Background: Controversy exists regarding the management of angiographically disease-free saphenous vein grafts at the time of redo coronary artery bypass grafting (CABG). Some authorities favor replacement of these disease-free grafts, arguing that occlusion is likely in the near future. Others believe that these grafts are "biologically privileged" and should not be replaced.
Methods: One hundred thirty-two consecutive patients (113 men, 19 women, aged 46 to 88 years, mean 67 years) underwent redo revascularization with one or more angiographically disease-free saphenous vein grafts at the time of redo CABG. Thirty-six patients had the disease-free grafts replaced (R) and 96 did not (NR). The mean interval from the first CABG was 9.25 years.
Results: Surgical mortality was comparable in the NR and R groups (5 of 96 or 5.2% versus 3 of 36 or 8.3%, respectively; p < 0.5). Survival at 1 and 3 years was higher in the NR group than the R group (98% versus 80%, and 95% vs. 66% respectively; p < 0.0001). Late myocardial infarction was less common in the NR group than in the R group (12 of 91 or 12.9% versus 12 of 33 or 36.4%; p < 0.003). Recurrent angina was less common in the NR than in the R group (21 of 91 or 23.1% versus 15 of 33 or 45.5%; p < 0.015). Cardiac hospitalization was required less commonly in the NR than in the R group (11 of 91 or 12.1% versus 12 of 33 or 36.4%; p < 0.002). In nondiseased grafts undergoing angiographic evaluation late after redo CABG, rate of new stenosis was lower in NR grafts than in R grafts (2 of 12 or 16.7% versus 2 of 3 or 66.7%; p < 0.05).
Conclusions: With a conservative approach that does not replace nondiseased saphenous vein grafts at redo CABG (1) there is no increase in operative mortality, (2) good late survival is obtained, (3) clinical ischemia related to the NR saphenous vein grafts is uncommon, and (4) NR grafts continue to be patent. We conclude that disease-free vein grafts may not require routine replacement at redo CABG. A randomized study is required for definitive resolution.