Idiopathic Ventricular Fibrillation Induced With Vagal Activity in Patients Without Obvious Heart Disease
Background Recently, idiopathic ventricular fibrillation (VF) has gained much attention. Although several subgroups have been described, its pathogenesis, mechanism, treatment, and prognosis remain unknown.
Methods and Results We studied six cases of idiopathic VF with transient late r′ waves and ST elevation (late r′/ST elevation) in leads V1 through V3. Late r′/ST elevation was augmented before and after VF episodes. Signal-averaged ECGs showed late potentials even when no late r′/ST elevation occurred. During late r′, a conduction delay was observed by use of body-surface maps at the anterior wall and outflow tract of the right ventricle without inhomogeneity of the repolarization phase. There was a decrease or total disappearance of late r′/ST elevation with isoproterenol, atropine, and exercise stress testing and induction or exacerbation with propranolol, edrophonium, and hyperventilation. VF was induced by programmed electrical stimulation in all cases but two, in which it was induced only after edrophonium injection. In two cases, VF was exacerbated by propranolol, and in all cases, it was uninducible with isoproterenol. Heart rate spectral analysis just before VF episodes showed a sudden rise in vagal activity in two cases. As the VF mechanism, a conduction delay exists at the anterior wall and outflow tract of the right ventricle that is possibly exacerbated by an abrupt rise in vagal activity, inducing random reentry that results in VF. Class I antiarrhythmic agents and β-blockers were ineffective for this VF. All subjects required implantable cardioverter-defibrillators.
Conclusions We propose this VF associated with late r′/ST elevation in the precordial leads and influenced by vagal activity as a new possible mechanism of idiopathic VF.
In recent years, idiopathic VF in patients without obvious underlying heart disease has gained much attention.1 2 3 4 5 6 7 8 Its incidence in Europe and the United States is considered to account for ≈1% of all out-of-hospital VF resuscitations, 3% to 9% of VF cases unrelated to myocardial infarction, and ≈14% of all VF resuscitations in patients under the age of 40. Therefore, idiopathic VF is not necessarily a rare occurrence, and it has become urgent to elucidate its pathogenesis, mechanism, treatment, and prognosis. Brugada and Brugada1 proposed VF cases that exhibited RBBB and persistent ST elevation in leads V1 through V3 as a new subtype of idiopathic VF, and Leenhardt et al2 proposed a new concept of torsade de pointes not exhibiting QT-interval prolongation and that arises from premature ventricular beats with short coupling intervals. However, the mechanism of idiopathic VF remains unclear. Therefore, we propose this newly described mechanism of idiopathic VF influenced by vagal activity in patients associated with transient late r′ waves and ST elevation in the V1 through V3 leads as described by Brugada and Brugada.9 In the present study, we will elucidate the characteristic ECG findings, clinical features, and mechanism involved and clarify the significance of these for treatment.
Of 11 idiopathic VF subjects, we selected 6 with late r′ waves and ST elevation transiently observed in leads V1 through V3 (which were not observed in the 5 subjects excluded from the study) in whom, under the conditions in which the episodes occurred, induction of VF appeared to be influenced by vagal activity. There were 5 men and 1 woman ranging in age from 22 to 58 years old (average, 39 years). Two case subjects had a family history of sudden death.
Idiopathic VF was defined as VF with no QTc prolongation (QTc <0.44 ms) and no underlying heart disease determined by 12-lead ECG, chest radiography, echocardiography, nuclear medicine examination, cardiac catheterization studies, coronary angiography, left ventriculography, right ventriculography, or myocardial biopsy. Furthermore, spontaneous episodes of sustained monomorphic VT or those induced by programmed electrical stimulation were excluded. Twelve-lead ECG, Holter ECG, exercise stress testing, body-surface mapping, signal-averaged ECG, and EPS were performed, and the mechanism of the VF was analyzed. A Corazonix Predictor was used for signal-averaged ECG recording based on methods described by Simson10 and Denes et al.11 The ECG was recorded during sinus rhythm using Frank X, Y, Z corrected orthogonal leads. Signals from 200 to 300 beats were amplified, digitized, averaged, and then filtered with a bidirectional filter with a high band-pass frequency of 40 Hz. The f-QRS, RMS40, and LAS40 were calculated via a computer algorithm. A late potential was considered to be positive when two of the three items (f-QRS ≥114 ms, RMS40 <20 μV, or LAS ≥38 ms) from the Kuchar diagnostic standards were met.
