The trusted source for
healthcare information and
Abstract & Commentary
Synopsis: SAECG is an effective predictor of mortality in patients with prior myocardial infarction and nonsustained VT.
Source: Gomes JA, et al. Circulation. 2001;104:436-441.
Gomes and colleagues report data on the predictive value of the signal-averaged ECG (SAECG) from the Multicenter Unsustained Tachycardia Trial (MUSTT). MUSTT was a study designed to test the prognostic significance of inducible sustained ventricular tachycardia at electrophysiologic study and to see if therapy guided by serial electrophysiologic studies would improve survival. The study enrolled 2202 patients for electrophysiologic study. SAECGs were obtained in 1925 of these patients. Tracings from 612 patients with preexisting bundle branch block or intraventricular conduction defects, ventricular pacing or unsatisfactory tracings were excluded. Of the remaining 1268 patients with good quality SAECG’s, 364 had inducible sustained ventricular tachycardia and were included in the randomized portion of the trial that compared electrophysiologically guided therapy to no antiarrhythmic therapy.1 The remaining 904 patients were followed in a registry.2 The primary end point of MUSTT was cardiac arrest or death from arrhythmia and the secondary end points were cardiac death and total mortality. Deaths were classified by an events committee after review of available records. In this study, various SAECG parameters were analyzed to test their relationship to arrhythmic death or cardiac arrest, cardiac death, or total mortality.
Among the 1268 patients included in this report, there were 230 cardiac arrests or arrhythmic deaths (18%), 341 cardiac deaths (27%), and 457 total deaths (36%). Filtered-QRS duration had an approximately linear relationship with cardiac arrest and arrhythmic death, and with cardiac mortality. There was also a relationship between the terminal root mean square (RMS) voltage but this was not a linear relationship. Duration of the low amplitude signal when analyzed as a continuous variable was not a significant predictor of end points. Since it has been traditional to characterize the SAECG as either normal or abnormal using various cut-off points, analyses using dichotomized variables were performed. A filtered QRS duration cut-off point of 114 msec was significant for all end points. A terminal RMS voltage cut-off of 20 mV was significant for cardiac death but not for arrhythmic death. Low amplitude signal duration dichotomized at 38 msec was significant for cardiac death but not for arrhythmic death. Events were also analyzed by subgroups. Patients with an abnormal SAECG were more likely to be male or white, to have a more remote myocardial infarction, have more severe coronary disease, and have more frequent induction of sustained VT. However, the predictive value of the filtered-QRS duration was seen in all subgroups including both treated and untreated patients. Since ejection fraction (EF) was also a powerful independent predictor of end points in MUSTT, EF and filtered-QRS duration were combined with the EF dichotomized at 30%. Patients with both values abnormal had the highest event rates. Those with either an EF below 30 or a filtered QRS greater than 114 msec had intermediate values and those with EF greater than or equal to 30% and a filtered-QRS duration less than 114 msec had the lowest event rates. Gomes et al conclude that the SAECG is an effective predictor of mortality in patients with prior myocardial infarction and nonsustained VT. The combination of the SAECG with EF permits selection of high risk in patients for intervention.
Comment by John P. DiMarco, MD, PhD
The signal averaged ECG was developed almost 20 years ago. It provides a highly amplified signal processed recording which can detect microvolt level electrical potentials in the terminal QRS complex. The potentials are called "late" potentials and have been shown to arise from areas of scarred myocardium. The SAECG’s primary use has been for risk stratification after myocardial infarction and for prediction of inducibility of ventricular tachycardia at electrophysiologic study. The value of the SAECG in patients with chronic coronary artery disease, poor left ventricular function, and nonsustained ventricular tachycardia was a planned substudy in MUSTT.
Unfortunately, this paper leaves many questions unanswered. Patients in MUSTT underwent a baseline electrophysiologic study. Patients who did not manifest an inducible arrhythmia were then followed over time with most of them not receiving antiarrhythmic therapy. These patients had a slightly but significantly lower mortality than did patients who had inducible ventricular tachycardia and were not treated. Among the patients with inducible ventricular tachycardia, the treated patients had a lower overall mortality but patients who received antiarrhythmic drugs had a slightly higher mortality. All of the benefit in the guided therapy group was seen in those patients who received implantable cardioverter defibrillators. In this paper, we are told that a multivariate analysis was performed and that the SAECG is a predictor for all patients with and without inducible ventricular tachycardia and with and without treatment, but it is hard to understand exactly how the SAECG should be used in the future. Is someone with both inducible ventricular tachycardia and a positive SAECG at a higher risk than someone with just inducible ventricular tachycardia? Is there a significant difference in patients without inducible ventricular tachycardia who have positive or negative SAECGs? Although these seem to be obvious questions that would be of value to the clinician, the data to answer them are not presented here even though they should be available.
It is also possible that an abnormal signal averaged ECG may be a predictor of a bad reaction to antiarrhythmic drugs. Most antiarrhythmic drugs prolong conduction and delayed conduction is the parameter measured by the SAECG. Thus, drug therapy might worsen the potential for arrhythmia and cardiac function and be responsible for some of the observations here. Patients treated with a defibrillator only might show a different relationship. We are not given the data to see if prescribing an antiarrhythmic drug to someone with an abnormal SAECG actually had a negative effect on outcome.
Hopefully, Gomes et al will do further analyses of their data in the future. The current data don’t really seem to answer the questions clinicians would like to ask.
1. Buxton A, et al. N Engl J Med. 1999;341:1882-1890.
2. Buxton A, et al. N Engl J Med. 2000;342:1937-1945.