Interventions for Perinatal Hypoxic–Ischemic Encephalopathy

Oxygen-free Radical Inhibitors and Scavengers

A free radical is an atom or a molecule that contains an uneven number of electrons in its outer most orbital. A free radical, such as ·O₂⁻ or ·OH⁻, can combine sequentially with nonradicals, the result of which are new free radicals. This characteristic enables free radicals to initiate and perpetuate chain reactions, the peroxidation of unsaturated fatty acids being a prominent example. In this regard, it is assumed that the manner in which oxygen-free radicals cause or contribute to brain damage relates to their ability to attack the fatty acid moiety of plasma and subcellular membranes.^15-17 Polyunsaturated fatty acids seem especially prone to peroxidative attack by free radicals, which initiate and perpetuate chain reactions within the hydrophobic core of the lipid bilayer, leading ultimately to membrane fragmentation. The brain is especially rich in polyunsaturated phospholipids and, accordingly, is susceptible to free radical attack. Free iron (Fe; iron not bound to protein) and nitric oxide (see below) are important contributors to oxidative injury, because they transform mildly reactive oxygen species to more toxic free radicals.^17,18 Prevention of the formation of these reactive products can be achieved by elimination of hydrogen peroxide or superoxide, or the catalyst of the reaction, specifically Fe.

Because all biological systems generate oxygen-free radicals even under physiologic conditions, enzymes are present within the cell to protect its constituents from the oxidizing effect of hydrogen peroxide and its metabolic products; these enzymes include superoxide dismutase, endoperoxidase, and catalase, which convert hydrogen peroxide to either water or stable oxygen. Additional defenses are provided by endogenous scavengers, which include cholesterol, α-tocopherol (vitamin E), ascorbic acid (vitamin C), and thiol-containing compounds, notable glutathione. Thus, cells including neurons are capable of rapidly destroying free radicals, once formed, via both enzymatic and nonenzymatic quenching.

Oxygen-free radicals are generated during and after hypoxia–ischemia in several ways. First, free radicals are produced within mitochondria when cytochrome oxidase is not fully saturated with oxygen, thereby liberating free radicals at more proximal steps. These oxygen-free radicals cannot be consumed further and leak out into the cytoplasm. Other sources of oxygen-free radicals during hypoxia–ischemia and especially during the reperfusion interval of recovery are as by-products in the synthesis of prostaglandins from arachidonic acid^19-21 and the conversion of hypoxanthine to xanthine and uric acid.^16,22,23 Therefore, drugs that inhibit the formation of oxygen-free radicals or that rapidly destroy free radicals, once formed, might be efficacious in reducing the severity of hypoxic–ischemic brain damage.

Free radicals and reactive oxygen species (superoxide and hydrogen peroxide) cause tissue injury only when the radicals exceed the brain's endogenous antioxidant defenses. The newborn human infant, especially the premature infant, might be particularly susceptible to free radical injury because of a relative deficiency in the brain's antioxidants, including the enzymes superoxide dismutase and glutathione peroxidase.^24,25 The premature infant also has low circulating levels of glutathione and a relative inability to sequester Fe because of low transferrin levels.²⁶

One therapeutic approach to the early destruction of oxygen-free radicals generated during and after hypoxia–ischemia has been the administration of specific enzymes known to degrade highly reactive radicals to nonreactive compounds. The administration of the antioxidant enzymes superoxide dismutase and catalase conjugated to polyethylene glycol has been shown to reduce hypoxic–ischemic brain damage and to maintain the stability of the blood–brain barrier.^27-29 The enzymes are conjugated to polyethylene glycol to prolong their half-life and to allow improved penetration into and across the endothelial cell layer of the blood–brain barrier. Unfortunately, these large molecules are restricted primarily to the vascular compartment, where they contribute to the destruction of reactive oxygen species generated within blood during reperfusion after hypoxia–ischemia. The prolonged interval required for the conjugated enzymes to penetrate the blood–brain barrier into brain parenchyma precludes their clinical usefulness in reperfusion injury. In this regard, the enzymes have a narrow therapeutic dosage range and are generally protective only when administered many hours before the hypoxic–ischemic insult.

