Electrical Storm Presages Nonsudden Death
The Antiarrhythmics Versus Implantable Defibrillators (AVID) Trial
Background—Electrical storm, multiple temporally related episodes of ventricular tachycardia (VT) or ventricular fibrillation (VF), is a frequent problem among recipients of implantable cardioverter defibrillators (ICDs). However, insufficient data exist regarding its prognostic significance.
Methods and Results—This analysis includes 457 patients who received an ICD in the Antiarrhythmics Versus Implantable Defibrillators (AVID) trial and who were followed for 31±13 months. Electrical storm was defined as ≥3 separate episodes of VT/VF within 24 hours. Characteristics and survival of patients surviving electrical storm (n=90), those with VT/VF unrelated to electrical storm (n=184), and the remaining patients (n=183) were compared. The 3 groups differed in terms of ejection fraction, index arrhythmia, revascularization status, and baseline medication use. Survival was evaluated using time-dependent Cox modeling. Electrical storm occurred 9.2±11.5 months after ICD implantation, and most episodes (86%) were due to VT. Electrical storm was a significant risk factor for subsequent death, independent of ejection fraction and other prognostic variables (relative risk [RR], 2.4; 95% confidence interval [CI], 1.3 to 4.2; P=0.003), but VT/VF unrelated to electrical storm was not (RR, 1.0; 95% CI, 0.6 to 1.7; P=0.9). The risk of death was greatest 3 months after electrical storm (RR, 5.4; 95% CI, 2.4 to 12.3; P=0.0001) and diminished beyond this time (RR, 1.9; 95% CI, 1.0 to 3.6; P=0.04).
Conclusions—Electrical storm is an important, independent marker for subsequent death among ICD recipients, particularly in the first 3 months after its occurrence. However, the development of VT/VF unrelated to electrical storm does not seem to be associated with an increased risk of subsequent death.
Compared with antiarrhythmic drugs, implantable cardioverter defibrillator (ICD) therapy reduces mortality in patients at high risk for arrhythmic death, especially those with depressed left ventricular function and ventricular fibrillation (VF) or ventricular tachycardia (VT).1 2 3 4 Recurrent arrhythmias occur in 40% to 60% of ICD recipients over the initial 3 years of follow-up, and multiple temporally related episodes of VT/VF occur in 10% to 20% of patients.5 6 However, the prognostic significance of recurrent ventricular arrhythmias,7 8 9 including multiple recurrent episodes,5 6 is unclear. Whether ICD recipients who experience electrical storm (≥3 distinct episodes of VT/VF within a 24-hour period) are at increased risk of subsequent death remains controversial. Further, it is unclear if electrical storm is a contributing factor to, or a marker of, increased mortality. Finally, whether the development of electrical storm identifies ICD recipients at a transiently higher risk of death and thus in whom therapies aimed at altering heart failure progression, recurrent ischemia, or recurrent arrhythmias should be considered10 11 12 13 14 15 16 17 18 19 is of obvious clinical significance.
The present analysis sought to determine the prognostic significance of electrical storm in patients receiving an ICD in the Antiarrhythmics Versus Implantable Defibrillators (AVID) trial. Specifically, we sought to assess the independent prognostic impact of electrical storm, both in the initial 3 months after its occurrence and long-term. The 3-month window was chosen a priori to identify whether a clinically meaningful window for intervention existed.
Patients presenting with VF, VT with syncope, or VT with hemodynamic compromise and a left ventricular ejection fraction ≤0.40 were eligible for enrollment in AVID, but those with stable VT, unexplained syncope and inducible VT, ventricular arrhythmias secondary to a correctable cause, prior VT or VF already treated with amiodarone, a limited life expectancy, New York Heart Association functional class IV symptoms, or contraindications to amiodarone or ICD therapy were not eligible.1 20 A total of 1016 patients were randomly assigned to ICD or antiarrhythmic drug therapy. Of the 507 patients assigned to receive an ICD, 457 (90%) were discharged alive from the hospital with a device that had advanced storage capability (Figure⇓). The survival of these 457 patients was evaluated through August 31, 1998.
