eMedicine Specialties > Neurology > Pediatric Neurology

Hypoxic-Ischemic Brain Injury in the Newborn

Marcio Sotero de Menezes, MD, Associate Professor, Department of Neurology, Division of Pediatric Neurology, Children's Hospital of Seattle, University of Washington
Dennis WW Shaw, MD, Professor, Department of Radiology, Department of Radiology, University of Washington School of Medicine; Consulting Staff, Children's Hospital and Regional Medical Center of Seattle

Updated: Apr 4, 2006

Introduction

Background

In the immediate newborn period, many factors can produce neurologic symptoms mimicking those of HIE, including prepartum and postpartum ischemia/hypoxemia, genetic factors, metabolic disease, and maternal and fetal drug use. Because the relationship between asphyxia and HIE cannot always be established, the term newborn encephalopathy (NE) was proposed as an alternative to remove the medicolegal implications of HIE. Newborn NE is a clinically defined syndrome of disturbed neurologic function in full-term infants that attempts to correlate symptoms in the neonatal period that have some relationship with neurologic outcomes in childhood. NE symptoms may or may not be causally linked to hypoxemia/ischemia. Far from fixing the problem, use of the term NE just removes from obstetric practitioners the unfair blame they receive for poor neonatal outcomes.

The National Collaborative Perinatal Project (NCPP), a prospective study of more than 50,000 pregnancies and 40,000 infants, was conducted to analyze the features of NE. Its results showed that the following were associated with increased morbidity on follow-up examination: decreased activity after the first day of life, need for incubator more than 3 days, feeding problems, poor suck, and respiratory difficulties.

Other factors not mentioned in the description of NE syndrome have been associated with postneonatal morbidity. Examples are static motor deficits (cerebral palsy [CP]), mental retardation, and epilepsy. These factors include neonatal seizures, low 10-minute Apgar scores, stupor, and coma.

Pathophysiology

General principles of perinatal ischemia and hypoxemia

In fetal life, hypoxic-ischemic disturbances are primarily a consequence of hypoperfusion because the arterial partial pressure of oxygen (PaO₂) normally is low. Despite this, severe hypoxemia can occur, leading to myocardial dysfunction with subsequent cerebral hypoperfusion or loss of cerebrovascular autoregulation. This cerebral hypoperfusion in turn may lead to neuronal ischemia.

Primary perinatal hypoxemia

In utero hypoxemia is usually the result of placental insufficiency, and infants who have experienced in utero hypoxemia often have clinically significant respiratory or cardiac failure after birth. By comparison, postnatal hypoxemia is the result of either respiratory or cardiac insufficiency, alone or in combination. Primary perinatal hypoxemia can derange the already fragile cerebrovascular autoregulation of the neonatal brain (see Pressure-passive cerebral circulation). This may explain why severe surfactant deficiency syndrome is associated with certain patterns of neuronal damage, such as periventricular leukomalacia (PVL).

Perinatal ischemia

Severe cardiac contractile dysfunction due to either major cardiac malformations or severe hypoxemia leads to cerebral hypoperfusion and loss of cerebrovascular regulation. Circulatory insufficiency can result from severe prenatal or postnatal hemorrhage or neonatal sepsis. In response to hypoxic-ischemic insults, circulatory rearrangement of the cardiac output occurs, with shunting of blood flow away from the liver, kidneys, gut, lungs, and skeletal muscle into the heart, brain, and adrenal glands of the infant. This shunting explains the coexistence of liver and kidney failure in cases of severe HIE-NE. This initial reaction to systemic hypoxia and ischemia in the newborn is one of lowering the heart rate and increasing blood pressure to maintain a near-normal cardiac output.

With progression of the hypoxic-ischemic state, the heart rate, blood pressure, and cardiac output decrease substantially; systemic metabolic acidosis increases partly because of the production of lactic acid.

In the neonatal period, autoregulation of cerebral circulation is impaired. This autoregulation is further impaired by both hypoxemia and hypercarbia, which leaves the cerebral circulation in a pressure-passive state. In this state, cerebral perfusion changes as intravascular pressure changes. The immature cerebral arteries have a limited ability to adapt to hypotensive episodes. This occurrence may be especially important in the genesis of the parasagittal pattern of cerebral injury observed after hypotension. The other extreme of the range of arterial pressure may also be problematic because the decreased upper limit of autoregulation may increase the risk of periventricular and intraventricular hemorrhage.

Poor perfusion of the depths of the sulci

In the neonatal period, relatively underperfusion affects the depths of the sulci, creating a zone of increased susceptibility to hypotensive insults. This effect may lead to gyri that resemble mushrooms owing to atrophy of their base near the deep parts of the sulci.

Excitatory neurotransmitter toxicity (excitotoxicity)

A large body of evidence indicates that certain excitatory neurotransmitters (ie, amino acids, especially glutamate) are excessively released at the synaptic cleft during conditions of hypoxia-ischemia. These excitotoxins may play a critical role in the neuronal damage observed during low-energy states. The evidence linking excitotoxins to ischemic cell death include the following: (1) Synaptic activity is necessary for hypoxic cell death. (2) Specific glutamate antagonists prevent hypoxic cell death. (3) Glutamate exposure mimics hypoxic cell death. (4) Glutamate accumulates extracellularly in vivo during hypoxia (because of increased release and decreased uptake). (5) The topography of neuronal death with hypoxia is similar to that of glutamate synapses. The regional distribution of glutamatergic neurons may explain some of the patterns of injury seen in the cerebrum after hypoxia-ischemia. Hypoglycemia and hyperthermia may potentiate the deleterious effect of ischemia on the CNS.

Increased sensitivity of white matter to ischemic damage in premature neonates (excitotoxicity)

A sensitivity to hypoxic-ischemic damage has been observed in premature infants. This pattern of brain damage has been traditionally attributed to developmental changes in the arterial watershed zones in immature fetuses (and premature infants). Over the past decade, the trajectory of these vessels producing a periventricular watershed zone has been questioned. On the contrary, new information shows that the oligodendrocyte and its precursors are susceptible to injury in the fetus. Some of that finding may be related to immature handling of oxygen radicals.

Mortality/Morbidity

See Prognosis.

Clinical

History

Overall, a patient with a history of problems during labor and delivery followed by difficulties during resuscitation in the delivery room (eg, low Apgar scores at 10 minutes, delayed respiratory effort), with subsequent development of a severely depressed level of consciousness (eg, stupor or coma), especially in association with ventilatory disturbances (eg, apnea, respiratory failure) and seizures, is likely to have long-term neurologic morbidity.

• Full-term infants
  • Many clinical features of neonatal HIE are nonspecific; as such, the diagnosis must be made with caution and only after careful examination of historical, physical-neurologic, and laboratory data. Antepartum factors that may be associated with neonatal HIE include the following:
    • Maternal diabetes
    • Pregnancy-induced hypertension
    • Intrauterine growth retardation (IUGR)
    • Maternal hypotension-shock
    • Severe bleeding
    • Placental insufficiency
    • Abruptio placentae
    • Cord prolapse
    • Prolonged expulsive period
    • Dystocia
  • NE is a clinically defined syndrome of disturbed neurologic function in full-term infants. It includes decreased activity after the first day of life, need for an incubator after 3 days, feeding problems, poor suck, and respiratory difficulties. These criteria for NE were used in the analysis of the NCPP.
  • In the NCPP, neonatal seizures were associated with high rates of mortality and morbidity, including CP, mental retardation, and epilepsy. In the same cohort, the number of days of neonatal seizures also was correlated with neurologic disability. The timing of neonatal seizures was associated with risk of death, not with risk of long-term motor deficits.
  • A history of low Apgar scores at 1 and 5 minutes is commonly used as an indicator of HIE and subsequent morbidity, but this index has several drawbacks (see Physical).
  • A history of birth trauma, especially in association with problems in the expulsive period of delivery (eg, severe dystocia), may play a role in HIE-NE. Nonetheless, when analyzed properly, obstetric complications (eg, birth trauma, dystocia, cord prolapse) are not useful in predicting the outcome of the newborn unless they are followed by low late Apgar scores, signs of NE, depressed level of consciousness, or neonatal seizures.
  • Extracorporeal membrane oxygenation (ECMO) is associated with clinically significant morbidity in term neonates with serious pulmonary disease.
• Preterm infants
  • In premature neonates, the patient's history is essential for diagnosis.
  • PVL is the most typical lesion in the preterm infant with NE. PVL is associated with the following historical findings: maternal hypotension, twin pregnancy, fetal or maternal arrhythmia, severe postnatal respiratory or cardiac disease associated with hypotension (<30 mm Hg), acidosis, and heart failure. Premature newborns who are small for their gestational age are particularly vulnerable to PVL. PVL has also been noted in small premature infants who had no evidence of hypotension.