Body-surface ECGs were simultaneously recorded from 87 electrodes by use of a VCM-3000 (Fukuda Denshi Co) recorder with Wilson’s central terminal as the reference during sinus rhythm. The leads were amplified by 53.7 dB, filtered, multiplexed, sampled at a rate of one sample per millisecond per channel for 8 seconds, digitized with a 12-bit analog-digital convertor, and stored in a 1.5-megabyte memory buffer.
Every 1 ms from the QRS-wave onset, QRS-T isopotential maps were obtained. The zero potential line that first caved in the positive area was defined as breakthrough.
By estimating the duration from the QRS-wave onset to the peak of the R wave, isochrone maps were computed, and the QRS-wave onset was determined with the use of Frank X, Y, Z leads and the spatial magnitude. The peak of the R wave was specified as the peak of the R (r) wave when only an R (r) wave was present or the peak of the R′ (r′) wave when both R (r) and R′ (r′) waves were present. The duration from the QRS onset to the onset of the QS wave was measured when there was a QS pattern and the R wave was absent in the lead.
The time integral of the QRST wave was calculated in each lead as the algebraic sum of all potentials from the QRS-wave onset to the T-wave offset and was displayed in microvolt·seconds for the QRST isointegral maps. To compare the similarity of two QRST isointegral maps, correlation coefficients were obtained by use of Yamada’s method12 between the two maps.
Twenty-four-hour Holter ECGs were recorded and tapes reviewed with the Marquette Laser SXP Holter Analysis System. Fluctuations in 2-minute consecutive RR intervals were converted into time-base–connected functions, which were then processed via fast Fourier transformation after being resampled to 469-ms equi-intervals, ie, 256 points. Resulting power spectra were segmented into two components comprising an LF component (0.04 to 0.15 Hz) and an HF component (0.15 to 0.40 Hz). Each was quantified for display as the square root of the area under the respective power spectrum. Hourly and 2-minute averages of LF, HF, and LF/HF ratio were calculated.
Programmed electrical stimulation was performed with a specifically designed programmable stimulator (San-ei Sokki).13 The electrical stimulation protocol consisted of single, double, and triple extrastimuli during paced rhythm with basic cycle lengths of 600, 500, and 400 ms; rapid pacing; and burst pacing. The site of stimulation was the RVA and the RVOT. Antiarrhythmic drugs were discontinued for a period of four drug half-lives before the test. After control studies were conducted, serial tests were performed after administration of the following: procainamide 10 to 20 mg/kg IV, disopyramide 2 mg/kg IV, mexiletine 2 mg/kg IV, propafenone 2 mg/kg IV, flecainide 2 mg/kg IV, propranolol 0.1 mg/kg IV, amiodarone 400 mg PO for 2 weeks, E-4031 6 to 9 μg/kg IV, verapamil 0.2 mg/kg IV, bepridil 4 mg/kg PO for 2 weeks, isoproterenol 0.01 to 0.02 μg·kg−1·min−1 IV, atropine 0.2 mg/kg IV, and edrophonium 10 mg IV.
In the control study, VF was induced and the drug was considered effective if VF was uninducible after drug administration or the polymorphic VT spontaneously terminated within 15 consecutive beats. We considered a drug to be ineffective when VF was inducible by use of the same stimulation protocol as before drug administration. We evaluated the exacerbation effect of the drugs by determining that a drug decreased by one the number of stimuli needed to induce VF. In the control study, if VF could not be induced, the drug was considered to have an exacerbation effect if VF could be induced by programmed electrical stimulation or if spontaneous episodes occurred after drug administration.
The automatic ICDs used were the Ventak-P 1600(CPI) in four cases and the PCD 7217 (Medtronic) in two cases. The average follow-up period was 63.5 months and ranged from 59 to 70 months.
Data are presented as mean±SD. Data are compared by use of the paired t test. Data were considered to be significant at a value of P<.05.