ICD Therapies and Electrical Storm
A group of experienced cardiac electrophysiologists characterized ICD therapies as appropriate or inappropriate using stored electrogram data, R-R interval data, and clinical history.21 Electrical storm was defined a priori as the occurrence of ≥3 separate episodes of VT/VF within a 24-hour period, each separated by ≥5 minutes.6 ICD interventions included antitachycardia pacing, low-energy shocks, and high-energy shocks. Only patients with ≥3 discrete VT/VF episodes were classified as having electrical storm. Nontemporally distinct clusters of VT/VF were not categorized as electrical storm.
A time-dependent analytic method was used22 wherein patients were categorized as surviving electrical storm, surviving VT/VF unrelated to electrical storm, or as either dying during an initial episode of VT/VF or having no VT/VF during follow-up. Because we sought to compare the risk of death during each time period (before VT/VF, after non-electrical storm VT/VF, and after electrical storm), patients who died during their initial episode of VT/VF and those who did not experience VT/VF were included in the same category because they contributed follow-up time to only the period before VT/VF. Remaining patients contributed follow-up time to the period before VT/VF and to the period after electrical storm and/or the period after non-electrical storm VT/VF, depending on whether they had VT/VF related or unrelated to electrical storm.
Categorization of Deaths
Deaths were categorized as cardiac or noncardiac, and cardiac deaths were subcategorized as arrhythmic or nonarrhythmic by an experienced committee that was masked to treatment assignment.23 Given the limitations inherent in clinical definitions of arrhythmic death,23 arrhythmic deaths are referred to as “sudden” in this analysis.
Window for Intervention
A 3-month window after the development of electrical storm was chosen a priori on the basis of data from prior studies5 6 to provide a clinically meaningful intervention period. Villacastin and colleagues5 found that patients who had clusters of ICD shocks had an excess risk of cardiac death beyond 12 months of follow-up, 2 months after the mean time to the development of these clusters (10 months). Likewise, Credner et al6 found a nonsignificant excess number of deaths beyond 8 months of follow-up, 4 months after the mean time to the development of electrical storm (4 months).
Continuous baseline characteristics are presented as mean±SD and were compared among the 3 groups using ANOVA. Pairwise comparisons were performed using a Student’s t test and correcting for multiple comparisons using the Bonferroni method. Group comparisons of categorical data were evaluated using an uncorrected χ2 test. Univariate Cox proportional hazards models were used to assess the significance of baseline variables with respect to outcome.24 Multivariate Cox models were used to adjust for important prognostic variables and baseline differences. To account for the fact that the occurrence of VT/VF, including episodes of electrical storm, are postrandomization events (ie, patients need to survive long enough to be at risk), time-dependent Cox models22 were used to evaluate the prognostic significance of VT/VF and the follow-up use of antiarrhythmic drugs. The Cox models were used to determine relative risk (RR) estimates and 95% confidence intervals (CI) surrounding these estimates. Analyses were performed using SIR (Scientific Information Retrieval Inc), SPSS 6.1 (SPSS Inc), and Stata: Release 6.0 statistical software.
Characteristics of patients who survived electrical storm, those who survived VT/VF unrelated to electrical storm, and the remaining patients are shown in Table 1⇓. Patients with electrical storm were more likely to have had VT versus VF as their index arrhythmia. Patients with electrical storm or VT/VF unrelated to electrical storm had lower mean ejection fraction values, were more likely to receive an angiotensin-converting enzyme (ACE) inhibitor or digoxin at baseline, and were less likely to have a revascularization procedure after their index arrhythmia compared with the remaining patients.
Most patients (274 of 457; 60%) had ≥1 appropriate ICD therapy for VT/VF during follow-up. Therapies for VT/VF unrelated to electrical storm accounted for two-thirds of these episodes (184 of 274; 67%); 90 patients (20%) developed electrical storm.
The development of electrical storm was a relatively late phenomenon, with the initial episode occurring 9.2±11.5 months after ICD implantation. The majority of initial episodes were attributed to multiple, temporally related episodes of VT (77 of 90 initial episodes; 86%). The remaining 14% of episodes were due to VF or a combination of VT and VF. The median number of VT/VF events in the initial electrical storm episode was 4 (range, 3 to 14). A minority of episodes (5 of 90 initial episodes; 6%) were associated with syncope. No episodes required external resuscitation.
