Noninvasive characterization of patients with monomorphic ventricular tachycardia and ventricular fibrillation using the signal-averaged electrocardiogram. A study on late potentials and spectral turbulence in patients with ischemic heart disease and in healthy subjects.

Abstract

Many people die suddenly from sustained ventricular tachyarrhythmias caused by ischemic heart disease, and the sudden death may be the first presentation of heart disease. If the patients at risk of ventricular tachyarrhythmia could be identified, some patients may be saved.The signal-averaged ECGs of 62 patients with spontaneous sustained monomorphic ventricular tachycardia (MVT), 64 survivors of cardiac arrest with ventricular fibrillation (VF), and 82 patients without ventricular arrhythmia who all had ischemic heart disease but no bundle branch block were compared. The patients with MVT frequently had a low ventricular ejection fraction and were significantly more late potential-positive and spectral turbulence-positive, which indicates that a more severe underlying substrate is necessary for development of MVT than for VF. The best variables of the signal-averaged ECG for identifying patients that had already suffered MVT vs VF were the width of the QRS complex measured as the filtered QRS duration and as the total QRS duration in the spectral analysis and the filtered QRS duration, respectively.The presence of a bundle branch block in patients with ischemic heart disease will per definition cause a wide QRS 3 120 ms and is known to further increase the risk of cardiac mortality. Patients with right bundle branch block (n=15) and with left bundle branch block (n=28) without spontaneous ventricular arrhythmias were late potential-positive in 27% and 50%, respectively, and spectral turbulence-positive in 80% and 79%, respectively. There were differences in the time domain analogous variables of the individual X, Y, Z orthogonal leads of the signal-averaged ECG between right and left bundle branch blocks. Patients with both spontaneous sustained MVT (n=13) or cardiac arrest with VF (n=6) and a bundle branch block were compared with control patients (n=38), matched 2:1 for each arrhythmia patient according to type of block, age, gender and previous myocardial infarction. There was no significant difference between patients with bundle branch block and VF and matched controls in any variable of the signal-averaged ECG, but the patients with MVT were more positive in variables of spectral analysis from individual X, Y, Z leads.In 121 subjects (55 men and 66 women; mean age 50 years) randomly selected without known heart disease and without bundle branch block, 7% were late potential-positive and 7% were spectral turbulence-positive. Almost all were completely normal at echocardiographic examination and during 24-hour ECG recording. The men had a significantly longer filtered QRS duration than women, which correlated with body surface area and intracardiac dimensions at echocardiography, but women had a longer QT interval in the 12 lead ECG. The numerical short-term reproducibility of the signal-averaged ECG variables was measured in the 121 healthy subjects. It was found to be good for time-domain and time-domain analogous variables but varied in spectral turbulence variables and was not different from short-term reproducibility in the 43 patients with bundle branch block or from long-term reproducibility. The reproducibility was affected by the noise level, but to a lesser extent by gender and age. If the spectral turbulence variable interslice-correlation standard deviation was replaced by the mean peaks per slice, the diagnostic reproducibility of spectral turbulence analysis improved substantially.