Twenty-five-year outcomes after multiple internal thoracic artery bypass.

OBJECTIVE: Coronary artery bypass grafting with multiple internal thoracic artery grafts is currently controversial. This study assessed single institutional outcomes with multiple internal thoracic artery grafting for guidance with future clinical decisions.
METHODS: In 19,482 patients undergoing multivessel coronary artery bypass grafting (1984-2009), baseline characteristics were recorded in a prospective databank, and follow-up was obtained by questionnaires, phone contact, or National Death Index. Outcomes examined were subsequent myocardial infarction, percutaneous coronary intervention, reoperative coronary artery bypass grafting, all-cause death, and a composite of the 4. Three groups were defined: (1) no internal thoracic artery graft (1874/19,482 or 9%); (2) single internal thoracic artery grafts and adjunctive venous conduits (single internal thoracic artery; 16,881/19,482 or 87%); and (3) multiple internal thoracic artery grafts (728/19,482 or 4%). Multivariable Cox modeling adjusted for differences in baseline characteristics, and comparisons were performed using area under the curve analysis.
RESULTS: Differences in baseline characteristics for the no internal thoracic artery graft, single internal thoracic artery, and multiple internal thoracic artery groups were as follows: median age 66, 64, and 59 years, respectively; congestive heart failure 22%, 18%, and 13%, respectively; ejection fraction 0.50, 0.52, and 0.51, respectively; reoperation 10%, 3%, and 7%, respectively; diabetes 27%, 30%, and 15%, respectively; and female gender 33%, 28%, and 20%, respectively. No differences existed in the median number of diseased vessels (3, 3, and 3, respectively) or number of grafts per patient (3, 3, and 3, respectively). Composite outcome improved with increasing internal thoracic artery grafts, whether assessing unadjusted or risk-adjusted data. Compared with no internal thoracic artery graft, the adjusted hazard ratio was 0.79 (confidence interval, 0.74-0.83) for single internal thoracic artery grafting and 0.70 (confidence interval, 0.62-0.80) for multiple internal thoracic artery grafting (both P < .001), reducing risk by 21% and 30%, respectively.
CONCLUSIONS: This study confirms improved patient outcomes with multiple internal thoracic artery grafting, achieving half again as much benefit as single internal thoracic artery grafting alone. The data suggest that increasing application of multiple internal thoracic artery grafting should be encouraged to mitigate the inherent risks and costs of long-term cardiac events.