Shocks alone were used to treat 41 of the 90 initial electrical storm episodes (46%), and the remaining 49 episodes (54%) were treated with antitachycardia pacing alone (n=25) or a combination of shocks and antitachycardia pacing (n=24). The therapies used to treat these episodes (shocks versus antitachycardia pacing versus shocks and antitachycardia pacing) were similar in patients with VT versus VF as their index arrhythmia (P=0.3).
Survival and Modes of Death
There were 34 subsequent deaths among the 90 patients (38%) who survived electrical storm, 28 subsequent deaths among the 184 patients (15%) who survived VT/VF unrelated to electrical storm, and 41 deaths among the remaining 183 patients (22%). Overall, most deaths were categorized as cardiac, nonsudden (46%); noncardiac (32%), sudden (21%), or unspecified (1%) deaths accounted for the remainder of deaths. Ten of the 34 deaths (29%) in patients who survived electrical storm occurred in the first 3 months after its development, and only 4 deaths (12%) occurred in the first month after electrical storm.
Before statistical adjustment, patients surviving electrical storm were at higher risk for death compared with all other patients (RR, 3.0; 95% CI, 2.0 to 4.6; P<0.0001). This was primarily related to a higher risk for cardiac, nonsudden death (RR, 3.8; 95% CI, 2.1 to 6.9; P=0.0001). The risk of death after electrical storm was most pronounced in the initial 3 months after electrical storm (RR, 5.6; 95% CI, 2.8 to 11.5; P<0.0001). In contrast, patients surviving VT/VF unrelated to electrical storm had only a somewhat higher risk of subsequent death compared with patients not in this category (RR, 1.6; 95% CI, 1.1 to 2.5; P=0.02).
The use of antiarrhythmic drugs during follow-up (RR, 1.9; 95% CI, 1.2 to 2.8; P=0.003), advancing age (RR, 1.7 per decade; 95% CI, 1.4 to 2.1; P<0.0001), a history of diabetes (RR, 1.9; 95% CI, 1.3 to 2.8; P=0.002), and a history of heart failure (RR, 1.8; 95% CI, 1.2 to 2.6; P=0.004) were individually associated with a higher risk of death. Higher ejection fraction values (RR, 0.7 per 0.1; 95% CI, 0.6 to 0.9; P=0.0003) were associated with a significantly lower risk of death. Male versus female sex (RR, 0.9; 95% CI, 0.6 to 1.5; P=0.8) and an index arrhythmia of VF versus VT (RR, 0.9; 95% CI, 0.6 to 1.3; P=0.4) were not significantly associated with an altered risk of death.
The development of electrical storm remained associated with a subsequent risk of death, independent of age, sex, ejection fraction, history of heart failure, the follow-up use of antiarrhythmic drugs, history of diabetes, and type of index arrhythmia (RR, 2.4; 95% CI, 1.3 to 4.2; P=0.003). Electrical storm remained independently associated with a 5.4-fold increase in the risk of death in the initial 3 months after its development (Table 2⇓). After the initial 3 months, the lingering risk of death was only marginally significant. Further, the risk of death during the initial 3 months after electrical storm was significantly higher than the risk beyond 3 months (χ2=5.25 for difference between models; P=0.02). The development of VT/VF unrelated to electrical storm was not independently associated with a significant alteration in the risk of subsequent death.
A secondary sensitivity analysis that evaluated the transient risk of death in the first month after electrical storm was also performed. The independent risks of death during (RR, 3.6; 95% CI, 1.1 to 12.2) and beyond (RR, 3.8; 95% CI, 0.1.6 to 9.6) the first month after electrical storm were similar. However, given the small number of deaths (n=4) in the first month after electrical storm, this analysis had limited power.
The occurrence of multiple clustered episodes of VT/VF independently identifies defibrillator recipients at increased risk of subsequent death. Most of this excess risk was attributable to cardiac, nonsudden mechanisms. The risk of death after electrical storm is greatest in the initial 3 months. Thus, the development of electrical storm identifies ICD recipients who warrant close follow-up, consideration of antiarrhythmic therapy, and/or consideration of other therapies aimed at reducing nonsudden, presumably nonarrhythmic, mechanisms of death.
