Prevention of Sudden Death in Patients with Coronary Artery Disease
abstract & commentary
Synopsis: Electrophysiologically guided antiarrhythmic therapy with implantable defibrillators but not antiarrhythmic drugs is effective in patients similar to those entered in the study.
Source: Buxton AE, et al. N Engl J Med 1999;341: 1882-1890.
The multicenter unsustained tachycardia Trial (MUSTT) was designed to test two hypotheses. The first was that electrophysiologic studies could be used to identify a high-risk subgroup among coronary artery disease patients with prior myocardial infarction (MI), depressed left ventricular function, and spontaneous nonsustained ventricular tachycardia. The second hypothesis was that electrophysiologically guided antiarrhythmic therapy would improve survival in patients identified to be at risk. This paper gives results relevant to the second hypothesis.
The trial enrolled patients between November 1, 1990, and October 31, 1996, at 85 study sites in the United States and Canada. The criteria for enrollment were as follows: known coronary artery disease, a left ventricular ejection fraction of 40% or less, and asymptomatic unsustained ventricular tachycardia. Patients could not have a prior history of sustained arrhythmias or syncope unless these occurred within the first 48 hours after an MI. All patients underwent an evaluation for ischemia prior to enrollment and patients with active ischemia were excluded until that had been treated. After enrollment, patients underwent an electrophysiologic study with programmed ventricular stimulation using a standard protocol. A positive study was defined as the reproducible induction of sustained monomorphic ventricular tachycardia induced by any method of stimulation or of sustained polymorphic ventricular tachycardia induced by one or two extrastimuli. Patients with an inducible arrhythmia were assigned in equal numbers to either receive antiarrhythmic therapy guided by serial electrophysiologic testing or to a control group that was not prescribed specific antiarrhythmic therapy. Patients who had no inducible arrhythmia, an arrhythmia that was not reproducible, or patients who refused randomization were also followed in a large registry. The antiarrhythmic drugs to be tested were chosen at the discretion of Buxton and colleagues, with the exception that amiodarone could not be the first drug tested. Once the patient had reached steady state on the drug, programmed ventricular stimulation was repeated. Suppression of inducible ventricular tachycardia on a drug was required for the patient to continue on that therapy. If no effective drug regimen could be identified, the patient could either be discharged on a regimen that was associated with hemodynamic stability during induced tachycardia or the patient could receive an implantable cardioverter defibrillator (ICD). Guidelines for ICD implant were made more liberal after enrollment of about one-half of the study group.
The primary end point of the study was cardiac arrest or death from arrhythmia. Total mortality was a secondary end point. Events were classified by a blinded events committee. Treatment groups were compared using an intention-to-treat analysis.
During the study, a total of 2202 patients were enrolled. Of these, 767 (34.8%) manifested an inducible sustained ventricular tachycardia during the baseline electrophysiologic study. Seven hundred four of these patients agreed to be randomized to either serial drug therapy or no therapy. The clinical characteristics of the group were typical for an ischemic heart failure population. The median age was 67 years, with 90% of the patients being male and 88% of the patients being Caucasian. The median ejection fraction was 30%. Patients were essentially equally distributed between New York Heart Association classes I, II, and III.
Among the 351 patients assigned to electrophysiologically guided therapy, only 158 were discharged on antiarrhythmic drugs. Twenty-six percent of the entire group received class I drugs, 10% received amiodarone, and 9% received sotalol. One hundred sixty-one of the 351 patients (46%) received defibrillators. In addition, 2% of the patients randomized to therapy died during their initial hospitalization and 7% of the patients refused antiarrhythmic drug therapy and did not receive a defibrillator. The median duration of follow-up was 39 months. Among the patients assigned to no antiarrhythmic therapy, the two-year and five-year rates of cardiac arrest or death from arrhythmia were 18% and 32%, respectively. The corresponding rates for patients assigned to electrophysiologically guided therapy were 12% and 25% (P = 0.04; RR, 0.73). Total mortality rates after two and five years were 28% and 48% for the patients assigned to no antiarrhythmic therapy compared to 22% and 42% for those assigned to electrophysiologically guided therapy (P = 0.06; RR, 0.80). A preplanned subgroup analysis comparing drug therapy and ICD therapy in the treated group showed that all of the benefit could be attributed to ICD therapy. The total mortality at five years was 24% among patients who received defibrillators and 55% among those who did not. This difference associated with ICD treatment remained significant even after a regression analysis in which adjustments for relevant clinical factors were made. Buxton et al conclude that electrophysiologically guided antiarrhythmic therapy with implantable defibrillators but not with antiarrhythmic drugs is effective in patients similar to those entered into the study.
Comment by John P. DiMarco, MD, PhD
Prevention of sudden death remains a major problem. Most cardiac arrest victims do not survive their initial event. Therefore, effective strategies for primary prevention of sudden death have long been sought. Unfortunately, trials with antiarrhythmic drugs either given empirically or when guided by Holter monitoring have been largely unsuccessful. A single trial that compared "conventional antiarrhythmic drug therapy" and ICD therapy, the Multicenter Automatic Defibrillator Implantation Trial showed benefit in the defibrillator group, but that trial was small and had no control group. The MUSST data reported here are important since they confirm that ICD therapy is superior to drug therapy in a population at high risk for sustained and life-threatening arrhythmias. However, a number of factors make it difficult to estimate from these data the magnitude of benefit that will be associated with ICD therapy. There are several reasons why the benefit of the ICD may be overestimated in this paper. First, most of the patients who received antiarrhythmic drugs were treated with class 1A agents. After the Cardiac Arrhythmia Suppression Trial showed an increased mortality with drug therapy and the Endocardial Stimulation Versus Electrocardiographic Monitoring Study showed sotalol to be superior to other agents, most electrophysiologists shifted to using class III drugs. Sotalol and amiodarone, however, were used in fewer than half of the drug-treated patients in MUSST. There was also a change in the rules for ICD implantation during the trial and device technology significantly improved during the study period. Therefore, it seems probable that a higher proportion of patients in the latter phases of the trial received an ICD. Since treatment for heart failure and ischemia also improved during the course of the trial, some of the benefits seen with ICD therapy may be attributable to better medical therapy. Finally, the improvement in total mortality was only 6% at five years, a difference that was only of borderline significance.
The first hypothesis of the trial—that the electrophysiologic study would help identify a high-risk subset—is really not discussed in this paper. However, in preliminary reports, Buxton et al have shown that the electrophysiologic study connotes only a relatively modest increase in risk over that seen among patients who do not have inducible arrhythmias. Although, due to the large size of the trial, a positive electrophysiologic study did turn out to be a significant predictor of arrhythmic events and mortality, the control group was also a relatively high-risk population.
The costs associated with ICD therapy in its present form are substantial. Several clinical trials have now indicated that the ICD is superior to antiarrhythmic drugs in different settings. We still need good studies to help identify in which groups of patients use of an ICD will be economically feasible and attractive.
Patients with CAD, reduced LVEF, and nonsustained ventricular tachycardia had the lowest cardiac death rate with:
a. electrophysiology testing guided drug therapy.
b. holter-monitor-guided drug therapy.
c. no specific antiarrhythmic therapy.
d. an internal cardioverter defibrillator.