Author + information
- Received March 25, 1998
- Revision received July 16, 1998
- Accepted August 6, 1998
- Published online December 1, 1998.
- Susanne C Credner, MDa,
- Thomas Klingenheben, MDa,
- Oliver Mauss, PhDa,
- Christian Sticherling, MDa and
- Stefan H Hohnloser, MD, FACCa,*
- ↵*Address for correspondence: Dr. Stefan H. Hohnloser, Division of Cardiology, Department of Medicine, J.W. Goethe University, Theodor-Stern-Kai 7, 60590 Frankfurt, Germany
Objectives. The purpose of this study was to determine the precise incidence, therapeutic options and prognostic implications of electrical storm in patients with transvenous implantable cardioverter-defibrillator (ICD) systems.
Background. Approximately 50% to 70% of patients treated with an ICD receive appropriate device-based therapy within the first 2 years. Most arrhythmic events require only one appropriate ICD firing for termination. However, some patients receive multiple appropriate shocks during a short period of time, a condition referred to as “arrhythmic or electrical storm.”
Methods. This prospectively designed observational study comprised 136 recipients of transvenous ICDs who were followed for 403 ± 242 days. Electrical storm was defined as ventricular tachycardia or fibrillation resulting in device intervention ≥3 times during a single 24-h period.
Results. During follow-up, 57/136 patients (42%) received appropriate ICD therapy. Electrical storm occurred in 14/136 patients (10%) at an average of 133 ± 135 days after ICD implantation. The mean number of arrhythmic episodes constituting electrical storm was 17 ± 17 (range: 3 to 50; median 8) per patient. In 12 patients, electrical storm required hospital admission. The arrhythmia cluster could be terminated by a combined therapy with β-blockers and intravenous amiodarone whereas class I antiarrhythmic drugs were only occasionally successful. The cumulative probability of survival as estimated by the Kaplan-Meier method showed that patients with an episode of electrical storm did not have a worse outcome compared to those without such an event.
Conclusions. Electrical storm represents a frequent event in patients treated with modern ICDs. It occurs most commonly late after ICD implantation and can be managed by combined therapy with β-blockers and amiodarone. Electrical storm does not independently confer increased mortality.
Approximately 50% to 70% of patients treated with an implantable cardioverter-defibrillator (ICD) receive appropriate device-based therapy within the first 2 years following implantation (1,2). In the majority of cases the total number of ICD shocks remains limited and most arrhythmic events require only one appropriate ICD firing for termination. However, some patients receive multiple appropriate shocks during a short period of time. This condition has been referred to as “arrhythmic or electrical storm” (3,4). The precise definition of this syndrome is still evolving, but a useful working definition has recently been proposed in several studies evaluating the efficacy of intravenous amiodarone in recurrent life-threatening ventricular tachyarrhythmias (5,6). Electrical storm was defined as recurrent ventricular tachycardia or fibrillation occurring two or more times in a 24-h period, and usually requiring electrical cardioversion or defibrillation (5,6).
At present, only a few studies have dealt with this clinical syndrome in ICD recipients. An earlier study by Kim and associates (7)suggested an approximately 20% incidence of electrical storm, which occurred predominantly during the first few days following implantation. In this study, the use of epicardial patch electrodes with thoracotomy devices was considered to be associated with greater electrical instability compared to nonthoracotomy devices. More recently published preliminary data have suggested an overall incidence of approximately 10% to 30% incidence of electrical storm among larger series of ICD patients (8–10). At present, however, several important questions regarding this potentially life-threatening syndrome have remained unanswered. For instance, the precise time of occurrence of arrhythmia clustering after ICD implantation, appropriate clinical management of these patients, and prognostic implications of electrical storm are ill defined. Thus, the present prospective observational study aimed to answer these questions in a large cohort of consecutive patients who had received contemporary transvenous ICDs in a single institution.