Physical

A full-term newborn with clinically significant intrapartum hypoxic-ischemic insults, enough to cause permanent neurologic sequelae should have abnormal findings on neurologic examination in the first week of life. However, an infant with significant prepartum injury to the brain in this period may be entirely asymptomatic in the neonatal period.

• Renal, cardiac, and pulmonary systems: Renal, cardiac, and pulmonary dysfunction often occur after a presumed hypoxemic insult. The true frequency of the co-occurrence of neurologic and other systemic (eg, renal, cardiac, pulmonary, liver) dysfunction is difficult to establish.
• Apgar scores: Apgar scores at 1 and 5 minutes are commonly used to predict late morbidity, despite their poor correlation with long-term outcomes.
• Mental status
  • As the patient wakes up, spontaneous movements and eye opening demonstrate his or her level of arousal.
  • Constant crying and irritability is evidence of an excessively aroused or hyperalert state.
  • Stupor and coma are poor prognostic signs in the absence of alternative reversible causes, such as high phenobarbital levels (>30-50 mcg/mL).
  • The patient's level of consciousness should be recorded as his or her best motor response.
  • The examiner also should detail the type of movement (eg, withdrawal, triple flexion, localization of pain) elicited as the patient's reaction to the stimulus.
• Cranial nerves
  • Lack of reflex activity mediated by the cranial nerves can indicate brainstem dysfunction.
  • Full-term infants should blink and sustain eye closure in response to a sustained light stimulus. Repeated flashes of light should produce habituation (eg, attenuated blinking) after 3-4 stimuli. Virtually all full-term newborns can track a ball of red wool, and the movement of stripes of at least one eighth of an inch or bigger can elicit opticokinetic nystagmus. Objects and pictures with round contours and facial appearances also make good targets for tracking in the newborn. Tracking is possible in infants with complete destruction of the occipital cortex by virtue of a subcortical pulvinar-collicular system. Retinal hemorrhages are commonly observed in the neonate after vaginal delivery. Pupillary reflexes are reliably present at term.
  • Neurologic examination may be difficult in the small and frail premature infant, but weakness of the lower extremities sometimes reflects the neuropathologic substrate of PVL. Over time, the patient with periventricular white-matter lesions develops spastic diplegia affecting the lower extremities more than the upper extremities.
  • Blinking to light starts at 26 weeks gestational age, sustained eye closure to light is seen around 32 weeks, and 90% of newborns track a ball of red wool by 34 weeks. Opticokinetic reflexes can be seen at 36 weeks. The pupil starts reacting to light around 30 weeks, but the light reflex is not consistently assessable until the gestational age of 32-35 weeks. Extraocular movements can be elicited by performing the doll's-eye maneuver at 25 weeks and by performing caloric stimulation at 30 weeks.
  • In infants aged 32-34 weeks, suck and swallow are reasonably coordinated with breathing, but the actions are not perfected until after term.
  • Patients with mild HIE-NE often have mydriasis. Progression of the disease may produce miosis (even in the dark) responsive to light, and in severe cases (stage 3 of Sarnat classification), the pupils are small or midpositioned and poorly reactive to light, reflecting sympathetic or parasympathetic dysfunction.
  • The lack of pupillary, eye movement, corneal, gag, and cough reflexes may reflect damage to the brainstem, where the cranial-nerve nuclei are located. Decreased respiratory drive or apnea can be from lesions of the respiratory center, which overlap with vagal nuclei (ambiguous and solitaire) or medullary reticular formation. Ventilatory disturbances in HIE may manifest as periodic breathing apnea (similar to Cheyne-Stokes respiration) or just decreased respiratory drive.
• Motor function
  • Begin the motor examination of an infant with suspected HIE-NE by qualitatively and quantitatively observing his or her posture and spontaneous movements. Asymmetry in the amount of movement and posture is a subtle sign of hemiparesis, but it may be the only focal feature of the examination. Slight stimulation (eg, gently touching the patient) can increase motor activity in the term neonate and may be helpful in demonstrating asymmetrical hemiparesis.
  • Eliciting the Moro reflex may be an excessive stimulus and mask a subtle asymmetry in limb movement. Asymmetry in the Moro reflex is seen in peripheral lesions (eg, those due to brachial plexus injury).
  • Total absence or paucity of spontaneous movements, especially if associated with no reaction to painful stimuli and generalized hypotonia indicates brainstem dysfunction or severe, diffuse, or multifocal cortical damage.
  • Specific patterns of motor weakness indicate cerebral injury patterns. Patients with parasagittal injury tend to have proximal-greater-than-distal and upper extremity–more-than–lower extremity weakness. A unilateral, focal infarct, especially one involving the middle cerebral artery, causes contralateral hemiparesis and focal seizures. Patients with selective neuronal necrosis may have severe hypotonia, stupor, and coma.
  • Motor examination of a newborn with large unilateral lesions may reveal mild hemiparesis and seizures in as many as 80%. The seizures often are partial (focal) and contralateral to the cortical lesion. Neonates with severe bilateral infarcts may have quadriparesis. Moro and tonic neck reflexes do not habituate, reflecting the lack of cortical modulation, which attenuates the response after repeated trials or sustained stimulus. Newborns with diencephalic lesions cannot regulate their temperature and have problems with sleep-wake cycles. The long-term sequelae of focal or multifocal cerebral necrosis include spastic hemiparesis and quadriparesis (eg, bilateral hemiparesis), cognitive deficits, and seizures.
  • Foot-ankle dorsiflexion or triple flexion (eg, foot-ankle dorsiflexion, knee and hip flexion) after plantar stimulation reflects only an intact spinal cord and sensory and motor nerves. Extensor movements (eg, arm elevation above the level of the shoulders) are more sophisticated motor actions than the dorsiflexion or triple flexion and require some cortical function.
  • A tonic neck reflex is performed by turning the patient's head to 1 side. The patient appears to be observing the arm and leg extension on the side to which the head is turned and flexion on the opposite side. The tonic neck reflex posture should go away after several seconds, and its persistence is a sign of cortical dysfunction.
  • Spasticity is a velocity-dependent increase in tone that is generally most prominent with limb extension in muscle groups with antigravitational action (arm flexion, plantar extension). This sign can be seen over time in infants with corticospinal tract damage caused by a hypoxic-ischemic insult. In the neonatal period, spasticity is commonly noted first and is most prominent in the distal parts of the extremities. All fingers are flexed with the thumb under the second to fifth fingers, a pattern commonly referred to as cortical thumbs. Fewer than 5-10 beats of ankle clonus may be present in healthy neonates, but infants with damage to the corticospinal tract may have sustained ankle clonus.
  • Hip abduction may be seen with increased tone and even with decerebrate posturing. Another manifestation of CNS dysfunction in the neonatal period is increased axial extensor tone with arching of the back and neck extension or opisthotonus. Many infants simultaneously have decreased axial flexor tone (eg, major head lag on arm traction maneuver) and increased axial extensor tone. In many cases, limb and axial hypotonia are present for several months before increased axial extensor tone or limb spasticity can be detected.
• Seizures
  • HIE is often reported to be the most frequent cause of neonatal seizures. They usually occur 12-24 hours after birth and are difficult to control with anticonvulsants. Large, unilateral infarcts occur with neonatal seizures in as many as 80% of patients. Seizures are often partial (focal) and contralateral to the cortical lesion. About two thirds of newborns with cerebral venous infarcts have seizures. Those with multiple or diffuse lesions and cerebral venous infarcts often have multifocal or migratory seizures. Seizures are observed during physical examination and may confirm the diagnosis. Observation often reveals clonic rhythmic contractions. When holding the limb affected by clonic seizures, the examiner's hand shakes or feels limb movement. Limb flexion or extension does not suppress the clonic activity, as it does in jitteriness and clonus.
  • Tonic unilateral or focal seizures consistently have an EEG signature. In the seizures, unilateral arm and leg posturing is often accompanied by ipsilateral trunk flexion. Generalized tonic posturing (eg, extension of the upper and lower extremities or extension of the legs and flexion of the arms) is related to an EEG seizure in 15% of affected neonates. Tonic seizures can be seen in neonates with local anesthetic intoxication. Although generalized tonic posturing is infrequently associated with electrical seizures, it is not a benign sign. Of neonates with tonic posturing and an abnormal EEG background, 13% have normal development. Mizrahi and Kellaway (1987) suggested the name brainstem release phenomena because tonic posturing and some subtle seizurelike motor automatisms (see below) are probably the result of primitive brainstem and spinal motor patterns liberated because the lack of inhibition from damaged forebrain structures.
  • Subtle seizures may be a part of the HIE-NE picture. Subtle manifestations of neonatal seizures are confirmed on EEG and include apnea; tonic eye deviation; sustained eye opening; slow, rhythmic, tongue thrusting; and boxing and swimming movements. Most still accept that some subtle seizures may be correlated with EEG results. However, publications since the late 1980s have shown that seizures are not as frequent as previously thought and that they are unusual in patients close to term. Several other patterns of subtle neonatal seizures are described without EEG confirmation. The lack of an EEG signature does not exclude CNS pathology because neonates with HIE often have motor automatisms without EEG seizures.
  • Seizures may be difficult to clinically diagnose in the premature neonate. Subtle seizures associated with ictal EEG changes are not rare in premature infants. The subtle patterns of neonatal seizures in the premature infant include sustained eye opening, oral-buccal-lingual movements (smacking, drooling, chewing), pedaling movements, grimacing, and autonomic manifestations.
• Grading systems for HIE
  • Several systems have been created to measure the severity and monitor the progress of the encephalopathic signs and symptoms in neonates after a presumed hypoxic-ischemic insult. These systems have some usefulness, but they are criticized because they were validated with retrospectively collected data. Most of these systems fail to consider the predictive values of each component separately. Some systems also fail to account for toxic, infectious, or metabolic factors masquerading as, or superimposed on, HIE.
  • Sarnat and Sarnat system: The system Sarnat and Sarnat created in 1976 is still one of the most popular and the basis for most modern systems. The authors advise that drugs and other conditions that may produce neonatal neurologic alterations similar to those of HIE-NE should be excluded. Their system also uses EEG to help predict the outcome.
    • Stage 1: Stage 1 relies on the finding of an early syndrome characterized by hyperalertness (eg, decreased sleep) and sympathetic activation (eg, eyes wide open, decreased blinking, mydriasis), excessive reaction to stimuli, weak suck with normal tone, and EEG findings. This stage lasts less than 24 hours, and patients who remain in stage 1 have normal neurologic outcomes.
    • Stage 2: Patients progressing to stage 2 have mild hypotonia, are lethargic or obtunded (eg, delayed and incomplete response to sensory stimuli), clinical seizures, parasympathetic activation with miosis (even on the dim light), slowed heart rate ( <120 bpm), increased peristalsis, and copious secretions. Early on, EEG shows relatively low-voltage amplitude (<25 µV in the slow theta and delta range). On the second day, the EEG demonstrates a bursting pattern during wakefulness or obtundation (which worsens during sleep) and multifocal, low-frequency (1- to 1.5-Hz) EEG seizures with a central or temporal predominance. If the EEG recovers within 5 days, it becomes completely normal. If the recovery takes more than 5 days, low-amplitude slowing is noted.Stage 2 lasts 2-14 days. Clinical and EEG recovery within 5 days is associated with a good prognosis. Periodic EEG may indicate a poor prognosis if interburst intervals are totally isoelectric or if the bursting frequency is less than every 6 seconds, with a bursting pattern (every 3-6 seconds) lasting more than 7 days.
    • Stage 3: Stage 3 is characterized by stupor with response to only strong stimuli, with withdrawal or decerebrate posturing, severe hypotonia (eg, flaccidity) and suppression of deep tendon reflexes, primitive (eg, Moro, tonic neck, suck) reflexes, and brainstem (eg, corneal, oculocephalic, gag) reflexes. Clinical seizures are less frequent in stage 3 than in stage 2. The patient has generalized sympathetic or parasympathetic autonomic dysfunction with abnormal respiration and small or midposition pupils that are poorly reactive to light. EEG shows a deepened periodic pattern with increased amplitude and decreased frequency of bursts (every 6-12 seconds). Further worsening of the picture leads to a very-low-voltage or isoelectric EEG.
  • Lipper postasphyxial index: Findings on neurologic examination can be scored by using the Lipper postasphyxial index, in which points are assigned to normal findings. The index, when calculated in the first 48 hours of life, is correlated with neurodevelopmental outcomes at 1 year of age.
• Miller system: Most recently, Miller and collaborators from the University of California at San Francisco validated a simple scoring system based on the typical signs and symptoms of NE. The maximum score from the first 3 days of life is used for prognostication. In their study, no scoring was done when the subjects were sedated or paralyzed. Table 1. Scoring in the Miller System to Stage HIE

