Association Between Ease of Suppression of Ventricular Arrhythmia and Survival
Background We tested the hypothesis that patients whose ventricular arrhythmias are easy to suppress have a lower rate of arrhythmic death, defined as arrhythmic death and nonfatal cardiac arrest, the primary end point in the Cardiac Arrhythmia Suppression Trials (CAST-I and CAST-II), than patients whose ventricular arrhythmias are hard to suppress. In addition, we evaluated the association between ease of suppression of ventricular arrhythmias and mortality of all causes.
Methods and Results CAST-I investigated the effect on arrhythmic death of ventricular premature depolarization (VPD) suppression achieved by three drugs, encainide, flecainide, and moricizine, at two different dose levels; CAST-II investigated the same effect, using moricizine alone at three dose levels. If suppression was achieved, patients were randomized to the effective active drug or corresponding placebo. To examine the independence of easily suppressed ventricular arrhythmias as a predictor of arrhythmic death, we adjusted statistically for other variables that were related both to ease of suppression and arrhythmic death. Patients with ventricular arrhythmias (n=1778) that were easy to suppress had fewer arrhythmic deaths during follow-up than those with ventricular arrhythmias that were hard to suppress (n=1173) (relative risk, .59; P=.003). Patients whose VPDs were easily suppressed were older and had a lower frequency of prior history of heart failure and myocardial infarction. They also had a higher incidence of anterior myocardial infarction, VPD frequency, and average ejection fraction. After adjusting for these variables, we found that easily suppressed ventricular arrhythmias were still significant predictors of arrhythmic death (relative risk, .66; P=.013).
Conclusions This study shows that the ease of VPD suppression identifies a subgroup of postmyocardial infarction patients who have low risk of arrhythmic death.
Ventricular arrhythmias are associated with increased mortality in patients who have experienced an acute myocardial infarction independent of left ventricular dysfunction.1 2 3 This finding led to the hypothesis that suppression of postinfarction ventricular arrhythmias with antiarrhythmic drugs would reduce mortality. Uncontrolled studies that assessed drug efficacy using Holter electrocardiography or electrophysiological testing suggested that suppression of ventricular arrhythmias is associated with improved survival.4 5 These studies also suggested that the suppression of ventricular arrhythmias was responsible for the improved survival. Contrary to this view, randomized trials of antiarrhythmic drugs have not shown an association between suppression of ventricular arrhythmias and improved survival.6 7 8 9 The Cardiac Arrhythmia Suppression Trial (CAST)8 9 10 11 randomized patients whose ventricular arrhythmias were suppressed by encainide, flecainide, or moricizine. Despite ventricular arrhythmia suppression, none of these drugs improved survival during long-term treatment.
Previous studies12 13 14 15 suggest that ventricular arrhythmia suppression is more easily achieved in patients who do not have heart failure or who have normal or only moderate decrease in ejection fraction. We were interested in investigating whether or not ease of suppression of itself could identify, in a group of post–myocardial infarction patients with prognostically important ventricular arrhythmias, a subset with low rates of arrhythmic death. The drug/dose titration strategy of CAST-I and CAST-II provided a unique opportunity to address this question.
Our primary objective was to test the hypothesis that patients whose ventricular arrhythmias are easy to suppress have a lower rate of arrhythmic death, defined as arrhythmic death and nonfatal cardiac arrest, the primary end point in CAST, than patients whose ventricular arrhythmias are hard to suppress. To examine the independence of easily suppressed ventricular arrhythmias as a predictor of arrhythmic death, we adjusted statistically for other variables that were related to both ease of suppression and arrhythmic death. In addition, we evaluated the association between ease of suppression of ventricular arrhythmias and mortality of all causes. Finally, we looked for a dose-response relation between ease of suppression and the rate of arrhythmic death to obtain evidence for or against causality.
Eligibility of Patients
Descriptions of CAST-I8 9 and CAST-II10 11 have been published previously. Patients were screened for CAST between 6 days and 2 years after a documented myocardial infarction. Qualifying arrhythmias were detected using ambulatory ECG recordings, with a minimum of 18 hours of analyzable data and at least six ventricular premature depolarizations (VPDs) per hour.
CAST-I was stopped in April 1989 due to an increased mortality associated with encainide and flecainide therapy. CAST-II was terminated on August 2, 1991, because of the lack of benefit from therapy and evidence of harm during the early administration of moricizine.