Consecutive patients with one or more documented episodes of sustained ventricular tachycardia or cardiac arrest due to ventricular fibrillation were considered for the present analysis if they had received a transvenous ICD system between July 1995 and January 1998 in our institution. All but 17 patients underwent preoperative electrophysiological testing using a standardized stimulation protocol with up to three extrastimuli at two right ventricular sites. Patients routinely underwent predischarge device testing where no patient failed to defibrillate successfully with the first discharge at a safety margin of 10 Joules below maximum output. Patients were followed for up to 935 days (mean 403 ± 242 days) after implantation at 2- to 3-month intervals. Details of subsequent deterioration, clinical symptoms or death were recorded. At each follow-up visit, the ICD was interrogated to obtain its present programmed status and data on its operation.
All patients received multifunctional third- or fourth-generation ICDs using a single transvenous lead in 116, an endocardial lead in conjunction with an additional subcutaneous array electrode in 2, or an additional superior vena cava electrode in 1 patient. A dual-chamber transvenous ICD system was used in the remaining 17 patients. All implanted devices were capable of storing RR cycle lengths and electrograms obtained from the endocardial leads except 20 Ventritex devices, which exclusively store electrograms. The following ICD devices were implanted: CPI PRX 3: n = 7; CPI Mini 1741: n = 1; CPI Mini: n = 11; CPI Mini II: n = 26; CPI Mini AV: n = 1; CPI Mini AVIIDR: n = 6; CPI Mini III: n = 3; Medtronic PCD 7220: n = 8; Medtronic PCD 7221: n = 25; Medtronic PCD 7223: n = 20; Medtronic PCD 7218C: n = 1; Medtronic PCD 7271 GEMDR: n = 7; Ventritex Cadet: n = 1; Ventritex Contour: n = 11; Ventritex Contour MD: n = 4; Ventritex Angstrom: n = 4.
Assessment of baroreflex sensitivity
Baroreflex sensitivity (BRS) is a marker of the capability to reflexly increase vagal activity and to decrease sympathetic activity in response to a sudden increase in blood pressure (11). In the present study, BRS was determined at the time of ICD implantation using the phenylephrine method devised by Smyth and co-workers (12). This method uses the intravenous (IV) injection of small doses of phenylephrine to increase arterial pressure. Changes in blood pressure and heart rate are recorded on a beat-by-beat basis and plotted against each other. The slope of this regression line (expressed as msec of increase in RR interval per mm Hg rise in pressure) allows quantification of the sensitivity of arterial baroreflex control of heart rate. Three consecutive measurements were performed, and the corresponding slopes were averaged to reduce measurement variability.
Definition of electrical storm
Because no generally accepted definition of electrical storm exists, this event was defined for the purpose of this analysis as the occurrence of ventricular tachycardia or fibrillation resulting in device intervention (antitachycardia pacing and/or shock delivery) three or more times within a 24-h period. For each arrhythmia episode, the appropriateness of ICD therapy was verified by device interrogation as described above.
All data were prospectively entered in a customized data base and subsequently analyzed using Statistical package for the Social Sciences (13). Measures are reported as mean ± SD. Patient characteristics were compared across the three patient groups using a Mantel-Haenszel test for categorical variables and analysis of variance for continuous variables. Individual groups were then compared using the Scheffe test (14). The cumulative probability of survival was determined by the Kaplan-Meier method (15), and differences in survival between groups were evaluated with the log-rank test (16). A p value of <0.05 was considered statistically significant.
This study is based on analysis of data from 136 consecutive patients previously resuscitated from cardiac arrest or with a history of sustained ventricular tachycardia who subsequently were treated with a multifunctional transvenous ICD system at our institution (Table 1). There were 110 men and 26 women with a mean age of 61 ± 14 years (range 23 to 79 years). The majority of patients suffered from coronary artery disease or from idiopathic dilated cardiomyopathy. Left ventricular ejection fraction averaged 36 ± 14% (range 15 to 74%). Thirty-seven percent of patients had a history of ventricular fibrillation and 53% of ventricular tachycardia. Both ventricular fibrillation and tachycardia had been previously documented in the remaining 13 patients (10%). There was no perioperative mortality among these patients.