Drugs potentially efficacious in favorably influencing the outcome of hypoxic–ischemic brain damage include agents that inhibit specific reactions in the production of prostaglandins and of xanthine, the formation of which involves the generation of oxygen-free radicals (see above). Both allopurinol and oxypurinol, which are xanthine oxidase inhibitors/scavengers, protect immature rats from hypoxic–ischemic brain damage even when the drugs are administered early during the recovery phase after resuscitation.^30-32 Indomethacin, a cyclooxygenase and phospholipase inhibitor, also has been shown to ameliorate ischemic brain damage, at least in adult animals,³³ and substantially reduces free radical generation during reperfusion from hypoxia–ischemia in newborn pigs.³⁴ Finally, members of a class of compounds called 21 amino-steroids (lazeroids) have been shown to be neuroprotective in both adult and immature animal models of cerebral hypoxia–ischemia.^35,36 The compounds apparently prevent Fe-dependent lipid peroxidation by scavenging peroxyl radicals, and their site of action is predominantly within cerebral blood vessels, thereby reducing reperfusion injury.

Recent evidence also suggests that circulating and endogenous inflammatory cells act as mediators of hypoxic–ischemic injury in the immature brain, presumably through the production of oxygen-free radicals.^37-39 Platelet-activating factor, a potent phospholipid inflammatory mediator, is synthesized in the brain, and its concentration is increased during cerebral ischemia.⁴⁰Furthermore, Liu et al⁴⁰ have shown in immature rats that the platelet-activating factor antagonist BN 52021 attenuates hypoxic–ischemic brain damage. The agent was efficacious even when given immediately after reperfusion and again at 2 hours after hypoxia–ischemia.

Excitatory Amino Acid Antagonists

Several lines of research in experimental animals have implicated a role for the excitatory amino acid glutamate in the production of hypoxic–ischemic brain damage in the immature and adult brain. First, glutamate is directly toxic to mature neurons in culture.⁴¹ Second, neurons in culture and hippocampal slices die on exposure to anoxia, but their death can be prevented by the presence of magnesium (Mg⁺⁺), which blocks glutamate receptors within the calcium (Ca⁺⁺) ion channel, or by specific glutamate antagonists.^42-44 Third, direct injection of glutamate or glutamate agonists into specific regions of brain in vivo produces neuronal injury identical to that seen after hypoxia–ischemia.^45-47 Fourth, deafferentation of the glutaminergic excitatory input into the hippocampus reduces the damage produced by hypoxia–ischemia.⁴⁸ These studies provide convincing evidence that excessive exposure of neurons to glutamate, as occurs during hypoxia–ischemia, leads to morphologic alterations characteristic of ischemic neuronal necrosis.