Episodes of VF
The age at which VF was confirmed ranged from 18 to 50 years (average, 35.5 years), and the duration from the first syncopal attack to the time of referral for admission to our facility ranged from 0.3 to 8 years (average, 3.6 years). Past history of cardiac resuscitations ranged from 1 to 4 times (average, 2.3). All subjects had a history of syncopal episodes, which were observed frequently in three subjects. The conditions under which the episodes occurred were during sleep or at rest in five subjects and during medical examination and during urination in one. During daily activities such as driving a car or gardening, episodes were observed in only three cases and were never observed in any of the patients during exercise (Table 1⇓).
In all cases, normal sinus rhythm exhibited a narrow QRS complex, and the QTc was 0.37 to 0.43 ms (average, 0.40 ms). Late r′ waves and ST elevation were transiently observed in the right precordial leads, occurring at random, which made it impossible to predict at what time of day they would occur. No consistency was observed concerning when they would appear in each subject or between subjects, but they were always present just before and/or immediately after VF episodes. Fig 1⇓ shows the late r′ waves and ST elevation fluctuating hourly and daily without any consistency over time in subject 1. With signal-averaged ECG, late potentials were positive in all subjects when late r′ waves were present. In four of the five subjects, late potentials were positive even when no late r′ waves were present (Fig 2A⇓, for subject 1). In the remaining subject, signal-averaged ECG was not performed; therefore, the presence of late potentials could not be determined in that subject. When the values when late r′ waves were present were compared with those when they were not, the f-QRS and LAS40 were significantly (P<.05) prolonged and the RMS40 was decreased (Fig 3⇓). When body-surface mapping was used in three cases and the appearance time of breakthrough (as determined via isopotential mapping) during the absence of late r′ waves was compared with that during the presence of late r′ waves, the latter was slightly prolonged by an average of 6.3 ms (range, 5 to 7 ms) (Fig 4⇓). However, in the V1 lead, the QRS duration was clearly prolonged by an average of 41.7 ms (range, 33 to 50 ms) (Fig 4⇓). When isochrone maps when late r′ waves were not present were compared with those when they were present, the former exhibited a normal distribution, and when a late r′ wave was present, excitation was delayed, extending from the anterior chest to a point more medially superior. The latest points were observed at line 5 and between rows E, D, and F, respectively (Fig 2B⇓, for subject 1). The isointegral maps recorded both during the absence and presence of late r′ waves displayed single dipole characteristics with correlation factors of .974, .964, and .992, respectively (Fig 2C⇓, for subject 1). Five subjects underwent treadmill testing, and the late r′ waves and ST elevation were inhibited in all. With Holter ECG monitoring, PVCs were observed in three subjects, of whom one had 2118 beats/d and the other two had only a few. However, in three other subjects, no PVCs whatsoever were observed. VF was recorded in three subjects. The QTc just before VF was 0.36 to 0.42 ms, and the coupling interval of the first beat of the VF was 360 to 400 ms. Although a fluctuation in late r′ waves and ST elevation was observed in the NASA lead (a lead similar to V1 used by NASA on space flights that is a bipolar lead with the negative electrode located at the superior edge of the sternum and the positive electrode at the inferior edge) in subject 1 throughout the day, late r′ waves and ST elevation were always observed just before the VF episodes (Fig 5⇓).
VF was inducible in the baseline state in four of six patients, with double extrastimuli from the RVA in two and the RVOT in one and triple ventricular extrastimuli at the RVOT in one. However, the reproducibility of VF induction was small, and VF was only inducible on the examination days closest to the spontaneous episodes. In the remaining two subjects, VF induction was not possible even with triple premature stimulation at either the RVA or RVOT and was only possible with triple stimulation at the RVOT after administering edrophonium 10 mg IV (Table 2⇓). The coupling interval of the last premature stimulus administered ranged from 170 to 210 ms (average, 191 ms), and all episodes of VF converted to sinus rhythm with defibrillation. In one subject, fragmentation after the ventricular potential (V wave) of the intracardiac electrogram at the septal side of the RVOT and a presystolic potential just before the V wave during polymorphic VT induced at the same site were observed. In the other five cases, no fragmentation or delayed potentials after the V wave were observed. In six subjects, MAP recordings were performed at >3 different locations, but no humps or MAP dispersions of the MAP duration were observed in any case.
