Author + information
- Received December 8, 2011
- Revision received February 2, 2012
- Accepted February 14, 2012
- Published online May 29, 2012.
- Yoshifusa Aizawa, MD⁎,⁎ (, )
- Akinori Sato, MD⁎,
- Hiroshi Watanabe, MD⁎,
- Masaomi Chinushi, MD⁎,
- Hiroshi Furushima, MD⁎,
- Minoru Horie, MD†,
- Yoshiaki Kaneko, MD‡,
- Tsutomu Imaizumi, MD§,
- Kimie Okubo, MD∥,
- Ichiro Watanabe, MD∥,
- Tsuyoshi Shinozaki, MD¶,
- Yoshiyasu Aizawa, MD#,
- Keiichi Fukuda, MD#,
- Kunitake Joo, MD⁎⁎ and
- Michel Haissaguerre, MD††
- ↵⁎Reprint requests and correspondence:
Dr. Yoshifusa Aizawa, Niigata University Graduate School of Medical and Dental Science, 1-754 Asahimachi-dori, Niigata 951-8510, Japan
Objectives This study evaluated the pause-dependency of the J-wave to characterize this phenomenon in idiopathic ventricular fibrillation (VF).
Background The J-wave can be found in apparently healthy subjects and in patients at risk for sudden cardiac death, and risk stratification is therefore needed.
Methods Forty patients with J-wave–associated idiopathic VF were studied for J waves with special reference concerning pause-dependent augmentation. J waves were defined as those ≥0.1 mV above the isoelectric line and were compared with 76 non-VF patients of comparable age and sex.
Results The J-wave was larger in patients with idiopathic VF than in the controls: 0.360 ± 0.181 mV versus 0.192 ± 0.064 mV (p = 0.0011). J waves were augmented during storms of VF (n = 9 [22.5%]), which was controlled by isoproterenol; they disappeared within weeks in 5 patients. In addition, sudden prolongation of the R-R interval was observed in 27 patients induced by benign arrhythmia, and 15 patients (55.6%) demonstrated pause-dependent augmentation (from 0.391 ± 0.126 mV to 0.549 ± 0.220 mV; p < 0.0001). In the other 12 experimental subjects and in the 76 control subjects, J waves remained unchanged. Pause-dependent augmentation of J waves was detected in 55.6% (sensitivity) but was specific (100%) in the patients with idiopathic VF with high positive (100%) and negative (86.4%) predictive values.
Conclusions Pause-dependent augmentation of J waves was confirmed in about one-half of the patients with idiopathic VF after sudden R-R prolongation. Such dynamicity of J waves was specific to idiopathic VF and may be used for risk stratification.
Early repolarization (ER) is defined as a slur or notch on the terminal part of the QRS complex with or without elevation of the ST-segment and is frequently observed in apparently healthy subjects (1–3). The prognosis of subjects with ER has been considered to be benign (4,5). However, J waves have been observed in association with idiopathic ventricular fibrillation (VF) (6,7), and recent studies have confirmed that ER is associated with idiopathic VF (8–10).
In population-based studies, Tikkanen et al. (11) and Haruta et al. (12) demonstrated that ER is a statistically significant risk for arrhythmic death, and a J-wave of a large amplitude (11) or a J-wave with a flat (horizontal or descending) ST-segment was shown to be a risk factor for sudden cardiac death (13). This risk was proven in cases of idiopathic VF (14). However, electrocardiogram (ECG) features that are able to distinguish “malignant” from “benign” J waves are still necessary for risk stratification.
Since our first reports of the association of the J-wave with idiopathic VF (6,7,15), we have studied pause-dependent augmentation of the J-wave and have been collecting case data regarding idiopathic VF. In this study, we analyzed the pause-induced dynamicity of the J-wave in patients with J-wave–associated idiopathic VF and compared this with control subjects to propose another characteristic of J waves in idiopathic VF.
Since 1992, we have collected data on 40 patients with J-wave–associated VF from 9 institutions, mainly from the Niigata University Hospital (Niigata, Japan). All of the patients met the following inclusion criteria for idiopathic VF: 1) documented episode of VF at the time of cardiac arrest; 2) absence of structural heart disease with normal cardiac function; 3) negative serological test result for inflammatory diseases; and 4) absence of coronary artery disease and a negative provocative test result for coronary spasms.
Patients with bundle branch block, intraventricular conduction delay, long or short QT interval (16,17), Brugada syndrome (18), or Wolff-Parkinson-White syndrome (19) were excluded. Pilsicainide was given to exclude Brugada syndrome, and coronary spasms in patients were excluded by a provocation test using acetylcholine or ergonovine maleate.