  Sign or Symptom	Score = 0	Score = 1
  Feeding	Normal	Gavage feeds, gastrostomy tube, or feeding by mouth not tolerated
  Alertness	Alert	Irritable, poorly responsive, or comatose
  Tone	Normal	Hypotonia or hypertonia
  Respiratory status	Normal	Respiratory distress (need for CPAP or mechanical ventilation)
  Reflexes	Normal	Hyperreflexia, hyporeflexia, or absent reflexes
  Seizure	None	Suspected or confirmed clinical seizure

  Note.—Total score = 0-6. CPAP = continuous positive airway pressure.

Differential Diagnoses

Other Problems to Be Considered

• Patients with CNS malformations commonly have perinatal depression and low Apgar scores.
• Maternal administration of CNS depressants or magnesium (fetal hypermagnesemia) can lead to a picture of decreased arousal, hypotonia, and respiratory failure, which may be confused with that of HIE.
• Neonatal bacterial sepsis, especially meningitis, may occur with most or all of the symptoms of HIE-NE. Enteroviral infection may also produce a picture of sepsis. In clinical neonatology practice, most patients with an HIE-like picture are evaluated for sepsis. This evaluation likely includes a blood culture, CBC determination, and CSF examination. Broad-spectrum antibiotic therapy is often started until the results of the evaluation are known and HIE is diagnosed with confidence.
• Hyperactivity and jitteriness suggesting mild HIE-NE, resembling Sarnat stage 1 disease, may occur in hypocalcemia and drug withdrawal syndrome. In hypocalcemia, the EEG may be abnormal, showing multifocal discharges and electrographic seizures.
• Myopathies (eg, myotonic dystrophy, congenital myopathies) or spinal muscular atrophy may cause hypotonia. Patients are often alert and responsive, and they do not have brainstem reflex depression; this feature easily differentiates this condition from HIE. Bifacial weakness may occur in congenital myopathies, and hypotonia may be noted in some inborn errors of metabolism, such as the peroxisomal disorders.
• Perinatal intoxication with local anesthetic may be difficult to differentiate from HIE. Mepivacaine administered as a paracervical block during delivery is the most common cause of an inadvertent injection of a local anesthetic. The anesthetic is commonly injected into the scalp of the fetus, and seizures are almost universal, beginning before 6 hours of life. Seizures are most often tonic, followed by apneic and multifocal types. EEG shows epileptiform discharges (temporal spikes) or may be normal. Many neonates have low 5-minute Apgar scores, apnea or hypoventilation, bradycardia, hypotonia, fixed and dilated pupils, and decreased eye movements to oculocephalic reflex (doll's eyes); this clinical picture is similar to that of HIE-NE.
  • Patients with local anesthetic intoxication become worse in the first 6 hours of life improve with supportive care alone. Pupillary and oculocephalic reflex abnormalities are seen in the first 12 hours and are presumably due to continued absorption of the drug from the subcutaneous tissue, which improves over the first and second days of life.
  • Patients with HIE have neurologic deterioration, which most often starts after 12 hours. Abnormalities in pupillary and oculocephalic reflexes are seen only after 12-24 hours.
  • The picture in patients with transplacental intoxication caused by local anesthetics can be difficult to characterize because hypoxic-ischemic injury may occur in this setting. The outcome of patients with local anesthetic intoxication caused by direct scalp injection is good if they are given appropriate supportive treatment.
• Pyridoxine dependency is associated with neonatal seizures and developmental delay responsive to high doses of pyridoxine. Some patients may have early-onset seizures (before 6 h of life) that often become intractable to regular anticonvulsants. Unless pyridoxine is administered, many patients have recurrent seizures that may lead to an encephalopathic picture, on both clinical evaluation and on EEG. Pyridoxine dependency is diagnosed by injecting 100-200 mg during EEG and checking for immediate improvement in epileptiform discharges and seizures.
• Several inborn errors of metabolism may occur in the neonatal period. Most patients do well initially but deteriorate 1-3 days after feedings begin, with a progressive decline in level of consciousness, seizures, and vomiting. A family history of neonatal death or seizures followed by neurologic deterioration suggests this condition. Many patients are depressed at birth and considered to have intrapartum hypoxia-ischemia. Unexplained vomiting should suggest this diagnosis. Affected neonates may be symptomatic. Examples of such errors are hyperammonemias due to urea-cycle defects, organic acidurias, lactic acidosis due to carbohydrate and mitochondrial metabolism dysfunction, and peroxisomal disorders.
  • Hyperammonemias due to urea-cycle defects commonly cause obtundation and neonatal seizures, which can be confused with those of HIE. High serum ammonia levels suggest the diagnosis. Newborns with urea-cycle defects do well initially and deteriorate 1-3 days after feedings start.
  • Organic acidurias may cause neonatal feeding difficulties, lethargy, vomiting, seizures, metabolic acidosis, hyperammonemia, and hyperglycinemia. Propionic acidemia and methylmalonic acidemia occur in the neonatal period. Nonketotic hyperglycinemia is a defect in the glycine-cleavage system, which often results in depressed mental status and seizures of early onset that are refractory to medical management. Myoclonus, hiccups, apnea, and lethargy are common symptoms. Diagnosis is based on verification of a high CSF-to-serum glycine ratio.
  • Neonates with pyruvate dehydrogenase complex deficiency often have the E1 or pyruvate decarboxylase deficiency and present with hypotonia, seizures, myoclonus, mild dysmorphic facies, and severe metabolic acidosis. The diagnosis is suspected when a metabolic acidosis is associated with elevated lactate pyruvate and alanine levels in the serum; however, urine organic acid values may be normal.
  • Generalized peroxisomal disorders can cause seizures, lethargy, hypotonia, and poor feeding in the neonatal period. Both Zellweger syndrome and neonatal adrenoleukodystrophy may occur in the immediate newborn period. The typical facial stigmata (high forehead, widow's peak, hypertelorism) and hepatomegaly are almost always present in Zellweger syndrome and may help in the diagnosis.