Prior studies assessing the prognostic significance of VT/VF unrelated to electrical storm have provided conflicting results.7 8 9 However, many of these studies were limited by relatively small sample sizes, an inability to distinguish between appropriate versus inappropriate shocks, insufficient adjustment for comorbid illness, or the use of inappropriate methods of analysis (ie, non–time-dependent). Further, most of these studies focused on shocks rather than any ICD therapy (ie, shocks and/or antitachycardia pacing).
In the present analysis, VT/VF unrelated to electrical storm was associated with a somewhat higher risk of death before adjustment for prognostic factors such as ejection fraction but no significant excess risk after adjustment. Thus, the development of VT/VF unrelated to electrical storm is not independently associated with an increased risk of subsequent death, but it may be a marker for comorbid illness. In contrast, the development of electrical storm was associated with a significantly elevated risk of death, both before and after adjustment for important prognostic variables. Moreover, the increased risk of subsequent death was most prominent in the first 3 months after electrical storm and approached baseline afterward.
Villacastin and colleagues5 assessed the prognostic significance of multiple, consecutive ICD discharges for VT/VF in 80 patients who were followed for 21±19 months. The 33% of patients who experienced single shocks were not at increased risk of subsequent death, but the 20% of patients surviving multiple, consecutive ICD discharges (≥2 shocks for a single arrhythmic episode) tended to have a greater risk of death (RR, 3.5; P=0.06).5 Further, the average time to the development of multiple, consecutive ICD discharges (10 months) was similar to the mean time to electrical storm in AVID (9.2 months).
More recently, Credner and colleagues6 assessed the prognostic significance of electrical storm in a group of 136 patients. Over 13±7 months of follow-up, 43 patients (32%) experienced ICD therapies unrelated to electrical storm, and 14 (10%) experienced electrical storm. Neither the patients who experienced VT/VF unrelated to electrical storm nor those who experienced electrical storm were significantly more likely to die compared with patients who did not have VT/VF during follow-up. Several important differences in their study versus the present analysis merit discussion. First, the present analysis included >3 times the number of patients and had substantially longer follow-up. Further, compared with AVID, the patients in Credner et al’s study had higher ejection fraction values (0.35±0.13), less ischemic heart disease (69%), and were more likely to receive a β-blocker (76%). Further, electrical storm occurred earlier in their patients (4±4 months) than in AVID (9.2±11.5 months). Finally, the method of analysis used in their study (ie, non–time-dependent) did not take into account the fact that patients need to survive long enough to be at risk for the development of VT/VF.22 Thus, on the basis of the strengths of the present analysis, electrical storm does seem to identify ICD recipients at higher risk of nonsudden mechanisms of death, particularly in the initial 3 months after its occurrence.
Direct Versus Indirect Role of Electrical Storm
It is unclear whether electrical storm plays an inciting, contributing, or bystander role in the observed excess mortality. Although older studies furnished conflicting evidence on the relationship between ICD shocks and myocardial injury,25 26 recent studies have shown that multiple shocks lead to elevations in cardiac troponin levels, which are indicative of minor degrees of myocardial injury,27 28 and pathological studies have demonstrated fibrosis and acute cellular injury in the hearts of patients who had received recent shocks.29 30 Recurrent episodes of VF are also associated with increased intracellular myocardial calcium levels,31 32 which in turn have been linked to progressive left ventricular dysfunction,33 cardiac apoptosis,34 and arrhythmia facilitation.32 Thus, the present analysis provides additional support to the notion that electrical storm contributes to the observed excess mortality. The similar risk of death in patients with electrical storm, before and after adjustment for other prognostic factors, further supports a causal role for electrical storm.
Window for Intervention
Because the risk of death is most prominent in the initial 3 months after electrical storm, prompt implementation or augmentation of therapies previously demonstrated to be efficacious in similar populations seems warranted. ACE inhibitors improve outcome in patients with left ventricular dysfunction,10 11 primarily via a reduction in heart failure progression. β-Blockers further reduce mortality in these patients12 13 through reductions in sudden death and heart failure progression. β-Blockers may be particularly effective in patients with electrical storm.6 Spironolactone has also been demonstrated to reduce mortality in patients with heart failure via a reduction heart failure progression.14 Coronary artery revascularization seems to be an important adjunct in patients with ischemic left ventricular dysfunction, both in terms of reducing the risk of future arrhythmias15 and in optimizing ventricular function.16 Finally, because electrical storm may be an inciting rather than a contributing factor for premature death, therapies specifically aimed at reducing recurrent arrhythmias17 18 19 may prove useful. Amiodarone and other antiarrhythmic drugs are often initiated in these patients, but there are insufficient prospective data to support or refute this approach.
