119: Perinatal Asphyxia

1. Perinatal asphyxia is a condition of impaired blood gas exchange that, if persistent, leads to progressive hypoxemia and hypercapnia. Hypoxic-ischemic encephalopathy (HIE), which is a subset of neonatal encephalopathy (NE), can result from perinatal asphyxia.

2. Neonatal encephalopathy (NE) is clinically defined as a disturbance in neurologic function demonstrated by difficulty in maintaining respirations, hypotonia, altered level of consciousness, depressed or absent primitive reflexes, seizures, and poor feeding. NE does not imply HIE. NE may represent a metabolic disorder, infection, drug exposure, or neonatal stroke, but it is the preferred terminology to describe a depressed newborn from any cause at the time of birth.

3. In order for an acute intrapartum hypoxic event to be considered a cause of cerebral palsy (CP), the American Academy of Pediatrics (AAP) and the American College of Obstetrics and Gynecology (ACOG) define 4 essential criteria that must be met:

   1. Evidence of a metabolic acidosis in fetal umbilical cord arterial blood obtained at delivery (pH <7 and base deficit ≥12 mmol/L).

   2. Early onset of severe or moderate neonatal encephalopathy in infants born at 34 or more weeks of gestation.

   3. CP of the spastic quadriplegic or dyskinetic type.

   4. Exclusion of other identifiable etiologies, such as trauma, coagulation disorders, infectious conditions, or genetic disorders.

4. Criteria that collectively suggest an acute intrapartum hypoxic event (within 0–48 hours to labor and delivery) but are individually nonspecific to asphyxial insults:

   1. A sentinel hypoxic event occurring immediately before or during labor

   2. A sudden and sustained fetal bradycardia or the absence of fetal heart rate variability in the presence of persistent, late, or variable decelerations, usually after a hypoxic sentinel event when the pattern was previously normal

   3. Apgar scores of 0–3 beyond 5 minutes

   4. Onset of multisystem involvement within 72 hours of birth

   5. Early imaging showing evidence of acute nonfocal cerebral abnormality

Data regarding perinatal asphyxia rates from 2001 indicate a prevalence rate of 25 per 1000 term live births and 73 per 1000 preterm live births, with 15 and 50% moderate to severe cases, respectively. The overall incidence of NE attributable to perinatal asphyxia was ∼1.6 per 10,000 live births in the absence of preconception or antepartum abnormalities. European data indicate an overall CP rate of 2.08 per 1000 live births for the period 1980–1990, with increased rates seen in infants with birthweight <1500 g. While earlier studies showed a stable rate of CP between the 1950s and 1990s, more recent data indicate a decline in the prevalence of CP for preterm infants born after 1980, possibly reflecting improved perinatal care. Only 8–17% of CP in term infants is associated with adverse perinatal events suggestive of asphyxia; the cause of ≥90% of cases remains unknown.

1. Hypoxic-ischemic injury. The asphyxial event results in cerebral ischemia, which precipitates an immediate drop in cellular high-energy phosphate levels, termed primary energy failure. Glutamate, an excitatory amino acid, is also released in substantial amounts due to cellular depolarization. N-methyl-d-aspartate (NMDA) receptors are subsequently overstimulated by glutamate and result in increased intracellular calcium and necrotic cell death. Cerebral blood flow is restored in the reperfusion period with normalization of cellular energy levels within 2–3 hours after the insult. A latent phase follows and lasts 6–15 hours, during which oxidative metabolism returns to baseline, but secondary inflammation and cellular apoptosis are initiated. Without intervention, the latent phase can progress to secondary energy failure, which is characterized by cellular excitotoxicity, oxidative damage, and neuronal death within 3 days. The severity of the insult determines the extent of brain injury. Neurodevelopmental outcomes have specifically been correlated with the degree of secondary energy failure. Current neuroprotective therapies are therefore designed to intervene within the latent phase and before the onset of secondary energy failure.

2. Adaptive responses of the fetus or newborn to asphyxia. The fetus and neonate are much more resistant to asphyxia than adults. In response to asphyxia, the mature fetus redistributes blood flow to the heart, brain, and adrenals to ensure adequate oxygen and substrate delivery to these vital organs.

   1. Impairment of cerebrovascular autoregulation. Results from direct cellular injury and cellular necrosis from prolonged acidosis and hypercarbia.

