BACKGROUND:

Reports of unsuspected, frequent and prolonged episodes of asymptomatic ST depression (ASTD) raised important issues for the evaluation and treatment of coronary disease. The traditional role of angina as the best indicator of myocardial ischemia had to be reevaluated, since several studies had shown that ASTD was much more frequent than symptomatic ischemia in many patients with coronary disease. If ASTD reflected clinically meaningful ischemia, then overt angina represented only 25 percent of the total ischemic burden. Moreover, incomplete data suggested that an increased and clinically unsuspected risk might be conferred on a patient with frequent episodes of ASTD. In 1988, studies were already in progress to find the most effective drug for treating ASTD. If no increased risk was associated with asymptomatic myocardial ischemia, there was little value in pursuing further investigations into its detection and therapy. But if the risk was increased, the detection and treatment of asymptomatic as well as symptomatic myocardial ischemia might completely alter therapy and the outlook for patients with coronary disease. The information in 1988 on the increased risk of ASTD was unfortunately scattered and sparse.

DESIGN NARRATIVE:

A total of 973 patients were enrolled two to six months after hospitalization for an acute coronary event. Patients were derived from geographically dispersed centers to enhance generalization of the results to the overall population at risk. Ambulatory 24-hour Holter electrocardiographic recordings were used to evaluate the frequency, severity, and duration of asymptomatic ST depression (ASTD) as a marker of silent myocardial ischemia during usual daily activities. This information was analyzed in conjunction with ST segment depression on treadmill exercise, reversible myocardial perfusion defects on stress thallium testing, and a limited number of prespecified clinical parameters to assess the contribution which ASTD added to prognostic information from then current diagnostic techniques used to evaluate myocardial ischemia. Patients were seen by the study coordinator every four months during the first year, then every six months thereafter. At each visit the study coordinator interviewed patients to determine interim history, functional status symptoms, and drug usage. At the four month visit a second 24-hour Holter recording was obtained. At the 12-month visit and at study termination, 12-lead electrocardiograms were recorded and sent to the ECG Core Laboratory for analysis. Data analyses were used: to assess the predictive usefulness and independence of ASTD in ambulatory patients with established coronary heart disease; to evaluate the interactions and associations between the frequency, severity, and duration of ASTD and other measures of myocardial ischemia; to obtain insight into suspected vasoactive and increased myocardial oxygen consumption mechanisms of myocardial ischemia; to better understand the interrelationship between myocardial ischemia and ventricular arrhythmias as it related to subsequent cardiac mortality and morbidity.

The study was renewed in 1993 for one year in order to fully analyze the prospectively accumulated data on the 973 enrolled patients in the Multicenter Study of Silent Myocardial Ischemia (MSSMI). The primary goal was to determine the usefulness, or the lack thereof, of ambulatory electrocardiographic (AECG) monitoring for identifying coronary patients with jeopardized ischemic myocardium at risk for ischemic cardiac events. The aims of the study were: 1) to evaluate the reproducibility of then current methods for detecting ischemic-type changes (ST depression) on AECG; 2) to improve the accuracy and reliability of contemporary scanner-interactive methodology that had been utilized for identifying ischemic-type ST-segment changes on the AECG; 3) to complete the development on a new, innovative, and reproducible computer-based technique for automatic (operator-independent), quantitative identification of beat-to-beat ischemic-type ST depression on the MSSMI AECG tapes; 4) to evaluate the clinical utility of the improved and new ST-segment analytic techniques to identify coronary patients at risk for natural cardiac events (unstable angina, non-fatal myocardial infarction, and cardiac death) during two year follow-up in the overall MSSMI population, in selected subgroups (gender, age), and in prespecified high-risk ischemic subsets; and 5) to determine the optimal cost- effective strategy for sequencing the assessment of silent, jeopardized ischemic myocardium by clinical variables and non-invasive tests to identify coronary patients at risk for subsequent ischemic cardiac events. This research utilized statistical techniques for evaluating reproducibility of detecting ST depression on AECG (kappa statistics, analyses of variance, Cochran's Q-test), for determining ischemic risk factors for time-to-cardiac events (Cox regression analyses), and for cost-effectiveness analyses (receiver-operator characteristic curves).

Twenty-four clinical centers In United States, Canada, Israel, and Japan participated in the study. There were also an ECG Core Lab, an Exercise Core Lab, a Holter Core Lab, a Thallium Core Lab, and a Coordinating Center.