In CAST-I, patients first underwent an open-label titration phase, during which up to three drugs (encainide, flecainide, and moricizine) at two oral doses each were evaluated. The doses of encainide were 35 mg three times daily (dose 1) and 50 mg three times daily (dose 2); of flecainide, 100 mg twice daily (dose 1) and 150 mg twice daily (dose 2); and of moricizine, 200 mg three times daily (dose 1) and 250 mg three times daily (dose 2). The titration was stopped as soon as a drug and dose were found that suppressed the arrhythmias. Patients then were randomized to that drug and dose or to a matching placebo. The criteria for suppression were ≥80% reduction of VPDs and ≥90% reduction of runs of unsustained ventricular tachycardia (less than 15 consecutive VPDs at a rate of 120 beats per minute or more) as measured by 24-hour ambulatory ECG recording 4 to 10 days after each dose was begun. When CAST-I was concluded, the study was continued as CAST-II using three doses of moricizine, the two used in CAST-I plus 300 mg three times daily if the lower doses did not suppress VPDs.
The primary end point for CAST-I and CAST-II was arrhythmic death defined as arrhythmic death and nonfatal cardiac arrest, using the criteria developed in the Cardiac Arrhythmia Pilot Study.16 The definition of arrhythmic death includes witnessed instantaneous death (less than 1 minute) in the absence of severe congestive heart failure or shock, unwitnessed death with no preceding change in symptoms and for which no other cause can be ascribed, and resuscitated cardiac arrest.
Ease of Suppression
To test the primary hypothesis of this study, ease of suppression was defined as 80% suppression of VPDs on the first dose of the first drug. To conduct an analysis exploring a possible dose response, we divided the hard-to-suppress group into those patients who were never suppressed during the two trials and those who were eventually suppressed with increasing doses or different protocol drugs, as previously described in this section.
A total of 3549 patients entered the study between June 15, 1987, and August 2, 1991. Our study population consisted of 2951 patients whose suppression status at first drug/first dose was known.
Univariate survival curves were computed by the Kaplan-Meier method and compared by the log rank statistic. For patients who experienced arrhythmic death, the exposure time for the survival analysis was defined as the period from the date of Holter evaluation of the first drug/first dose to the date of death. For patients who did not experience an end point, exposure was computed from the Holter date above to the date of withdrawal for patients who withdrew from CAST, or to August 2, 1991, for patients who were still alive then. Baseline comparisons of the patients who were easily suppressed versus those who were not easily suppressed were performed by t tests and χ2 tests. Adjustment for baseline variables that were significantly associated with mortality and significantly related to ease of suppression was carried out using the Cox regression model. The selected baseline variables were forced into the Cox regression model, and then ease of suppression and the interaction between ease of suppression and the first drug attempted were allowed to enter the model using a stepwise procedure. Kaplan-Meier survival curves for the three groups, the easily suppressed, eventually suppressed, and never-suppressed patients, were compared using the log rank statistic.
Patients who were randomized to long-term active therapy presented a methodological problem, since the active therapy was shown to be harmful and a higher proportion of patients who were easily suppressed were randomized to active therapy, compared with those who were not easily suppressed. The data were not stratified by assignment to active/placebo therapy, which therefore increases the risk of those in the easily suppressed group.
A total of 555 deaths and nonfatal cardiac arrests occurred during follow-up. Table 1⇓ shows the crude event rate that occurred during the differing duration of follow-up for the two groups. Arrhythmic deaths occurred in 9.7% of the patients whose arrhythmia was easily suppressed and in 12.4% of those whose arrhythmia was not easily suppressed.
Of the 2946 patients included in this study, 1778 patients (60%) had their arrhythmias easily suppressed by the first drug/first dose and 1173 (40%) did not. The patients whose arrhythmias were easily suppressed had a better survival in regard to arrhythmic deaths over the 3 years of follow-up (P=.003). The relative risk of arrhythmic mortality at 12 months was .59, with the easily suppressed group experiencing a 95.3% survival compared with 92.0% for the not-easily suppressed group (Fig 1⇓). Similar findings were observed in regard to total survival over the 3 years of follow-up (P<.001, Fig 2⇓); at 12 months, the total survival for the patients whose arrhythmias were easily suppressed was 92.6% compared with 88.4% for those not easily suppressed (Fig 2⇓).
The better survival rates for patients with easily suppressed arrhythmias can be partially accounted for by differences in baseline characteristics (Table 2⇓). In univariate analysis, the group with easily suppressed arrhythmias had a greater incidence of anterior infarctions, higher ejection fractions, a lower prevalence of heart failure, previous myocardial infarction, and smoking, were more frequently receiving concomitant β-blockers, and were less frequently receiving diuretics. On the other hand, patients whose arrhythmias were hard to suppress were also slightly younger and had a lower VPD frequency.