   2. Majority of neuronal disintegration. Occurs after termination of the asphyxial insult because of persistence of abnormal energy metabolism and low adenosine triphosphate (ATP) levels (primary energy failure).

   3. Major circulatory changes during asphyxia (reperfusion phase):

      1. Loss of cerebrovascular autoregulation under conditions of hypercapnia, hypoxemia, or acidosis, cerebral blood flow (CBF) becomes “pressure passive,” leaving the infant at risk for cerebral ischemia with systemic hypotension and cerebral hemorrhage with systemic hypertension.

      2. Increase in cerebral blood flow (occurs in phase of secondary energy failure) secondary to redistribution of cardiac output, initial systemic hypertension, loss of cerebrovascular autoregulation, and local accumulation of vasodilator factors (H⁺, K⁺, adenosine, and prostaglandins).

      3. In prolonged asphyxia, there is a decrease in cardiac output, hypotension, and a corresponding fall in CBF. In general, brain injury occurs only when the asphyxia is severe enough to impair CBF.

      4. The postasphyxia newborn is in a persistent state of vasoparalysis and cerebral hyperemia, whose severity is correlated with the severity of the asphyxial insult. Cerebrovascular hemorrhage may occur on reperfusion of the ischemic areas of the brain. However, when there has been prolonged and severe asphyxia, local tissue recirculation may not be restored because of collapsed capillaries in the presence of severe cytotoxic edema.

3. Neurophysiology

   1. Cerebral edema is a consequence of extensive cerebral necrosis rather than a cause of ischemic cerebral injury.

   2. Regional central nervous system (CNS) vulnerability changes with postconceptional age (PCA) and as the infant matures.

      1. Periventricular white matter is most severely affected in infants <34 weeks' PCA. The “watershed” areas between the anterior and middle cerebral arteries and between the middle and posterior cerebral arteries are predominantly involved in term infants.

      2. Areas of brain injury in profound asphyxia correlate temporally and topographically with the progression of myelination and of metabolic activity within the brain at the time of the injury. White matter, therefore, is more susceptible to hypoxic injury.

      3. The topography of brain injury observed in vivo corresponds closely to the topography of glutamate receptors.

      4. When cerebral blood flow is increased in response to asphyxia, regional differences exist such that there is relatively more blood flow to the brainstem than to higher cerebral structures.

4. Neuropathology. Experimental models in animal studies have been used extensively in the study of human asphyxia to establish the basic physiology of the CNS injury. Findings in humans include the following:

   1. Cortical edema, with flattening of cerebral convolutions, is followed by cortical necrosis until finally a healing phase results in gradual cortical atrophy, and may result in microcephaly.

   2. Selective neuronal necrosis is the most common type of injury observed in neonatal HIE. The pathogenesis most likely involves hypoperfusion and reperfusion with injury promulgated by glutamate.

   3. Other findings in infants include status marmoratus, a marbled histopathological appearance caused by hypermyelination of the basal ganglia and thalamus, and parasagittal cerebral injury (bilateral and usually symmetric) with the parieto-occipital regions affected more often than regions anteriorly.

   4. Periventricular leukomalacia (PVL) is hypoxic-ischemic necrosis of periventricular white matter resulting from cerebral hypoperfusion and the vulnerability of the oligodendrocyte within the white matter to free radicals, excitotoxin neurotransmitters, and cytokines. Injury to the periventricular white matter is the most significant problem contributing to long-term neurologic deficit in the premature infant, although it does occur in sick full-term infants as well. The incidence of PVL increases with the length of survival and the severity of postnatal cardiorespiratory disturbances. PVL involving the pyramidal tracts usually results in spastic diplegic or quadriplegic CP. Visual perception deficits may result from involvement of the optic radiation.

   5. Porencephaly, hydrocephalus, hydranencephaly, and multicystic encephalomalacia may follow focal and multifocal ischemic cortical necrosis, PVL, or intraparenchymal hemorrhage.

   6. Brainstem damage is seen in the most severe cases of hypoxic-ischemic brain injury and results in permanent respiratory impairment.

Perinatal asphyxia can occur in the antepartum, intrapartum, or postnatal period.

1. Antepartum risk factors such as maternal trauma, maternal hypotension, and uterine hemorrhage account for 20% of HIE cases.

2. About 70% of cases are estimated to occur in the intrapartum period due to factors (sentinel events) such as placental abruption, umbilical cord prolapse, uterine rupture, and conditions associated with placental vascular insufficiency (maternal diabetes, intrauterine growth restriction, preeclampsia, and multiple gestation).

