Events in the cardiac arrhythmia suppression trial (CAST): Mortality in the entire population enrolled

To test the hypothesis that suppression of ventricular arrhythmias by antiarrhythmic drugs after myocardial infarction improves survival, the Cardiac Arrhythmia Suppression Trial (CAST) was initiated. Suppression was evaluated before randomization during an open label titration period. Patients whose arrhythmias were suppressed were randomized in the main study and those whose arrhythmias were partially suppressed were randomized in a substudy. Overall survival and survival free of arrhythmic death or cardiac arrest were higher in patients treated with encainide or flecainide than in patients treated with placebo. However, the death rate in patients randomized to placebo therapy was lower than expected. This report describes the survival experience of all patients enrolled in CAST and compares it with mortality in other studies of patients with ventricular arrhythmias after myocardial infarction.As of April 18, 1989, 2,371 patients had enrolled in CAST and entered prerandomization, open label titration: 1,913 (81%) were randomized to double-blind, placebo-controlled therapy (1,775 patients whose arrhythmias were suppressed and 138 patients whose arrhythmias were partially suppressed during open label titration); and 458 patients (19%) were not randomized because they were still in titration, had died during titration or had withdrawn. Including all patients who enrolled in CAST, the actuarial (Kaplan-Meier) estimate of 1-year mortality was 10.3%.To estimate the “natural” mortality rate of patients enrolled in CAST, an analysis was done that adjusted for deaths that might be attributable to encainide or flecainide treatment either during prerandomization, open label drug titration or after randomization. Because the censoring procedure excluded patients treated with encainide or flecainide after randomization, the mortality estimate will be less than the unadjusted mortality estimate of 10.3%. A minimal 1-year “natural” mortality rate of 6.3% was estimated by assuming that all deaths during open label titration with encainide or flecainide were due to drug treatment and that after randomization all excess mortality in the group randomized to active treatment was also due to drug effect. A maximal 1-year “natural” mortality rate of 8.4% was estimated by assuming that no deaths during open label titration with encainide or flecainide were due to drug treatment and that after randomization all excess mortality in the group randomized to active treatment was due to drug effect. Thus, the “natural” overall mortality rate in CAST, falling somewhere between 6.3% and 8.4%, is similar to rates observed in other postinfarction natural history studies, suggesting that the low mortality rate in the group randomized to placebo is largely due to the process of selecting only those enrolled patients for the randomized phase of CAST whose ventricular arrhythmias were suppressed in the prerandomization open label titration phase