ± 18.7† 24 ± 14** 40 ± 22 0.007Early night QT/RR 0.1 ± 0.03 0.125 ± 0.04 0.153 ± 0.04 0.068Mid-night QT/RR 0.076 ± 0.02 † 0.092 ± 0.04* 0.117 ± 0.04 0.004Early morning QT/RR 0.074 ± 0.03 ¶ 0.079 ± 0.02 ‡ 0.118 ± 0.04 <0.001

Values are expressed as mean ± SD.P values refer to the total significance of differences (ANOVA) among the 3 groups.¶P < 0.001, †P < 0.01: ERS versus control using ANOVA and post hoc Tukey’s test.‡P < 0.001, **P < 0.01, *P < 0.05: BrS versus control using ANOVA and post hoc Tukey’s test.±P < 0.05: BrS versus ERS using ANOVA and post hoc Tukey’s test.ERS = early repolarization syndrome; BrS = Brugada syndrome; Tp–Te = the interval from the peak to the end of the T wave; HR = heart rate.

TABLE 3

QT Indices in the ERS, ERP, and Control Groups

ERS (n = 11) ERP (n = 50) Control (n = 100) P

Age (years) 51.2 ± 12 50.9 ± 19 50.8 ± 18 0.267HR (beat/min) 57.4 ± 7.9 66.8 ± 9.6 67.5 ± 10.2 0.84QT (milliseconds) 413 ± 29 396.6 ± 30 399.8 ± 29 0.142QTc (milliseconds) 405.5 ± 26.5 418.7 ± 18 420.5 ± 23 0.297Tp–Te (milliseconds) 81.2 ± 9 69.7 ± 11.2† 70.9 ± 7.4* <0.0524-hour QT/RR 0.103 ± 0.01 0.158 ± 0.02‡ 0.156 ± 0.03 § <0.001Early night QT/RR 0. 091 ± 0.02 0.136 ± 0.04 0.153 ± 0.04* 0.032Mid-night QT/RR 0. 073 ± 0.02 0.115 ± 0.04† 0.117 ± 0.04* 0.012Early morning QT/RR 0. 079 ± 0.02 0.119 ± 0.04†† 0.118 ± 0.04** 0.003

Values are expressed as the mean ± SD.P values refer to the total significance of differences (ANOVA) among the 3 groups.§P < 0.001, **P < 0.0, 1*P < 0.05: control group versus ERS patients, using ANOVA and post hoc Turkey’s test.‡P < 0.001, ††P < 0.01, †P < 0.05: ERP group versus ERS patients, using ANOVA and post hoc Turkey’s test.ERS = early repolarization syndrome; ERP = early repolarization pattern.

(version 11.2, Synstat Software, Inc., San Jose, CA, USA).A P value of <0.05 was considered significant.

Results

There was no significant age difference among the groups(P = 0.365). Table 1 shows the patient characteristics. Thebaseline ECG in all patients exhibited a normal PR intervalwith no significant difference between the ERS and BrS pa-tients. The majority of the patients presented with VF and 1ERS patient presented with an electrical storm, which waseffectively controlled by deep sedation. After an ICD im-plantation and during the follow-up of 7.3 ± 4.1 (from 1.2to 14) years, 3 (2 ERS and 1 BrS) patients developed electri-cal storms that were effectively controlled by an intravenousisoproterenol infusion in 2 patients and by deep sedation in1 patient. Furthermore, 1 BrS patient developed 1 appropri-ate ICD discharge. As shown in Table 2, although there wasno significant difference in the average heart rate, QT, andQTc intervals among the ERS, BrS, and control groups, theQT/RR slopes were significantly smaller in the ERS and BrSgroups than in the control group.

It is known that in healthy individuals, the QT intervalexhibits a circadian variation with a longer QT interval duringnighttime.9 To investigate the QT characteristics in ERS andBrS, we separately analyzed the daytime and nighttime QTintervals and found no significant difference in the daytime aswell as nighttime QT intervals among the groups; however,the day–night difference in the QT interval was significantlysmaller in the BrS and ERS groups than that in the controlgroup, suggesting a loss of the day–night modulation of theQT interval in the former groups (Table 2).