3. Postnatal risk factors, which account for 10% of HIE cases, are due to cardiorespiratory failure and congenital heart disease.

Perinatal asphyxia can result in CNS injury alone (16% of cases), CNS and other end-organ damage (46%), isolated non-CNS organ injury (16%), or no end-organ damage (22%).

1. CNS injury in severe cases of HIE manifest with variable clinical signs that evolve over time:

   1. Birth–12 hours. Deep stupor or coma, respiratory failure or periodic breathing, diffuse hypotonia, intact pupillary and oculomotor responses, and subtle or focal clonic seizures by 6–12 hours in term infants. Preterm infants can present with generalized tonic seizures.

   2. 12–24 hours. The level of alertness can appear to improve in less critical cases of brain injury. However, severe seizures, marked jitteriness, and apnea also present at this time. Term infants can present with weakness of the proximal upper limbs, while preterm infants have lower extremity weakness.

   3. 24–72 hours. The consciousness level worsens leading to deep stupor and coma, leading to respiratory failure. Pupillary and oculomotor disturbance are now present due to brainstem involvement. Death due to HIE most often occurs at this time with a median time of 2 days. Preterm infants who die at this time often have severe intraventricular hemorrhage (IVH) and periventricular hemorrhagic infarction.

   4. After 72 hours. Mild to moderate stupor may persist, but the overall level of alertness improves. Diffuse hypotonia may persist or hypertonia can become evident. Feeding difficulties become obvious due to abnormal sucking, swallowing, and tongue movements.

2. Non-CNS multi-organ dysfunction can present as follows:

   1. Renal. Acute tubular necrosis can present with hematuria or renal insufficiency or failure.

   2. Pulmonary. Respiratory failure and meconium aspiration are due to fetal distress and persistent pulmonary hypertension.

   3. Cardiac. Myocardial dysfunction and congestive heart failure may result in arrhythmias and hypotension.

   4. Hepatic. Abnormal liver enzymes, elevated serum bilirubin, and decreased coagulation factors secondary to hepatic dysfunction.

   5. Hematologic. Thrombocytopenia due to bone marrow suppression and decreased platelet survival add to the coagulopathy.

   6. Gastrointestinal. Paralytic ileus or necrotizing enterocolitis (NEC) are due to decreased end-organ perfusion.

   7. Metabolic. Acidosis (elevated lactate), hypoglycemia (hyperinsulinism), hypocalcemia (increased phosphate load, correction of metabolic acidosis), and hyponatremia/syndrome of inappropriate antidiuretic hormone secretion (SIADH).

1. Maternal data

   1. History. A thorough maternal history (prior pregnancy loss, thyroid disease, fever, drug use, infection) and family history (thromboembolic disorders, seizure disorder) can help identify causes of NE other than HIEs.

   2. Fetal heart rate (FHR) patterns. The following FHR tracings may serve as antepartum indictors of uteroplacental insufficiency or fetal compromise. Reactive FHR and subsequent prolonged FHR deceleration is suggestive of a sudden catastrophic event (pattern of acute asphyxia). A reactive FHR, which, during labor, becomes nonreactive, is associated with rising FHR baseline and repetitive late decelerations (pattern of intrapartum asphyxia). A persistent nonreactive FHR tracing with a fixed baseline rate, from admission until delivery, is suggestive of prior neurologic injury. This FHR pattern is often associated with reduced fetal movement, old passage of meconium, oligohydramnios, and abnormal fetal pulmonary vasculature (persistent pulmonary hypertension). FHR patterns are not always specific and have a substantial false-positive rate.

   3. Umbilical cord blood gases. Umbilical cord gases provide objective evidence regarding the intrapartum metabolic status of the fetus. An arterial cord pH <7 and base deficit ≥12 mmol/L are consistent with fetal metabolic acidosis. Neonatal morbidity increases as umbilical arterial pH falls below 7.0. The metabolic component (base deficit and bicarbonate) is more important than the respiratory component (Pco₂). Isolated respiratory acidosis is not typically associated with neonatal complications. The precise value that is required to define damaging acidemia is not known. A pH <7.0 realistically represents clinically significant acidosis. Acidemia alone does not establish that hypoxic injury has occurred. Umbilical artery Po₂ levels are not predictive of adverse neonatal outcome.

