Closed circuit xenon delivery for 72h in neonatal piglets following hypoxic insult using an ambient pressure automated control system: Development, technical evaluation and pulmonary effects

Background: Therapeutic hypothermia (TH) for 72h is the standard treatment following neonatal encephalopathy (NE). However, one-third do not benefit and adjunctive therapies are urgently needed. Xenon enhances neuroprotection with TH when administered at 50% concentration within 5hours of hypoxia in experimental studies. Delayed initiation (~10 hours of age) of 30% xenon for 24 hours during TH did not improve early adverse biomarkers in a clinical trial of Xenon+TH vs TH. After hypoxia-ischemia, excitotoxic injury via N-methyl-D-aspartate receptor overactivation lasts days. Since xenon partially inhibits this receptor, we hypothesised that giving 50% xenon throughout the entire 72h TH and rewarming periods would enhance neuroprotection. Xenon costs $30/litre, so a closed-circuit breathing system is desirable with automated fresh gas delivery. Methods: Seven mechanically ventilated newborn pigs were randomized to receive 50% inhaled xenon for 72h during hypothermia (rectal-temperature 35°C) and subsequent rewarming following a global hypoxic-ischemic insult (XeHT, N = 4) or under normothermia for 72h (rectal-temperature 38.5°C) following sham insult (XeNT, N = 3). An automated fresh gas delivery system injected oxygen/air/xenon boluses into a closed-circuit based on measured gas concentrations. Results and discussion: Median (IQR) xenon consumption was 0.31 L/h (0.18, 0.50) and 0.34L/h (0.32, 0.49) for hypothermic and normothermic groups respectively, 0.34L/h (0.25, 0.53) overall. 92% of 9626 xenon and 69% of 9635 oxygen measurements were within 20% variation from targets. For xenon concentration, the median absolute performance errors for the XeHT and XeNT groups were 6.14% and 3.84% respectively and 4.31% overall. For oxygen these values were 13.42%, 15.05% and 12.4% respectively. There were no adverse pulmonary pathophysiology findings. Clinical problems over the total period included three related to sensors, seven breathing system leaks, ten partial and one complete tracheal tube occlusion episodes. Conclusion: The automated controller functioned as intended maintaining an inhaled xenon concentration close to the 50% target for 72-78h at a xenon cost of $11.1/h.