Non-additive effects of adjunct erythropoietin therapy with therapeutic hypothermia after global cerebral ischaemia in near-term fetal sheep.

KEY POINTS: Recombinant human erythropoietin (rEpo) is neuroprotective in immature animals, but it is unclear whether the combination of high-dose rEpo therapy with therapeutic hypothermia can further improve outcomes. Hypothermia and rEpo independently improved neuronal survival, with greater improvement with hypothermia, and similarly reduced numbers of caspase-3 positive cells and reactive microglia after 7 days recovery. Hypothermia, but not rEpo, was associated with markedly improved EEG power, whereas both interventions improved recovery of EEG frequency. There was no significant improvement in any outcome after combined rEpo and hypothermia compared with hypothermia alone, and of concern, the combination was associated with increased numbers of cortical caspase-3-positive cells compared with ischaemia-hypothermia. These data suggest that the mechanisms of neuroprotection with hypothermia and rEpo overlap and, thus, high-dose rEpo infusion does not appear to be an effective adjunct therapy for therapeutic hypothermia. ABSTRACT: Therapeutic hypothermia for hypoxic-ischaemic encephalopathy (HIE) provides incomplete neuroprotection. Recombinant human erythropoietin (rEpo) is neuroprotective in immature animals, but it is unclear whether adjunct rEpo therapy with therapeutic hypothermia can further improve outcomes. Near-term fetal sheep received sham-ischaemia (n = 9) or global cerebral ischaemia for 30 min (ischaemia-vehicle, n = 8), followed by intravenous infusion of rEpo (ischaemia-Epo, n = 8; 5000 U/kg loading dose, then 833.3 U/kg/h), cerebral hypothermia (ischaemia-hypothermia, n = 8), or rEpo plus hypothermia (ischaemia-Epo-hypothermia, n = 8), from 3 to 72 h post ischaemia. Fetal brains were collected 7 days after cerebral ischaemia. Cerebral ischaemia was associated with severe neuronal loss and microglial induction in the parasagittal cortex and subcortical regions. Hypothermia reduced overall neuronal loss, cortical caspase-3 and reactive microglia in the striatum and cortex, with greater recovery of electroencephalographic (EEG) power and spectral edge (SEF) from 48 h onwards. rEpo independently improved neuronal survival in the parasagittal cortex, hippocampal CA4 and thalamus, and reduced cortical caspase-3 and activated microglia in striatal and cortical areas, with greater SEF from 120 h onwards. However, ischaemia-Epo-hypothermia did not further improve outcomes compared with ischaemia-hypothermia and was associated with increased numbers of cortical caspase-3-positive cells. These findings suggest that although delayed, prolonged treatment with both hypothermia and rEpo are independently neuroprotective, they have overlapping anti-inflammatory and anti-apoptotic mechanisms, such that the delayed, high-dose rEpo infusion for 3 days did not materially augment neuroprotection with therapeutic hypothermia.

Extending the duration of hypothermia does not further improve white matter protection after ischemia in term-equivalent fetal sheep.

A major challenge in modern neonatal care is to further improve outcomes after therapeutic hypothermia for hypoxic ischemic encephalopathy. In this study we tested whether extending the duration of cooling might reduce white matter damage. Term-equivalent fetal sheep (0.85 gestation) received either sham ischemia followed by normothermia (n = 8) or 30 minutes of bilateral carotid artery occlusion followed by three days of normothermia (n = 8), three days of hypothermia (n = 8) or five days of hypothermia (n = 8) started three hours after ischemia. Histology was assessed 7 days after ischemia. Ischemia was associated with loss of myelin basic protein (MBP) and Olig-2 positive oligodendrocytes and increased Iba-1-positive microglia compared to sham controls (p < 0.05). Three days and five days of hypothermia were associated with a similar, partial improvement in MBP and numbers of oligodendrocytes compared to ischemia-normothermia (p < 0.05). Both hypothermia groups had reduced microglial activation compared to ischemia-normothermia (p < 0.05). In the ischemia-five-day hypothermia group, but not ischemia-three-day, numbers of microglia remained higher than in sham controls (p < 0.05). In conclusion, delayed cerebral hypothermia partially protected white matter after global cerebral ischemia in fetal sheep. Extending cooling from 3 to 5 days did not further improve outcomes, and may be associated with greater numbers of residual microglia.

Subclinical decelerations during developing hypotension in preterm fetal sheep after acute on chronic lipopolysaccharide exposure.