Importantly, a comparison of the intranocturnal QT/RRslopes among the ERS, BrS, and control groups revealed nosignificant difference in the early night QT/RR slope amongthe groups, but the mid-night and early morning QT/RRslopes were significantly smaller in the ERS and BrS groupscompared to those in the control group. Because the clinicaland mechanistic issues of benign and malignant ER are stillunresolved issues,2,10 we also sought to examine the QT be-havior in ERS patients and age- and gender-matched healthyERP and non-J wave control subjects. Although there wasno significant difference in the heart rate and QT and QTcintervals among the groups, the QT/RR slope was signifi-cantly smaller in the ERS group compared to the ERP group,

4 Journal of Cardiovascular Electrophysiology Vol. No.

400 600 800 1000 1200 1400 1600 1800300

320

340

360

380

400

420

440

460

480

400 600 800 1000 1200 1400 1600 1800300

320

340

360

380

400

420

440

460

480

400 600 800 1000 1200 1400 1600 1800300

320

340

360

380

400

420

440

460

480

400 600 800 1000 1200 1400300

320

340

360

380

400

420

440

460

400 600 800 1000 1200 1400300

320

340

360

380

400

420

440

460

400 600 800 1000 1200 1400300

320

340

360

380

400

420

440

460

QT ( TQ)sm ( TQ)sm ( ms)

RR ( ms) RR ( RR)sm ( ms)

QT ( ms) QT ( TQ)sm ( ms)

RR ( RR)sm ( ms) RR ( ms)

early night mid-night early morning

early night mid-night early morning

ERS

BrS

Figure 1. Linear regression of the QT/RR for all ERS (upper panel) and BrS (lower panel) patients during early night, mid-night, and early morning togetherwith the average regression (bold black) line of each group.

whereas there was no significant difference in either of theslopes between the ERP and control groups (Table 3). Simi-larly, an intranocturnal analysis of the QT/RR slope revealeda significant reduction in the QT/RR slopes of the ERS pa-tients at mid-night and early morning, while there were nosignificant differences in the intranocturnal QT/RR slopesbetween the ERP and control groups (Table 3).

Finally, comparing the Tp–Te interval among all ourgroups, we found that BrS patients had the longest Tp–Teinterval (Table 2). Although BrS patients had a longer Tp–Teinterval than the ERS patients, the latter had a longer Tp–Te interval than those with ERP or non-ERP control groups(Tables 2 and 3). Figure 1 summarizes the linear regressionslopes of the QT/RR for all ERS and BrS patients duringearly night, mid-night, and early morning together with theaverage regression line of each group.

Figure 2 shows examples of the QT/RR slopes in a controlsubject, ERS patient, and patient with BrS in whom VF, fol-lowed by the appearance of large J waves after resuscitation,was documented. ERS and BrS patients had smaller QT/RRslopes than the control subjects.

Figure 3 shows a representative ECG of an ERS malepatient whose baseline ECG showed small, almost nondiag-nostic, J waves in the limb leads I and aVL. Two years afteran ICD implantation, he developed an electrical storm at 5:20in the morning. The storm was not terminated by several ICD

discharges and was finally controlled by deep sedation. TheECG monitor showed augmented J waves just before the VF,which was initiated by a premature ventricular contractionwith a short coupling interval. A QT/RR analysis in this pa-tient showed a QT/RR slope comparable to that of an age-and gender-matched uneventful ERP subject during the earlynighttime; however, during mid-night and early morning, theQT/RR slopes of the ERS patient were smaller than those ofthe ERP subject (Fig. 4).

Discussion

Main Findings

In the present study, we demonstrated that: (1) the day-night modulation of the QT interval, a physiological phe-nomenon seen in healthy ERP and control subjects, is lost inERS and BrS patients, and (2) ERS and BrS patients havea QT heart rate maladaptation, leading to an abnormal tem-poral repolarization which may predispose them to phase2 re-entry. This maladaptation was most prominent at mid-night and early morning (i.e., 0–6 am), which may explain, inpart, the propensity of ERS and BrS patients to develop SCDduring heart rate decelerations during such critical periods.Although we have already demonstrated impaired QT-ratedependence in a small cohort of ERS patients, to the best of

Talib et al. Repolarization Indices in ER and Brugada Syndromes 5

A

B

0 200 400 600 800 1000 1200 1400 1600 1800200

300

400

500

600

0 200 400 600 800 1000 1200 1400 1600 1800200

300

400

500

600

0 200 400 600 800 1000 1200 1400 1600 1800200

300

400

500

600

QTe/RR slope = 0.160 701.0=epolsRR/eTQ390.0=epolsRR/eTQ

SrBSRElortnoC

C

V1

V2

V3

V4

V5

V6

aVR

aVL

aVF

QT ( ms) QT ( ms) QT ( ms)

