Nonadditive Neuroprotection With Early Glutamate Receptor Blockade and Delayed Hypothermia After Asphyxia in Preterm Fetal Sheep
Background and Purpose—Hypothermia induced after perinatal hypoxia–ischemia is partially protective. This study examined whether early treatment with the noncompetitive N-methyl-d-aspartate receptor antagonist, dizocilpine, can augment neuroprotection with delayed hypothermia after severe asphyxia in preterm fetal sheep at 0.7 weeks gestation (equivalent to 28–32 weeks in humans).
Methods—Fifty minutes after umbilical cord occlusion for 25 minutes, fetuses were randomized to either dizocilpine (2 mg/kg estimated fetal weight intravenously, then 0.07 mg/kg/h for 4 hours) and then after 5.5 hours to whole-body cooling to 3°C below baseline, or sham cooling, until 72 hours, and euthanized 7 days after umbilical cord occlusion.
Results—Delayed hypothermia was associated with improved neuronal survival (P<0.02) and reduced microglia (P=0.004) and caspase-3-positive cells (P<0.01) compared with umbilical cord occlusion. Dizocilpine was associated with reduced microglia (P<0.05) but no effect on caspase-3 induction and improved survival only in CA1/2 (P<0.05) with no apparent additive effect with delayed hypothermia.
Conclusions—Early N-methyl-d-aspartate blockade and a clinical regime of delayed whole-body hypothermia provide nonadditive neuroprotection in the preterm brain.
There is increasing evidence from postmortem and imaging studies that acute subcortical neuronal injury associated in part with perinatal hypoxia–ischemia in preterm infants is an important contributor to long-term neurodevelopmental disability.1 Although hypothermia has not been evaluated in preterm infants, at term, induced mild cerebral hypothermia partially improves recovery from acute hypoxia–ischemia.2 Early anticonvulsant treatment in the P7 rat can extend the window of opportunity for delayed treatment with hypothermia.3,4 However, these studies used relatively short, less effective intervals of hypothermia. Thus, in this study, we evaluated whether early treatment with dizocilpine could augment neuroprotection with a clinical protocol of delayed, but prolonged, whole-body hypothermia after severe asphyxia in preterm fetal sheep. At this age brain development is comparable to humans at 28 to 32 weeks of gestation.5
A detailed description of methods is provided in the online-only Data Supplement.
Animal Groups and Surgery
All procedures were approved by the Animal Ethics Committee of the University of Auckland, Auckland, New Zealand. Romney/Suffolk fetal sheep (Annadale, New Zealand) were instrumented at 97 to 98 days of gestation (term=147 days), including placement of catheters for intravenous infusions and a cooling coil wrapped over the back and sides of the thorax. At 103 to 104 days gestation, fetuses received either sham umbilical cord occlusion (UCO, n=5) or UCO for 25 minutes. UCO fetuses were randomized to either UCO+vehicle (n=9), UCO+dizocilpine (Diz, n=7), UCO+hypothermia (Hypo, n=7), or UCO+Diz+Hypo (n=7).
Dizocilpine (dizocilpine hydrogen maleate; Sigma-Aldrich Pty Ltd, Sydney, Australia) or the same volume of vehicle was given in a 2 mg.kg−1 bolus followed by a 0.07 mg.kg.h−1 constant infusion through the brachial vein from 15 minutes after UCO until 4 hours.
Moderate whole body hypothermia (a 3°C reduction in esophageal temperature) was induced from 5.5 hours to 72 hours after the end of UCO by circulating cold water (10°C) through the cooling coil. After 72 hours, fetuses were allowed to rewarm over 2 hours and then euthanized 7 days after UCO.
Plasma samples were taken during and after UCO. Fetal extradural and esophageal temperatures, mean arterial blood pressure, and parietal electroencephalograms were recorded continuously; electroencephalographic seizures were analyzed offline.6
Cell counts of brain sections stained for NeuN (neuronal survival), cleaved caspase-3 (a measure of apopotosis), and Isolectin B4 (IB4, activated microglia) were performed in the caudate nucleus and putamen and in the CA1/2 and CA3 regions of the hippocampus using stereological principles (online-only Data Supplement Figure I).
Data were analyzed using mixed model analysis of variance followed by the Tukey post hoc test if a significant effect was found. Data are mean±SEM.
There were no significant differences in baseline blood gases, pH, glucose, or lactate concentrations before UCO (online-only Data Supplement Table I). UCO was associated with hypoxemia and severe mixed respiratory and metabolic acidosis, which resolved after release of UCO. Subsequently, PaO2 was higher in the UCO groups than sham controls. Hypothermia was associated with a small increase in pH and glucose levels that resolved after rewarming.
Extradural and esophageal temperatures increased during dizocilpine infusion from 1 to 5 hours after UCO (P<0.05; online-only Data Supplement Figure II). Cooling was associated with significantly reduced temperatures from 5.5 to 72 hours (P<0.05).