Coronary Angiography, Cardiac Catheterization Studies, and Myocardial Biopsy
Coronary angiography was performed on all subjects, with normal results in all cases. In one case, coronary artery spasms (via left coronary artery injection, the left anterior descending artery was 100% occluded and was associated with ST elevation in V1 through V5, and via right coronary artery injection, the right posterior descending artery was 90% occluded and was associated with ST elevation in leads III and aVF) were induced by intracoronary artery acetylcholine injection (20 to 60 μg), resulting in ST elevation and chest pain; however, VF was not induced. Coronary artery spasms were not inducible by acetylcholine (60 to 100 μg) loading in the other five subjects. All subjects exhibited normal results per right and left ventriculography. Myocardial biopsy was performed in all subjects, and no abnormal findings were observed.
Role of the ANS
We investigated the role of the ANS in transient late r′ waves and ST elevation and in VF induction and spontaneous VF episodes by investigating the influence of ANS agonists and antagonists on late r′ waves and ST elevation in five subjects (this was not performed in subject 6). Isoproterenol caused inhibition of late r′ waves and ST elevation in all five subjects; propranolol caused an exacerbation in two and no change in three; edrophonium caused an exacerbation in four and no change in one; atropine caused inhibition in three and no change in two; and hyperventilation caused an exacerbation in three, no change in one, and was not attempted in one other subject (Table 3⇓). Fig 6⇓ shows the serial ECGs of the ECG changes observed after administering isoproterenol, propranolol, edrophonium, and atropine and with hyperventilation testing in subject 1.
During Holter ECG monitoring when no VF episodes occurred, the level of ANS activity, as determined by the hourly HF and LF/HF ratio values over a 24-hour period, was within normal limits for all subjects according to age and sex by use of the method described by Matsuda et al.14 In subject 1, three VF episodes were recorded during the first Holter monitoring session and none during the second session. The hourly HF and LF/HF ratios in the first and second Holter sessions were within normal limits (within an average value of ±1 SD for the HF value of normal subjects matched by age and sex14 ) (Fig 7A⇓). However, as shown in Fig 7B⇓ and 7C⇓ for the first VF episode, the HF and LF/HF ratio values recorded every 2 minutes before the VF episodes showed an abrupt rise in the HF (Fig 7B⇓) and a decrease in the LF/HF ratio for the VF episode (Fig 7C⇓). The second and third VF episodes in subject 1 and the three episodes in subject 4 showed a tendency similar to this.
In the control group, in the four subjects in whom VF was induced by VF induction testing during EPS, VF was not inducible with continuous intravenous infusion of 0.01 to 0.02 μg·kg−1·min−1 isoproterenol. In one subject, VF was exacerbated by intravenous injection of 4 mg of propranolol. In this case, in the control state, VF was inducible with triple extrastimuli; however, after the administration of propranolol, it was inducible only with single extrastimuli. VF that was uninducible in the control state became inducible after the intravenous injection of edrophonium in two subjects.
Evaluation of the Efficacy of Antiarrhythmic Medications
In regard to efficacy in preventing spontaneous episodes of VF, the class I antiarrhythmic drugs (disopyramide, procainamide, and mexiletine) were ineffective in four subjects, as evidenced by the spontaneous occurrence of VF. The class II antiarrhythmic drug propranolol caused an exacerbation in one subject. In subject 6, VF occurred 5 days after the administration of propranolol was begun. Before propranolol was used, that subject had no episodes for 6 months, and it was considered that the propranolol caused an exacerbation and induced VF. The class IV antiarrhythmics verapamil and diltiazem were ineffective in two subjects. In the drug efficacy evaluation via programmed stimulation, disopyramide was ineffective in two subjects, as evidenced by the induction of VF with the same induction protocol used before drug administration; mexiletine was ineffective in one subject and caused an exacerbation in one, and propranolol caused an exacerbation in another, as evidenced by the induction protocol being decreased from triple extrastimuli before administration of mexiletine or propranolol to single stimulation after administration; flecainide was ineffective in one subject; and procainamide was ineffective in one subject (Table 2⇑). All subjects underwent ICD implantation and are presently alive.