During follow-up (403 ± 242 days), 57 patients (42%) received at least one appropriate ICD therapy. The time from implantation to the first device therapy averaged 124 ± 145 days.
Fourteen of 136 consecutive patients (10%) (Table 2)experienced an episode of electrical storm during follow-up. This arrhythmia clustering occurred at an average of 133 ± 135 days after ICD implantation. Following the first electrical storm episode, three patients subsequently presented with a second episode. Table 3shows the comparison of clinical characteristics of patients with and without episodes of electrical storm. Except the difference in age, this analysis did not reveal any other significant difference.
Careful analysis of ICD-stored electrograms demonstrated that electrical storm consisted of the occurrence of repeated episodes of ventricular tachycardia in 10 and ventricular fibrillation in 4 patients. The mean number of arrhythmic episodes constituting electrical storm was 17 ± 17 (range: 3 to 50; median 8) per patient. The ICD interrogation revealed no evidence for device malfunctioning or inappropriate shock delivery for supraventricular arrhythmias. Ten patients experienced syncope or presyncope during the event, whereas no hemodynamic compromise was apparent in four patients. Three patients had minor symptoms such as palpitations or light-headedness.
In seven patients, electrical storm represented the first arrhythmic episode resulting in appropriate ICD therapy after implantation. Seven patients had had prior episodes of ventricular tachyarrhythmias terminated by their ICD before they developed electrical storm.
Precipitating factors for electrical storm could be identified in only five patients (26%). Hypokalemia (serum potassium ≤3.5 mmol/l) was present in three patients. Another patient developed acute myocardial ischemia resulting in subsequent recurrent myocardial infarction 2 days after cessation of electrical storm. One patient developed acute congestive heart failure and was treated with IV diuretics at which time he developed electrical storm despite normal serum potassium levels. In the remaining patients, careful analysis of symptoms, electrocardiograms (ECGs), and laboratory values revealed no apparent triggering factors for the electrical storm.
Electrical storm and vagal reflex activity
The BRS was assessed in 105/136 patients; BRS could not be assessed due to atrial fibrillation in 15 patients and to pacemaker dependence in 7 patients. In five patients, BRS could not be analyzed owing to technical problems, whereas four patients declined to undergo BRS assessment. The BRS averaged 4.5 ± 5.0 msec/mm Hg for the total patient cohort and was significantly different across the three patient groups (p = 0.05). Patients without ICD interventions during follow-up had a better preserved BRS (5.4 ± 6.1 msec/mm Hg) than did patients with appropriate ICD therapy (3.1 ± 2.4 msec/mm Hg; p = 0.08). Patients with electrical storm had a similarly depressed BRS of 3.1 ± 2.2 msec/mm Hg; owing to the smaller number of patients in this group, however, this difference did not reach the level of statistical significance.
Management of electrical storm
Nine of 12 patients with electrical storm requiring in-hospital pharmacological therapy were treated in our institution according to a prespecified therapeutic regimen. Antiadrenergic therapy by means of intensifying or starting β-blocker treatment was commenced by IV administration of 2.5 to 10 mg metoprolol. This was followed by IV administration of a class I antiarrhythmic substance (ajmaline 0.5 to 1.0 mg/kg). When electrical storm did not subside following these therapeutic measures, IV administration of amiodarone was initiated (bolus infusion of 300 mg for 1 h followed by 1.2 g/24 h). Whereas the combined therapy with β-blocker and the class I drug terminated the arrhythmia cluster in only three patients, all other patients could be stabilized during the infusion of amiodarone. Except in two patients, this averaged 10 ± 9 h (range 1 to 24 h; median 3.5 h). In the remaining two patients, amiodarone infusion had to be continued for 133 and 432 h, respectively, until no further breakthrough arrhythmia was observed. All of these patients continued to receive oral therapy with amiodarone at the time of hospital discharge.
