We analyzed whether baseline parameters of time-domain and spectrotemporal analysis of a signal-averaged ECG or their changes during
Mexiletine therapy can predict the antiarrhythmic efficacy of the
drug. On 60 post-MI patients with > 100 ventricular
premature beats per hour, signal-averaged ECGs were recorded before and after a constant infusion of
Mexiletine (7 mg/kg) for 1 h and again after 4 days of oral
Mexiletine therapy (
Mexiletine SR, 360 mg twice daily). Spectrotemporal analysis was performed on a fixed analyzed signal duration of QRS-complex and ST-segment of X-, Y-, Z-leads using the temporal window of the rectangular type, measuring signals between 10-20 Hz. Intravenous and oral
Mexiletine did not produce significant changes in mean values of any time-domain parameters. However, using informative variables of spectra of the signal-averaged ECG, we managed retrospectively to predict antiarrhythmic efficacy in 92% of the patients. Only certain frequency bands (from the range of the spectra at baseline, 10-120 Hz) were predictive for intravenous
Mexiletine efficacy: 40-55 Hz in lead Y (P = 0.0116); 55-70 Hz in leads X and Z (P = 0.0063 and P = 0.0269, respectively); 70-85 Hz in lead Z, (P = 0.0227). When the treatment with intravenous
Mexiletine was effective, the baseline power spectrum density was lower than when the
drug was ineffective, and vice versa. Moreover, the efficacy of oral
Mexiletine can be predicted by power density spectrum at baseline (10-25 Hz in lead Z, P = 0.0210; 70-85 Hz in lead Y, P = 0.0254) and by one of the possible (increased, decreased, unchanged) effects of intravenous
Mexiletine on the spectra at frequency bands (70-85 Hz in lead X, P = 0.0432 and 40-120 Hz in lead Z, P = 0.0156). These results show the value of spectrotemporal signal-averaged ECG in selecting a subgroup of post-
myocardial infarction patients that may benefit from
Mexiletine therapy.