Multivariate analysis of factors associated with arrhythmic death identifies a number of baseline characteristics that might explain the improved survival of patients with easily suppressed arrhythmias. Ejection fraction >30% had a strong beneficial effect on both arrhythmic and total deaths, whereas ejection fraction <30%, more advanced age, anterior infarction, and heart failure had a negative effect on both survival end points. However, ease of arrhythmia suppression remains a significant predictor of arrhythmic deaths (relative risk, .66; P=.013; Table 3⇓) after adjusting for age, ejection fraction, heart failure, history of myocardial infarction, β-blocker use, VPD frequency, diuretic use and anterior Q-wave index infarction, and the drug used for the first attempt of suppression. A similar effect of ease of suppression is observed for total deaths in the Cox regression model (relative risk, .83; P=.039). The interaction term between specific first drug used and ease of suppression was not significant and hence not included in the model.
To further examine the relation between ease of suppression and arrhythmic survival, easily suppressed, eventually suppressed, and never-suppressed patients were compared over the 3 years of follow-up. The observed order of the three groups was as expected; the easily suppressed had the lowest event rate, with the eventually suppressed and never-suppressed experiencing a progressive increase in event rates. The log rank statistic for an overall difference among the three Kaplan-Meier curves for arrhythmic deaths was highly significant (P<.001). The unadjusted 1-year relative risk for easily suppressed versus eventually suppressed was .75 (P=.12) and for the easily suppressed versus never-suppressed was .47 (P<.0001) (Fig 3⇓).
This study indicates that the ease of suppression of VPDs identifies in high-risk post–myocardial infarction patients with high-frequency VPDs a subgroup that is less likely to experience arrhythmic death during the 2 years of observation. There are also clinical characteristics that distinguish patients whose ventricular arrhythmias are most easily suppressed. Patients who were easy to suppress were older, had a greater incidence of anterior infarctions, and had less evidence of left ventricular dysfunction as evidenced by a lower frequency of a history of heart failure, previous myocardial infarction, and a higher ejection fraction. They also had a higher frequency of VPDs and concomitant β-blocker therapy. After adjusting for these variables, we found that ease of suppression of ventricular arrhythmias still had a statistically significant association with improved mortality.
The suppression of ventricular arrhythmias with the first dose and first drug was used as a measure of ease of suppression. The active and placebo patients are both included in this analysis. The fact that more patients in CAST were easily suppressed attests to the fact that the drugs used in this study were indeed extremely effective in VPD suppression. Since active therapy was harmful in CAST, including the patients on active therapy in this analysis, makes it harder to find the effect we observed in this analysis. The hard-to-suppress patients spent a longer time in titration on active drug compared with the easily suppressed group; however, a higher percentage of the easily suppressed patients received long-term active therapy—50% versus 35% for the hard-to-suppress group. Thus, the effect of active therapy in our analysis is likely to increase the death rate for the easily suppressed group more than for the hard-to-suppress group. This result is a conservative estimate of the size of the difference in the event rates between the two groups and has a conservative P value.
CAST focused on the overall benefit of drug suppression of ventricular arrhythmias on arrhythmic deaths. It reported that encainide and flecainide, although effective in suppressing VPDs, were associated with a higher rate of arrhythmic and total mortality when compared with the placebo group.8 9 For many years, cardiologists presumed that pharmacological suppression of ventricular ectopy was beneficial and led to improved survival.7 The strength of this presumption made it difficult to study the unique characteristics of those patients who achieve antiarrhythmic drug suppression, measured either by ambulatory ECG recording or electrophysiological testing. In considering the benefits of electrophysiological testing, Friedman and Yusuf17 raised the issue of the importance of suppression itself as a predictor of survival.
Previous investigations, however, demonstrated that suppression is more readily achieved in certain patients. Meissner et al13 examined predictors of successful medical therapy in 201 patients with chronic recurrent sustained ventricular arrhythmias evaluated by 560 electrophysiological tests. They observed that younger patients and those with ejection fractions >50% and with hypokinesia as their only ventricular dysfunction and without chronic heart disease were most likely to respond to antiarrhythmic therapy. Kuchar et al18 observed that suppression of sustained ventricular tachycardia induced at electrophysiological testing with oral antiarrhythmic agents was more readily achieved if it was rendered noninducible after intravenous procainamide or if the ventricular ejection fraction was >40%. The Cardiac Arrhythmia Pilot Study12 observed that likelihood of VPD suppression was directly related to ejection fraction. In a study of patients with symptomatic ventricular arrhythmias in which electrophysiologically guided arrhythmia suppression was compared with empiric antiarrhythmic drug administration, no difference was observed in mortality rates.14 However, in that study, arrhythmia suppression did identify a group of patients at low risk of arrhythmia recurrence or sudden death.