Two patients without hemodynamic compromise or only minor symptoms during their arrhythmia cluster did not consult a physician; accordingly, their arrhythmic episodes and consecutive ICD interventions were only detected on the next regular follow-up visit in the outpatient clinic. Because electrical storm had subsided at the time of their presentation, no changes in concomitant therapy were deemed necessary.
Prognostic implications of electrical storm
Patients with electrical storm were followed in the arrhythmia outpatient clinic for the remaining observation period of 306 ± 215 days (range: 38 to 626 days) after cessation of the arrhythmia cluster. Two patients had to be hospitalized again for another episode of electrical storm 13 and 115 days after their first event, respectively. Eventually, both cases could be managed successfully. Two patients with a previous episode of electrical storm died 257 and 182 days after cessation of electrical storm due to progressive heart failure. In the group of patients with only single appropriate ICD interventions (n = 43; follow-up 500 ± 245 days), two patients died of progressive heart failure and one died of cancer. There were three deaths, all due to progressive heart failure, in the group of patients without any ICD intervention during an average follow-up of 352 ± 234 days. Figure 1shows the cumulative probability of survival of these three patient groups from the time of ICD implantation. As demonstrated, there was no significant difference in survival for patients with electrical storm compared to those without electrical storm during their clinical course.
The present study reveals several new findings. It demonstrates that a substantial proportion of patients treated with modern ICD devices experience at least one episode of multiple adequate ICD therapies during a single 24-h period. In this series, 14/136 patients (10%) showed such an episode of electrical storm. These 14 patients represent one-fourth of all patients who received at least one appropriate ICD therapy during follow-up. Electrical storm required in-hospital therapy in 12/14 patients (86%), thus significantly impacting on ICD-related treatment costs. However, extended follow-up of these patients demonstrated that electrical storm does not independently confer increased mortality.
Definition of electrical storm
At present, there is no commonly agreed definition of electrical storm. It represents a serious but treatable clinical syndrome of recurrent severe ventricular arrhythmias most commonly observed in patients with advanced structural heart disease. A useful working definition has been recently proposed in two prospective studies (5,6)where electrical storm was defined as recurrent ventricular tachycardia or ventricular fibrillation occurring more than two times in a 24-h period. Moreover, it usually requires electrical cardioversion or defibrillation (5,6).
This entity should be differentiated from so-called incessant ventricular tachycardia. Although there has been no consistency in the definition of this term as well, incessant ventricular tachycardia implies that a patient is in monomorphic ventricular tachycardia for a significant portion of the day. For instance, Lemery et al. (17)have defined it as the presence of ventricular tachycardia for at least one half of each of 3 days during a continuous observation period. A similar definition implying the presence of sustained and nonsustained tachycardia resulting in total ventricular ectopic beats in 24 h more than total sinus beats has been used by Hariman and associates (18).
Incidence and time of occurrence of electrical storm
There is only sparse data in the literature concerning the incidence of electrical storm in ICD populations as defined in the present study. Most of these data stem from patient samples evaluated in the late 1980s or early 1990s, implying that most of these patients had been treated with thoracotomy ICD devices. For instance, O’Donoghue and co-workers (9)reported on 130 ICD recipients of whom 12 (9%) showed at least one episode of electrical storm defined as 10 or more appropriate ICD discharges during a 48-h period. More recently, a preliminary report (10)demonstrated an incidence of 23% (35/149 patients) of cases of electrical storm; although not specified in the paper, it can be assumed that many of those patients had received modern nonthoracotomy ICD devices. These numbers are in excellent agreement with the present finding of a 10% incidence of electrical storm in patients exclusively treated with modern nonthoracotomy ICDs.
Another important question relates to the precise time of occurrence of electrical storm after ICD implantation. In the beginning of widespread ICD therapy, most episodes of increased electrical instability were reported to occur in the immediate postoperative period (7). For instance, Kim and associates (7)observed 9/52 patients who developed recurrent ventricular tachycardias requiring intervention during the immediate postoperative phase after implantation of thoracotomy ICDs. The same authors noted postoperative unstable ventricular tachycardia in 4/59 recipients of a nonthoracotomy device. Accordingly, these authors concluded that an exacerbation of ventricular arrhythmia is common during the immediate postoperative phase, particularly in patients undergoing thoracotomy for placement of epicardial patch electrodes. Unfortunately, there is a lack of data concerning the occurrence of electrical storm during follow-up in such patients.