Although Viskin and Belhassen3 reviewed 45 subjects with idiopathic VF, they observed abnormal findings of incomplete RBBB, intraventricular conduction defects, left-axis deviation, and nonspecific T-wave abnormalities in the abnormal ECG findings at rest in only 5 subjects. In 1992, Brugada and Brugada1 proposed a form of idiopathic VF in subjects in whom RBBB and persistent ST elevation (>0.1 mV) in the right precordial leads were observed in the ECG findings during periods free from VF episodes. Recently, they reported cases with an intermittent pattern of ST-segment elevation change.9 Our subjects fall into a subgroup of Brugada’s, as evidenced by intermittent late r′ waves and ST elevation in the right precordial leads. However, the mechanism of the Brugada-type idiopathic VF is not yet clear. In 1994, Leenhardt et al2 proposed a new concept of torsade de pointes that arises from a PVC with a short coupling interval (<300 ms) and in which there is no QT prolongation observed in the ECG findings just before the event. For all our subjects, the coupling interval just before VF occurrence, emphasized by Leenhardt et al,2 was >300 ms. The mechanism of VF that we observed differs from the mechanism described by Leenhardt et al,2 which suggested a greater depression of vagal activity than of sympathetic activity as a potential mechanism of VF. Therefore, we will now discuss the mechanism, especially with regard to the influence of vagal activity, and treatment of this idiopathic VF associated with transient late r′ waves and ST elevation induced by that vagal activity.
The average age at onset of VF was 34.6 years old, and males predominated, with a male to female ratio of 5:1. Among the blood relations of two of the subjects, sudden death was observed in two. In the eight subjects studied by Brugada and Brugada,1 two brothers were included among the subjects, and sudden death was observed in the blood relatives of two other subjects, suggesting a genetically related cause.
In our study, transient late r′ waves and ST elevation were observed in leads V1 through V3 in six subjects and either appeared or increased while the VF episodes were occurring. In particular, in two cases in which VF events were recorded during Holter ECG recordings, an increase in the late r′ waves and ST elevation was observed just before the VF and decreased or disappeared during periods in which no events occurred. Therefore, a relationship is suggested between the late r′ waves and ST elevation and the onset of the VF. Goldberger15 described two types of ST-elevation patterns in the right precordial leads. They are the coved and saddle-back types. In our study, we also observed the coved pattern in leads V1 through V3, depending on the subject. On the Holter ECG just before the VF episodes, we observed the saddle-back pattern in the NASA lead. In this case, the ST-elevation pattern in the NASA lead transformed from a coved pattern to a saddle-back pattern just before the VF episode. Furthermore, both patterns were observed simultaneously in different leads in the same patient and in the same lead over time. Therefore, it appears that the type of pattern (coved or saddle-back) was not an essential factor in Brugada’s syndrome. Although the NASA lead has limitations, one must consider the continuous spectrum from the coved pattern to the saddle-back pattern that is observed in the precordial leads.
The mechanism involved is thought to be a delayed localized abnormality or an abnormality during the repolarization process. According to Morace et al,16 late r′ waves can be observed in youths and disappear with sympathetic β-stimulation or exercise workloads. This may be due to the fact that during rest, there is inhomogeneity in the repolarization process of the left ventricular anterior and posterior walls, which disappears when accelerated sympathetic activity exists. Recently, Miyazaki et al17 showed that ST elevation was augmented by selective stimulation of α-adrenoceptors or muscarine receptors or by class IA drugs but was mitigated by β-adrenoceptor stimulation or α-adrenoceptor blockade. They proposed that those responses might be explained by the presence of an area of early repolarization or a delayed localized activation area in the ventricle causing ST-segment elevation in cases of Brugada syndrome. Although they suggested in that study that the ST elevation was caused by an early repolarization or by a delayed localized activation area, they failed to determine which of these theories was most likely to be the true cause. In the present study, by signal-averaged ECG and body-surface mapping, we were able to show that the late r′ waves and ST elevation were a result of a delayed localized activation area. Furthermore, although the study by Miyazaki et al17 showed the effects of ANS activity on ST elevation, it did not show the effects of vagal activity on the induction of VF, as did the present study. The present study suggests, regardless of the 12-lead ECG findings, that an intraventricular conduction delay exists and that the degree of delay fluctuates during the day and from day to day, because late potentials were positive via signal-averaged ECG even when late r′ waves and ST elevation were not present. Additional interesting findings are suggested from body-surface maps recorded during periods in which late r′ waves are and are not present. The breakthrough appearance time on the isopotential map represents conduction time in the main trunk of the RBB. Regardless of the fact that the width of the QRS was significantly prolonged, the delay in breakthrough appearance is mild during periods when late r′ waves are present. Therefore, the conduction delay is thought to be mainly a delay of excitation propagation further distal to the RBB. It is suggested by isochrone mapping that during the presence of late r′ waves, the excitation in the RVOT is remarkably delayed. The QRST integral map does not rely on the depolarization phase and solely reflects the repolarization phase. In our subjects, the repolarization phase did not change during periods when late r′ waves and ST elevation were present or absent, and no inhomogeneity was observed in the repolarization phase. Therefore, the late r′ wave and ST elevation reflect a conduction delay localized in the RV anterior wall and RVOT, and it is thought that an abnormality in the repolarization process may not play a role in this.