RR ( ms) RR ( ms) RR ( ms)

Figure 2. Representative QT/RR slopes in a healthy control subject, ERS patient, and patient with Brugada syndrome (A) whose ECG showed ST-elevationin leads V1–V2 (B). The patient was successfully resuscitated out of hospital, where his AED monitor showed large J-waves just after the VF termination (C).For a high quality, full color version of this figure, please see Journal of Cardiovascular Electrophysiology’s website: www.wileyonlinelibrary.com/journal/jce

our knowledge, this is the first study to analyze the nocturnalQT indices in detail in both ERS and BrS patients.

QT Day–Night Modulation and QT-Rate Dependence

It is known that the QT interval duration is longer atnight than during the day, even when comparisons are madeduring identical heart rates, and this phenomenon is relatedto the circadian variations in the sympathovagal balance.11

On the other hand, it is reported that patients with short-QT syndrome (SQTS) have very little change in the QTinterval with heart rate changes.12 In our study population,the above-mentioned physiological phenomenon was seen inthe uneventful ERP and control subjects; however, the ERSand BrS patients had an impaired QT day–night modulation,suggesting that these groups have a similar phenomenon tothat seen in SQTS.12

The relationship between the QT-interval duration andthe immediately preceding RR interval, the so-called repo-larization dynamics, is a surrogate of the action potentialelectrophysiology. It reflects the influence of several factorsacting on myocytes such as the heart rate and autonomicnervous system, as well as the structure and function of re-polarizing currents.13-15 The reproducibility of the modelingof the QT/RR dynamics has been well demonstrated.16 The

lack of dependence of the QT interval on the heart rate underbaseline conditions is a striking feature of SQTS.17

Indeed, Bjerregaard and Gussak12 suggested a smallQT/RR to be a useful way to distinguish between patientswith a short-QT interval at risk of lethal arrhythmic eventsand normal subjects with a QT interval at the lower levelof normal. Antzelevitch et al.18 showed that the QT/heartrate relationship was reduced in BrS patients with L-type Cachannel mutations compared to the controls. Recently, wehave found that compared to a control group, ERS patientshave impaired QT-rate dependence suggesting that such a re-polarization heterogeneity may provide a baseline substratefor reentry in ERS patients.5

Experimentally, Jeyaraj et al.19 provided molecular ev-idence that the cardiac ion-channel expression and QT-interval duration exhibit endogenous circadian rhythmicityunder the control of a clock-dependent oscillator, Kruppel-like factor 15 (Klf15), which transcriptionally controls therhythmic expression of Kv channels interacting with pro-tein 2 (KChIP2), a critical subunit required for generatingthe transient outward potassium current (Ito). A deficiencyor excess of Klf15 causes a loss of the rhythmic QT varia-tion, abnormal repolarization, and enhanced susceptibility toventricular arrhythmias. Further, Klf15–Tg mice (which ex-hibit a substantial increase in the Ito fast density) have a dra-matic shortening of the action potential duration, persistently

6 Journal of Cardiovascular Electrophysiology Vol. No.

aVR

aVL

aVF

V1

V2

V3

V4

V5

V6

A

B

Figure 3. Representative electrocardio-gram showing almost nondiagnostic Jwaves in leads I and aVL in amiddle-aged male patient with earlyrepolarization syndrome (A) who wasadmitted to our intensive care unit dueto an electrical storm. J-wave augmen-tation was seen just before the develop-ment of VF, which was initiated by an ex-trasystole with a short coupling interval(B). For a high quality, full color versionof this figure, please see Journal of Car-diovascular Electrophysiology’s website:www.wileyonlinelibrary.com/journal/jce

short-QT intervals with no rhythmic day–night variation, andST-segment changes indicative of early repolarization.19

Intranocturnal QT Dynamics

In ERS and BrS patients, the J-wave circadian modula-tion has been assessed.20,21 However, nocturnal QT indiceshave not been assessed in both ERS and BrS patients in de-tail. To the best of our knowledge, the present study is thefirst to demonstrate the relationship between the QT and theRR intervals during daily life in both ERS and BrS patientsin comparison to control subjects with and without J-waveelevation.