J waves were defined as: 1) notches or slurs at the terminal portion of the QRS complexes; and 2) amplitude ≥0.1 mV above the isoelectric line in at least 2 contiguous leads. The location was classified as inferior (II, III, or aVF), left precordial (V4 to V6), right precordial (V1 to V3), or high lateral (I or aVL) sites. The amplitudes of J waves were measured after 5-fold magnification in the leads to reveal maximal amplitude, by 2 cardiologists who were blinded to the clinical findings (19).
To investigate the instantaneous dynamicity of J waves, the amplitude of the J-wave was measured in the beat immediately after a pause and compared with the mean J-wave amplitude measured in the 2 to 3 beats preceding the pause (Fig. 1). A pause represented sudden prolongation of the R-R interval that was induced by benign arrhythmias such as sinus arrest, sinoatrial block, atrioventricular block, or atrial or ventricular premature beats. If possible, the J-wave amplitude was measured in the beat after the pause to identify temporary changes. Concomitant changes in the ST- and T-wave morphology with J-wave augmentation were analyzed.
J waves were observed after admission until discharge, and if VF developed in storms, isoproterenol was given.
As the control, the dynamicity of J waves was analyzed in 76 subjects who had J waves in the 12-lead ECG. They visited our hospitals for cardiac or noncardiac diseases but had no syncope or symptoms suggestive of serious arrhythmias such as ventricular tachycardia or VF. None had a family history of sudden cardiac death. Heart failure (New York Heart Association functional class >II) or organic heart diseases were excluded by ECG and echocardiography as well as clinical history. Other exclusion criteria were the same as in the experimental group. The dynamicity of J waves was analyzed on the standard ECGs or 12-lead Holter ECGs.
Patients were divided into 2 groups according to the presence of pauses. In the patients with pauses, the dynamicity of J waves and concomitant changes in the ST-segment were evaluated. The amplitudes of J waves were compared among the pre-, post-, and the beat next to the post-pause (Fig. 1). Temporary changes of the J waves were observed to the time of discharge. When VF recurred, the effects of isoproterenol were evaluated. Finally, the sensitivity, specificity, and predictive values of the pause-dependent J-wave augmentation were calculated.
Numerical values are presented as mean ± SD, and categorical variables are expressed as absolute numbers or percentages. The differences between groups were analyzed by using Wilcoxon or Mann-Whitney-Wilcoxon tests for continuous variables and the Pearson's chi-square test for categorical variables. Statistical analyses were performed with SPSS version 12.0 (SPSS Inc., Chicago, Illinois). A 2-sided p < 0.05 was considered statistically significant.
The study was approved by the ethics committee of Niigata University School of Medicine.
J-wave in idiopathic VF
Forty patients displayed J waves: slurs or notches ≥0.1 mV in ≥2 contiguous leads. The mean age of the patients was 38 ± 14 years, and 37 (92.5%) were males. The QT and QTc intervals were all within normal ranges: 384 ± 25 ms and 401 ± 40 ms1/2. The mean J-wave amplitude was 0.360 ± 0.181 mV. The J waves were located in the inferior region in 28 (70.0%), left precordial region in 19 (47.5%), right precordial region in 4 (10.0%), and high lateral region in 9 (22.5%) patients. Twenty (50.0%) patients exhibited J waves at >1 site (Table 1). Brugada syndrome was excluded by ECGs in all patients and by drug testing in 32 patients.
Pause-dependent changes in J waves could be analyzed in 27 (67.5%) of the 40 patients who experienced sudden prolongation of the R-R interval due to arrhythmias, and J-wave accentuation was observed in 9 (22.5%) patients before VF episodes. Isoproterenol was effective in controlling VF (Fig. 2). In 5 patients (12.5%), J waves disappeared within weeks. Of these, 3 patients had exhibited no J waves in the ECGs recorded 3 to 6 months previously. VF occurred between 8:00 pm and 6:00 am in 26 (65.0%) patients, between 6:00 am and 8:00 pm in 12 (30.0%) patients, and at both time intervals in 2 patients (5.0%). In the other patients (5.0%), VF developed during exercise in the daytime.
Pause-dependency of the J-wave
Among these 27 patients with pauses by benign arrhythmias, 15 (55.6%) demonstrated significant augmentation of the J waves, as shown in Table 2 and Figures 3 and 4: from 0.391 ± 0.126 mV to 0.549 ± 0.220 mV (p < 0.0001); the R-R interval was prolonged suddenly from 802 ± 204 ms to 1,450 ± 572 ms (p < 0.0001). The changes in J-wave amplitude were 0.185 ± 0.129 mV and ranged from 0.05 to 0.43 mV (0.5 to 4.3 mm).