Given the premise that excessive stimulation of neuronal surface receptors by glutamate promotes cellular death, and that glutamate release from the axon terminal into the synaptic cleft occurs during hypoxia–ischemia,^41,49-51 it has been rational to search for pharmacologic agents that would either inhibit glutamate release or block its postsynaptic action. Inhibitors of glutamate release from the nerve terminal (eg, baclofen) have not been investigated as potential neuroprotective drugs, whereas antagonists of glutaminergic cell surface receptors have undergone extensive study in experimental animals.^52,53 Of the several available antagonists, those that affect the NMDA and AMPA/QA receptors or the ion channels they subserve have received the most attention. Available compounds include phencyclidine, dextromethorphan, ketamine, MK-801, and NBQX, among others. These compounds have been found efficacious in reducing the extent of hypoxic–ischemic brain damage in adult animals even when administered up to 24 hours after the metabolic insult.⁵⁴Experimental studies in immature animals have shown that, as in adult animals, glutamate receptor antagonists are capable of reducing the severity of hypoxic–ischemic brain damage. First, systemically administered MK-801 prevents the tissue necrosis produced by the direct injection of the glutamate agonist NMDA into the striatum of 7-day postnatal rats.⁵⁵ Second, MK-801 protects against hypoxic–ischemic brain damage in immature rats even when the drug is administered during the course of or up to 1 hour after the metabolic insult.^55-59 Pretreatment of the animals entirely prevents tissue injury.⁵⁸ The protective effect of MK-801 or MBQX, which exceeds by far the effect of other glutamate receptor antagonists, appears greater in immature rats compared with their adult counterparts.^55,60 The age-related difference in the efficacies of the glutamate receptor antagonists favoring the immature brain appears to reside in the sensitivity of the developing brain to excitatory neurotransmitter toxicity, which in turn arises from developmental alterations in the density and distribution of glutamate receptor subtypes,^61-63 in glutamate binding to its receptors, or in transmembrane biochemical events initiated by receptor activity. Indeed, it has been suggested that the NMDA and AMPA receptor antagonists presently are the most potent drugs available to ameliorate the potential devastating effect of cerebral hypoxia–ischemia.⁵³

As noted previously, the divalent cation Mg⁺⁺ acts as a glutamate receptor antagonist to the extent that it blocks the neuronal influx of Ca⁺⁺ within the ion channel. In this regard, magnesium sulfate has been shown to reduce the severity of hypoxic–ischemic brain damage in immature rats.^64-66However, the solution does not improve neuropathologic outcome when administered before and during the course of asphyxia produced by umbilical cord occlusion in near-term fetal lambs.⁶⁷ In this model, magnesium caused no greater alterations in systemic blood pressure, heart rate, or cerebral blood flow than that seen in nontreated fetuses subjected to asphyxia. In the clinical setting, a retrospective study has suggested that premature fetuses whose mothers received magnesium sulfate for the treatment of preeclampsia or as a tocolytic agent are less likely to develop cerebral palsy compared with a gestational age-matched group of fetuses not exposed to the drug.⁶⁸ Based on the investigation, Nelson and Grether speculated that magnesium sulfate might provide a protective effect against brain damage in immature fetuses and newborn infants. Indeed, Levene et al currently are conducting a controlled, randomized trial to determine the protective effect of magnesium sulfate in asphyxiated full-term infants.⁶⁹

Calcium Channel Blockers

Because of its multiple functions, calcium (Ca⁺⁺) often is considered an intracellular second messenger. The divalent cation is intimately involved as a cofactor in numerous biochemical reactions, thereby acting as a regulator of cellular metabolic homeostasis. Accordingly, a disruption of intracellular free Ca⁺⁺ concentrations has wide-ranging deleterious effects on neuronal function.^15,49

The mechanisms by which alterations in Ca⁺⁺ balance that occur during cerebral hypoxia–ischemia contribute to brain damage relate to disturbances in those biochemical reactions subserved by the cation.⁷⁰ Ca⁺⁺ activates numerous intracellular reactions, the continued stimulation of which by elevated concentrations of free Ca⁺⁺ compromises the viability of the neurons. These reactions include the activation of several lipases, proteases, and endonucleases, all of which attack the structural integrity of the cell. Ca⁺⁺ also activates phospholipase C, which promotes a progressive breakdown in the phospholipid components of the plasma and subcellular membranes. Ca⁺⁺ also contributes to the formation of oxygen-free radicals via the formation of xanthine and prostaglandins (see above). Finally, increased concentrations of intracellular Ca⁺⁺ lead to an uncoupling of oxidative phosphorylation within mitochondria, because the energy formed during recovery from hypoxia–ischemia is consumed immediately in an attempt to reverse and then maintain the electrochemical (ion) gradient across the mitochondrial membrane. This futile cycling of ions restricts the production and transfer of ATP into the cytosol to be used for structural repair and reestablishment of ion gradients across the plasma membrane. Taken together, the toxic effects of excessive free Ca⁺⁺ accumulation are adequate to cause membrane disintegration and death of the neuron.^49,70

