Clinical trials of hypothermia after pediatric cardiac arrest have not seen robust improvement in functional outcome, possibly because of the long delay in achieving target temperature. Previous work in infant piglets showed that high nasal airflow, which induces evaporative cooling in the nasal mucosa, reduced regional brain temperature uniformly in half the time needed to reduce body temperature. The mouth is kept open to allow the high nasal airflow to easily exit. Here, we evaluated whether initiation of hypothermia with high transnasal airflow (32 L/min) provides neuroprotection without adverse effects in the setting of asphyxic cardiac arrest. Anesthetized, mechanically ventilated piglets (approximately 2-weeks-old) underwent sham-operated procedures (Group 1) or asphyxic cardiac arrest (Groups 2-6). The asphyxic insult consisted of reducing the inspired oxygen from 30% to 9.5-10% for 45 minutes (hypoxia period), then briefly increasing the inspired oxygen to 21% for 5 min (to improve the later success of cardiac resuscitation), and then completely stopping ventilation for 7 minutes. Cardiopulmonary resuscitation (CPR) commenced by re-establishing ventilation, performing chest compression, and injecting epinephrine as needed. The five cardiac arrest groups were further divided into those with normothermic recovery (38.5°C; Group 2), with mild hypothermia (34°C) initiated by surface cooling at 10 minutes (Group 3) or 120 minutes (Group 5) after resuscitation, or with mild hypothermia (34°C) initiated by transnasal cooling initiated at 10 minutes (Group 4) or 120 minutes (Group 6) after resuscitation. In the two transnasal cooling groups, the high nasal airflow continued for 2 hours and was then stopped; thereafter, surface cooling was used to maintain hypothermia. In all four groups with induced hypothermia, rectal temperature was sustained at the targeted temperature of 34°C with surface cooling until 20 hours after resuscitation, followed by 6 hours of gradual rewarming and cessation of fentanyl/70% nitrous oxide anesthesia. At four days of recovery, the piglets were euthanized and their brains were analyzed for the density of morphologically intact neurons in putamen, sensorimotor cortex, ventrolateral thalamus, and prefrontal cortex. The data sheet shows the density of viable neurons in these 4 brain regions for the 45 piglets that completed the study. The data sheet also shows the serial measurements of rectal temperature, mean arterial blood pressure, heart rate, and arterial blood measurements of the partial pressure of oxygen (PO2) and carbon dioxide (PCO2), oxyhemoglobin saturation, and pH obtained at baseline, during the period of hypoxia, at 4 minutes of ventilation with 21% O2, during the period of asphyxia, and during the first 24 hours of recovery. The piglets are assigned the same unique identifier number, labelled 1-45, for each set of measurements. Transnasal cooling initiated at 10 minutes after resuscitation was able to significantly rescue neurons in the highly vulnerable putamen without adverse effects.

• Body temperature was measured with a rectal thermometer.
• Arterial pressure was measured with a pressure transducer connected to a fluid-filled catheter in the femoral artery.
• Heart rate was derived from the frequency of the arterial pressure pulses.

Blood samples were periodically drawn from the femoral artery and analyzed for pH, PCO2, PO2, oxyhemoglobin saturation, and hemoglobin concentration on a Radiometer ABL800 Flex blood gas analyzer, with values corrected for body temperature.

For histopathology, deeply anesthetized piglets underwent transcardial perfusion with cold phosphate-buffered saline, followed by ice-cold 4% paraformaldehyde. The brains were post-fixed overnight, then carefully removed from the skull. Each brain was bisected along the mid-sagittal plane and sectioned into a 20-mm coronal slab, ranging from -2 to +18 mm from the bregma. This process ensured the inclusion of key areas: the basal ganglia, sensorimotor cortex, and thalamus. The slab was further divided into three slices, with an additional slice reserved for the prefrontal cortex, and all sections were embedded in paraffin. For histological analysis, 10-μm sections were stained with hematoxylin and eosin. The examination was conducted under oil immersion at 1000× magnification by a researcher blinded to the treatment details. Viable neurons were identified as round or oval cell bodies with a thin rim of cytoplasm, an open nucleus, and evident nucleolus. Neurons were considered to have an ischemic morphology if they had a shrunken, pyknotic nucleus with eosinophilic cytoplasm. Neuron counts were obtained in seven nonoverlapping fields in the putamen and in nine nonoverlapping fields per slide in layers II-VI of the sensorimotor cortical gyrus. Three slides representing the entire length of the putamen were used. A similar process was applied to the prefrontal cortex, where neurons were counted in nine nonoverlapping fields on the para-sagittal cortical gyrus. Additionally, in the most caudal block, viable neurons in the ventrolateral posterior thalamus were counted in seven distinct fields. Data are presented as the number of neurons with a viable morphology per mm^2 area.