Although a delayed potential was observed in one subject on the septal side of the RVOT by endocardial mapping, absolutely no abnormal potentials were noted in five subjects. This may be due to incomplete mapping of the endocardium; however, we assume that it is due to the conduction delay being located toward the epicardium.
VF was induced in all the subjects with programmed stimulation and exhibited a transition from polymorphic VT to VF, and monomorphic VT was not inducible in any subject. Furthermore, VF was not inducible in four subjects at the RVA but was inducible when the same premature stimulation was used at the RVOT. These results suggest the possibility that random reentry is occurring at the RV anterior wall and RVOT.
When the effects of ANS agonists and exercise stress testing on late r′ waves and ST elevation were viewed, the late r′ waves and ST elevation decreased or disappeared with intravenous injection of isoproterenol and exercise stress testing and were either induced or exacerbated with intravenous injection of propranolol. This is consistent with the results of Morace et al.16 Furthermore, in our study, intravenous injection of the parasympathomimetic agent edrophonium and hyperventilation resulted in an induction or exacerbation of the late r′ waves and ST elevation, whereas intravenous injection of the parasympatholytic agent atropine caused a decrease or disappearance. This suggests that the contribution of the ANS, in particular the increase of vagal activity and the decrease in sympathetic activity, plays a role in the late r′ waves and ST elevation.
Much attention has been drawn to the role of the ANS as a trigger for idiopathic VF. Viskin and Belhassen3 state that ventricular arrhythmias are exacerbated by psychological and physical stress, and because β-blockers are effective, Brugada and Brugada1 state that this indicates the role of the ANS. In the present study, the contribution of the ANS, especially the PNS, was suggested by the fact that in all subjects, the VF episodes occurred at times such as during sleep, rest, drinking, and urinating and never occurred during exercise. Therefore, we analyzed the role of the ANS by observing the effects of ANS agonists/antagonists on VF induction via programmed electrical stimulation and the activity level of the ANS via Holter recordings. In four subjects with inducible VF by programmed stimulation in the control state, VF was not inducible after the administration of the sympathomimetic agent isoproterenol, and in one of three subjects, VF induction was exacerbated by intravenous injection of the β-blocker propranolol. In two subjects in whom VF was uninducible in the control state, VF became inducible with the parasympathomimetic agent edrophonium. These results suggest the possibility that in VF induction, the PNS has an augmenting effect and the sympathetic nervous system has an inhibitory effect.
From the evaluation of the spectral analysis of RR intervals via Holter monitoring, no difference was observed in the 24-hour activity level of the ANS in the absence of VF events from that observed in age- and sex-matched normal subjects, and thus it was believed that no abnormality in the function of the ANS existed. From Holter ECG nonspectral analysis and power spectral analysis in the absence of events, Leenhardt et al2 observed abnormalities in the ANS that consisted of a decrease in vagal activity greater than that during sympathetic activity. In two of our subjects, we recorded episodes of VF during Holter recording, and for the 30-minute period just before each of those events, we analyzed the level of ANS activity every 2 minutes. For each of those episodes, the HF for the 2-minute period just before the episode abruptly increased and the LF/HF ratio clearly decreased. These results suggest that the sudden rise in vagal activity plays an important role in the triggering of VF. It is also possible that it may not be due solely to vagal activity but may also be dependent on sympathetic activity occurring at the same time. During such circumstances, VF would be the most inducible. Sato et al18 performed a similar analysis on one idiopathic VF subject and observed a relative increase in the function of the sympathetic nervous system as a relative decrease in the function of the PNS. Their cases have completely opposite presentations from ours, and this suggests the complexity of the role played by the ANS in idiopathic VF. Although reports evaluating the activity level of the ANS just before VF events are rare at present, it is believed that important findings will be attained with the accumulation of other such cases in the future.