Our findings are based on a retrospective analysis. Although we adjusted for potentially confounding differences, we cannot exclude the possibility that other factors may have contributed to our findings. Further, because serial ejection fraction data are not available, it is not possible to determine whether electrical storm resulted in reduced ventricular function or vice versa. Thus, a causal role for electrical storm cannot be proven. Likewise, although the prompt implementation of therapies demonstrated to be efficacious in altering heart failure progression and/or reducing the risk of future arrhythmias in other populations would be anticipated to be effective in reducing mortality in patients surviving electrical storm, confirmatory data are lacking. With these limitations in mind, the data from the present analysis provide firm evidence that patients experiencing multiple, temporally related episodes of VT/VF are at higher risk of death, particularly in the initial 3 months after electrical storm.
Apart from confirming that electrical storm is an independent risk factor for death in defibrillator recipients, the present analysis provides insights into improving their outcome. Because most of the excess mortality was attributable to nonsudden mechanisms, interventions designed to optimize ventricular function are likely to prove useful in reducing mortality. Finally, the demonstration of a discrete, 3-month window for intervention provides an opportunity to implement and evaluate established or new therapies in this population.
Supported by the National Heart, Lung, and Blood Institute (N01 HC-25117). Dr Exner is supported by the Canadian Institutes of Health and Research and the Alberta Heritage Foundation for Medical Research.
See Reference 1 for a complete list of participating investigators and sites.
- Received November 6, 2000.
- Revision received January 31, 2000.
- Accepted January 31, 2000.
- Copyright © 2001 by American Heart Association
Antiarrhythmics Versus Implantable Defibrillators (AVID) Investigators. A comparison of antiarrhythmic-drug therapy with implantable defibrillators in patients resuscitated from near-fatal ventricular arrhythmias. N Engl J Med. 1997;337:1576–1583.
Connolly SJ, Gent M, Roberts RS, et al. Canadian Implantable Defibrillator Study (CIDS): a randomized trial of the implantable cardioverter defibrillator against amiodarone. Circulation. 2000;101:1297–1302.
Moss AJ, Hall WJ, Cannom DS, et al. Improved survival with an implanted defibrillator in patients with coronary disease at high risk for ventricular arrhythmia: Multicenter Automatic Defibrillator Implantation Trial Investigators. N Engl J Med. 1996;335:1933–1940.
Buxton AE, Lee KL, Fisher JD, et al. A randomized study of the prevention of sudden death in patients with coronary artery disease: Multicenter Unsustained Tachycardia Trial Investigators N Engl J Med. 1999;341:1882–1890.
Villacastin J, Almendral J, Arenal A, et al. Incidence and clinical significance of multiple consecutive, appropriate, high-energy discharges in patients with implanted cardioverter-defibrillators. Circulation. 1996;93:753–762.
Credner SC, Klingenheben T, Mauss O, et al. Electrical storm in patients with transvenous implantable cardioverter-defibrillators: incidence, management and prognostic implications. J Am Coll Cardiol. 1998;32:1909–1915.
Levine JH, Mellits ED, Baumgardner RA, et al. Predictors of first discharge and subsequent survival in patients with automatic implantable cardioverter-defibrillators. Circulation. 1991;84:558–566.
Grimm W, Flores BT, Marchlinski FE. Shock occurrence and survival in 241 patients with implantable cardioverter-defibrillator therapy. Circulation. 1993;87:1880–1888.
Pacifico A, Ferlic LL, Cedillo-Salazar FR, et al. Shocks as predictors of survival in patients with implantable cardioverter-defibrillators. J Am Coll Cardiol. 1999;34:204–210.
SOLVD Investigators. Effect of enalapril on survival in patients with reduced left ventricular ejection fractions and congestive heart failure. N Engl J Med. 1991;325:293–302.
SOLVD Investigators. Effect of enalapril on mortality and the development of heart failure in asymptomatic patients with reduced left ventricular ejection fractions [published erratum in N Engl J Med. 1992;327:1768]. N Engl J Med. 1992;327:685–691.