Workup

Laboratory Studies

• Acid-base measurements
  • Acid-base disturbance, assessed by means of umbilical-artery pH measurement, are correlated with neonatal seizures and death only in extreme cases when the pH is <7.04.
  • In studies, the overwhelming majority of patients with extreme acidosis and pH <7 had no seizures and did not die.
  • Low umbilical-artery pH may be due to causes other than ischemia, such as sepsis.
  • The association of low pH and long-term neurologic outcome is also weak.
  • In summary, most neurologically symptomatic neonates are not markedly acidotic, and most newborns with acidosis are not neurologically symptomatic.
• The serum concentration of brain-specific creatine kinase isoenzyme BB (CK-BB) appears to be somewhat correlated with the outcome after an episode of HIE-NE; however, the correlation for CSF CK-BB, especially with short-term outcome, is better.
• Other serum factors studied include the following:
  • Blood lactate
  • Hypoxanthine
  • Aspartate-aminotransferase
  • Erythropoietin beta-endorphin
• Factors measured in the CSF include the following:
  • Lactate
  • Neuron-specific enolase
  • Lactate dehydrogenase
  • Hydroxybutyrate dehydrogenase
  • Fibrinogen degradation products
  • Ascorbic acid
• The utility of most of these blood and CSF determinations in predicting long-term neurologic outcomes have not been validated in large, well-controlled studies.

Imaging Studies

• Head ultrasonography
  • Ultrasonography is most useful for the detection of PVL, and experienced technicians may detect lesions of the basal ganglia as well. Selective neuronal injury and parasagittal or watershed lesions are frequently missed on sonography.
  • Ultrasonography has poor specificity in differentiating increased echogenicity due to ischemic or hemorrhagic lesions.
  • Focal and multifocal ischemic lesions, especially small cortical infarcts, may be missed on sonography but detected on CT or MRI.
  • Early, large ischemic infarcts (eg, large infarct of the middle cerebral artery) may be observed on ultrasonography before it is apparent on CT.
• Cranial CT
  • CT depicts focal, multifocal, and generalized ischemic lesions. In the first few days after a severe hypoxic-ischemic insult, bilateral hypoattenuations are seen and probably reflect both neuronal injury and edema. CT neuropathologic studies show that areas of edema are correlated with hypoattenuation lesions when autopsy is performed within 10 days of CT, but generalized edema may obscure focal ischemic lesions. Diffuse cortical injury is not initially detected on CT. After days to weeks, diffuse hypoattenuation may appear, with loss of the gray matter–white matter differentiation. Diffuse cerebral atrophy with ex vacuo ventricular dilatation due to severe hypoxemic insult may take several weeks to develop. Atrophy is a consequence of cortical and white-matter destruction.
  • Areas of hypoattenuation can be challenging to interpret in premature infants, and autopsy studies show poor correlation between this finding and neuropathologically documented ischemic damage. CT scanning can be performed to help reliably diagnose generalized edema in the premature newborn.
  • After 48 hours, CT may depict focal ischemic infarcts well. On the first day after a focal thromboembolic event, the ischemic area may not be visible on CT. A CT scan depicting hypoattenuation in the distribution of the left middle cerebral artery in the first day of life suggests prenatal-onset of ischemia. Symptoms of a focal infarct (usually seizures) on the first day of life with normal CT findings and with hypoattenuation developing over the first week suggest perinatal-onset ischemia. Hemorrhagic conversion of a focal ischemic lesion is uncommon in the neonatal period, but CT can depict it easily.
  • A CT scan demonstrating generalized, diffuse hypoattenuation after a hypoxic-ischemic event is predictive of both neonatal death and long-term severe disability, whereas normal CT findings are predictive of mild disability or a normal outcome. Interpret normal results with caution because hypoattenuation may take a few weeks to develop.
  • CT may depict hemorrhagic lesions, which are seen in 10-25% of patients with HIE-NE. These lesions include intraparenchymal, intraventricular, and subarachnoid hemorrhages. Basal ganglia–thalamic lesions and selective neuronal injury can be detected on CT, but they are more reliably visualized on MRI than on CT.
  • On CT, PVL can be visualized around the frontal horns or posteriorly around the trigonal area of the lateral ventricles. PVL appears as a region of decreased attenuation, occasionally intermixed with areas of increased attenuation due to secondary hemorrhage. Periventricular hypoattenuations should be interpreted carefully because maturation and myelination processes increase the lipid and protein content but the water content of the white matter. These changes explain the findings of hypoattenuations in neonates with normal development.
    • The long-term changes seen on a CT scan of patients with PVL are thinning of the periventricular white matter (especially in the region of the trigone) and ventriculomegaly with an irregular outline and deep sulci on the wall of the lateral ventricles.
    • Areas of white-matter necrosis may become calcified over time.
  • In thrombosis of the sagittal sinus, CT may depict the delta sign (clot in the sinus) or the empty delta sign (partially recanalized clot in the sinus).The sagittal sinus may be hyperattenuating, but CT scans often do not show cerebral venous thrombosis in the neonatal period; this thrombosis is better visualized on MRI than on CT.
• MRI in term HIE
  • MRI is the imaging modality of choice in the assessment of HIE, and it is sensitive in detecting focal and multifocal ischemic lesions.
  • In general, diffusion-weighted imaging (DWI) is the most sensitive technique for detecting ischemia, though it can delay the detection of ischemia in neonates compared with adults. Changes on DWI are correlated with clinical outcomes and have been reported within 6-8 hours of life in neonates who had presumed HIE. Nonetheless, these changes may not be reliably seen until after 24 hours.
  • The radiologic classification of ischemic lesions is simpler than the pathologic one. Barkovich and Triulzi divide the lesions presumably due to HIE in term neonates into 3 categories that somewhat depend on the severity and the mechanism of the insult, as follows:
    • Mild-to-moderate insult, primarily hypotensive injuries: Lesions are in the arterial watershed zone between major brain arteries (also referred to as a parasagittal pattern).
    • Severe insult, primarily energy-failure injuries due to a combination of decreased oxygen delivery and local relatively higher metabolic rate: Bilateral abnormalities are seen primarily in the lateral thalami, posterior putamina, hippocampi, and perirolandic cortices leading to lesions of the corticospinal tract. The dorsal mesencephalon is involved less frequently than the other areas.
    • More severe insult: A third and most severe pattern with diffuse cortical abnormalities can be seen in dramatic cases.
  • In a 2005 multicenter study, the watershed pattern was most prevalent, followed by the basal ganglial–perirolandic pattern. Compared with other patients, those with the abnormalities of the deep gray matter required more aggressive resuscitation, they had more severe NE and seizures, and they had worse motor and cognitive outcomes.
  • The location, pattern of abnormalities and order of appearance on the various sequences are summarized on table 1. Initially only DWI changes may be seen on the first 24-72 hours. Some investigations have found measurement of the apparent diffusion coefficient (ADC), which is calculated from the DWI, to be more sensitive than visual inspection of the DWI sequence early on.
  • In the first 2 days of life after birth injury a transient decrease T1 signal may be seen particularly in the lateral thalami and posterior putamen. This T1 signal rapidly becomes hyperintensity, attributed to lipid break down or mineralization. At the same time an increased T2 signal may be seen.
  • Abnormalities if the basal ganglia, thalamus, and internal capsule have been correlated with motor dysfunction on the first 3 years of life. Motor impairments appear to be most severe when these lesions also have high lactate-to-choline: ratios on magnetic resonance spectroscopy (MRS) (see below).
  • The watershed or parasagittal pattern of insult is associated with late cognitive impairment.
  • MRI is also the imaging modality of choice to demonstrate parasagittal injuries, which appear as areas of increased signal intensity in the watershed distribution on coronal and axial T2-weighted images.
  • In patients with early MRI changes suggestive of cortical necrosis, follow-up examinations may show cortical atrophy and multicystic encephalomata.
• Magnetic resonance spectroscopy
  • Proton MRS may reveal indirect evidence of neuronal damage by showing a decreased ratio of N -acetylaspartate (NAA) to choline and by showing elevated lactate peaks. These findings are somewhat correlated with subsequent neurologic deficits.
  • High lactate-to-choline ratios with basal ganglial and thalamic abnormalities appear to be correlated with poor neurologic function after the neonatal period.
  • Phosphorus-31 MRS in patients with presumed hypoxic-ischemic exposure shows a pattern of elevation inorganic phosphate levels associated with decreased phosphocreatine values. This finding belies a loss of high-energy phosphates. A decreased adenosine triphosphate (ATP) level is associated with death in the neonatal period. Changes on³¹ P MRS occur mostly in the first 24-72 hours, with return to normal in subsequent days.
• Timing of MRI and MRS changes
  • During the first 24 hours, DWIs may still be normal. Proton MRS may show an increase in lactate, which may be seen during the first day of life. After the first 24 hours, proton MRS also shows a decreased NAA-to-choline ratio.
  • Restricted diffusion of water molecules can reliably be seen on DWI after 1-3 days. The reason for this slow onset of DWI changes in neonates compared with older children and adults is unknown. Table 2. Summary of MRI and MRS Findings in HIE