The amplitude of the J waves in the beat next to the post-pause beat was measureable in 6 of 15 patients and was smaller than those of the baseline J waves as well as the augmented J waves: 0.325 ± 0.092 mV (p = 0.0406 and p = 0.0065, respectively). When J waves were augmented, the ST-segment was depressed from 0.10 ± 0.39 mV at baseline to –0.24 ± 0.53 mV after pauses (p = 0.0015). VF occurred in 5 (33.3%) of 15 patients in storms after a short-long sequence. Isoproterenol was effective in controlling VF (Fig. 2).
In the remaining 12 (44.4%) of 27 patients, the J-wave remained unchanged (<0.05 mV), as shown in Figure 3: 0.192 ± 0.079 mV versus 0.196 ± 0.080 mV (p = 0.8377) when the R-R interval was prolonged from 809 ± 137 ms to 1,156 ± 175 ms, as summarized in Table 2 (p < 0.0001). In 4 (33.3%) patients, VF developed and was controlled by isoproterenol.
The patients with pause-dependent augmentation of the J-wave amplitude were similar in age, sex, and J-wave locations to those without (Table 2). The baseline R-R intervals and their changes were similar between the 2 groups, but the pre- and post–J-wave amplitudes were larger in the patients with pause-dependent augmentation of the J-wave compared with those without (p < 0.0001).
In the 76 control subjects, sex and age were comparable to the 40 patients (Table 1). The locations of J waves were as follows: 65 (85.5%) in the inferior region, 25 (33.9%) in the left precordial region, 10 (13.2%) in the right precordial region, 12 (15.8%) in the high lateral region, and 36 (47.4%) at >1 site. The distribution pattern did not differ between the 2 groups.
The baseline R-R interval and the J-wave amplitude were different from the patient group (Table 1). When the R-R interval was prolonged from 941 ± 138 ms to 1,352 ± 342 ms by arrhythmias (n = 17), there was no augmentation of the J-wave amplitude (Figs. 3 and 4).
Sensitivity, specificity, and predictive values
Pause-dependent augmentation of the J waves was observed in 15 of 27 VF patients with a sensitivity of 55.6%, or 37.5% of the original 40 patients with J waves. Pause-dependent augmentation of J waves was observed only in patients with idiopathic VF. Both the specificity and the positive predictive values were 100%: the negative predictive value was 86.4% of the 27 patients, or 75.2% of the original 40 patients. The presence of pause-dependent augmentation of J waves was highly diagnostic for idiopathic VF.
Fifty-four patients with idiopathic VF were admitted to Niigata University Hospital, and J waves were observed in 24 (44.4 %); another 16 similar patients were recruited from 8 other institutions.
Of the 40 patients, we were able to assess the instantaneous dynamicity of J waves after pauses in 27 patients (67.5%), and pause-dependent augmentation of the J waves was observed in 15 (55.6%) of 27 patients but not in any of the control patients. Augmentation of J waves was associated with depression of the ST-segment or inversion of the T-wave, and the beats just after the post-pause beat revealed attenuated J waves. The pause-dependent augmentation of J waves was highly specific and had highly predictive value. Effects of isoproterenol were reconfirmed.
The age and sex of the patients with idiopathic VF were similar to those reported previously (8–10). Prevalence of the J-wave was also similar to that reported by earlier researchers varying from 31% to 65%. This prevalence was higher than that of the Japanese (11% to 27%) (19) compared with the non-Japanese control subjects (5.0% to 34%) (5,8–12). Because of the frequent presence of J waves in the general population, it is urgent to establish a standard to distinguish “benign J waves” from “malignant” ones. So far, several characteristics of J waves involving idiopathic VF have been reported.
Extensive location or a large amplitude of J waves was statistically shown to be a risk factor for arrhythmic death (11); however, overlapping was remarkable between the patients with idiopathic VF and the control.
In addition, marked fluctuation of the J-wave might be observed in idiopathic VF especially before VF episodes (8,10,20–22), and the J-wave might develop new or disappear over time. This seems to be in contrast to ER or J waves in healthy individuals: ≥80% subjects exhibited the same type of J-wave after an interim of 5 years (11,23,24). Whether attenuation of the J-wave amplitude by isoproterenol or quinidine is specific to J waves with idiopathic VF needs be established (8,10,25–27).