Subclinical (shallow) heart rate decelerations occur during neonatal sepsis, but there is limited information on their relationship with hypotension or whether they occur before birth. We examined whether subclinical decelerations, a fall in fetal heart rate (FHR) that remained above 100 bpm, were associated with hypotension in preterm fetal sheep exposed to lipopolysaccharide (LPS). Chronically-instrumented fetal sheep at 0.7 gestation received continuous low-dose LPS infusions (n = 15, 100 ng/kg over 24 h, followed by 250 ng/kg/24 h for 96 h) or saline (n = 8). Boluses of 1 µg LPS or saline were given at 48 and 72 h. FHR variability (FHRV) was calculated, and sample asymmetry was used to assess the severity and frequency of decelerations. Low-dose LPS infusion did not affect FHR. After the first LPS bolus, 7 fetuses remained normotensive, while 8 developed hypotension (a fall in mean arterial blood pressure of ≥5 mmHg). Developing hypotension was associated with subclinical decelerations, with a corresponding increase in sample asymmetry and FHRV (p < 0.05). The second LPS bolus was associated with similar but attenuated changes in FHR and blood pressure (p < 0.05). In conclusion, subclinical decelerations are not consistently seen during prenatal exposure to LPS, but may be a useful marker of developing inflammation-related hypotension before birth.

Non-additive effects of delayed connexin hemichannel blockade and hypothermia after cerebral ischemia in near-term fetal sheep.

Hypothermia is partially neuroprotective after neonatal hypoxic-ischemic encephalopathy. Blockade of connexin hemichannels can improve recovery of brain activity and cell survival after ischemia in near-term fetal sheep. In this study, we investigated whether combining delayed hypothermia with connexin hemichannel blockade with intracerebroventricular infusion of a mimetic peptide can further improve outcomes after cerebral ischemia. Fetal sheep (0.85 gestation) received 30 minutes of cerebral ischemia followed by a 3-hour recovery period before treatment was started. Fetuses were randomized to one of the following treatment groups: normothermia (n=8), hypothermia for 3 days (n=8), connexin hemichannel blockade (50 µmol/L intracerebroventricular over 1 hour followed by 50 µmol/L over 24 hours, n=8) or hypothermia plus hemichannel blockade (n=7). After 7 days recovery, hypothermia was associated with reduced seizure burden, improved electroencephalographic (EEG) power, and a significant increase in neuronal and oligodendrocyte survival and reduced induction of Iba1-positive microglia. In contrast, although hemichannel blockade reduced seizure burden, there was no effect on EEG power or histology (P<0.05). There was no further improvement in outcomes with combined hypothermia plus hemichannel blockade. In conclusion, these data show that there is no additive neuroprotection with combined hypothermia and hemichannel blockade after cerebral ischemia in near-term fetal sheep.

How long is too long for cerebral cooling after ischemia in fetal sheep?

Therapeutic hypothermia can partially reduce long-term death and disability in neonates after hypoxic-ischemic encephalopathy. The aim of this study was to determine whether prolonging the duration of cooling from 3 days to 5 days could further improve outcomes of cerebral ischemia in near-term fetal sheep. Fetal sheep (0.85 gestation) received 30 minutes bilateral carotid artery occlusion followed by 3 days of normothermia (n = 8), 3 days of hypothermia (n = 8), or 5 days of hypothermia (n=8) started 3 hours after ischemia. Sham controls received sham ischemia followed by normothermia (n = 8). Cerebral ischemia was associated with profound loss of electroencephalography power and spectral edge, with greater and more rapid recovery in both hypothermia groups (P<0.05). Ischemia was associated with severe loss of neurons in the cortex, hippocampus and thalamus (P < 0.05), with a significant improvement in both hypothermia groups. However, the ischemia-3-day hypothermia group showed greater neuronal survival in the cortex and dentate gyrus compared with ischemia-5-day hypothermia (P < 0.05). Ischemia was associated with induction of iba1-positive microglia, which was attenuated in both hypothermia groups (P < 0.05). Extending the duration of delayed therapeutic hypothermia from 3 to 5 days did not improve outcomes after severe ischemia, and was associated with reduced neuronal survival in some regions.

Spontaneous pre-existing hypoxia does not affect brain damage after global cerebral ischaemia in late-gestation fetal sheep.

There is considerable evidence that a mild, non-injurious insult can protect (precondition) against a subsequent injurious insult. Typically, protection is seen when the gap between insults is several days to a week. However, the effect of mild but persistent hypoxia is unknown. In this study we examined the hypothesis that mild pre-existing hypoxia (PaO2<17 mm Hg) would reduce neural injury in chronically instrumented late-gestation (0.85 gestation) fetal sheep exposed to 30 min of global cerebral ischaemia induced by bilateral carotid artery occlusion (normoxia: n=9 vs. pre-existing hypoxia: n=9) or normoxia plus sham ischaemia (sham controls: n=9). Histopathology was assessed after 7 days of recovery. Fetuses with pre-existing hypoxia had lower PaO2 values (16.1±0.6 vs. 26.0±1.1 mm Hg) and were lighter at post-mortem (4,033±412 vs. 5,261±238 g) compared to normoxic fetuses. Cerebral ischaemia was associated with secondary cortical oedema and seizures, reduced final EEG power, loss of sleep state cycling, and significant loss of neurons and oligodendrocytes, with no significant effect of pre-existing hypoxia. Pre-existing hypoxia was associated with a significantly attenuated rise in mean arterial pressure between 18 and 36 h and slower resolution of cortical oedema between 96 and 150 h after ischaemia. These data suggest that chronic hypoxia is not associated with a significant preconditioning effect.