Determinants of Mid-term Failure in Coronary Artery Bypass Grafts: Insights from 5 Angiographic RCTs and Control Angiographies of 2069 Conduits
Mario Gaudino1, Brian Buxton2, Stephen Fremes3, David L. Hare4, Philip Hayward5, Giuseppe Nasso6, Neil Moat7, Suk-Won Song8, Giuseppe Speziale6, David P. Taggart9.
1Weill Cornell Medical Center, New York, NY, USA, 2Victorian Heart Centre, Richmond, Australia, 3Sunnybrook Health Science Center, Toronto, ON, Canada, 4University of Melbourne, Melbourne, Australia, 5Austin Hospital, Melbourne, Australia, 6Anthea Hospital, Bari, Italy, 7Royal Brompton & Harefield Trust, London, United Kingdom, 8Gangnam Severance Hospital, Seoul, Korea, Democratic People's Republic of, 9University of Oxford, Oxford, United Kingdom.
OBJECTIVE: We aimed to evaluate the determinants of midterm failure of left internal thoracic artery (LITA) grafts to the left anterior descending (LAD) and radial artery (RA) or saphenous vein grafts (SVG) non-LAD coronary arteries using a pooled database of 5 angiographic randomized controlled trials (RCT) that compared the patency of different conduits.
METHODS: The patients-level data of 5 angiographic RCTs (RAPCO, RAPS, RSVP, Stand-in-Y and Yoo trials) were joined in a common database. Angiographic results were evaluated overall (Overall Analysis) and in relation to the target vessel (LAD and non-LAD Analysis). Cox regression analysis was used to evaluate independent predictors of graft occlusion. Variables tested were: age, gender, diabetes, hypertension, previous myocardial infarction (MI), renal failure, location of target vessel, percentage of target vessel stenosis ≥90%, type of conduit used and location of proximal anastomosis (for free grafts).
RESULTS: Angiographic data from 2069 conduits (911 LITA, 680 RA, 543 SVG) were analyzed. Mean follow-up time was 5.0 ± 2.8 years. Overall patency rate was 89.6% (91.5% LAD, 88.2% non-LAD). Results of the Cox regression are summarized in the table. In the Overall Group, age and previous MI were independent predictors of graft failure. Female sex and anastomosis on the LAD were associated with lower graft failure rate. In the LAD Group, age was an independent predictor of occlusion. In the non-LAD group, previous MI and use of a saphenous vein instead of a radial artery graft were independently associated with graft occlusion. RA and SVG anastomoses to the right coronary artery were associated with higher risk of graft failure, but this did not reach statistical significance. The severity of stenosis of the target vessel was not associated with graft occlusion in any model.
CONCLUSIONS: Based on this large angiographic dataset, the determinants of failure are slightly different for grafts directed to the LAD and non-LAD distributions.
Back to 2017 Program