Prediction of late arrhythmic events after acute myocardial infarction from combined use of noninvasive prognostic variables and inducibility of sustained monomorphic ventricular tachycardia.

A combined use of noninvasive techniques and electrophysiologic study in the prediction of arrhythmic events was prospectively evaluated in 303 surviving patients of acute myocardial infarction (AMI). The most powerful combination of noninvasive prognostic variables in identifying patients suitable for invasive strategies was also assessed. Patients who had > or = 2 variables among left ventricular ejection fraction < 0.4, ventricular late potentials and repetitive ventricular premature complexes (VPCs) were considered eligible for programmed ventricular stimulation. After 15 +/- 7 months of follow-up, 19 patients (6%) had an arrhythmic event. Left ventricular dyskinesia (p < 0.00001) and ejection fraction < 0.4 (p < 0.000001), late potentials (p < 0.001), filtered QRS duration > or = 106 ms (p < 0.00001), VPCs/hour > 6 (p < 0.05), paired VPCs (p < 0.01), > or = 2 runs of unsustained ventricular tachycardia (VT) per monitoring (p < 0.001), heart rate variability index < or = 29 (p < 0.00001) and mean RR interval < or = 750 ms (p < 0.01) were found to be significant univariate predictors of events. At multivariate analysis, only low left ventricular ejection fraction, prolonged filtered QRS duration, reduced heart rate variability index and detection of > or = 2 runs of unsustained VT per monitoring had an independent relation to late arrhythmic events. Of 67 eligible patients, 47 (70%) consented to undergo programmed stimulation. A positive electrophysiologic study was found to be the strongest independent predictor of events among patients preselected by noninvasive techniques. With a good sensitivity (81%), a combined use of noninvasive tests and electrophysiologic study selected a group of post-AMI patients at sufficiently high risk (event rate 65%) to be considered candidates for interventional therapy. The combination of > or = 2 variables among left ventricular ejection fraction < 0.4, filtered QRS duration > or = 106 ms and > or = 2 runs of unsustained VT was superior to the other ones in identifying high-risk subjects (positive and negative predictive values for arrhythmic events of 44 and 99%, respectively). On the basis of the data, this scheme appears to be the most appropriate for selecting patients suitable for electrophysiologic testing and invasive strategies after AMI.