Time-domain T-wave alternans measured from Holter electrocardiograms predicts cardiac mortality in patients with left ventricular dysfunction: a prospective study.

Background: Time-domain T-wave alternans (TWA) is useful for identifying patients at risk for serious events after myocardial infarction.
Objective: The purpose of this study was to prospectively evaluate the utility of time-domain TWA measured from Holter ECGs in predicting cardiac mortality in patients with left ventricular (LV) dysfunction.
Methods: Two hundred ninety-five consecutive patients with LV dysfunction were enrolled in the study. Patients were divided into two groups: the ischemic group (n = 195) and the nonischemic group (n = 100). Time-domain TWA was assessed using the modified moving average method from routine 24-hour Holter ECGs recorded during daily activity. The maximal time-domain TWA voltage at heart rate <or=120 bpm in either lead V(5) or V(1) was derived and its value defined as positive when the voltage was >or=65 microV. The primary end-point was defined as cardiac mortality.
Results: Mean maximal time-domain TWA voltage was 54 +/- 16 microV. During follow-up of 390 +/- 212 days, 27 patients (17 in the ischemic group and 10 in the nonischemic group) died of cardiac causes. Fifty-three patients (18%) were time-domain TWA positive and 242 (82%) were time-domain TWA negative. Univariate Cox proportional hazards analyses revealed that older age, New York Heart Association functional class III or IV, diabetes, renal dysfunction, nonsustained ventricular tachycardia, and time-domain TWA were associated with cardiac mortality. In multivariate analysis, time-domain TWA had the most significant value (hazard ratio = 17.1, P <.0001). This index also was significant in both subgroups (ischemic group: hazard ratio = 19.0, P <.0001; nonischemic group: hazard ratio = 12.3, P = .002).
Conclusion: Time-domain TWA measured from 24-hour Holter ECGs predicts cardiac mortality in patients with ischemic and nonischemic LV dysfunction.