It is known that the incidence of SCD shows circadianvariations22 and there is a corresponding temporally co-ordinated increase in the ventricular refractoriness,23 QTinterval,9 and QT dispersion.24 Experimentally, Yamashitaet al. reported that gene expression of Kv4.2 channels, whichencode Ito, exhibit a unique circadian variation, with an in-creased gene expression at night.25 This would shorten theaction potential duration, and in addition, an elevated night-time muscarinic stimulation would also enhance the actionpotential shortening by suppressing the L-type Ca2+ currentand/or inducing acetylcholine-sensitive K+ currents.25

More recently, an association of a short QTc with J-point elevation and enhanced susceptibility to ventriculararrhythmias has been experimentally reported.19 Althoughwe could not perform any genetic testing in most of our pa-tients, our finding of a lack of a QT-rate dependence couldbe explained by an increase in the net outward current due toeither a reduction in the inward depolarizing currents such asIca, or an augmentation of the outward repolarizing currentssuch as Ito. Indeed, both of these channelopathies have beenreported in ERS and BrS patients.2

Our finding of a close overlap of SQTS with BrS and ERSis supported by many facts including: (1) the shorter QTcinterval in ERS patients compared to the control,3 (2) theability of quinidine to prevent VT/VF in all groups,17,18,26

(3) the efficacy of isoproterenol in the acute management ofVF in ERS,26 BrS,27 and SQTS28 patients, (4) the genderdifference with a male predominance in all 3 groups2,29,30 inaddition to the known effect of anabolic androgenic steroidsand testosterone in QT interval shortening,31 (5) the timingof cardiac arrest (during rest or sleep) in the majority of casesin the 3 groups,2,30 and (6) a high prevalence of ER in SQTSpatients.32

From a clinical point of view, the average QT/RR forthe ERS and BrS patients was consistently less than 0.13 andsuch a finding could not be found in either the ERP or controlgroups. The ability to test such a threshold in other cohortswould be useful.

Finally, the mechanism of VF in SQTS is known to bedue to short effective refractoriness leading to R on T phe-nomenon in addition to QT abbreviation resulting in a verysignificant accentuation of the transmural dispersion of therepolarization (TDR).33

In BrS cohort, Castro Hevia et al. reported, as a measureof the TDR, the interval from the peak to the end of theelectrocardiographic T wave was significantly increased inthose who have a high incidence of recurrent cardiac events.34

Recently, we reported enhanced TDR in a small cohort ofERS and BrS patients.8 In this study, and in a larger cohortof patients, we also found an increased Tp–Te interval in BrSand ERS patients when compared to the age- and gender-matched ERP and control groups.

Enhanced TDR, in addition to impaired QT prolongationduring heart decelerations at mid-night and early morning,might provide a susceptible window for lethal arrhythmiaswhen combined with pro-arrhythmic elements, such as emo-tional stress and sympathovagal imbalance, and may lead toSCD during such critical periods.

Limitations of the Study

This study had some limitations, the most significant ofwhich is the sample size, which reflects the disease rarity. Our

Talib et al. Repolarization Indices in ER and Brugada Syndromes 7

0 200 400 600 800 1000 1200 1400 1600 1800200

300

400

500

600

0 200 400 600 800 1000 1200 1400 1600 1800200

300

400

500

600

0 200 400 600 800 1000 1200 1400 1600 1800200

300

400

500

600

0 200 400 600 800 1000 1200 1400 1600 1800200

300

400

500

600

0 200 400 600 800 1000 1200 1400 1600 1800200

300

400

500

600 QTe/RR slope 0.12

QTe/RR slope 0.13 QTe/RR slope 0.06

QTe/RR slope 0.05 QTe/RR slope 0.11

0 200 400 600 800 1000 1200 1400 1600 1800200

300

400

500

600QTe/RR slope 0.14

RR ( ms) RR ( ms)

RR ( RR)sm ( ms)

RR ( RR)sm ( ms)

QT ( ms)

QT ( ms)

QT ( ms)

QT ( ms)

QT ( ms)

QT ( ms)

Figure 4. Representative slopes showing the early night, mid-night, and early morning QT/RR slopes in the ERS patient described in (A) and an age-and gender-matched healthy male with an uneventful ERP (B). During the early night, the QT/RR slope of the ERS patient was comparable to that of theuneventful ERP subject, then the QT/RR slope of the former became flatter than that of the latter during mid-night and early morning, indicating loss ofQT-rate dependence in such critical periods.

