It has long been held that the safe duration of hypothermic circulatory arrest (HCA) is at least 25 to 30 minutes. However, this belief is based primarily on clinical outcomes research and has not been systematically investigated using more sensitive brain imaging and neurocognitive assessments.

This exploratory substudy of the randomized Cognitive Effects of Body Temperature During Hypothermic Circulatory Arrest (GOT ICE) trial, which compared outcomes for deep vs moderate hypothermia during aortic arch surgery, investigated the frequency of neurocognitive and structural and functional magnetic resonance imaging (MRI) deficits with HCA of short (<20 minutes) duration. Neurocognitive deficit was defined as ≥1 SD decline in ≥1 of 5 cognitive domains on neurocognitive testing.

Of 228 GOT ICE participants with complete 4-week cognitive data, 74.6% (n = 170 of 228) had HCA durations <20 minutes, including 59 patients randomized to deep hypothermia (<20.0 °C), 55 patients randomized to low-moderate (20.1-24.0 °C) hypothermia, and 56 randomized to high-moderate (24.1-28.0 °C) hypothermia. Of these participants, cognitive deficit was detected 4 weeks postoperatively in ∼40% of patients in all 3 groups (deep hypothermia, 22 of 59 [37.3%]; low-moderate hypothermia, 23 of 55 [41.8%]; and high-moderate hypothermia, 24 of 56 [42.9%]). Furthermore, in a subset of patients with complete MRI data (n = 43), baseline to 4-week postoperative right frontal lobe functional connectivity change was inversely associated with HCA duration (range, 8-17 minutes; P for familywise error rate < .01).

Even short durations of HCA result in cognitive deficits in ∼40% of patients, independent of systemic hypothermia temperature. HCA duration was inversely associated with frontal lobe functional MRI connectivity, a finding suggesting that this brain region may be preferentially sensitive to HCA. Surgeons should be aware that even short durations of HCA may not provide complete neuroprotection after aortic arch surgery.