Aortic arch surgery : pros and cons of selective cerebral perfusion. A multivariable analysis for cerebral injury during hypothermic circulatory arrest

Thirty-five consecutive patients with aortic arch aneurysm who required surgical reconstruction were operated on with the aid of extracorporeal circulation between February 1985 and December 1993. Nineteen patients (54.3%) were treated with hypothermic circulatory arrest (HCA) (Group A) and 16(45.7%) (Group B) with HCA and selective cerebral perfusion (SCP) through the carotid arteries. Preoperative characteristics didn't show any significant differences between the two groups: mean age was 58.7 ± 12 vs 62.1 ± 7, p = ns, male sex 73.6% vs 75%, p = ns; atherosclerotic aneurysms were 57.8% vs 43.7%, p = ns; Type A dissections 42.2% vs 56.3%, p = ns and emergency operation were 68.4% vs 43.7%, p = ns in Groups A and B respectively. For SCP, blood was infused initially at a rate of 200-300 ml/min, maintaining the 30-40% of cerebral blood from in normothermia, successively, with the aid of transcranial Doppler sonography (TDS) SCP-flow was improved to 500-1000 ml/min. The MHz pulsed TDS was used to measure the middle cerebral artery flow velocity in deep hypothermia before the arrest, in order to adjust the SCP flow during the HCA. In all patients we used open aortic anastomosis; in two cases an extraanatomical ascending-descending aorta was required, and in other two the 'elephant trunk' technique was used in case of combined aortic arch and descending aneurysms. The HCA times were similar in the two groups 47.5 ± 22 vs 47.7 ± 78, p = ns. Early deaths occurred in 5 patients of the Group A (26.3%) and in 3 patients of the group B (18.7%), p = ns. Permanent and transient cerebral complications did not differ in the two groups: 15.7% vs 12.5%, p = ns, and 10.5% vs 18.7%, p = ns, in Groups A and B respectively. A multivariate analysis was performed in order to identify preoperative and operative risk factors for neurologic morbidity and mortality. Presence of aortic dissection (p = 0.029), emergency procedure (p = 0.032), age p = 0.038) and length of HCA p = 0.049) mere isolated as predicting variables of cerebral injury, while SCP was not found as a protective factor (p = 0.56). In summary, despite a slightly lower incidence of irreversible cerebral damage with the adjunct of SCP during HCA, this technique was not found in our hands free from complications. We think that SCP, which remains and interesting technique when HCA times exceed 45-50 min, still shows some controversial features and its risk/benefit ratio has not been completely cleared.