data were from a single institution and the vast majority of ourpatients were survivors from aborted SCD due to a rare dis-ease; however, our results provided findings that are worthyof further investigation in a larger multicenter study. Second,we faced some technical difficulties regarding the T-waveoffset determination; therefore, we excluded 23 subjects fromthe study including one patient and 22 ERP and controlsubjects due to poor quality and/or markedly low-amplitudeT waves (<0.1 mV), which further limited the sample size.Third, we did not investigate the reproducibility of the datain all participants; however, in several patients and subjects,

the QT/RR value was consistent. Importantly, the ERS andBrS patients exhibited the same pattern in repeated record-ings at 2 different times. Furthermore, the reproducibility ofthe QT/RR measurement has been demonstrated.16 Fourth,genetic testing could only be performed in 1 BrS patient;this patient was found to have an SCN5A mutation. Finally,it remains difficult to explain the pathophysiology of SCDby marked QT-rate maladaptation alone; however, we havealso demonstrated enhanced TDR in ERS and BrS cohortsand the combination of these 2 mechanisms may provide apotential substrate for VF.

8 Journal of Cardiovascular Electrophysiology Vol. No.

Conclusion

Detailed analysis of the QT indices in BrS and ERS pa-tients revealed 2 characteristics commonly seen in SQTS,which are blunted day–night QT modulation and lack of QTdependence on the heart rate, of which the latter may, at leastin part, provide a substantial baseline substrate for phase 2 re-entry. This lack of QT-rate dependence was most prominentat mid-night and early morning during heart rate decelera-tions. Such a finding may explain the propensity of ERS andBrS to develop SCD during these critical periods.

References

1. Haıssaguerre M, Derval N, Sacher F, Jesel L, Deisenhofer I, de RoyL, Pasquie JL, Nogami A, Babuty D, Yli-Mayry S, De Chillou C,Scanu P, Mabo P, Matsuo S, Probst V, Le Scouarnec S, Defaye P,Schlaepfer J, Rostock T, Lacroix D, Lamaison D, Lavergne T, EnglundA, Anselme F, O’Neill M, Hocini M, Lim KT, Knecht S, VeenhuyzenGD, Bordachar P, Chauvin M, Jais P, Coureau G, Chene G, Klein GJ,Clementy J: Sudden cardiac arrest associated with early repolarization.N Engl J Med 2008;358:2016-2023.

2. Antzelevitch C, Yan GX: J wave syndromes. Heart Rhythm2010;7:549-558.

3. Rosso R, Kogan E, Belhassen B, Rozovski U, Scheinman MM, ZeltserD, Halkin A, Steinvil A, Heller K, Glikson M, Katz A, Viskin S: J-pointelevation in survivors of primary ventricular fibrillation and matchedcontrol subjects: Incidence and clinical significance. J Am Coll Cardiol2008;52:1231-1238.

4. Tikkanen JT, Junttila MJ, Anttonen O, Aro AL, Luttinen S, KerolaT, Sager SJ, Rissanen HA, Myerburg RJ, Reunanen A, Huikuri HV:Early repolarization: Electrocardiographic phenotypes associated withfavorable long-term outcome. Circulation 2011;123:2666-2673.

5. Talib AK, Sato N, Asanome A, Myojo T, Nishiura T, Yamaki M,Nakagawa N, Sakamoto N, Ota H, Tanabe Y, Takeuchi T, Kawa-mura Y, Hasebe N: Impaired ventricular repolarization dynamics inpatients with early repolarization syndrome. J Cardiovasc Electrophys-iol 2013;24:556-561.

6. Kim SH, Nam GB, Baek S, Choi HO, Kim KH, Choi KJ, Joung B,Pak HN, Lee MH, Kim SS, Park SJ, On YK, Kim JS, Oh IY, ChoiEK, Oh S, Choi YS, Choi JI, Park SW, Kim YH, Lee MY, Lim HE,Lee YS, Cho Y, Kim J, Lee DI, Cho DK, Kim YH: Circadian and sea-sonal variations of ventricular tachyarrhythmias in patients with earlyrepolarization syndrome and Brugada syndrome; analysis of patientswith implantable cardioverter-defibrillator. J Cardiovasc Electrophys-iol 2012;23:757-763.

