Neuropsychometric outcome following aortic arch surgery: a prospective randomized trial of retrograde cerebral perfusion.
Harrington DK., Bonser M., Moss A., Heafield MT., Riddoch MJ., Bonser RS.
BACKGROUND: Aortic surgery requiring hypothermic circulatory arrest is associated with a high incidence of brain injury. However, knowledge of neuropsychometric outcome is limited. Retrograde cerebral perfusion has become a popular adjunctive technique to hypothermic circulatory arrest. The aim of this study was to assess neuropsychometric outcome and compare the 2 techniques. METHODS: In a prospective randomized trial, 38 patients requiring elective aortic arch surgery were allocated to either hypothermic circulatory arrest plus retrograde cerebral perfusion or hypothermic circulatory arrest alone. Neuropsychometric testing was performed preoperatively, and at 6 weeks and 12 to 24 weeks postoperatively. Deficit was defined as a 20% decline in 2 tests or more. Standardized Z scores were calculated for each patient and test. Eighteen patients underwent hypothermic circulatory arrest and 20 patients underwent hypothermic circulatory arrest plus retrograde cerebral perfusion. The mean cardiopulmonary bypass, hypothermic circulatory arrest, and retrograde cerebral perfusion durations were 169, 30, and 25 minutes, respectively. RESULTS: There were 2 deaths and 2 neurological deficits. At 6 weeks postoperatively, 77% of the hypothermic circulatory arrest group and 93% of the hypothermic circulatory arrest plus retrograde cerebral perfusion group had a deficit (P =.22). At 12 weeks this was reduced to 55% and 56%, respectively (P =.93). There was a worse total Z test score in the hypothermic circulatory arrest plus retrograde cerebral perfusion group at 12 weeks (P =.05). Neuropsychometric change did not correlate with hypothermic circulatory arrest duration, presence of aortic atheroma, cannulation technique, or procedure. CONCLUSIONS: Hypothermic circulatory arrest plus/minus retrograde cerebral perfusion is associated with a high incidence of neuropsychometric change despite ostensibly normal clinical outcomes and apparently safe arrest duration. Retrograde cerebral perfusion did not improve outcome in this small study.