   4. Placental pathology. Important information regarding the etiology of NE may be gleaned from abnormalities seen on the maternal or fetal side of the placenta. Pathological umbilical cord lesions, such as velamentous or marginal cord insertion or cord hematoma or tears, may indicate a disruption in the fetal vascular supply. Chorioamnionitis and funisitis may indicate an infectious etiology of NE, while fetal thrombotic vasculopathy can point to a genetic coagulopathic disorder.

2. Neonatal data

   1. Apgar scores. While the previous AAP/ACOG definition of perinatal asphyxia mandated an Apgar score of ≤3 at 5 minutes, it is not part of the updated criteria. The most recent AAP/ACOG policy statement concludes that while a low 5-minute Apgar score may be associated with neonatal mortality, low 1- and 5-minute Apgar scores are not definite markers of an asphyxial event. An infant with an Apgar score of 0–3 at 5 minutes, improving to ≥4 by 10 minutes, has >99% chance of not having CP at 7 years of age; 75% of children who develop CP have normal Apgar scores at birth.

   2. Physical examination. Findings determine grading of HIE severity based on the Sarnat scale.

      1. Stage 1 (mild). Hyperalertness, normal muscle tone, weak suck, low threshold Moro, mydriasis, and absence of seizures.

      2. Stage 2 (moderate). Lethargic or obtunded, mild hypotonia, weak or absent suck, weak Moro, miosis, and focal or multifocal seizures.

      3. Stage 3 (severe). Stupor, flaccid muscle tone, intermittent decerebration, absent suck, absent Moro, and poor pupillary light response.

   3. Laboratory studies. Complete blood count with differential, blood culture, serum electrolytes, blood urea nitrogen, creatinine, cardiac enzymes, liver enzymes, a coagulation panel, and blood gases should be obtained at time of admission and serially monitored as indicated.

   4. Imaging other than cranium. An echocardiogram can be obtained to evaluate cardiac ventricular function. A renal and hepatic ultrasound may provide further information regarding end-organ damage.

   5. Conventional electroencephalogram (cEEG). Multiple neonatal cEEG classification systems, with varying criteria, exist focusing on the following variables:

      1. Amplitude. Described as isoelectric (maximally depressed or flat), low voltage, or mild voltage suppression.

      2. Symmetry. Classified as mildly, moderately, or severely abnormal.

      3. Continuity. Categorized as burst suppression pattern, which has the worst prognosis; persistent discontinuity (moderately or severely abnormal); or mild discontinuity (mildly abnormal).

      4. Sleep-wake state. The presence of sleep-wake cycling (SWC), as expected in a normal EEG, may be poorly defined or absent in a neonate affected by HIE with the time from birth to onset of normal SWC predictive of outcome.

      5. Frequency. Background frequency or pattern may be described as diffuse delta pattern (moderately or severely abnormal), poorly developed background rhythms (moderately abnormal), mild disturbance in background rhythm (mildly abnormal), and presence of seizures (moderate or severely abnormal).

   6. Amplitude-integrated EEG (aEEG). This mode of monitoring utilizes a bedside cerebral function monitor, which records and amplitude integrates a single-channel EEG from biparietal electrodes. aEEG has the advantage over cEEG of not requiring extensive formal training for interpretation. A meta-analysis has shown aEEG to be useful in predicting long-term neurodevelopmental outcomes in term infants with HIE. The appearance of SWC in the aEEG within the first 36 hours is indicative of a good prognosis. The following classification scheme is suggested to describe the aEEG findings (see Figure 16–1):

      1. Continuous normal voltage. Continuous activity with lower (minimum) amplitude around (5 to) 7–10 μV and maximum amplitude of 10–25 (to 50) μV.

      2. Discontinuous normal voltage. Discontinuous background with minimum amplitude variable, but below 5 μV, and maximum amplitude above 10 μV.

      3. Burst suppression. Discontinuous background with minimum amplitude without variability at 0–1 (2) μV and bursts with amplitude >25 μV.

      4. Continuous extremely low voltage. Continuous background pattern of very low voltage (around or below 5 μV).

      5. Inactive, flat trace. Primarily inactive (isoelectric tracing) background below 5 μV.

   7. MRI of the brain. It is important to note that HIE infants with a normal MRI may still be at risk for neurodevelopmental dysfunction. Diffusion tensor magnetic resonance imaging (MRI) (vs T1- and T2-weighted MRI) may be more accurate at demarcating areas of brain injury due to the presence of cerebral edema in the first 48–72 hours.