MERIT-HF Investigators. Effect of metoprolol CR/XL in chronic heart failure: Metoprolol CR/XL Randomised Intervention Trial in Congestive Heart Failure (MERIT-HF). Lancet. 1999;353:2001–2007.
CIBIS-II Investigators and Committees. The Cardiac Insufficiency Bisoprolol Study II (CIBIS-II): a randomised trial. Lancet. 1999;353:9–13.
Pitt B, Zannad F, Remme WJ, et al. The effect of spironolactone on morbidity and mortality in patients with severe heart failure: Randomized Aldactone Evaluation Study Investigators. N Engl J Med. 1999;341:709–717.
O’Rourke RA. Role of myocardial revascularization in sudden cardiac death. Circulation. 1992;85:I-112–I-117.
Alderman EL, Fisher LD, Litwin P, et al. Results of coronary artery surgery in patients with poor left ventricular function (CASS). Circulation. 1983;68:785–795.
Sim I, McDonald KM, Lavori PW, et al. Quantitative overview of randomized trials of amiodarone to prevent sudden cardiac death. Circulation. 1997;96:2823–2829.
Pacifico A, Hohnloser SH, Williams JH, et al. Prevention of implantable-defibrillator shocks by treatment with sotalol. d, l-Sotalol: Implantable Cardioverter-Defibrillator Study Group. N Engl J Med. 1999;340:1855–1862.
Stevenson WG, Friedman PL, Kocovic D, et al. Radiofrequency catheter ablation of ventricular tachycardia after myocardial infarction. Circulation. 1998;98:308–314.
AVID Investigators. Antiarrhythmics Versus Implantable Defibrillators (AVID): rationale, design, and methods. Am J Cardiol. 1995;75:470–475.
Swerdlow CD, Chen PS, Kass RM, et al. Discrimination of ventricular tachycardia from sinus tachycardia and atrial fibrillation in a tiered-therapy cardioverter-defibrillator. J Am Coll Cardiol. 1994;23:1342–1355.
Fisher LD, Lin DY. Time-dependent covariates in the Cox proportional-hazards regression model. Annu Rev Public Health. 1999;20:145–157.
AVID Investigators. Causes of death in the Antiarrhythmics Versus Implantable Defibrillators (AVID) trial. J Am Coll Cardiol. 1999;34:1552–1559.
Cox DR. Regression models and life-tables. J Royal Stat Soc B. 1972;34:187–220.
Avitall B, Port S, Gal R, et al. Automatic implantable cardioverter/defibrillator discharges and acute myocardial injury. Circulation. 1990;81:1482–1487.
Allan JJ, Feld RD, Russell AA, et al. Cardiac troponin I levels are normal or minimally elevated after transthoracic cardioversion. J Am Coll Cardiol. 1997;30:1052–1056.
Hurst TM, Hinrichs M, Breidenbach C, et al. Detection of myocardial injury during transvenous implantation of automatic cardioverter-defibrillators. J Am Coll Cardiol. 1999;34:402–408.
Joglar JA, Kessler DJ, Welch PJ, et al. Effects of repeated electrical defibrillations on cardiac troponin I levels. Am J Cardiol. 1999;83:270–272.
Singer I, Hutchins GM, Mirowski M, et al. Pathologic findings related to the lead system and repeated defibrillations in patients with the automatic implantable cardioverter- defibrillator. J Am Coll Cardiol. 1987;10:382–388.
Epstein AE, Kay GN, Plumb VJ, et al. Gross and microscopic pathological changes associated with nonthoracotomy implantable defibrillator leads. Circulation. 1998;98:1517–1524.
Jones DL, Narayanan N. Defibrillation depresses heart sarcoplasmic reticulum calcium pump: a mechanism of postshock dysfunction. Am J Physiol. 1998;274:H-98–H-105.
Zaugg CE, Wu ST, Barbosa V, et al. Ventricular fibrillation-induced intracellular Ca2+ overload causes failed electrical defibrillation and post-shock reinitiation of fibrillation. J Mol Cell Cardiol. 1998;30:2183–2192.
Swynghedauw B. Molecular mechanisms of myocardial remodeling. Physiol Rev. 1999;79:215–262.
Marks AR. Intracellular calcium-release channels: regulators of cell life and death. Am J Physiol. 1997;272:H-597–H-605.