    Time	Finding
    24 h	Increased lactate peak
    24-72 h	Increased NAA-to-choline ratio and DWI signal intensity
    >72 h	Increased T2-weighted signal intensity
    1-3 wk	Generalized atrophy (ex vacuo hydrocephalus), cystic changes (polycystic encephalomata)

• MRI in pre-term infants - PVL
  • In the premature infant, an acute PVL lesion can be demonstrated on MRI. MRI can be more sensitive than sonography, particularly to noncystic PVL. The sequelae of PVL are visualized better on MRI than on CT. PVL and its sequelae are often seen in premature infants who have severe respiratory and other medical complications leading to poor oxygenation, decreased blood pressure, or both.
  • MRI signs of remote PVL include thinning and increased T2 signal intensity of the white matter, particularly in the peritrigonal area. This pattern has been correlated with spastic diplegia in formerly premature infants with a history of PVL, but it is also frequently observed in term infants after HIE.

Other Tests

• Electroencephalograph
  • EEG allows for the diagnosis of neonatal seizures and helps in determining the prognosis for infants with HIE-NE.
  • EEG studies of neonates with HIE showed that low-voltage (5- to 15-mV) activity, electrocerebral inactivity (voltage, <5 mV), and burst-suppression patterns are predictive of a poor outcome on follow-up neurodevelopmental examination. Normal EEG activity and maturational delay were not associated with excess morbidity on follow-up. Some data suggest that EEG performed in the first 2 weeks of life may be better than physical-neurologic examination because it increases the specificity for predicting abnormal outcomes.
  • One pattern that portends a poor prognosis is the burst-suppression pattern. This pattern contains bursts of high-voltage activity composed of a mixture of delta-theta rhythms and spikes and sharp waves of 1-10 seconds alternating with low amplitude (background suppression) with <5 V. During the bursts, no age-appropriate activity is seen. The burst-suppression pattern is associated with a grim prognosis.
    • EEGs may show burst-suppression during sleep but a continuous tracing when the patient wakes up. In these patients, burst-suppression is seen during most of the recording and only vigorous stimulation wakes the patient.
    • The outcomes of patients with reactive burst suppression are somewhat better than those of neonates with nonreactive burst suppression. Approximately 20% of patients with reactive burst suppression have severe disability on follow-up, and the rest have mild-to–moderately severe sequelae.
    • Nonreactive burst suppression is associated with an 86-100% risk of death or severe sequelae on follow-up. Use caution and serial EEGs in premature infants born at less than 33 weeks' gestational age before confirming the diagnosis of a burst-suppression pattern.
  • Besides HIE, the following have been associated with a burst-suppression pattern:
    • Acquired and congenital infections
    • Inborn errors of metabolism (eg, nonketotic hyperglycinemia)
    • Chromosomal abnormalities
    • Brain dysgenesis
    • Intraventricular or periventricular hemorrhage
    • PVL
    • Focal cerebral infarcts
    • Pontosubicular necrosis
  • Separating the prognostic value of EEG seizure patterns from the EEG background is difficult. In some studies, EEG seizures had no independent prognostic value above that of the background abnormalities. However, low-frequency (1- to 1.5-Hz) discharges in a suppressed background are correlated with a poor prognosis.
  • EEG background patterns of low voltage (<15 mV), burst suppression, and the ominous isoelectric EEG (<2 mV) are associated with a poor outcome in HIE. Neonatal seizures confirmed on EEG are associated with a poor prognosis, especially if the seizures are accompanied by background abnormalities. Normal EEG findings are correlated with a normal neurologic outcome, unless signs of severe brainstem damage are noted after an episode of complete ischemia.
• Amplitude-integrated EEG (aEEG)
  • Although the bulk of the studies have used standard neonatal EEGs, the body of evidence in regard to aEEG and HIE has been growing in the last 10 years.
  • Integrating the amplitude domain (generally by using fast-Fourier transformation) makes the EEG easier to interpret than before so that pediatric house staff without previous training in neonatal EEG can both read the amplitude and detect neonatal seizures. Learning to read neonatal EEG takes 1-2 years.
  • aEEG studies have used criteria that take in consideration both the upper margin of the aEEG band and the lower margin or lowest amplitude of the EEG voltage.
    • Normal is an upper range of >10 µV and a lower margin of >5 µV.
    • Moderate abnormalities have an upper range of >10 µV and a lower margin of 5 µV or less.
    • Severe abnormalities have an upper range of <10 µV and a lower margin of <5 µV, usually with a burst-suppression pattern.
  • An experienced neonatal encephalographer immediately notices the potential for overlapping of normal and moderate ranges because the lowest possible voltage is less reliable than the maximum voltage in neonatal EEG. The transition from non–rapid-eye-movement (REM) sleep to REM sleep (or REM-2) occurs when neonates behaviorally appear to be asleep, sometimes for over an hour. The start of REM-2 is associated with a notable reduction in background amplitude.
  • Although this method is promising, further validation of the accuracy of aEEG in trials of patients undergoing both aEEG and standard EEG are needed. Also needed is a demonstration that aEEG adds to the accuracy of other modalities (eg, DWI, MRS, evoked-potential testing).
• Evoked-potential testing
  • In several studies of neonates at high risk for poor neurologic outcomes, both somatosensory evoked potentials (SSEPs) and visual evoked potentials (VEPs) have utility in predicting the long-term outcome of HIE-NE. Persistently and bilaterally absent SSEP cortical potentials in high-risk term neonates are correlated with adverse neurologic sequelae after HIE-NE.
  • SSEPs obtained by the end of the first week of life appear to have the highest predictive value. Bilateral absence or prolonged latencies of cortical potentials in premature infants are predictive of an adverse neurodevelopmental outcome. However, preterm infants with normal SSEPs may have a decreased likelihood of a normal outcome. SSEP can be abnormal owing to lesions of the median nerve or of the median nerve fibers as they go through the brachial plexus or posterior cervical roots (C6-T1). Spinal-cord lesions, especially posterior lesions, affecting the proprioceptive large-fiber system and diseases that affect the peripheral myelin in the neonatal period (eg, congenital hypomyelinating neuropathy) also cause abnormal SSEPs.
  • High-risk neonates with a clinical picture suggestive of HIE and abnormal VEPs (obtained on days 3-7 of life) have a high risk of dying in the neonatal period or of having severe neurologic deficits on follow-up. Normal VEPs have negative predictive values lower than those of SSEPs. (A normal VEP does not guarantee a normal outcome.) VEPs that are initially abnormal but that improve over time still tend to indicate poor prognosis. Bilateral eye and optic-nerve lesions invalidate VEP results. Retinal function can be assessed by simultaneously registering the electroretinogram during recording of the VEP. When both the VEP and the electroretinogram are absent in a neonate, prognostication regarding HIE-NE is invalid.
  • In neonates with an HIE-NE picture, SSEPs and VEPs are best for predicting the outcome by the end of the first week of life. The combination of normal VEPs and SSEPs is a strong predictor of a normal outcome.
  • Scalp edema, subdural hematomas, and epidural hematomas can produce falsely abnormal results with either SSEPs or VEPs.
  • Brainstem auditory evoked potentials (BAEPs) may be useful in evaluating potential brainstem injury, but they lack the predictability of VEPs and SSEPs.
  • The combination of SSEP, VEP, and BAEP neurophysiologic tests is ideal to predict the neurologic outcome after HIE.
• Near-infrared spectroscopy: Near-infrared spectroscopy appears to be a promising modality for monitoring ischemic events in the CNS. It is used to monitor cerebral oxygenated hemoglobin levels, and it can depict clinically significant changes that occur during acute hypoxemic events.
• Intracranial pressure (ICP) monitoring: In HIE, brain edema is a consequence of severe cerebral necrosis, and it is not a common cause of ischemic cerebral injury. ICP is generally increased in severely asphyxiated term newborns after 24-48 hours of life. However, elevated ICP is not common in HIE of the premature infant unless concomitant posthemorrhagic hydrocephalus is present. ICP is not routinely monitored in neonates with HIE because it rarely leads to changes in their care.
• Technetium scanning: This technique, now rarely used, reflects the increased uptake of radionucleotides by damaged brain tissue that occurs after the blood-brain barrier is disrupted.

