P5602Comparison of innominate vs axillary artery cannulation for cerebral protection on neurocognitive outcomes in aortic surgery: a pre-specified analysis of the ACE CardioLink-3 randomised trial

Abstract

Abstract Background Success after aortic surgery depends on avoidance of neurocognitive dysfunction, thus novel adjuncts to proximal aortic surgery must be evaluated for efficacy of cerebral protection during circulatory arrest. We report the primary neurocognitive results from the ACE CardioLink-3 randomised controlled trial comparing innominate to axillary artery cannulation for cerebral protection (NCT02554032). Methods The primary safety endpoint was the proportion of patients with new radiologically severe ischaemic cerebral lesions found on post-operative versus pre-operative diffusion weighted magnetic resonance imaging (DW-MRI). Neurocognitive outcomes were assessed using the Mini-Mental State Exam (MMSE), and the Montreal Cognitive Assessment (MoCA). Continuous and binary outcomes were analysed using ANCOVA (controlling for baseline score) and chi-square/Fisher's exact tests. Results Of the 111 patients randomised, 102 patients were included in the primary safety per-protocol analysis. The primary safety outcome (significant new ischaemic lesions on DW-MRI) occurred in 34% in the innominate group and 38.8% in the axillary group (OR 0.81; 0.41 to 1.60; P=0.0009 for non-inferiority). Rates of post-operative stroke/transient ischaemic attack, seizure, delirium, and encephalopathy were similar between groups. The rate of patients with a post-operative MoCA score less than 26 was 44.9% and 39.1% in the innominate and axillary groups respectively (P=0.807). A post-operative MMSE score of less than 24 was observed in 2% vs. 6.5% of the patients in the innominate and axillary groups respectively (P=0.866). A >1-point decrease in the MoCA score from pre-operatively to post-operatively was seen in 32.7% and 34.8% in the innominate and axillary groups respectively (P=0.962). A >1-point decrease in the MMSE score from pre-to post-operative was observed in 20.4% in the innominate artery group compared with 30.4% in the axillary group (P=0.346). Conclusion Post-operative neurocognitive dysfunction and DW-MRI incidence of severe ischaemic lesions did not differ in patients randomised to innominate artery cannulation vs, conventional axillary artery cannulation, though the burden of new severe ischaemic lesions is high in both groups.