From the results obtained in our six subjects, the possibility is suggested that the mechanism involved consists of a conduction delay existing at the anterior wall and RVOT and that this conduction delay is aggravated by accelerated vagal activity. With additional accelerated vagal activity, random reentry might be induced, leading to VF. However, reports to date have indicated the importance of accelerated sympathetic activity, and we believe that from the viewpoint of the role of the ANS, our subjects represent a new entity. We propose that the mechanism of VF involved in our subjects is one in which the functional conduction delay is mainly located from the anterior wall to the septal region of the RVOT but may also be located epicardially. Furthermore, it is an exacerbation of the conduction delay by the ANS, in particular the PNS, that triggers the VF. Recently, Brugada and Brugada9 described idiopathic VF cases with transient RBBB. It is possible that some of those cases may have been induced by vagal activity, as was experienced in our subjects.
Myocardial disease and coronary vasospasms as the pathology for idiopathic VF have been reported. Martini et al4 observed a pathological abnormality of the RV via myocardial biopsy and autopsy, indicating a role played by RV myocardial disease. In our six subjects, no unique findings were obtained from myocardial biopsy. However, from the positive late potentials on signal-averaged ECG when no late r′ waves could be found and from the isochrone maps obtained during late r′ wave episodes, it is suggested that pathological myocardial tissue may be localized in the RVOT, which may possibly play a role. Therefore, this suggests the possibility of our cases being in an early stage of arrhythmogenic RV dysplasia with a very localized lesion, similar to the conduction delay distal to the RBB described by Fontaine et al.19
From the ST elevation observed in the right anterior chest leads and the inducibility of VF during hyperventilation, it can be considered that coronary vasospasms may play a role. However, in five subjects, induction testing via intracoronary arterial injection of acetylcholine proved to be negative, and although coronary vasospasms accompanied by chest pain and ST elevation were induced in one subject, VF was uninducible. This suggests the possibility that factors such as hyperventilation that influence coronary artery tone play a role in VF episodes without being mediated by coronary vasospasms.
Mortality in idiopathic VF has not been clearly elucidated; however, in the review by Viskin and Belhassen,3 of 45 subjects, 11% had recurrence, and in the review by Priori et al,5 death or ICD shocks were observed in 12 of 100 subjects. Thus, treatment for idiopathic VF is a major concern. According to Viskin and Belhassen,3 in 15 subjects to whom class Ia antiarrhythmic drugs were administered, there were no episodes of sudden death or syncope, whereas in those dosed with β-blockers or amiodarone, VF recurrences were observed. Belhassen et al6 emphasized that class Ia drugs were effective in electrophysiological drug evaluation and in long-term dosing. According to Leenhardt et al,2 verapamil was very effective in the short term, but with long-term dosing, 5 of 14 subjects had sudden death episodes. Also, β-blockers are reported to be effective in some cases. In contrast, in our five subjects, class I drugs were ineffective in all, and class Ib drugs caused exacerbations in two of four. Furthermore, it should be noted that in two subjects, β-blockers caused an exacerbation, meaning that not only does the idiopathic VF that we propose express unique ECG findings that elucidate its mechanism, but this idiopathic VF also has great significance for its treatment. Hojo et al20 reported an interesting idiopathic VF case in which long-term dosing of sympathomimetic agents was effective. We believe it should be noted that although β-blockers are becoming contraindicated for idiopathic VF, there may still be cases in which sympathomimetic agents and parasympatholytic agents could possibly be indicated.
In our study, we were able to clearly show that within idiopathic VF, there is not only catecholamine-induced VF but also vagal activity–induced VF. However, because we studied only a small number of subjects and because it remains to be clarified whether this mechanism applies to all cases with Brugada-type VF and/or cases with other types of idiopathic VF, additional research is necessary to confirm the clinical significance and prevalence of vagally induced VF.