The hypothesis that a healthy responder phenomenon manifested by drug-induced arrhythmia suppression may exist in nonrandomized clinical trials was proposed by Hallstrom et al.19 They developed a theoretical model that suggested that this type of response could lead to misinterpretation of seemingly beneficial drug efficacy. They postulated that higher suppression rates to antiarrhythmic therapy by healthier patients would lead to a substantial overestimate of benefit. The observations in this investigation support that hypothesis.
In this study, it is clear that a number of characteristics are related to the ease of VPD suppression. Three well-known predictors of increased mortality in post–myocardial infarction patients, ejection fraction <30%, a history of heart failure, and prior myocardial infarction, also identified a subgroup of patients whose arrhythmia was difficult to suppress. Both the presence of an ejection fraction >30% and the concomitant use of β-blockers were associated with improved survival. However, even when adjustments were made for these and other characteristics of the patient population, ease of suppression still had a significant association with both decreased arrhythmic and total mortality.
The precise mechanism by which ease of suppression is related to mortality in this study is not clear. It is possible that pharmacodynamic properties of these drugs may play a role. Blood levels of these drugs, however, were not systematically measured during the study. Whether ease of suppression is an expression of multiple hitherto undetermined factors or one specific factor remains to be identified. The observation that patients who were eventually suppressed also differed from those who were never suppressed provides further validation of the importance of suppression as a predictor of arrhythmic death. Although these observations relate specifically to patients in CAST, in which ambulatory ECG monitoring was used to measure suppression, they have implications for other methods of drug evaluation, including electrophysiological testing. In any case, this analysis shows that the use of uncontrolled studies to assess the benefits of antiarrhythmic drugs is likely to lead to inappropriate conclusions.
This study was supported by contracts with the National Heart, Lung, and Blood Institute, Department of Health and Human Services, Bethesda, Md.
Reprint requests to Sidney Goldstein, MD, Heart and Vascular Institute, Henry Ford Hospital, 2799 W Grand Blvd, Detroit, MI 48202.
- Received June 17, 1994.
- Accepted September 28, 1994.
- Copyright © 1995 by American Heart Association
Bigger JT Jr, Fleiss JL, Kleiger R, Miller JP, Rolnitzky LM. Multicenter Postinfarction Research Group: the relationships among ventricular arrhythmias, left ventricular dysfunction, and mortality in the 2 years after myocardial infarction. Circulation. 1984;69:250-258.
Mukharji J, Rude RE, Poole WK, Gustafson N, Thomas LJ, Strauss HW, Jaffe AS, Muller JE, Roberts R, Raabe DS, Croft CH, Passamani E, Braunwald E, Willerson JT, and the MILIS Study Group. Risk factors for sudden death after myocardial infarction: two-year follow-up. Am J Cardiol. 1984;54:31-36.
Lown B, Wolf M. Approaches to sudden death from coronary heart disease. Circulation. 1971;44:130-142.
Hine LK, Laird NM, Hewitt P, Chalmers TC. Meta-analysis of empirical long-term antiarrhythmic therapy after myocardial infarction. J Am Coll Cardiol. 1989;262:3037-3040.
Echt DS, Liebson PR, Mitchell LB, Peters RW, Obias-Manno D, Barker AH, Arensberg D, Baker A, Friedman L, Greene HL, Huther ML, Richardson DW, and the CAST Investigators. Mortality and morbidity in patients receiving encainide, flecainide, or placebo. N Engl J Med. 1991;324:781-788.
Anderson JL, Hallstrom AP, Griffith LS, Ledingham RB, Reiffel JA, Yusuf S, Barker AH, Fowles RE, Young JB, for the Cardiac Arrhythmia Pilot Study (CAPS) Investigators. Relation of baseline characteristics to suppression of ventricular arrhythmias during placebo and active antiarrhythmic therapy in patients after myocardial infarction. Circulation. 1989;79:610-619.
Greene HL, Richardson DW, Barker AH, Roden DM, Capone RJ, Echt DS, Friedman LM, Gillespie MJ, Hallstrom AP, Verter J, and the CAPS Investigators. Classification of deaths after myocardial infarction as arrhythmic or nonarrhythmic (the Cardiac Arrhythmia Pilot Study). Am J Cardiol. 1989;63:1-6.
Friedman L, Yusuf S. Does therapy directed by programmed electrical stimulation provide a satisfactory clinical response? Circulation. 1986;73(suppl II):II-59-II-66.