Given the potential neurotoxicity of Ca⁺⁺ when free intracellular concentrations increase to dangerous levels, drugs have been developed that inhibit Ca⁺⁺ influx into neurons. Of the numerous calcium channel blockers currently available for experimental and clinical research, flunarizine and nimodipine appear most efficacious in reducing the extent of hypoxic–ischemic brain damage, at least in adult animals.^71-73 Furthermore, several investigators independently have shown an amelioration of neuropathologic alterations in immature rats subjected to hypoxia–ischemia and pretreated with the Ca⁺⁺ channel blocker flunarizine.^74-76 However, the neuroprotective effect of calcium channel blockers is not nearly as great as that of the excitatory amino acid antagonists in the experimental setting, and any efficacy of the calcium channel blockers in the clinical setting has been marginal at best.⁵³ Also in the clinical setting, Levene et al⁷⁷ administered the calcium channel blocker nicardipine to four severely asphyxiated newborn infants. Heart rate increased in all four infants, whereas mean arterial blood pressure decreased in three. Two infants had a sudden and dramatic fall in systemic blood pressure. As a result of their findings, the investigators cautioned against the use of these drugs in asphyxiated infants.

Inhibitors of Nitric Oxide Production

Recently, experiments suggest that the free radical gas nitric oxide (NO) is involved in the cascade of metabolic events that causes or contributes to the occurrence of hypoxic–ischemic brain damage.^78-80 NO is produced in selective neurons of the brain, and the pathway for its synthesis involves the direct conversion of l-arginine to citrulline by the catalytic, cytosolic enzyme NO synthase. NO production is linked to the activation of glutamate cell surface receptors, especially NMDA receptors, the activation of which leads to Ca⁺⁺ influx into neurons and its binding to calmodulin (see above). Once formed, NO influences numerous metabolic events, primarily through an activation of the second messenger enzyme guanylate cyclase with the formation of cyclic GMP. In excessive concentrations, NO can act as a neurotoxic agent and might constitute a final common pathway for amino acid excitotoxicity in the brain, as has also been proposed for Ca⁺⁺.¹⁵ Experiments in adult animals suggest that NO mediates neuronal death after cerebral ischemia,^81,82 and that the severity of neuronal loss can be reduced by the previous administration of inhibitors of NO synthase activity.^83-85 A similar protective effect also has been observed in the immature rat subjected to cerebral hypoxia–ischemia,^86,87 but NO synthase inhibition appears to accentuate neuronal injury in fetal sheep.⁸⁸

The mechanisms whereby NO functions as a neurotoxin are numerous. Being a free radical, NO can react with other free radicals to form even more reactive species, including the hydroxyl free radical.¹⁸ NO also activates the glycolytic enzyme glyceraldehyde 3-phosphate dehydrogenase.⁸⁹ Paralysis of the glycolytic pathway would curtail the cytosolic production of ATP and prevent the production of reducing equivalents available to mitochondria for additional energy production. NO also inhibits components of the mitochondrial electron transport chain as well as the tricarboxylic acid cycle enzyme aconitase.^90,91 Thus, the neurotoxic effect of NO in part relates to its capacity to disrupt oxidative metabolism.

