Identifying Candidates for ICD Therapy
Abstract & Commentary
By John P. DiMarco, MD
Source: Chow T, et al. Does microvolt T-wave alternans testing predict ventricular tachyarrhythmias in patients with ischemic cardiomyopathy and prophylactic defibrillators? J Am Coll Cardiol. 2008;52:1607-1615.
The Microvolt T-wave Alternans Testing for risk Stratification of Post-Myocardial Infarction Patients (MASTER) trial tested the hypothesis that microvolt T-wave alternans (MTWA) testing could successfully predict arrhythmia occurrence in patients receiving an ICD for primary prevention of sudden cardiac death.
The trial was conducted in 50 US centers. All patients were scheduled to receive an ICD for primary prevention, according to "MADIT-II criteria." This meant that they were required to have a history of a prior myocardial infarction and an ejection fraction of ≤ 30%. Baseline microwave T-wave alternans (MTWA) testing was required for all patients. Treadmill testing was the preferred method, but if patients could not exercise, dobutamine infusion or cardiac pacing could be used. An MTWA test was considered positive if greater than 1.9 mV of alternans was recorded, with an onset heart rate ≤ 110 bpm during a two-minute recording period that was free from significant artifacts. If the test was indeterminant, the MTWA test was repeated. MTWA results were overread by a Core laboratory blinded to patient characteristics. Patients with positive and indeterminant MTWA results were combined as a non-negative group and compared against the true negatives for outcome analysis. MTWA testing was repeated every 12 months after enrollment until trial completion. Patients could receive a single-chamber, dual-chamber, or resynchronization ICD.
Arrhythmia detection was programmed as follows: ventricular fibrillation detection: 320 m/sec, VF number of intervals to detect: 24 of 32, ventricular tachycardia detection interval: 370 m/sec, and ventricular tachycardia number of intervals to detect: 16. +Supraventricular tachycardia discriminators for programmed and pre-storage electrogram storage were both programmed on. Antitachycardia pacing therapy was not used. Patients were followed at six-month intervals. The primary endpoint was a ventricular tachycardia (VT) event, defined as either sudden death or an appropriate ICD discharge by device programmed in accordance with the trial protocol. VT episodes were adjudicated by an independent clinical events committee.
A total of 575 patients met inclusion and exclusion criteria and had a technically adequate MTWA test. There were 361 (63%) MTWA non-negative patients, of whom 293 were MTWA-positive and 68 MTWA-indeterminant. The clinical characteristics were typical for a primary prevention ICD group. The average age was 65, with 84% male. The mean left ventricular ejection fraction was 24%. MTWA non-negative patients were older, less frequently Caucasian, less likely to be taking beta-blocker drugs, and had wider QRS durations. During 2.1 ± 0.9 years of follow-up, there were seven sudden cardiac deaths and 63 appropriate-device therapies for a ventricular tachyarrhythmia. There was no significant difference in annual event rate between MTWA non-negative and negative patients (6.3% vs 5.0%, HR: 1.26; 95% CI: 0.76 to 2.09; p = 0.37). After adjustment for baseline variables, MTWA continued to be non-predictive of ventricular tachyarrhythmias. However, total mortality was significantly greater in MTWA non-negative patients (HR: 2.04; 95% CI: 1.10 to 3.78; p = 0.02). This remained true after multivariate adjustment. Other predictors of ventricular tachyarrhythmic episodes were identified. Beta-blocker treatment was associated with a decreased frequency. Spironolactone treatment was associated with an increased frequency. Ventricular tachyarrhythmic events were more common among those with lower ejection fractions.
Chow et al concluded that MTWA is a predictor for total mortality, but not ventricular tachyarrhythmic events in patients receiving an ICD for primary prevention of sudden death.
The Multicenter Automatic Defibrillator Implantation trial and the Sudden Cardiac Death-Heart Failure trial form the primary bases on which guidelines for ICD implantation were established. Many cardiologists, however, have looked at the data from these trials and have noted that most patients who received an ICD will not have a ventricular tachyarrhythmic event during the first five years after device placement. Therefore, many investigators have sought more specific ways to risk stratify the patients so that only truly high-risk patients would receive an ICD. Microvolt T-wave alternans testing was previously shown to be frequently abnormal in patients with sustained or inducible ventricular arrhythmias, but as shown in this report, prospective use of the study for risk stratification has been disappointing.
One striking observation in this study is the relatively low incidence of events during the course of the study. Despite the fact that the mean ejection fraction was 24% and that 50% of the patients had a QRS duration of greater than 120 m/sec, the annual total mortality was only 4.9%, with a 5.2% annual appropriate ICD discharge rate. These mortality and arrhythmia rates are far below those observed in the early ICD trials such as MADIT-I or MADIT-II and even below the rates seen in the more recently published SCD-HeFT Trial. The reasons for this are not totally certain.
Clearly, pharmacologic management of patients with heart failure has improved dramatically over the last several years. It is also likely that physicians enrolling patients in MASTER selected those most likely to be able to do treadmill exercise testing. These patients may well have overall better health than those entered in prior trials.
The data presented here from the MASTER trial, and other data presented recently from the SCD-HeFT trial on micro-volt T-wave alternans cast doubt on this test's ability to effectively risk stratify patients before they receive an ICD for primary prevention of sudden death. The search for a truly effective risk stratifier continues.