Spontaneous variability in ventricular ectopic activity during chronic antiarrhythmic therapy

Previous determinations of variability in frequency of ventricular arrhythmias have been based on repeated recordings obtained in the absence of therapy. We evaluate variability during "effective" treatment with antiarrhythmic drugs. Variability in the percent suppression of premature ventricular complexes (PVCs) was determined in 55 patients with chronic arrhythmias who underwent multiple ambulatory electrocardiographic recordings during evaluation of chronic therapy with antiarrhythmic drugs initially determined to be effective, which was defined as 70% or more reduction in total PVC frequency or 90% or more reduction in repetitive forms. During chronic therapy, total PVCs were suppressed by 92%, averaged after a logarithmic transformation step, and repetitive beats were suppressed by 88%. Variability in suppression was substantial. The one-sided 95% confidence intervals required a fall in suppression of total PVCs to 40% or less to exceed limits of spontaneous variability and of repetitive PVCs to 66% or less. Suppression declined at least once during therapy to less than 60% for total PVCs in 24 of 55 patients (44%) and to less than 80% for repetitive PVCs in 13 of 33 patients (39%); nine patients (16%) showed increases in PVC frequency at least once to levels above pretreatment baseline. Seven subgroups were analyzed for their effects on variability and loss of suppression: age, gender, disease etiology, cardiac function, baseline PVC frequency, use of beta-blockers, and class of antiarrhythmic drug. Differences in confidence bounds and loss of suppression were found to be determined in a complex way by subgroup differences in variability and in initial levels of PVC suppression. Variability was greater for patient subgroups with greater PVC frequency, beta-blocker therapy, and non-coronary artery disease. However, clinical loss of suppression was more common only in more elderly patients and those with worse cardiac function. In summary, substantial variability in arrhythmia frequency occurs during effective antiarrhythmic therapy, and the 95% confidence limits of spontaneous variability are broad and determined in a complex way. Careful consideration should be given before concluding on the basis of a single Holter test that changes (increases) in arrhythmia frequency, especially in certain subgroups, are caused by treatment failure.