7. Takigawa M, Noda T, Shimizu W, Miyamoto K, Okamura H, SatomiK, Suyama K, Aihara N, Kamakura S, Kurita T: Seasonal and circa-dian distributions of ventricular fibrillation in patients with Brugadasyndrome. Heart Rhythm 2008;5:1523-1527.

8. Karim Talib A, Sato N, Sakamoto N, Tanabe Y, Takeuchi T, Saijo Y,Kawamura Y, Hasebe N: Enhanced transmural dispersion of repolar-ization in patients with J wave syndromes. J Cardiovasc Electrophysiol2012;23:1109-1114.

9. Browne KF, Prystowsky E, Heger JJ, Chilson DA, Zipes DP: Prolonga-tion of the Q-T interval in man during sleep. Am J Cardiol 1983;52:55-59.

10. Benito B, Guasch E, Rivard L, Nattel S: Clinical and mechanisticissues in early repolarization of normal variants and lethal arrhythmiasyndromes. J Am Coll Cardiol 2010;56:1177-1186.

11. Murakawa Y, Inoue H, Nozaki A, Sugimoto T: Role of sympatho-vagal interaction in diurnal variation of QT interval. Am J Cardiol1992;69:339-343.

12. Bjerregaard P, Gussak I: Short QT syndrome. In: Gussak I, Antzele-vitch C, eds. Electrical Diseases of the Heart. Genetics, Mechanism,Treatment, Prevention. 1st ed. Philadelphia: WB Saunders, 2009, pp.555-563.

13. Zareba W, Bayes de Luna A: QT dynamics and variability. Ann Non-invasive Electrocardiol 2005;10:256-262.

14. Merri M, Moss AJ, Benhorin J, Locati EH, Alberti M, Badilini F:Relation between ventricular repolarization duration and cardiac cy-cle length during 24-hour Holter recordings. Findings in normal pa-tients and patients with long QT syndrome. Circulation 1992;85:1816–1821.

15. Chevalier P, Burri H, Adeleine P, Kirkorian G, Lopez M, LeizoroviczA, Andre-Fouet X, Chapon P, Rubel P, Touboul P; Groupe d’Etude duPronostic de l’Infarctus du Myocarde: QT dynamicity and sudden deathafter myocardial infarction: results of a long-term follow-up study. JCardiovasc Electrophysiol 2003;14:227-233.

16. Batchvarov VN, Ghuran A, Smetana P, Hnatkova K, Harries M,Dilaveris P, Camm AJ, Malik M: QT-RR relationship in healthy sub-jects exhibits substantial intersubject variability and high intrasubjectstability. Am J Physiol Heart Circ Physiol 2002;282:H2356-H2363.

17. Wolpert C, Schimpf R, Giustetto C, Antzelevitch C, Cordeiro J, Du-maine R, Brugada R, Hong K, Bauersfeld U, Gaita F, Borggrefe M:Further insights into the effect of quinidine in Short QT syn-drome caused by a mutation in HERG. J Cardiovasc Electrophysiol2005;16:54-58.

18. Antzelevitch C, Pollevick GD, Cordeiro JM, Casis O, Sanguinetti MC,Aizawa Y, Guerchicoff A, Pfeiffer R, Oliva A, Wollnik B, Gelber P,Bonaros EP Jr, Burashinikov E, Wu Y, Sargent JD, Schickel S, Ober-heiden R, Bhatia A, Hsu LF, Haıssaguerre M, Schimpf R, BorggrefeM, Wolpert C: Loss-of-function mutations in the cardiac calciumchannel underlie a new clinical entity characterized by ST-segmentelevation, short QT intervals, and sudden cardiac death. Circulation2007;115:442-449.

19. Jeyaraj D, Haldar SM, Wan X, McCauley MD, Ripperger JA, Hu K,Lu Y, Eapen BL, Sharma N, Ficker E, Cutler MJ, Gulick J, SanbeA, Robbins J, Demolombe S, Kondratov RV, Shea SA, Albrecht U,Wehrens XH, Rosenbaum DS, Jain MK: Circadian rhythms governcardiac repolarization and arrhythmogenesis. Nature 2012;483:96-99.

20. Mizumaki K, Fujiki A, Tsuneda T, Sakabe M, Nishida K, Sugao M,Inoue H: Vagal activity modulates spontaneous augmentation of STelevation in the daily life of patients with Brugada syndrome. J Cardio-vasc Electrophysiol 2004;15:667-673.

