Hypothermic circulatory arrest is not a risk factor for neurologic morbidity in aortic surgery: a propensity score analysis

OBJECTIVE: Hypothermic circulatory arrest has been an important tool in aortic arch surgery, even though its use has recently been discussed controversially. We sought to clarify the role of hypothermic circulatory arrest as a risk factor for mortality and neurologic morbidity in aortic surgery by using a propensity score-matching analysis. METHODS: Five hundred eleven patients (60 +/- 13 years, 349 male patients) who underwent replacement of the ascending aorta with (n = 273) or without (n = 238) arch involvement were analyzed by means of multivariate analysis. Using propensity score matching, we identified comparable patient groups: HCA(+) group and HCA(-) group (n = 110 each). For aortic arch replacement, hypothermic circulatory arrest was used with a mean duration of 14 +/- 9 minutes: 12 +/- 7 minutes or 26 +/- 8 minutes for partial or total arch replacement, respectively. RESULTS: In the entire cohort multivariate analysis identified acute dissection and duration of cardiopulmonary bypass as significant predictors for hospital death. Predictors for stroke were acute dissection, diabetes mellitus, peripheral arterial disease, and concomitant mitral valve surgery, and predictors for temporary neurologic dysfunction were peripheral arterial disease and age. After propensity score matching, the incidence of death (HCA+]: 0.9% vs HCA-]: 2.7%), stroke (0% vs 1.8%, respectively), and temporary neurologic dysfunction (15.5% vs 13.6%, respectively) was comparable between the 2 groups. Multivariate analysis identified age, diabetes mellitus, peripheral arterial disease, and concomitant coronary artery bypass grafting as the independent risk factors for temporary neurologic dysfunction. CONCLUSIONS: In a standard clinical setting (hypothermic circulatory arrest of <30 minutes and nasopharyngeal temperature of <20 degrees C), hypothermic circulatory arrest constitutes no significant risk for mortality or neurologic morbidity and thus appears clinically safe. Patient-related risk factors primarily determine clinical outcome.