Signal-averaged electrocardiography in children with anthracycline-induced cardiomyopathy.

The aim of the present study was to determine if signal-averaged ECG of patients with anthracycline-induced left ventricular dysfunction could differentiate between patients with anthracycline-induced cardiotoxicity and those without. Sixteen children with anthracycline-induced cardiomyopathy, aged 6.5 to 15.5 years (anthracycline dose = 198-737 mg/m2), and 31 patients aged 5.0 to 16.7 years, who received anthracyclines without evidence of left ventricular dysfunction (anthracycline dose = 120-517 mg/m2), were studied with signal averaged ECG. The two groups were comparable in age, body surface area, and time since completion of chemotherapy. Signal averaged ECG parameters of the patients were compared with data obtained from 530 healthy children. These parameters were converted to z-scores to account for growth-related changes in signal averaged ECG recordings. Z-scores for filtered QRS duration and low amplitude terminal signal < 40 microV were significantly lower (p = 0.002 and p = 0.015, respectively), and Z-score for root mean square voltage of the last -30 ms of filtered QRS tended to be higher (p = 0.06) in patients with left ventricular dysfunction. Filtered QRS duration lower than -1.5 SD was found in 4 of 16 patients with left ventricular dysfunction and in only 1 of 31 patients without (p < 0.05) yielding a sensitivity of 25% and a specificity of 97% to detect left ventricular dysfunction. Only 1 patient had late potentials; his left ventricular function was normal. Left ventricular mass index tended to be lower in patients with left ventricular dysfunction (p = 0.07), whereas left ventricular diastolic diameter was similar in the two groups. The mechanism that accounted for the difference in signal averaged-ECG between the two groups of patients could be linked with the decrease in left ventricular mass in patients with left ventricular dysfunction. In conclusion, children with left ventricular dysfunction following anthracycline therapy have a SA ECG different from those without left ventricular dysfunction, which is mainly characterized by a lower filtered QRS duration. A prospective study is needed in order to determine if this modification of SA ECG recordings precedes alteration of left ventricular function, and, therefore, if it could help in early detection of cardiac toxicity of anthracyclines.