G iven the high prevalence of coronary artery disease in Western countries, selection of the optimal therapeutic strategy for individual cardiac patients has become an issue of utmost cost-benefit importance. During the past two decades, large clinical trials have established the respective roles of medical treatment versus coronary artery bypass surgery in the management of patients with coronary disease. In 1978, Grintzig et al's report of their early experience with coronary angioplasty1 opened the era of interventional cardiology, and since then, angioplasty has emerged as a major tool in the armamentarium for treating coronary artery disease. Initially, the use of angioplasty was restricted to selected patients with angina pectoris refractory to medical treatment. However, as interventionists became more skilled and equip-ment more sophisticated, indications for angioplasty were rapidly extended to includepatients with moderate or mildangina. This led to an exponential growth in the number of angioplasties performed over the past decade, and in 1991, more than 300,000 procedures were performed in the United States alone at an estimated cost of $7.2 billion, representing roughly 1% of the total annual expenditure for health care. See p 1489

To the outside observer this growth may appear puzzling, given some well-known limitations of this procedure. Although balloon dilatation is successful in up to 90% of cases for immediately improving the coronary artery luminalcaliber, and it is associated with acceptably low mortality and morbidity rates, 2 several important problems remain. First, the restenosis rate remains at a disappointing 30-50% at 6 months, and pharmacological, dietary, and mechanical interventions have as yet failed to significantly reduce this rate. 34 Second, the abrupt closure rate remains at 2-5%, although the introduction into clinical practice of new techniques such as stents, directional atherectomy, and laser balloonangioplasty may, in some cases, restore vessel patency. 5-7 Third, although highlysuccessful in reducing the percent stenosis ofa given coronary lesion, angioplasty has not yet been shown to reduce the subsequent risk ofplaque rupture and infarction, events that represent an important cause of morbidity and