Monosialogangliosides

Monosialogangliosides or glycosphingolipids are found in high concentrations in the brain and are important constituents of cellular membranes. When administered systemically, the monosialoganglioside GM₁ crosses the blood–brain barrier and is incorporated into neuronal cell membranes.⁹² Recently, it has been demonstrated that both pre- and posttreatment with GM₁ protects the near-term sheep fetus from hypoxic–ischemic brain damage.^93,94 When given as an infusion over 6 hours beginning immediately after reperfusion, the ganglioside improves recovery of cerebral edema and reduces neuronal injury, especially in cerebral cortex, hippocampus, and striatum. The exact mechanism whereby GM₁ protects the brain from hypoxic–ischemic injury is unknown, but possibly its incorporation into cellular membranes results in a stabilization of membrane integrity and function.⁹⁴

Growth Factors

Many factors are important for normal growth and maturation of the brain. Accordingly, it is not surprising that such growth factors would be altered by cerebral hypoxia–ischemia, and that their administration might be neuroprotective. Thus far, nerve growth factor has been shown to reduce the severity of hypoxic–ischemic brain damage in the immature rat.⁹⁵ Other growth factors might be similarly efficacious.

Glucocorticosteroids: A Special Case

As briefly mentioned previously, glucocorticosteroids have been used previously in human infants, children, and adults to reduce the cerebral edema that arises from hypoxia–ischemia. However, controlled clinical studies on the use of either low- or high-dose steroids, predominantly in adult patients suffering traumatic coma, indicate that these drugs fail to attenuate increases in intracranial pressure or improve ultimate neurologic outcome.^96-98Experimental studies also suggest that glucocorticosteroid therapy is ineffective in reducing either the cerebral edema or the ultimate neuropathologic alterations that accompany ischemic stroke in adult animals.^99-101 Relevant to the perinatal brain, Altman et al¹⁰² administered dexamethasone (40 mg/kg) to immature rats immediately before the onset of cerebral hypoxia–ischemia. The results showed that not only was the extent of brain damage not different in the steroid-treated animals compared with nontreated controls, but also that mortality in the treated rat pups was greater than that in the control animals (but see Reference 103). Accordingly, based on both clinical and experimental data, glucocorticosteroids administered either shortly before or after cerebral hypoxia–ischemia does not improve neurologic or neuropathologic outcome and might increase morbidity and mortality.

Remote or chronic administration of glucocorticoids now appears to have a protective benefit. Barks et al¹⁰⁴ found a protective effect of dexamethasone on hypoxic–ischemic brain damage in the immature rat when the drug was administered ≥24 hours before the metabolic insult. A single low dose (0.1 mg/kg) given 24 hours before hypoxia–ischemia prevented cerebral infarction. A single dose given 0 to 3 hours before hypoxia–ischemia was not effective (see also Reference 102). In a more recent study, Chumas et al¹⁰⁵showed that pretreatment of immature rats with dexamethasone 6 hours before hypoxia–ischemia also offered protection with no infarction. Additional studies by the same research group have shown that the improved neuropathologic outcome afforded to the dexamethasone-treated immature rats is not the result of improved cerebral blood flow during hypoxia–ischemia.¹⁰⁶ Furthermore, the mild hyperglycemia observed in the steroid-treated animals does not account for the cerebral protection, and there also is no induction of antioxidant enzymes.¹⁰⁷

The mechanism(s) by which glucocorticosteroid therapy protects the developing brain from hypoxic–ischemic brain damage has yet to be elucidated. Given the fact that the drug is most efficacious when given ≥24 hours before cerebral hypoxia–ischemia in the immature rat, the interval between the treatment and the onset of the metabolic stress is adequate to allow for some form of molecular or cellular adaptation that offers protection to occur. A somewhat analogous situation occurs with hypoxic preconditioning (see below). Additional experiments hopefully will elucidate the mechanism whereby glucocorticosteroids provide such a dramatic protective effect on the immature brain subjected to hypoxia–ischemia. However, it must be kept in mind that an interval of 1 day in the developing rat brain is probably equivalent to 1 month in the human fetus or newborn infant. On the other hand, glucocorticosteroids reduce the incidence and severity of respiratory distress syndrome when administered to premature newborn infants within 24 hours of delivery and—directly or indirectly—reduce the risk of periventricular/intraventricular hemorrhage.^108-111