Another striking feature of the J-wave in patients with idiopathic VF is an accentuation of J waves after a sudden prolongation of the R-R interval: that is, the pause dependency. This phenomenon was described in our first report in 1992 (7). Since then, data from similar cases have been collected. Among the 27 patients who were able to be analyzed with respect to pause-dependency, the pause-dependent augmentation of J waves was confirmed in 37.5% (15 of 40) patients with J waves, but such pause-dependent augmentation of J waves was not observed in any of the non-VF subjects. Notably, we would be able to provoke pauses by programmed stimulation during electrophysiological study and analyze the pause-dependency (15,28).
To date, only a few such cases have been reported (22,25). Although rare, a striking pause-dependent augmentation of J waves in the inferolateral leads has been observed in a patient with Brugada syndrome (29). During the follow-up of 2 years, he developed electrical storms from VF.
The presence of early repolarization with a horizontal/descending ST-segment was found to be able to predict arrhythmic death in a large population study (13). The presence of J waves was associated with a history of idiopathic VF with an odds ratio of 4.0, but the combination of J waves and a horizontal/descending ST-segment yielded an odds ratio of 13.8 for patients with idiopathic VF (14). In the present study, the ST-segment or T waves became more negative when J waves were accentuated after a pause (Fig. 2). Regarding the discordant relationship between the J-wave amplitude and ST-segment, delayed epicardial repolarization causing the epicardium to repolarize after endocardium seems to be responsible (30,31).
Experimentally, J waves are well explained by the transmural voltage gradient of the myocardial cells in phase 1 of the action potential of myocardial cells, which is created by transient outward currents (Ito) (32,33). Accentuated Ito is shown to result in accentuation of J waves on the surface ECG and Ito is known to be augmented at a slower rate. The bradycardia-dependent augmentation of the J-wave amplitude in idiopathic VF could be well explained by Ito (32,33). Delayed activation due to phase 4 block (block at a slow rate) may mimic J waves appearing at a slower rate and can be differentiated from the J-wave (34,35). Phase 4 block can be affected by the recovery characteristics of Ito, which permit a greater action potential overshoot and amplitude, and therefore a greater source current early in diastole (36).
At this point, we may be able to characterize the J-wave associated with idiopathic VF as follows: 1) large amplitude (often >0.2 mV) (11); 2) recent appearance (7,8); 3) remarkable fluctuation without any apparent cause (8,10); 4) extensive distribution (8); 5) response to isoproterenol or quinidine (8,10,24–27); 6) a concomitant horizontal/descending ST-segment (13,14); and 7) pause-dependent augmentation (7).
This was a small case series, and our conclusions must be confirmed in a larger number of patients. However, this study represents 20 years of experience. A worldwide survey is needed for further elucidation.
The response of the J-wave was analyzed in one-half of the patients because some did not have their situation complicated by benign arrhythmias. However, the pause-dependency of J waves might be evaluated in a systematic manner during electrophysiological study after giving electrical stimuli to either the atrium or the ventricle, inducing pauses (29). The causes of the waxing and waning nature of the J-wave were not determined. The possible role of a latent pathological process such as myocarditis as the cause of the J-wave appearing for weeks or longer should be excluded.
Finally, genetic abnormalities have been noted in some patients with idiopathic VF (37–40), but the present study did not include genetic screening. A systematic survey is of great importance in this respect.
Regarding the dynamicity of the J-wave in idiopathic VF, the pause-dependent augmentation was highly specific with high predictive values. This simple phenomenon may be used for the risk stratification of J waves.
This work was partly supported by a grant from the Daiwa Health Foundation, Japan. The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- early repolarization
- ventricular fibrillation
- Received December 8, 2011.
- Revision received February 2, 2012.
- Accepted February 14, 2012.
- American College of Cardiology Foundation
- Rosso R.,
- Kogan E.,
- Belhassen B.,
- et al.
- Haruta D.,
- Matsuo K.,
- Tsuneto A.,
- et al.
- Tikkanen J.T.,
- Junttila M.J.,
- Anttonen O.,
- et al.
- Rosso R.,
- Glickson E.,
- Belhassen B.,
- et al.
- Moss A.J.,
- Schwartz P.J.,
- Crampton R.S.,
- Locati E.,
- Carleen E.
- Brugada P.,
- Brugada J.
- Nam G.B.,
- Ko K.H.,
- Kim J.,
- et al.
- Haïssaguerre M.,
- Sacher F.,
- Nogami A.,
- et al.
- Yan G.X.,
- Antzelevitch C.
- Elizari M.V.,
- Nau G.J.,
- Levl P.J.,
- Lazzar J.O.,
- Halpern M.S.,
- Rosenbaum M.B.
- Litovsky S.H.,
- Antzelevitch C.