Optimum therapeutic hypothermia temperature after perinatal asphyxia: a magnetic resonance spectroscopy biomarker and immunohistochemistry study in the newborn piglet

Abstract

BackgroundTherapeutic hypothermia for neonatal encephalopathy improves survival with normal neurological outcome. Despite treatment, however, 50% of infants have adverse outcomes; customising cooling with more precision might provide further benefit. The specific brain temperature for optimum neuroprotection is unknown. We have previously shown in the piglet that optimum neuroprotection by delayed cooling occurs at different temperatures in the cortical (33°C) and deep grey (35°C) matter. Magnetic resonance biomarkers, including lactate:creatine ratio (Lac/Cr), predict neurodevelopmental outcome after neonatal encephalopathy. We aimed to assess the optimum temperature for regional neuroprotection in a proton magnetic resonance system (1H MRS). Methods28 male piglets aged less than 24 h were anaesthetised with isoflurane and underwent trachaeal intubation, ventilation, continuous physiological monitoring, and intensive care treatment to maintain sedation and physiological homoeostasis. Within the bore of our 9·4 T magnetic resonance system all piglets underwent a hypoxic-ischaemic insult for 12·5 min, measured with real-time phosphorous MRS, before resuscitation. After this insult piglets were randomised (seven in each group) to normothermia (38·5°C), hypothermia (35°C), hypothermia (33·5°C), and hypothermia (30°C). Serial MRS assessments were acquired before, during, and up to 48 h after transient cerebral hypoxia-ischaemia. Areas under the curve (AUC) for the 1H MRS Lac/Cr peak-area ratio in ventromedial forebrain (vmFB, predominantly grey matter) and dorsal subcortical (dsc, predominantly white matter) voxels were calculated. Programmed cell death at 48 h was measured with terminal deoxynucleotidyl transferase dUTP nick-end labelling (TUNEL) staining on paraffin-embedded tissue in corresponding regions. FindingsCompared with normothermia, cooling to 35°C and 33·5°C produced consistent 50–75% reduction in density of TUNEL-positive cells with significant decreases in insular cortex, hippocampus, subcortical white matter, thalamus, and putamen (p<0·05 in ANOVA and post-hoc Tukey test). Cooling to 30°C did not further reduce TUNEL-positive cell density compared with 33·5°C and 35°C in dorsal parietal cortex (dCTX); more cell death was seen in the deep grey matter (thalamus and putamen) at 30°C. MRS biomarker analysis did not show an increase in Lac/Cr AUC in the vmFB voxel at 30°C compared with 33·5° and 35°C (p=0·07). Raised Lac/Cr and increased cell death at 30°C was not seen in the dsc voxel or in dCTX. InterpretationCooling to 30°C worsened hypoxic-ischaemic injury in the thalamus and basal ganglia compared with cooling to 33·5°C and 35°C. We showed a higher threshold optimum temperature for neuroprotection in the deep grey matter than in white matter/dCTX on the basis of immunohistochemical markers of cell death. Systemic effects of cooling to 30°C might have exacerbated these detrimental effects on deep grey matter cell death. These data support previous work in the developing brain and are relevant to clinical practice for optimum therapeutic hypothermia in newborn infants with birth asphyxia. FundingUK Medical Research Council, National Institute for Health Research.