Procedures

• Although findings on lumbar puncture are almost never diagnostic of HIE-NE, they help exclude other entities, such as meningitis and hemorrhage. In addition, the opening CSF pressure may help in detecting increased ICP.
• Several CSF markers of HIE-NE are currently under investigation for their usefulness and validity in predicting long-term outcomes. Those most studied are CK-BB, lactate, and neuron-specific enolase.

Histologic Findings

Neuropathologic patterns and pathogenesis of neonatal HIE

According to Volpe, 5 major neuropathologic patterns described in patients are believed to be related to perinatal insults due to HIE. They are parasagittal cerebral injury, PVL, selective neuronal necrosis, status marmoratus of the basal ganglia and thalamus, and focal and multifocal ischemic brain necrosis. Most cases of neonatal HIE result from antenatal injuries. Nonetheless, the exact timing of these perinatal insults is often hard to pinpoint. In addition, some neuropathologic patterns of injury, like PVL, are related to the gestational age of the newborn infants.

Parasagittal cerebral injury

Parasagittal cerebral injury is a major ischemic lesion of the term infant. The timing of injury is primarily perinatal. The characteristic distribution of the injury, as indicated by the name, is in the parasagittal or superomedial aspect of the cerebral convexities. This topographic distribution corresponds to the locations of arterial end zones and border zones of the anterior, medial, and posterior cerebral artery territories. The pattern results because systemic hypotension preferentially affects the watershed areas. Injury usually is bilateral and symmetric, and the posterior portions of the cerebral hemispheres are affected more than the anterior parts.

The most extensive injury seen in the posterior cerebrum, probably because this region represents the watershed of the anterior, middle, and posterior cerebral arteries. In addition, posterior cortex is metabolically more active and therefore more vulnerable than other areas.

In the neonatal period, a parasagittal cerebral injury clinically manifests as weakness in the proximal extremities, arms more than legs. The long-term correlates of the parasagittal cerebral injury include proximal greater-than-distal and arm-more-than-leg involvement of spastic quadriparesis. Specific cognitive deficits, such as language dysfunction and visuospatial or visuomotor impairment, have been described and are probably due to temporoposterior parieto-occipital involvement.

Periventricular leukomalacia

PVL is most common in premature infants who survive at least a few days, and it is considered the main ischemic lesion in this population. Infants with low birth weight (<1.5-2 kg) and with a history of cardiorespiratory disturbances, especially those requiring ventilatory support, have an increased incidence of PVL. The offending ischemic insults can be either prenatal or postnatal.

PVL is characterized by necrosis of the periventricular white matter, dorsolateral to the external angles of the lateral ventricles, that occasionally becomes hemorrhagic. Hemorrhagic PVL in association with intraventricular hemorrhage may be indistinguishable from the periventricular hemorrhagic venous infarct that commonly accompanies intraventricular bleeding. Two common sites for PVL are around the anterior horns of the lateral ventricles (ie, white matter around the foramen of Monro) and around the trigones at the level of the parieto-occipital junction or optic radiation. A potentially deleterious effect of PVL is destruction of subplate zone neurons, which may interfere with cortical organization and thalamic and cortical connectivity.

The pathogenesis of PVL is related to a combination of vascular, metabolic, local, and systemic circulating factors. Before a gestational age of 32 weeks, and especially before 28 weeks, the long, penetrating cerebral arteries have few anastomoses with the short penetrators, which are few. Between 24 and 28 weeks gestational age, the periventricular arterial end zone, including areas of the white matter relatively distant from the immediate periventricular zone, is relatively large.

In addition, the pressure-passive cerebral circulation and the increased sensitivity of oligodendroglia cells to injury in premature neonates increase their susceptibility to ischemic injury. For this reason, extremely premature infant brains are most susceptible to even mild degrees of hypotension. From 28 weeks' gestational age to term, the periventricular vasculature increases progressively, decreasing the neonate's susceptibility to PVL.

In the premature infant, many common systemic problems have been associated with PVL. These include severe respiratory distress, apnea, myocardial failure, patent ductus arteriosus, hypotensive episodes (<30 mm Hg), hypovolemia, acidosis, and hypocarbia. About one third of patients with a patent ductus arteriosus that has retrograde flow during diastole develop PVL. Premature infants who are small for their gestational age are particularly vulnerable to PVL.

During the neonatal period, identifying a specific pattern of neurologic deficit in the patient with PVL may be difficult. On occasion, weakness in the legs may be observed in a neonate with PVL. The major long-term neurologic correlate of PVL is spastic diplegia involving the lower extremities more than the upper extremities. This pattern of injury to the corticospinal tract is thought to be a direct consequence of damage to the leg fibers as they go through the zone of periventricular necrosis. Corticospinal-tract fibers bound to the trunk, arm, face, and mouth have a relatively direct trajectory as they go from the cortex to the centrum semiovale and internal capsule, often avoiding the PVL area; however, large lesions with lateral extension can affect these fibers.

Patients with clinically significant spasticity involving the upper and lower extremities are most likely to have intellectual deficits. On the contrary, patients with sonographically depicted noncavitary lesions may have substantial improvement of spastic diplegia in the first several years of life. Patients with PVL may also have visuomotor and visual-field disturbances, which are probably related to damage to the optic radiations and visual association fibers.