      More than the severity of injury, the pattern of brain involvement identified by MRI correlates with neurodevelopmental outcomes for infants with HIE:

      1. Watershed predominant pattern. Involves the vascular watershed in white matter. Involvement of the cortical gray matter can be seen in severe HIE. This pattern results from partial prolonged asphyxia and is associated with cognitive impairment.

      2. Basal ganglia/thalamus predominant pattern. Affects the deep gray nuclei and perirolandic cortex. The total cortex can be involved in severe HIE. This pattern results from acute profound asphyxia and is associated with severe cognitive and motor deficiencies.

   8. Magnetic resonance spectroscopy (MRS). Used to assess the concentration of cerebral metabolites, thereby providing information regarding the biochemical changes in the brain secondary to HIE. Specifically a high ratio of N-acetylaspartate to choline and a low ratio of lactate to choline are predictive of better neurodevelopment outcomes.

   9. Diffusion weighted imaging (DWI) and diffusion tensor imaging (DTI). Provide important information regarding the directionality and magnitude of water diffusion in the brain. The apparent diffusion coefficient (ADC) reflects the rate of diffusion and is reduced rapidly after an ischemic insult. A low ADC translates into a high-intensity signal on diffusion imaging. While restricted diffusion has the highest sensitivity for early detection, it can lead to underestimation of injury if obtained within the first 24 hours. The diffusion abnormalities peak at 3–5 days and then normalize.

The management of infants with HIE begins with the identification of perinatal patients at high risk for asphyxia and optimal resuscitation in the delivery room. As many cases of perinatal asphyxia are unanticipated and unpreventable, clinical care mostly focuses on providing supportive care to prevent further exacerbation of injury and specific neuroprotective therapies targeted at the therapeutic window prior to the onset of irreversible secondary energy failure. The ethical and medicolegal aspects of care also need to be considered.

1. Supportive care

   1. Resuscitation. The 2011 Neonatal Resuscitation Program guidelines (Kattwinkel et al, 2010) recommend initiating resuscitation with room air or blended oxygen with a targeted preductal Spo₂ of 60–65% by 1 minute of life and 80–85% by 5 minutes of life in all term and preterm infants. There are no current guidelines specific to neonates with HIE. While resuscitation with 100% O₂ more rapidly restores CBF and perfusion in animal studies, hyperoxia should be avoided, as oxidative damage from oxygen-free radicals can further exacerbate hypoxic-ischemic brain injury.

   2. Ventilation. Assisted ventilation may be required to maintain Pco₂ within the physiologic range. While hypercarbia exacerbates cerebral intracellular acidosis and impairs cerebrovascular autoregulation, hypocarbia (Paco₂ <20–25 mm Hg) decreases CBF and is associated with PVL in preterm infants and late-onset sensorineural hearing loss in full-term infants.

   3. Perfusion. Arterial blood pressure should be maintained in the normotensive range for gestational age and weight. Due to the loss of cerebrovascular autoregulation, volume expanders and inotropic support should be used cautiously in order to avoid rapid shifts between systemic hypotension and hypertension.

   4. Acid-base status. The base deficit is thought to increase in the first 30 minutes of life due to an initial washout effect secondary to improved perfusion and transient increase in lactic acid levels. However, acidosis normalizes in the majority of infants by 4 hours of life, regardless of bicarbonate therapy. The rate of recovery from acidosis is reflective of HIE severity but not duration, and is not predictive of outcomes. Sodium bicarbonate therapy is not recommended as it causes a concomitant rise in intracellular Pco₂ levels, negating any changes in pH, and is associated with increased rates of intraventricular hemorrhage and mortality. One must consider an inborn error of metabolism if the degree of acidosis seems out of proportion to history and presentation and if the metabolic acidosis persists despite vigorous therapy.

   5. Fluid status. Initial fluid restriction is recommended as HIE infants are predisposed to a fluid overload state from renal failure secondary to acute tubular necrosis (ATN) and SIADH. The avoidance of volume overload helps avert cerebral edema. A single dose of theophylline (8 mg/kg) may be considered within the first hour to increase glomerular filtration by blocking adenosine-mediated renal vasoconstriction.

   6. Blood glucose. Initial hypoglycemia (<40 mg/dL) in the context of HIE amplifies the risk of progression from moderate to severe encephalopathy. Timely and frequent monitoring of blood glucose levels is therefore essential.