Phenobarbital: Another Special Case

Numerous investigations have indicated that barbiturate pretreatment and even early posttreatment of adult animals subjected to cerebral hypoxia–ischemia reduces the severity of ultimate brain damage.^112-115 Experiments also suggest that barbiturates protect the fetus and newborn animal against asphyxia by prolonging survival and by preventing or reducing the subsequent development of hypoxic–ischemic brain injury.^116-118 The type of barbiturate used in all of these experimental studies have been of the short-acting variety, specifically, thiopental or pentobarbital. The mechanism(s) of the neuroprotection relates predominantly to an overall suppression of cerebral oxidative metabolism, as reflected in a reduction in oxygen consumption and a slower depletion of energy stores during hypoxia–ischemia or ischemia.^119-121 Barbiturates might also blunt cerebral excitotoxicity by depressing glutamate responses within the brain.¹²²

High-dose barbiturates have been administered to newborn infants sustaining cerebral hypoxia–ischemia before or at the time of delivery.^123-125 In a controlled clinical study, Goldberg et al¹²⁴ randomly assigned 32 full-term, severely asphyxiated newborn infants to barbiturate-treated and control groups. All newborn infants exhibited evidence of hypoxic–ischemic encephalopathy and required mechanical ventilation. Thiopental was begun at a mean postnatal age of 2 hours and was given as an infusion for 24 hours. Despite barbiturate therapy, no significant difference in the frequency of seizures or in elevations in intracranial pressure was noted between the two groups. Early treatment with thiopental did not improve neonatal mortality or neurologic morbidity at 12 months of age. Of additional importance was the fact that systemic hypotension occurred significantly more often in the treated group, requiring greater vasopressor support in these infants. The investigators concluded that barbiturate therapy offers little benefit to the previously asphyxiated newborn infant and might actually perpetuate existing cardiovascular derangements.

However, a recent clinical investigation by Hall et al¹²⁶has demonstrated a beneficial effect of high-dose phenobarbital on neurologic outcome in severely asphyxiated full-term newborn infants. Twenty asphyxiated newborn infants received an intravenous infusion of phenobarbital (40 mg/kg) between 1 and 6 hours after birth. Twenty infants served as controls. Thirty-one infants successfully completed the study (15 treated, 16 controls). Control infants received phenobarbital (20 mg/kg) only if and when clinically apparent seizures occurred. No difference in the frequency of seizures was seen in the two groups. In addition, no adverse effects on heart rate, respiratory rate, blood pressure, or arterial blood gases were observed in the high-dose phenobarbital-treated group. Three-year follow-up revealed normal neurologic outcome in 10 of 15 infants in the treatment group but in only 3 of 16 infants in the control group (P < .05). From the findings, the authors concluded that early high-dose phenobarbital therapy to severely asphyxiated full-term newborn infants appears to be safe and is associated with significant improvement in neurologic outcome at 3 years of age.

The clinical investigation of Hall et al¹²⁶ is important in several respects. First, phenobarbital is a drug frequently used by neonatologists for the treatment of seizures in newborn infants. Second, no adverse systemic physiologic effects were noted. Third, the drug appeared to be efficacious in ultimately reducing the severity of hypoxic–ischemic brain damage, at least from a functional perspective. A controlled, prospective study with a larger number of asphyxiated newborn infants would be worthwhile.

