Abstract 262 Critical Care Platform, Monday, 5/3

Background: Mild hypothermia (HT) starting after hypoxia-ischemia (H-I) is neuroprotective in piglets who are anaesthetized during HT. In the 7 day old rat unilateral carotid occlusion model we have shown both short and long term protection by hypothermia, but when the rats were stressed by restraint the protection was lost. Newborn infants suffering from birth asphyxia will often be selfventilating and not necessarily sedated. In a clinical study it is important to know whether HT protection is dependent on simultaneous sedation. We aimed to test whether mild HT lasting 24 hours and started after H-I was neuroprotective if the animals were not sedated.

Subjects: 36 piglets <48 h were anaesthetised and ventilated and subjected to a 45 min global insult by reducing FiO2 (∼6%) until EEG was low amplitude (LAEEG <7µV). On reoxygenation 18 were maintained normothermic (NT 39.0°C) for 72h and 18 were cooled to 35°C (HT) for 24 h followed by 48 h of NT. The piglets were extubated when clinically able to selfventilate (for NT: 6.3±9.5 hrs, for HT: 1.8±1.5).

Measurements: Continuous EEG, intermittent neurologic assessment and biochemical markers were documented throughout. The brain was perfusion fixed and areas examined for degree of damage using light microscopy and our 9 step scoring (0.0-4.0) for this model (Thoresen, Ped Res 1996; 40:738-48).

Results: The insult duration severity was similar in the NT and HT groups (duration of LAEEG was 28 vs. 25±9.8 min and arterial pH at the end of the insult was 6.98±0.14 vs 6.99 ±0.11). 6 NT and 7 HT piglets developed posthypoxic seizures. The distribution and degree of damage within the brain were similar in the NT and HT groups (cortex/white matter 2.2±1.1 vs. 2.1±1.5 hippocampus 2.3±1.3 vs. 2.3±1.6, cerebellum 1.8±1.0 vs. 1.8±1.2, basal ganglia 1.2±1.3 vs. 1.6±1.6, thalamus 0.4±0.6 vs. 0.7±0.9). Other predictors of damage like EEG background amplitude at 3 h and neurologic score at 6 h were also similar. Plasma cortisol, however was significantly higher in the HT group during the hypothermic period, peak values after 24 hrs were 766±277 vs. 244 ± 144 in the HT and NT group respectively. Unpaired t-tests with Bonferoni corrections as indicated were used and p<0.05 regarded as significant. Values are mean ± SD.

Conclusion: Mild post-insult hypothermia for 24 hrs is not protective if the subjects are unsedated. This may be due to the stress of shivering or unknown mechanisms related to being cold and awake.