Selective neuronal necrosis

Selective neuronal necrosis is a prominent injury pattern in infants who have hypoxic-ischemic injury in the postnatal period, and it is often associated with other patterns of HIE injury. Although the neuropathologic injury pattern seems to mirror the regional distribution of glutamatergic neurons (which can mediate neurotoxicity), some regional vascular factors may also play a role, as the neuronal injury is most prominent in the depths of sulci and watershed zones. Neuronal injury occurs at specific sites of the cerebral cortex, such as the Sommer sector of the hippocampus and, in severe cases, the calcarine and central cortices.

Other affected sites in the CNS include the diencephalon, brainstem, Purkinje cells of the cerebellum, and spinal cord. In the less commonly affected premature infants, sites of predilection include the subiculum in the hippocampus, basis pontis, internal granular layer of the cerebellum, and inferior olivary nuclei.

During the neonatal period, selective neuronal necrosis produces a clinical picture of stupor and coma or of hypotonia with oculomotor, sucking, and swallowing disturbances, as well as seizures and abnormal tongue movements. The major long-term sequelae are cognitive deficits, spastic quadriparesis, seizure disorder, ataxia, bulbar and pseudobulbar palsies, hyperactivity, and inattention.

In rare cases, presumed perinatal total asphyxia is followed by a lack of body movements and no brainstem, cranial-nerve, or mediated reflexes. In these cases, pupillary constriction to light is often absent, as are facial and eye movements (spontaneous or induced by an oculocephalic maneuver). The vulnerable pontine area is the watershed zone between the paramedian and short and long circumferential branches of the basilar artery. In the medulla, the vulnerable area is between the anterior and posterior branches of the spinal and vertebral arteries. Structures in the pontomedullary watershed zone, which hypoperfusion may affect, are parts of the respiratory center. As a result, the clinical correlates of this lesion include decreased respiratory drive, apnea, and decreased arousal.

Status marmoratus

Status marmoratus is the least common pattern of HIE-related neuropathologic injury, and it is more common in term infants than preterm infants. This injury is thought to be of postnatal onset, though prenatal factors may play a role in some patients. The terms status marmoratus and état marbré refer to the marbled appearance of the deep nuclear structures, especially the putamen, caused by hypermyelination surrounding the astrocytic fibers.

Neuronal injury primarily occurs in the putamen (particularly the dorsal part), globus pallidus, and thalamus (ie, ventral, medial, lateral nuclei). Marked changes in the thalamus occur in 80-90% of patients. Microscopic features include neuronal loss, astrogliosis, and hypermyelination around astrocytic fibers. The physiopathology also invokes the neurotoxicity mechanism, since the distribution of lesions is related to the distribution of N -methyl D-aspartate (NMDA)–glutamate receptors.

During the neonatal period, no clear-cut pattern of neurologic deficit can be attributed to status marmoratus. Long-term survivors have movement disorders or cognitive deficits. The movement disorder is often occult until the child is aged 1-4 years, and it may be preceded by a history of delayed motor development and hypotonia. However, in many instances, motor development may be normal until the movement disorder appears. In rare cases, the onset of choreoathetosis and dystonia may be delayed until the child is aged 7-14 years. Spastic quadriparesis occurs in about one third of patients, more commonly among patients with dystonia and less commonly in those with choreoathetosis. Cognitive deficits can occur, but normal intelligence is possible, especially in patients with delayed-onset movement disorder.

Focal and multifocal ischemic brain necrosis

An estimated 20% of infants with HIE have focal or multifocal lesions. These lesions are most common in term infants with postnatal ischemic insults. Some experts prefer to exclude focal lesions (eg, single vessel occlusion) from HIE-NE because, in many cases, focal cerebral infarcts are not associated with systemic ischemia or hypoxemia. However, in many cases both focal (eg, embolic) and generalized (eg, heart failure) ischemia coexist, as in the case of sepsis with disseminated intravascular coagulation causing thromboembolic phenomena. The hallmark of focal and multifocal ischemic brain necrosis is the destruction (necrosis) of all cellular elements in the distribution of a single vessel or several vessels (ie, single or multiple infarcts, respectively).

When these infarcts develop cavitation due to the dissolution of brain parenchyma, a porencephalic cyst (if single), multicystic encephalomalacia (if multiple), or even hydranencephaly (if multiple and extensive) forms. These cavitated lesions may communicate with the ventricular system. The tendency of the fetal neonatal brain to develop cavitation after ischemic insults is likely related to its high water content, low numbers of myelinated fibers, and poor astroglial response. Distinction between this focal or multifocal necrosis, parasagittal injury, and PVL may be difficult, and, in some cases, these injury patterns coexist.

The incidence of arterial occlusion varies with gestational age. It is rare in infants younger than 28 weeks and gradually increases with advancing maturation, affecting about 15% of full-term infants in the autopsy series. The middle cerebral artery is affected in more than one half of patients, and the lesions are usually unilateral.

Focal or multifocal infarcts may be the result of cerebral venous thrombosis. The superior sagittal sinus is involved in about 85% of patients (most often posteriorly), and the rest affect the lateral sinuses and the galenic system deep veins. Hemorrhagic conversion is common in cases of venous infarcts.

The pathogenesis involves thromboembolic vessel occlusion, vasculopathy, vasospasm due to vasoconstricting drugs (eg, cocaine, methamphetamine), or a vascular malformation.

• Embolic phenomena may be due to placental fragments, an embolus originating from a dead twin (detritus), catheterized vessels, congenital heart disease with right-left shunting, fragments of cardiac tumors (eg, atrial myxoma and rhabdomyoma in tuberous sclerosis), involuting vessels (eg, umbilical vein, ductus arteriosus), and isoimmune neonatal thrombocytopenia. Isoimmune neonatal thrombocytopenia may be associated with obstruction of the middle cerebral artery in utero, perhaps due to endothelial injury. Pathologically documenting an intra-arterial embolus is often difficult.
• A thrombotic tendency that may lead to vessel occlusion is seen in hypercoagulable states, disseminated intravascular coagulation, dehydration, trauma, meningitis (eg, arteritis or phlebitis), and neck or cranial trauma (eg, vascular injury). Hypercoagulable states with neonatal manifestations include protein C and S deficiencies, antithrombin III deficiency, antiphospholipid antibody syndrome, and polycythemia with hyperviscosity.
• ECMO is a multifactorial cause of focal, multifocal, and generalized ischemia. Most lesions are hemorrhagic, with or without superimposed ischemia; however, isolated ischemic lesions represent 40% of the imaging abnormalities in patients who received ECMO. Explanations for ECMO-related cerebrovascular insults include anticoagulation with heparin (eg, hemorrhagic conversion of a previous ischemic lesion), jugular-vein ligation, decreased blood-flow velocity in the venous system, retrograde flow in the right vertebral artery, increased blood-flow velocity in the left internal carotid artery, and changes in blood-flow direction in the circle of Willis.
• Approximately one third of focal ischemic lesions have evidence of a generalized-systemic circulatory insufficiency of either prenatal or postnatal origin. However, the pathophysiology of the focal nature of the resulting lesion is not clear.
• Despite progress in evaluating focal cerebrovascular ischemia, the etiology in more than one half of patients remains undiagnosed.

Neonatal seizure is the presenting symptom for at least 80% of patients with a focal unilateral cerebral infarct. Mild hemiparesis may be noted in the newborn infant with large unilateral lesions. The long-term sequelae of focal or multifocal cerebral necrosis include spastic hemiparesis and quadriparesis (ie, bilateral hemiparesis), cognitive deficits, and seizures. Follow-up examination of patients with unilateral neonatal infarction shows hemiparesis in only 55%, but a lower incidence of residual weakness has been reported. This relatively low frequency of residual hemiparesis is probably related to both the severity of the lesions and the subsequent reorganization of cortical function. Cognitive dysfunction is reported in about 30% of patients with focal infarcts. Patients with cerebral venous infarcts frequently have seizures as a presenting symptom.

