Medical compared with surgical management of unstable angina. 5-year mortality and morbidity in the Veterans Administration Study.

Abstract

We evaluated medical in comparison to surgical plus medical (surgical) treatment of unstable angina using a prospective randomized protocol that stratified patients by clinical presentation and by invasive evaluation of left ventricular (LV) function. Clinical presentations were as follows--type 1: progressive or new onset angina relieved by medication; type 2: prolonged bouts of angina poorly or incompletely relieved by medication. Abnormal LV function was arbitrarily defined as ejection fraction less than 0.50 or LV end-diastolic pressure 16 mm Hg or more. Of 468 patients, 237 were assigned to medical and 231 to surgical therapy. There were 374 type 1 and 94 type 2 patients. LV function was normal in 334 and abnormal in 134 patients. Compared with results at 24 months, this 60-month follow-up study showed important differences in survival for patients with three-vessel disease: 75% for medical and 89% for surgical patients (p less than 0.02). The cumulative 5-year rate of repeat hospitalizations for cardiac reasons was less with surgical patients for either clinical presentation. For type 1, medical patients had a 56% rate, and surgical patients had a 42% rate (p = 0.004). For type 2, medical patients had a 62% rate, and surgical patients had a 43% rate (p = 0.05). Overall mortality did not differ between the two treatments, and this remained true in type 1 versus type 2 patients and in those with normal versus abnormal LV function. However, regression analysis of medical and surgical groups with ejection fraction as a continuous variable showed that mortality of medical patients depended on ejection fraction (p = 0.004), whereas the mortality of surgical patients did not (p = 0.76), and survival in the surgical group was higher in the lowest ejection fraction tercile-73% for medical and 86% for surgical patients, p = 0.03. We conclude that surgery improves survival in patients with three-vessel disease and leads to fewer subsequent hospitalizations for cardiac reasons. An impaired ejection fraction had an adverse impact on survival of medical patients but not on surgical patients, and mortality in surgical patients was improved compared with medical patients in the lowest ejection fraction tercile.