Some cases of idiopathic VF are induced by a sympathomimetic state and therefore can be prevented by β-blockers. On the other hand, our cases, which fall into a subgroup of Brugada syndrome, were induced by vagal activity and exacerbated by propranolol. Therefore, for the treatment of idiopathic VF, we recommend that idiopathic VF subjects be classified according to the mechanism of induction, especially whether it is induced by vagal activity or by a sympathomimetic tone (catecholamine-induced VF).
We propose this mechanism of idiopathic VF induced by vagal activity and associated with transient late r′ waves and ST elevation in the precordial leads as a new mechanism of idiopathic VF.
Selected Abbreviations and Acronyms
|ANS||=||autonomic nervous system|
|f-QRS||=||total filtered QRS duration|
|LAS||=||duration of the low-amplitude electric potential component (<40 μV) of the terminal portion of the QRS complex|
|MAP||=||monophasic action potential|
|PNS||=||parasympathetic nervous system|
|PVC||=||premature ventricular complex|
|RBB||=||right bundle branch|
|RBBB||=||right bundle-branch block|
|RMS40||=||root mean square voltage of the 40-ms terminal portion of the QRS complex|
|RV||=||right ventricle, right ventricular|
|RVA||=||right ventricular apex|
|RVOT||=||right ventricular outflow tract|
We would like to thank John Martin for his linguistic advice and assistance.
- Received August 21, 1996.
- Revision received November 4, 1996.
- Accepted November 23, 1996.
- Copyright © 1997 by American Heart Association
Leenhardt A, Glaser E, Burguera M, Nurnberg M, Maison-Blanche P, Coumel P. Short-coupled variant of torsade de pointes: a new electrocardiographic entity in the spectrum of idiopathic ventricular tachyarrhythmias. Circulation. 1994;89:206-215.
Priori SG, Borggrefe M, Camm AJ, Hauer RN, Klein H, Kuck KH, Schwartz PJ, Touboul P, Wellens HJJ. Unexplained cardiac arrest: the need for a prospective registry. Eur Heart J. 1992;13:1445-1446.
Belhassen B, Shapiral I, Shoshani D, Paredes A, Miller H, Laniado S. Idiopathic ventricular fibrillation: inducibility and beneficial effects of class I antiarrhythmic agents. Circulation. 1987;75:809-816.
Wellens HJJ, Lemery R, Smeets JL, Brugada P, Gorgels AP, Cheriex EC, de Zwaan C. Sudden arrhythmic death without overt heart disease. Circulation. 1992;85(suppl I):I-95-I-97.
Simson MB. Use of signals in the terminal QRS complex to identify patients with ventricular tachycardia after myocardial infarction. Circulation. 1981;64:235-242.
Denes P, Santarelli P, Hauser RG, Uretz EF. Quantitative analysis of the high-frequency components of the terminal portion of the body surface QRS in normal subjects and in patients with ventricular tachycardia. Circulation. 1983;67:1129-1138.
Matsuda N, Kasanuki H, Yamada T, Hosoda S. Power spectral analysis of heart rate variability using ambulatory electrocardiogram in normal subjects. Ther Res. 1991;12:3144-3149.
Goldberger AL. Myocardial Infarction: Electrocardiographic Differential Diagnosis. St Louis, Mo: Mosby Yearbook; 1991:200.
Sato N, Akasaka K, Kawashima E, Ishii Y, Ogawa Y, Matsuhashi H, Kawamura Y, Yamashita H, Kikuchi K, Ohe T, Ohmiya H, Maruyama J. A case of idiopathic ventricular fibrillation with possible mechanism of autonomic dysfunction [in Japanese]. Jpn J Electrocardiol. 1994;14:206-217.
Fontaine G, Frank R, Guiraudon G, Pavie A, Tereau Y, Chomette G, Grosgogeat Y. Signification des troubles de conduction intraventriculaires observés dans la dysplasie ventriculaire droite arhythmogène. Arch Mal Coeur. 1984;77:872-879.
Hojo Y, Yamasawa M, Ichida M, Ebata H, Ikeda U, Nakayama T, Shimada K. Polymorphic recurrent ventricular tachycardia with bizarre QRS wave prevented by sympathomimetic agents [in Japanese]. Shinzoh. 1994;26:540-545.