Hyperglycemia

Several investigations have indicated that hyperglycemia superimposed on cerebral hypoxia–ischemia or isolated ischemia accentuates brain damage in adult experimental animals and humans.^127-130 Such is not the case in perinatal animals, at least in the immature rat. In this regard, Vannucci and Mujsce¹³¹ demonstrated that hyperglycemia to blood glucose concentrations in the range of 600 mg/dL entirely prevents the occurrence of brain damage in an immature rat model of cerebral hypoxia–ischemia. Hyperglycemia with blood glucose concentrations ranging from 300 to 400 mg/dL has no beneficial effect, nor is it deleterious to the brain subjected to hypoxia–ischemia.¹³²Additional investigations have shown that hyperglycemia superimposed on cerebral hypoxia–ischemia results in an enhanced anaerobic glycolytic flux compared with normoglycemic controls.¹³³ The enhanced glycolysis, in turn, leads to better preservation of cerebral high-energy reserves in the hyperglycemic animals, thus accounting for the greater resistance of these animals to hypoxic–ischemic brain damage. However, hyperglycemia has been shown to increase hypoxic–ischemic brain damage in newborn pigs,¹³⁴ in contrast to the findings in immature rats. In addition, glucose supplementation during recovery after resuscitation from cerebral hypoxia–ischemia appears to accentuate brain damage in immature rats.¹³⁵

In contrast to hyperglycemia, mild insulin-induced hypoglycemia is detrimental to immature rat brain subjected to hypoxia–ischemia.¹³⁶ However, if hypoglycemia is induced by fasting the animals for 12 hours, a high degree of protection is afforded to the brain during hypoxia–ischemia. The fasting is associated with increased concentrations of blood β-hydroxybutyrate and acetoacetate concentrations, which presumably serve as alternate substrates to the immature brain, thereby protecting it from hypoxic–ischemic damage.¹³⁶ It has been shown that ketone bodies enter the immature brain more readily than glucose;¹³⁷ their ready availability for oxidative metabolism even under conditions of cerebral hypoxia–ischemia provides the necessary reducing equivalents to preserve high-energy phosphate reserves during metabolic stress.

Based on the experimental studies described above, it appears prudent that blood glucose concentrations be maintained within a physiologic range during and after cerebral hypoxia–ischemia in fetuses and newborn human infants. Hypoglycemia appears to be deleterious, whereas the data regarding any potential beneficial effect of hyperglycemia during and after hypoxia–ischemia presently are controversial. Clinical investigations are necessary to determine whether an infusion of ketone bodies might protect the perinatal brain from hypoxic–ischemic injury.

Carbon Dioxide

Recent clinical investigations suggest that premature infants who require mechanical ventilation to prevent or minimize hypoxemia arising from respiratory distress syndrome are at increased risk for the development of periventricular leukomalacia if hypocapnia occurs during the course of respiratory management.^138-140 A causal relationship between low Paco₂ and hypoxic–ischemic brain damage was not established in any of the studies, but given the multiple effects of carbon dioxide (co₂) on hemodynamics and cerebral metabolism, a cause and effect relationship is certainly plausible. The question remained as to the contribution of hypocapnia to hypoxic–ischemic brain damage and whether hypercapnia is neuroprotective.

To resolve the issue, Vannucci et al¹⁴¹ subjected immature rats to cerebral hypoxia–ischemia with or withoutco₂ added to the hypoxic gas mixture to which the animals were exposed. The results showed that normocapnic (Pco₂ = 39 mm Hg) cerebral hypoxia–ischemia is associated with less severe brain damage than hypocapnic (Pco₂ = 26 mm Hg) hypoxia–ischemia, and that mild hypercapnia (Pco₂ = 54 mm Hg) is more protective than normocapnia. Additional studies were conducted that determined that mild hypercapnia is associated with greater preservations of cerebral blood flow, glucose utilization, and high-energy phosphate reserves than seen during either hypocapnic or normocapnic cerebral hypoxia–ischemia.¹⁴² From their findings, the investigators recommended additional corroboration in other animal models as well as as a clinical reappraisal of the ventilatory management strategies of sick newborn human infants, as regard especially the prevention of hypocapnia and the occurrence of permissive mild hypercapnia.

