Veterans Affairs (VA) Coronary Artery Bypass Surgery Cooperative Study Group - VA-CABG


CABG vs. medical therapy for mortality in stable angina.


CABG may be more effective in the treatment of stable angina pectoris than medical therapy alone.

Study Design

Study Design:

Patients Screened: Not given
Patients Enrolled: 686
NYHA Class: III or IV
Mean Follow Up: 16.8 years

Patient Populations:

Stable angina for more than six months with at least a 3-month trial of medical management.
ECG evidence of previous myocardial infarction (MI) or ischemic changes at rest or with exercise.
At least one major coronary artery with a 50% or greater stenosis and a graftable distal segment.


Not given

Primary Endpoints:

Mortality (all causes)

Secondary Endpoints:

Incidence of MI and severity of angina

Drug/Procedures Used:

CABG vs medical therapy.

Concomitant Medications:

Nitroglycerin, long-acting nitrates, beta-blockers

Principal Findings:

Overall 18-year survival rates were 33% for medicine compared with 30% for surgery (p = 0.60).
Survival rates for high-risk patients without left main disease, which had shown a significant advantage for surgical therapy up to 11 years, were 23% medicine vs. 24% surgery for patients with three-vessel disease and impaired left ventricular function (p = 0.49) and 22% vs. 25% for those with multiple clinical risk factors (p = 0.12).

For patients with two-vessel disease, who had significantly better survival with medical therapy at 11 years, rates were similar at 18 years in the two treatment groups (34% medicine vs. 30% surgery, p = 0.09).

Cumulative 18-year myocardial infarction rates (fatal plus nonfatal) were 41% in medical and 49% in surgical patients (13% perioperative infarction rate), p = 0.15.

Nonfatal infarction rates were lower with medical than with surgical therapy (32% vs. 44%, p = 0.015), but fatal infarction rates were similar (14% medicine vs. 13% surgery, p = 0.62).

The combined rate of myocardial infarction or death was also lower with medical therapy (75% vs. 82%, p = 0.016). In contrast, surgery reduced mortality after myocardial infarction by 35% at 10 years (p less than 0.001) but only by 13% at 18 years (p = 0.09).
The percent of medical and surgical patients who were angina-free was 3% vs. 22% (p less than 0.001) at 1 year and 4% vs. 12% (p less than 0.001) at 5 years compared with rates of 6% vs. 5% (p greater than 0.50) at 10 years and 3% vs. 4% (p greater than 0.50) at 15 years.


The benefits of coronary artery bypass surgery on survival, symptoms, and postinfarction mortality were transient and lasted fewer than 11 years. The benefits began to diminish after 5 years, when graft closure accelerated. Surgery was effective in reducing mortality only for patients with a poor natural history. Low-risk patients, who had a good prognosis with medical therapy, derived no survival benefit with surgical therapy at any time during the follow-up period. Regardless of risk, surgery also did not reduce the incidence of myocardial infarction or the combined incidence of infarction or death. This study is now 10 years old and the rates of LIMA and free arterial grafts have increased, and there is now a growing option to treat vein graft failures with intracoronary stents. Newer agents such as glycoprotein 2b3a inhibitors and Plavix are now also available to improve these PCI outcomes. Thus, the findings in this study may not hold true in the current era.


1. Circulation 1992;86:121-130. Final results

Keywords: Myocardial Infarction, Follow-Up Studies, Ventricular Function, Left, Angina, Stable, Ticlopidine, Risk Factors, Constriction, Pathologic, Electrocardiography, Stents, Survival Rate, Nitrates, Coronary Artery Bypass, Nitroglycerin

< Back to Listings