The present observations indicate that by far the majority of electrical storm episodes occur late after ICD implantation. In fact, of the 14 episodes, only 3 were noted during the first 48 h after surgery. In the remaining 11 patients, electrical storm occurred on average 169 ± 130 days after ICD placement. This figure is in accordance with the preliminary data provided by Greene et al. (10), who reported an average of 259 ± 353 days for their patient population. These findings clearly indicate that electrical storm in recipients of modern nonthoracotomy ICD devices is not related to surgical factors such as mechanical irritation or inflammatory processes. Moreover, the fact that electrical storm represented the first episode of appropriate ICD therapy delivery in 7/57 cases (12%) in this series indicates that electrical storm reflects an episode of enhanced electrical instability that can occur at any time in individuals with a history of life-threatening ventricular tachyarrhythmias.
Precipitating factors for electrical storm
A number of well-known precipitating factors significantly increase the electrical instability of the heart. Among them, myocardial ischemia is probably one of the most important factors. In this series, there was one patient in whom myocardial ischemia was proven to be the trigger for electrical instability. This patient eventually developed recurrent myocardial infarction several days following the cessation of electrical storm and died of extensive left ventricular damage as a result. Other triggering factors include the development of electrolyte disturbances such as profound hypokalemia, which was demonstrated in three of our patients. Finally, episodes of acute congestive heart failure with resulting increase in sympathetic tone may give rise to enhanced electrical instability upon which drug treatment for congestive heart failure often acts to further increase electrolyte disturbances. Proarrhythmic side effects of antiarrhythmic drugs may also constitute a precipitating factor for recurrent ventricular tachycardia or fibrillation.
Our results also provide new data on the role of the autonomic nervous system in the genesis of electrical storm. Several previous reports have demonstrated that vagal reflex activity is depressed in patients with life-threatening ventricular arrhythmias (19–21). The present study demonstrates for the first time a diminished BRS in patients with compared to those without appropriate ICD interventions during follow-up. This result extends previous findings (19–21)and further corroborates the decisive role of the autonomic nervous system in the genesis of sudden cardiac death (22). The second important aspect of our analysis of BRS is that patients with electrical storm do not differ in terms of their capability to reflexly increase vagal activity from individuals with only single ICD interventions. Thus, it is intriguing to speculate that further impairment in the ability to activate vagal reflexes appears not to be a prerequisite for triggering electrical storm in ICD recipients.
Prognostic implications of electrical storm
One of the most important aspects of the current study relates to potential prognostic implications of electrical storm. At present, there is a paucity of information on this clinically important issue. A recent report of Villacastin and co-workers (23)on 80 ICD recipients indicated that multiple consecutive high-energy ICD discharges were an independent predictor of subsequent cardiac and arrhythmic mortality. On first glance, this appears to be in contrast to our findings, which demonstrate that outcome of patients with multiple discharges was similar to that of individuals with only single appropriate ICD discharges or no ICD therapy at all during follow-up. There are two explanations for this apparent discrepancy. In the study of Villacastin et al. (23), patients with electrical storm had a significantly worse left ventricular function (LVEF 26 ± 4%) compared to patients with single appropriate ICD discharges (39 ± 3%) or no ICD therapy (43 ± 2%). Because the degree of left ventricular dysfunction is the most important single prognostic factor in all patients with a history of life-threatening ventricular arrhythmias (24–26), occurrence of electrical storm in the study of Villacastin et al. (23)may be considered an epiphenomen of the impairment of LVEF.