Mean Arterial Blood Pressure
UCO was associated with profound hypotension followed by rebound hypertension from 1 to 2 hours compared to sham controls (P<0.05; online-only Data Supplement Figure II) with no subsequent between-group differences.
Electroencephalographic Power (dB) and Electrographic Seizures
UCO was associated with suppressed electroencephalographic power compared with sham controls from 0.5 to 72 hours after UCO (P<0.05) followed by progressive recovery (online-only Data Supplement Figure II), which was similar between groups. High-amplitude, low-frequency evolving seizures developed after UCO (online-only Data Supplement Table II). Delayed hypothermia and dizocilpine were associated with a significant delay before onset of seizures and reduced mean amplitude but no significant difference in the total number or duration of events.
UCO was associated with severe loss of NeuN+ neurons in the striatal nuclei and the hippocampus (P<0.001). Delayed hypothermia was associated with improved neuronal survival in the striatum (P=0.02) and hippocampus (P=0.01; Figures 1 and 2). In contrast, dizocilpine infusion had no independent main effect in either the striatum or hippocampus (P=0.91 and P=0.97, respectively) with no additive effect. In the hippocampus, but not the striatal nuclei, there was a significant interaction of region with dizocilpine infusion (P<0.05), and post hoc analysis suggested significant protection with dizocilpine alone in CA1/2 Figure 3A–B P<0.05).
Induction of Microglia and Apoptosis in Striatal Nuclei
Delayed hypothermia was associated with reduced numbers of microglia (P=0.004) and caspase-3+ cells (P<0.01) compared with UCO. Dizocilpine infusion was associated with reduced numbers of microglia (P<0.05) but not caspase-3+ cells with no interaction with delayed hypothermia.
Experimental hypothermia is neuroprotective if initiated as soon as possible in the first 6 hours after hypoxia–ischemia.7 In practice, cooling is typically initiated 4 to 4.5 hours after birth.2 It is likely that this delay is a major factor underlying the modest clinical improvement in normal survival.2 Thus, there has been considerable interest in whether early drug therapy could extend the window of opportunity for treatment with hypothermia.3,4 In the present study, mild whole-body cooling in preterm fetal sheep, delayed by 5.5 hours and continued for 72 hours, was associated with partial neuroprotection. Although there was a striking improvement in neuronal survival in CA1/2, to near sham control values, CA3 and striatal nuclei were only partially protected. This improvement was associated with marked suppression of the inflammatory response to asphyxia and reduced cleaved caspase-3 expression.
In contrast, initiation of N-methyl-d-aspartate receptor blockade from 15 minutes after asphyxia was independently neuroprotective, but only in the CA1/2 region of the hippocampus. This was a more modest effect than previously reported after hypoxia–ischemia in neonatal rodents and preterm fetal sheep.6,8,9 The dizocilpine dose in the present study was titrated to compensate for transplacental passage and was associated with electroencephalographic suppression and delayed onset of seizures. Intriguingly, microglial activation was suppressed, consistent with evidence that activated microglia express the N-methyl-d-aspartate receptor and, in vitro, posthypoxic dizocilpine reduces microglial activation and expression of inflammatory factors.10
The finding that early drug treatment did not augment protection with delayed hypothermia contrasts with previous findings that treatment with anticonvulsants such as topiramate and phenobarbital started 15 minutes after hypoxia–ischemia in P7 rats synergistically increased neuroprotection with hypothermia started up to 3 hours later.3,4 It is possible that the contrasting findings may reflect differences in species, insult, or receptor-specific effects. Furthermore, we used a longer interval of delayed hypothermia, consistent with clinical protocols,7 which may have reduced the scope to demonstrate improvement with combined therapy. Earlier and mildly deeper cooling would be more protective and, given the increase in brain temperature with dizocilpine,7 might improve synergism, but at the cost of reduced scope to show combined effects.
In conclusion, these data suggest that early N-methyl-d-aspartate blockade and a clinical regime of delayed whole-body hypothermia provide nonadditive neuroprotection.
Source of Funding
This study was funded by the Health Research Council of New Zealand. Dr Barrett received the William Georgetti Scholarship.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.112.671982/-/DC1.
- Received July 25, 2012.
- Accepted July 30, 2012.
- © 2012 American Heart Association, Inc.
- 1.↵Committee on Understanding Premature Birth and Assuring Healthy Outcomes. Preterm birth: causes, consequences, and prevention. Available at: http://books.nap.edu//catalog/11622.html#toc. Institute of Medicine, The National Academies Press; 2007. Accessed June 1, 2010.
- Edwards AD,
- Brocklehurst P,
- Gunn AJ,
- Halliday H,
- Juszczak E,
- Levene M,
- Strohm B,
- Thoresen M,
- Whitelaw A,
- Azzopardi D
- Liu Y,
- Barks JD,
- Xu G,
- Silverstein FS
- Hattori H,
- Morin AM,
- Schwartz PH,
- Fujikawa DG,
- Wasterlain CG