A closed-circuit neonatal xenon delivery system: a technical and practical neuroprotection feasibility study in...

Asphyxia accounts for 23% of the 4 million annual global neonatal deaths. In developed countries, the incidence of death or severe disability after hypoxic-ischemic (HI) encephalopathy is 1-2/1000 infants born at term. Hypothermia (HT) benefits newborns post-HI and is rapidly entering clinical use. Xenon (Xe), a scarce and expensive anesthetic, combined with HT markedly increases neuroprotection in small animal HI models. The low-Xe uptake of the patient favors the use of closed-circuit breathing system for efficiency and economy. We developed a system for delivering Xe to mechanically ventilated neonates, then investigated its technical and practical feasibility in a previously described neonatal pig model approximating the clinical scenario of global HI injury, prolonged Xe delivery with and without HT as a potential therapy, subsequent neonatal intensive care unit management, and tracheal extubation.

Sixteen newborn pigs underwent a global 45 min HI insult (4%-6% inspired oxygen reducing the electroencephalogram amplitude to <7 microV), then received 16 h 50% inspired Xe during normothermia (39.0 degrees C) or HT (33.5 degrees C). A conventional neonatal ventilator provided breaths of oxygen to a lower chamber compressing a hanging bag within. This bag communicated with the upper closed part of the breathing system containing soda lime, unidirectional valves, Xe/oxygen analyzers, and a tracheal tube connection. At each end-inspiration, this bag emptied fully and a bolus of oxygen, the driving gas, crossed from the lower to upper chamber via an additional valve. This mechanically substituted the gas uptake from the circle during the previous breath cycle (oxygen + small volume of Xe) with an equivalent volume of oxygen creating a slow-rising inspired oxygen concentration. This was offset by manual injection of Xe boluses, infrequently at steady state, due to the low-Xe uptake of the patient.

Total mean Xe usage was 0.18 (0.16-0.21) L/h with no differences between Xe-HT and Xe-NT groups, which had weights of 1767 (1657-1877) g and 1818 (1662-1974) g, respectively (95% CI). HT reduced heart rate in the cooled animals; 180 (165-195) vs 148 (142-155) bpm (P < 0.0001) with no differences in arterial blood pressure, oxygen saturation, arterial carbon dioxide tension, or weaning times between these groups.

We describe a closed-circuit Xe delivery system with automatic mechanical oxygen replenishment, which could be developed as a single use device. Gas exchange was maintained while Xe consumption was minimal (<$2/h at $10/L*). We have shown it is both feasible and cost-efficient to use this Xe delivery method in newborn pigs for up to 16 h with or without concurrent cooling after a severe HI insult.