It must be emphasized that the apparent neuroprotective effect of mild hypercapnia seen in immature rats occurs in the setting of cerebral hypoxia–ischemia extending over 2.5 hours.¹⁴¹ It is also likely that the deleterious effect of hypocapnia seen in human premature infants occurs in the setting of cerebral hypoxia–ischemia extending over ≥1 hours. It remains to be determined whether hypercapnia is at least partially protective in the setting of acute intrapartum asphyxia occurring in full-term human newborn infants. In this regard, Low et al¹² found no evidence of newborn complications, including encephalopathy, in full-term fetuses in whom only a respiratory acidosis was apparent on umbilical blood acid-base analysis. A metabolic acidosis was associated with a high rate of newborn complications. Goodwin et al¹¹ showed that full-term newborn infants with respiratory acidemia at birth did not differ significantly with respect to newborn multiorgan injury compared with infants with similar umbilical blood pH, but there was a trend toward a lower incidence of hypoxic–ischemic encephalopathy (P = .06). In contrast, van den Berg et al¹⁴³ found more newborn complications, including encephalopathy, in a group of premature and full-term newborn infants born with combined respiratory and metabolic acidosis than in infants with a metabolic acidosis alone. Accordingly, it remains to be determined whether hypercapnia in the setting of acute asphyxia accentuates or reduces perinatal hypoxic–ischemic brain damage.

Hypothermia

Systemic or focal cooling of the brain by as little as 3° to 6°C has been shown to reduce the extent of tissue injury that follows several cerebral insults in adult experimental animals, including stroke, trauma, and hypoglycemia.^144-146 In contrast, hyperthermia to 38° to 39°C accentuates ischemic brain damage.^147-149 Hypothermia also protects the perinatal brain from hypoxic–ischemic damage. In this regard, Yager et al¹⁵⁰ demonstrated in immature rats that a reduction in systemic temperature by 3°C during cerebral hypoxia–ischemia provides partial benefit, and a 6°C decrease completely protects the brain from injury (see also Reference 151). Presumably, hypothermia reduces cerebral energy demands, such that during hypoxia–ischemia, high-energy phosphate reserves are maintained at relatively normal levels.¹⁵² Systemic hypothermia during the first 3 hours of reperfusion after hypoxia–ischemia also protects the immature rat brain from damage,¹⁵³ although this observation has not been universal.¹⁵⁰ Similar findings have been observed in newborn pigs.^154,155 Selective cooling of immature rat brain to temperatures comparable with that of mild systemic cooling (see above) also affords protection from hypoxic–ischemic injury.¹⁵⁶ It has been known for decades that systemic cooling of human infants to temperatures ranging from 16° to 24°C protects their brains from ischemic damage during total circulatory arrest for up to 90 minutes; this intervention is the mainstay of the operative correction of congenital heart defects.^157-159

Hypoxic Preconditioning

Recently, it has been demonstrated that immature rats subjected to cerebral hypoxia–ischemia sustain less brain damage if they were exposed previously to systemic hypoxia alone compared with animals not exposed previously to hypoxia.¹⁶⁰ The finding is comparable with the protective influence of ischemic preconditioning on subsequent ischemic brain damage in adult animals.^161-163 The underlying mechanism(s) for the neuroprotection of hypoxic preconditioning in the immature animal has not been elucidated but probably involves the induction of genes or proteins that favorably influence metabolic events that occur either during the course of hypoxia–ischemia or in the reperfusion period after resuscitation. That hypoxic preconditioning protects the immature brain from subsequent hypoxic–ischemic brain damage is reminiscent of the physiologic hypoxia to which the human newborn infants is exposed during fetal life. Whether the physiologic fetal hypoxia is a preconditioning event, protecting the fetus from subsequent hypoxic–ischemic brain injury, is conjectural. However, Vannucci and Duffy¹⁶⁴ have demonstrated that term fetal rats exposed to a total nitrogen atmosphere (anoxia) at birth survive twice as long as newborn rats similarly exposed despite an age difference of <24 hours.