21. Mizumaki K, Nishida K, Iwamoto J, Nakatani Y, Yamaguchi Y,Sakamoto T, Tsuneda T, Kataoka N, Inoue H: Vagal activity mod-ulates spontaneous augmentation of J-wave elevation in patients withidiopathic ventricular fibrillation. Heart Rhythm 2012;9:249-255.

22. Muller JE, Ludmer PL, Willich SN, Tofler GH, Aylmer G, KlangosI, Stone PH: Circadian variation in the frequency of sudden cardiacdeath. Circulation 1987;75:131-138.

23. Kong TQ Jr, Goldberger JJ, Parker M, Wang T, Kadish AH:Circadian variation in human ventricular refractoriness. Circulation1995;92:1507-1516.

24. Mayuga KA, Thattassery E, Taneja T, Karha J, Subacius H, GoldbergerJ, Kadish A: Circadian and gender effects on repolarization in healthyadults: A study using harmonic regression analysis. Ann NoninvasiveElectrocardiol 2010;15:3-10.

25. Yamashita T, Sekiguchi A, Iwasaki YK, Sagara K, Iinuma H, HatanoS, Fu LT, Watanabe H: Circadian variation of cardiac K channel geneexpression. Circulation 2003;107:1917-1922.

26. Haıssaguerre M, Sacher F, Nogami A, Komiya N, Bernard A, ProbstV, Yli-Mayry S, Defaye P, Aizawa Y, Frank R, Mantovan R, CappatoR, Wolpert C, Leenhardt A, de Roy L, Heidbuchel H, DeisenhoferI, Arentz T, Pasquie JL, Weerasooriya R, Hocini M, Jais P, DervalN, Bordachar P, Clementy J: Characteristics of recurrent ventricularfibrillation associated with inferolateral early repolarization: Role ofdrug therapy. J Am Coll Cardiol 2009;53:612-619.

27. Jongman JK, Jepkes-Bruin N, Ramdat Misier AR, Beukema WP,Delnoy PP, Oude Lutttikhuis H, Dambrink JH, Hoorntje JC, ElvanA: Electrical storms in Brugada syndrome successfully treated withisoproterenol infusion and quinidine orally. Neth Heart J 2007;15:151-155.

28. Bun SS, Maury P, Giustetto C, Deharo JC: Electrical storm in short-QT syndrome successfully treated with isoproterenol. J CardiovascElectrophysiol 2012;23:1028-1030.

29. Shimizu W, Matsuo K, Kokubo Y, Satomi K, Kurita T, Noda T, NagayaN, Suyama K, Aihara N, Kamakura S, Inamoto N, Akahoshi M, To-moike H: Sex hormone and gender difference. Role of testosterone onmale predominance in Brugada syndrome. J Cardiovasc Electrophysiol2007;18:415-421.

30. Mazzanti A, Kanthan A, Monteforte N, Memmi M, Bloise R, NovelliV, Miceli C, O’Rourke S, Borio G, Zienciuk-Krajka A, Curcio A,Surducan AE, Colombo M, Napolitano C, Priori SG: Novel insightinto the natural history of short QT syndrome. J Am Coll Cardiol2014;63:1300-1308.

31. Charbit B, Christin-Maıtre S, Demolis JL, Soustre E, Young J, Funck-Brentano C: Effects of testosterone on ventricular repolarization inhypogonadic men. Am J Cardiol 2009;103:887-890.

Talib et al. Repolarization Indices in ER and Brugada Syndromes 9

32. Watanabe H, Makiyama T, Koyama T, Kannankeril PJ,Seto S, Okamura K, Oda H, Itoh H, Okada M, Tanabe N, YagiharaN, Kamakura S, Horie M, Aizawa Y, Shimizu W: High prevalence ofearly repolarization in short QT syndrome. Heart Rhythm 2010;7:647–652.

33. Extramiana F, Antzelevitch C: Amplified transmural dispersionof repolarization as the basis for arrhythmogenesis in a canine

ventricular-wedge model of short-QT syndrome. Circulation2004;110:3661-3666.

34. Castro Hevia J, Antzelevitch C, Tornes Barzaga F, Dorantes SanchezM, Dorticos Balea F, Zayas Molina R, Quinones Perez MA, FayadRodrıguez Y: Tpeak-Tend and Tpeak-Tend dispersion as risk factorsfor ventricular tachycardia/ventricular fibrillation in patients with theBrugada syndrome. J Am Coll Cardiol 2006;47:1828-1834.