In contrast, in the present study there was no significant difference in the degree of left ventricular dysfunction in patients with or without electrical storm. Accordingly, the subsequent clinical course of these patients was similar. The second explanation is related to the fact that in the study of Villacastin et al. (23), another definition for electrical storm was used; these investigators examined the prognostic value of repeated ICD shocks delivered for termination of one arrhythmia episode, unlike in our study where the occurrence of multiple arrhythmia episodes within a single 24-h time period was evaluated. Moreover, our findings are supported by preliminary data of Greene and co-workers (10), which also failed to demonstrate prognostic implications of electrical storm.
Acute management of electrical storm
The majority of patients with electrical storm in this series who were in the need of in-hospital treatment were admitted to the same institution and were managed by adhering to a prospectively defined therapeutic regimen. This fact allows some comments on acute handling of such patients despite the uncontrolled and nonrandomized nature of antiarrhythmic drug therapy. In all patients, attention was paid to correct potential trigger factors such as electrolyte disturbances before specific antiarrhythmic drug therapy was instituted. According to the hypothesis that myocardial ischemia and an increase in sympathetic tone may be major precipitating factors, therapy with β-adrenergic blocking substances was initiated or intensified whenever possible. The importance of antiadrenergic therapy has recently been illustrated in a patient who had 76 appropriate ICD shocks for recurrent ventricular fibrillation and could be successfully managed by IV metoprolol (4). Nademanee and Singh reported a similar experience some years ago (27).
Finally, when specific antiarrhythmic drug therapy was necessary due to recurrence of arrhythmic episodes (mean number of events constituting electrical storm was 17 ± 17), IV administration of amiodarone was initiated. Utilizing this drug regimen, the majority of patients could be stabilized within a relatively short period of time (median 3.5 h). This experience is in agreement with recent reports on the utility of IV amiodarone in the treatment of recurrent life-threatening ventricular arrhythmias (5,6,28). Furthermore, amiodarone is not only effective in suppressing recurrent ventricular tachycardia or fibrillation but also carries a low proarrhythmic potential (29).
Electrical storm occurs frequently in ICD recipients even when contemporary nonthoracotomy devices are used. In many instances, electrical storm represents the first arrhythmic event after ICD implantation resulting in appropriate device therapy. The capability to reflexly increase vagal tone is diminished both in patients with single ICD interventions and with electrical storm to a similar extent. In patients with frequent ICD shocks, IV amiodarone therapy is effective in suppressing further arrhythmic episodes. As indicated by the results of this study, electrical storm is not necessarily a harbinger of poor prognosis in these patients.
The authors wish to express their gratitude to H. Ackermann, PhD, Department of Biostatistics, J.W. Goethe University, Frankfurt, for his expertise in helping to perform the statistical analysis of this data.
- Received March 25, 1998.
- Revision received July 16, 1998.
- Accepted August 6, 1998.
- American College of Cardiology
- Block M,
- Breithardt G
- Zipes DP, Roberts D, for the Pacemaker-Cardioverter-Defibrillator Investigators
- Tavernier R,
- Derycke E,
- Jordaens L
- Kowey P.R,
- Levine J.H,
- Herre J.M,
- et al.
- Scheinman M.M,
- Levine J.H,
- Cannom D.S,
- et al.
- Korte T,
- Jung W,
- Ostermann G,
- et al.
- O’Donoghue S,
- Platia E.V,
- Waclawski S,
- Mispireta L
- Greene M,
- Geist M,
- Paquette M,
- et al.
- Smyth H.S,
- Sleight P,
- Pickering G.W
- ↵(1988) SPSS-X User’s Guide (SPSS, Chicago), 3rd ed.
- Scheffe H
- Prentice R.L
- Lemery R,
- Brugada P,
- Della Bella P,
- Dugernire T,
- van den Dool A,
- Wellens H.J.J
- Hohnloser S.H,
- Klingenheben T,
- van de Loo A,
- et al.
- Landolina M,
- Mantica M,
- Pessano P,
- et al.
- Schwartz P.J,
- La Rovere M.T,
- Vanoli E
- Villacastin J,
- Almendral J,
- Arenal A,
- et al.
- Powell A.C,
- Fuchs T,
- Finkelstein D.M,
- et al.