   7. Seizures. Seizure activity is both a consequence and determinant of brain injury. A Cochrane review showed no reduction in death, neurodevelopmental disability, or combined outcome with the prophylactic use of anticonvulsant therapy. Phenobarbital therapy is recommended as the first-line agent for prolonged or frequent clinical seizures. The use of prophylactic phenobarbital in conjunction with hypothermia has shown a reduction in clinical seizures but not neurodevelopmental outcome. Phenobarbital levels in asphyxiated infants should be carefully monitored because hepatic and renal dysfunction, as well as hypothermia, can increase the drug's half-life and plasma concentration.

2. Neuroprotective strategies

   1. Hypothermia. (See Chapter 39.) Therapeutic hypothermia attenuates secondary energy failure by decreasing cerebral metabolism, inflammation, excitotoxicity, oxidative damage, and cellular apoptosis. Hypothermia is now emerging as standard of care for perinatal asphyxia. Early identification of neonates with perinatal asphyxia and their timely referral to tertiary care centers for hypothermia therapy is therefore crucial. Hypothermia protocols and recommended temperature regulation prior to admission (such as passive cooling or active cooling on transport) are institution specific and must be clarified with the accepting facility at the time of referral. To date, 3 large multicenter trials of cerebral hypothermia for HIE, initiated within 6 hours of birth and continued for 72 hours, have been completed. In a recent review (Wu, 2011), direct comparisons of selective head cooling and whole body cooling are lacking, but both appear to have similar safety and effectiveness. Whole body cooling is preferred in most centers in the United States due to ease of administration and easier access to the scalp for EEG monitoring.

      1. The “Cool Cap” trial utilized selective head cooling with mild systemic hypothermia (34–35°C) and an aEEG screen as part of entrance criteria. In a subset of patients with less severe aEEG findings, an improvement in survival without severe neurodevelopmental disability was demonstrated.

      2. The National Institute of Child Health and Human Development (NICHD) trial subjected infants to whole body cooling with moderate systemic hypothermia (33.5°C) without any aEEG entrance criteria. They demonstrated a reduction in the combined endpoint of death or moderate or severe disability.

      3. The Total Body Hypothermia for Neonatal Encephalopathy (TOBY) trial also cooled infants to a body temperate of 33.5°C but used aEEG entrance criteria. Cooling did not reduce the combined rate of death or severe disability, but improved neurodevelopmental outcomes were seen among survivors.

      4. Selective head cooling (SHC) increases the temperature gradient across the brain from the central to peripheral regions. This is in contrast to whole body cooling, which maintains a uniform temperature gradient across the brain. In a systematic review of 13 studies (Shah, 2010), systemic hypothermia, but not SHC, was associated with a reduction in outcomes of cognitive delay, psychomotor delay, and cerebral palsy. The reduction in mortality or neurodevelopmental disability among survivors was similar between both modes of cooling. There are no clinically significant adverse effects from therapeutic hypothermia, and the mode of cooling does not have any differential impact on multiorgan system dysfunction in asphyxiated infants.

   2. Pharmacotherapy. While investigation of potentially neuroprotective drugs is ongoing, these drugs are not in wide clinical use. Antioxidant enzymes such as superoxide dismutase and catalase were effective when given in animals prior to brain injury. Free-radical inhibitors such as allopurinol were effective in asphyxiated infants when given soon after resuscitation. Deferoxamine, a free-radical inhibitor, was beneficial when given during the reperfusion phase of injury in animal studies. Magnesium, an N-methyl-d-aspartate (NMDA) glutamate receptor antagonist, may have a beneficial effect on preventing cerebral palsy, based on retrospective clinical data, but neonatal animal studies are equivocal. While the prophylactic use of calcium channel blockers, such as flunarizine, was beneficial in animal studies, their use in infants is currently contraindicated due to adverse cardiovascular effects. Erythropoietin has been shown to improve outcomes for term infants with mild to moderate HIE by modulating neuronal injury and promoting neural regeneration.

3. Ethics. Most cases of perinatal asphyxia are unanticipated and families are often unprepared to deal with the complexities of HIE. Direct and timely communication between the medical team and the neonate's family is therefore essential to foster shared decision making with difficult medical and emotional decisions such as discontinuation of life support. A multidisciplinary team approach is essential, as severely depressed infants may have multiple complex medical needs (see also Chapter 21). A neurological consult is helpful to provide parents with important prognostic information based on the infant's neurological assessment. A palliative care consult may offer families support while optimizing the quality of the life for the infant.

