Review of the Veterans Administration Cooperative Study
A trial of saphenous-vein-bypass grafting in chronic stable angina
Circulation 1976; 54:Suppl:III-107-III-117
Background: By the 1970's, coronary artery bypass surgery was performed to improve the quality of life in patients with chronic stable coronary artery disease.
The Veterans Administration Cooperative Study was the first notable randomized controlled trial to test the hypothesis that saphenous-vein-bypass grafting improves survival in patients with chronic stable angina.
Patients: Eligible patients were men with stable angina pectoris for at least six months, who received medical therapy for at least three months and had at least one of the following features: electrocardiographic evidence of prior Q wave myocardial infarction, ST-T wave changes at rest suggestive of ischemia or a positive exercise stress test. Patients were excluded if they had myocardial infarction in the previous six months, unstable angina, congestive heart failure in the previous three weeks, uncontrolled diabetes, or conditions that limited long term survival.
Angiographic criteria were a 50% or greater reduction in the luminal diameter of at least one of the major coronary arteries, with a distal segment that was potentially graftable. Patients were excluded if they had ventricular aneurysm, poor left ventricular systolic function or marked elevation in left ventricular end diastolic pressure, although cut off points were not defined.
Baseline characteristics: The baseline characteristics reported here refer to the group without left main disease. There were 310 patients in the medical arm and 286 in the surgical arm. The average age of patients was 51 years. Approximately, 62% had prior myocardial infarction, 29% had hypertension, 7% had congestive heart failure, 3% had prior cerebrovascular accident and 13% had diabetes. More patients in the medical arm had serum cholesterol > 249 mg/dL; 32% vs 21%.
Procedures: Patients were randomly assigned to receive medical therapy alone or medical therapy plus saphenous-vein-bypass graft surgery. Medical therapy was prescribed at the discretion of the treating physician without a standardized protocol but included nitrates, beta-blockers and anti-arrhythmic drugs.
Patients were followed up every three months for the first two years then every six months.
Endpoints: The primary end point was all-cause death. For patients who were lost to follow up, the BIRLS (Beneficiary Information Retrieval and Locator System in Austin, Texas) was used to determine survival status.
The results provided in the main manuscript were based on the treatment assigned analysis, defined as considering patients who crossed over as lost to follow up at the time of treatment change. Sample size calculations and power analysis were not provided.
Results: The authors reported the results of patients with left main disease vs no left main disease separately. We report both results here since they are part of the same study.
Among 5,538 patients who were screened, 1,015 (18.3%) were randomized. In their methodology paper, the authors outlined the reasons for excluding certain patients from the trial. Of note, we found inconsistencies in the number of patients reported. For example, the methodology paper reported the randomization of 1,015 patients, however, we found outcomes data on only 686 patients. Additionally, some papers indicated that the number of patients with left main disease was 113, while others reported it as 90 or 91.
Adherence to the assigned treatment group was 98.6% early after randomization and 90.9% at a minimal follow up time of 21 months. On average, patients received 2 grafts. Vein graft angiography was performed in 79% of the patients after approximately 10-15 months. The graft patency rate was 69% with 87% of the patients having at least one functioning graft and 11% without any functioning graft.
Among patients without left main disease. The 30-day operative mortality was 5.6%. At 36 months, mortality was 13% in the medical arm and 12% in the surgical arm and this was not statistically significant. Patients with three vessel disease did not derive significant benefit with surgery.
Among patients with left main disease and after an average of 30 months follow up, mortality was lower in the surgery arm 20% vs 36%; p< 0.06.
No formal subgroup analysis or interaction testing was provided.
Conclusion: In patients with chronic stable angina pectoris and 50% or greater reduction in the luminal diameter of at least one of the major coronary arteries, saphenous-vein-bypass graft surgery did not improve outcomes compared to medical therapy alone. The benefit with surgery seen in patients with left main stenosis should be viewed as hypothesis generating only as they represented a small number of patients in the trial.
This trial has limited external validity in the current era due to advances in both medical and surgical management over the past several decades. For example, statins are widely used now in patients with obstructive coronary artery disease. Surgical outcomes have also improved and the left internal mammary artery, which has higher patency rates compared to vein grafts, is now typically used over a saphenous vein graft to bypass the left anterior descending artery.
Another concern with the study results is that, although 1,015 patients were randomized, outcome data were provided for only about two-thirds of them. Additionally, while the authors provided reasons for excluding certain patients, they did not define cut-off points for exclusion based on left ventricular ejection fraction or left ventricular end diastolic pressure. This lack of clarity makes it difficult for physicians to determine which patients would qualify for the trial.
Finally, the authors conducted a “modified intention-to-treat analysis” by considering patients who crossed over to the other treatment arm as lost to follow-up at the time of crossover. This method is not standard practice and is no longer used. It also introduces selection bias, as patients who switch from one treatment to another likely differ from those who adhere to the assigned treatment.
In summary, this trial does not provide evidence that coronary artery bypass surgery is superior to medical therapy in patients with three vessel or left main coronary artery disease.
Sounds like their “modified ITT” would now be called an on-treatment analysis, since they censored all crossovers.
Not only has medical and surgical therapy come a long way in the last 50 years, but trial methods, reporting, and statistical analysis have too.