   The use of therapeutic hypothermia has also raised some important ethical issues. The concern that hypothermia therapy would simply increase the survival rate of severely disabled infants but not actual neurologic outcomes was addressed by the deliberate use of a compound primary outcome of death or survival without disability in the research design. Another concern regarding whether hypothermia therapy would hinder discussion regarding the withdrawal of care in profoundly affected infants was not found to be valid in the 3 major cooling trials.

4. Medicolegal aspects

   1. Obstetric care. Fetal status must be assessed with electronic fetal monitoring at the time of maternal admission to identify and categorize risk for intrapartum fetal distress. A reactive fetal heart rate (FHR) pattern is a reliable sign of fetal well-being, while a nonreactive pattern indicates a higher probability of intrapartum fetal distress and adverse fetal outcome. The early recognition of fetal heart rate patterns that are associated with brain-damaged infants and timely intervention may potentially alleviate brain injury. It is important to note, however, that FHR patterns may actually be reflective of preexisting or underlying neurological impairment that existed prior to the perinatal period. Also, only cerebral palsy of the spastic quadriplegic type is thought to have perinatal asphyxia as an etiological component. Caution must therefore be used in identifying a specific event as the cause for an adverse outcome, as the baseline fetal brain status is often unknown. One cannot state with a reasonable degree of medical certainty that CP in a given child was due to intrapartum asphyxia merely because the physician can find no other explanation.

   2. Neonatal care. Many cases of perinatal asphyxia are unanticipated, and an experienced resuscitation team may not always be readily available. In an anticipated high-risk birth, clear communication between members of the obstetric and neonatal teams prior to delivery is imperative. The resuscitation team must be thoroughly prepared for the potential for vigorous resuscitation per current AAP Neonatal Resuscitation Program guidelines. Careful attention must be paid to avoiding hypoglycemia, hypotension, and hypocarbia in the postresuscitation stabilization period. Early referral for evaluation for hypothermia therapy is crucial, as transport and admission to the tertiary center is ideally achieved within 6 hours after birth. Documentation of the resuscitation and stabilization process as well as informed consent for transport is an important part of the medical record. Informed consent for hypothermia is institution specific.

The presence of neonatal encephalopathy is considered an essential etiologic link between perinatal events and permanent brain damage. Mild cases of neonatal HIE have a favorable outcome, while severe cases have a high rate of mortality and neurodevelopmental disability. Predictions of outcome can be categorized by the following time points:

1. 0–6 hours after birth. One model used the following 3 factors: chest compression >1 minute, onset of respiration after age 20 minutes, and base deficit >16 mmol/L to predict outcome in patients with moderate or severe HIE. Severe adverse outcome rates were 46% with none of the 3 predictors, 64% with any 1 predictor, 76% with any 2 predictors, and 93% with all 3 predictors present. Abnormal aEEG tracings also have predictive value for death or moderate to severe disability.

2. 6–72 hours after birth. The clinical examination focusing on the Sarnat stages of encephalopathy, presence of seizures, spontaneous activity, and brainstem function are all predictive of outcome.

   The following findings on MRI and MRS are predictive of a poor outcome: absence of normal hyperintensity in the posterior limb of internal capsule, a watershed or basal ganglia/thalamus predominant pattern, a high lactate/choline ratio or a low N-acetylaspartate/choline ratio. Abnormalities seen on diffusion tensor imaging do not correlate with long-term outcome. Amplitude-integrated EEG findings of burst suppression or discontinuous background are predictive of death or severe disability.

3. Prior to discharge. A normal neurological examination at 1 week of age is highly correlated with a normal outcome. Establishment of oral feedings is a good prognostic indicator. Lesions on neuroimaging may evolve or become more apparent as cerebral edema resolves.

4. Postdischarge follow-up. Microcephaly at 3 months of age or an abnormal neurologic examination at 12 months of age predict poor neurodevelopmental outcome at 5 years of age. A decrease in head circumference (HC) ratios (actual HC/mean HC for age × 100%) of >3.1% between birth and 4 months of age is highly predictive of the eventual development of microcephaly before 18 months of age. Suboptimal rate of head growth associated with moderate cerebral white matter changes on MRI may be a better predictor of poor neurodevelopmental outcome.