Staging

• Stage 1
  • Hyperalertness
  • Decreased sleep
  • Uninhibited reflexes
  • Excessive reaction to stimuli
  • Weak suck but normal tone
  • Sympathetic overactivity - Eyes wide open, decreased blinking, mydriasis, and EEG normal
  • Duration less than 24 hours
  • Good prognosis - No long-term neurologic sequelae
• Stage 2
  • Lethargy or obtundation (ie, delayed and incomplete response sensory stimuli)
  • Mild hypotonia
  • Cortical thumbs
  • Suppressed primitive reflexes
  • Seizures
  • Hypotonia
  • Lethargy
  • Parasympathetic activation with miosis (even on dim light), heart rate less than 120 beats per minute, increased peristalsis, and copious secretions)
  • EEG early (first day): Relatively low voltage, less than 25 microvolts (slow theta and delta)
  • EEG late (second day): Bursting pattern (awake or obtunded) and multifocal low-frequency (1-1.5 Hz) electrographic seizures
  • Good prognosis if clinical and EEG recovery within 5 days
  • Poor prognosis if periodic EEG with interburst intervals totally isoelectric, bursting frequency less than every 6 seconds, bursting pattern (every 3-6 seconds) lasting more than 7 days
• Stage 3
  • Stupor response only to strong stimuli with withdrawal or decerebrate posturing only
  • Rarely coma
  • Severe hypotonia (ie, flaccidity)
  • Suppression of deep tendon and primitive (ie, Moro, tonic neck, oculocephalic, suck) reflexes
  • Suppression of brainstem reflexes (corneal or gag)
  • Clinical seizures less frequent than stage 2
  • Deep, periodic EEG pattern with high amplitude and frequency of bursts less than every 6-12 seconds, very-low-voltage or isoelectric EEG
  • Major neurologic sequelae, including microcephaly, mental retardation, CP, seizures (all cases)

Treatment

Medical Care

The primary goal of medical care is to prevent prematurity and other prenatal and postnatal causes of HIE. In most cases, this means prevention of intrauterine ischemia and hypoxemia. Good prenatal care and management of medical conditions, such as maternal diabetes, are the most important means of reducing the risk of HIE and preterm labor early. The detection of fetal distress is important late in the pregnancy; however, methods for detecting fetal distress are far from perfect. Large-scale studies have not shown that fetal heart-rate monitoring is effective in preventing late motor disability or CP. The principles of prevention and management of HIE are described below.

• Prevention and management of HIE
  • Provide good prenatal care and detect and manage the mother's medical conditions, as well as promptly recognize and appropriately treat clinically significant fetal distress.
  • Resuscitate and stabilize the depressed neonate in the delivery room.
  • Perform early serial neurologic evaluation of the depressed infant.
  • Supportive care: Avoid hypertension-hypoperfusion, provide adequate oxygenation and ventilation, and correct metabolic abnormalities (eg, hypoglycemia, acidosis, electrolyte imbalance).
  • Other measures: In addition to the measures listed above, prevent secondary CNS insults; control seizures; administer anticonvulsants (eg, phenobarbital, phenytoin, benzodiazepine); correct hypoglycemia, hyponatremia (slowly correct), hypocalcemia, and hypomagnesemia; control brain edema (increased ICP); and prevent fluid overload. Use of mannitol and dexamethasone is not indicated. No effective therapy to control neurotoxicity is available at this time.
• After the diagnosis of HIE has been established, treatment is largely supportive and symptomatic. The utility of pharmacologic agents to prevent perinatal brain damage has not been established. The use of glutamate antagonists to reduce secondary hypoxic-ischemic damage is currently under investigation.
• In the delivery room, appropriate resuscitation of the depressed infant should be initiated. Specific guidelines for neonatal resuscitation are beyond the scope of this article. Sensible guidelines from the American Academy of Pediatrics and the American Heart Association are contained in the Neonatal Resuscitation Program. In addition, information is available in Neonatal Resuscitation.
  • Appropriate evaluation of the newborn in the delivery room is also crucial. Apgar scores should be determined at 1 and 5 minutes and every 5 minutes thereafter if the score is less than 7.
  • Neurologic examination should be part of the initial assessment. Include an assessment of the patient's degree of arousal and alertness, cranial-nerve or brainstem function (eg, pupils, eye and facial movements, gag and suck reflexes), axial and limb tone, and motor activity. Do not wait for the neurologist because the findings may change spontaneously or after medication.
• Fluid management is initially aimed at establishing normal tissue perfusion. After the first 24 hours, fluid overload must be avoided, especially in patients with the oliguric phase of renal failure or cerebral edema. Although arterial hypotension may worsen ischemic lesions, hypertension increases the risk of hemorrhagic complications, especially in previously ischemic tissue. Therefore, extremes in blood pressure should be avoided if possible.
  • Brain swelling is not the primary event in HIE and usually occurs after the first or second day of life in association with cerebral necrosis in full-term infants. Although ICP increases in as many as 22% of neonates with HIE, decreased cerebral perfusion pressure (ICP minus mean arterial pressure) is rare and often due to systemic arterial hypotension. In addition, transtentorial and tonsillar herniation is rare in the neonate with cerebral edema. Therapy with dexamethasone or mannitol to decrease ICP is not recommended because it does not appear to improve the outcome in HIE. The main strategy to address brain swelling is the prevention of fluid overload.
  • The syndrome of inappropriate antidiuretic hormone secretion (SIADH) may complicate the care of patients with severe HIE. This syndrome is due to the decreased excretion of free water and its consequences; therefore, it should be treated with careful fluid restriction. SIADH is characterized by hyponatremia, low serum osmolality, and high urinary osmolality with continued urinary excretion of sodium despite fluid overload with bulging fontanel. In cases of SIADH, seizures are probably due to low sodium levels or fluid-electrolyte shifts.
• Respiratory support is often necessary in neonates with HIE, even those without seizures or high doses of sedating medications. Respiratory treatment of patients with persistent pulmonary hypertension may be particularly challenging.
  • Appropriate oxygenation is the goal of respiratory support because both hypoxemia and hyperoxia may accentuate neuronal damage. In addition, hyperoxia can substantially reduce cerebral blood flow.
  • Carbon-dioxide disturbances may be deleterious to the brain. Hypercarbia may induce acidosis or cerebrovascular dilatation, increasing the risk of hemorrhage and steal phenomenon. Steal phenomenon is characterized by decreased blood flow to areas of reversible ischemia (infarct penumbra), which are due to vasodilation of arteries in the nonischemic regions. These reversibly ischemic areas often surround regions with complete infarct. Hypocarbia can worsen ischemic injury by inducing vasoconstriction.
• Neonatal seizures are commonly a consequence or comorbid condition of HIE-NE and need to be treated appropriately.
• Hypocalcemia is often seen in HIE-NE. Hypocalcemia with or without secondary hypomagnesemia may exacerbate or even trigger seizures; therefore, treat it when detected. Ionized calcium measurements are preferred to monitor the treatment of hypocalcemia, as this represents the physiologically active fraction of the total serum calcium value. Elevated serum magnesium levels (often due to maternal administration) may produce hypotonia and even respiratory failure. Hypoglycemia should be corrected when present, but the indiscriminate use of glucose is not recommended, as data from animal models of ischemia suggest that raising the glucose above normal levels may increase neuronal damage.

Consultations

If the patient survives the acute physiologic derangements of the first several days of life, he or she will likely require long-term care and follow-up.

• Patients with an HIE-NE picture are at risk for long-term neurologic morbidity; therefore, they require close monitoring at a developmental follow-up clinic in consultation with a developmental specialist and a neurologist.
  • Although the seizures in some patients stop after the neonatal period, they often recur after a few months or years.
  • Most often, the seizures that occur after HIE-NE are difficult to manage, and the patients may develop epileptic encephalopathies, such as infantile spasms or myoclonic epilepsies.
• Patients with motor, speech, or cognitive disabilities need careful follow-up, serial testing, and therapy by occupational, physical, and speech therapists to optimize their developmental outcome. Early-intervention programs (at 0-3 y) are helpful in providing these services in the United States.
• If irreversible muscle-tendon contractures develop, consultation with an orthopedic surgeon is advisable.
• Careful testing by a neuropsychologist is instrumental in placing these children in the appropriate school environment.

Medication

Neonatal seizures may be a consequence or comorbid condition of HIE-NE. Neonatal seizures increase the cerebral metabolic rate, worsening the after-effects of ischemia. Seizures also are expected to increase cerebral blood flow, potentially increasing the risk of intracerebral hemorrhage. Another theoretical consideration is induction of seizure-mediated neurotoxicity. These factors have led clinicians to treat neonatal seizures. A complete discussion of the treatment of neonatal seizures is beyond the scope of this review.

Phenobarbital is one of the most popular choices in the conventional management of neonatal seizures. The initial infusion stops the seizures in as many as one third of patients. If patients do not respond to the initial infusion, the administration of 1-2 additional doses raises the serum level and controls the seizures in 70-80%. If the seizures persist despite the achievement of therapeutic levels of phenobarbital, adding a second drug (eg, fosphenytoin, lorazepam) may be advisable. In many centers, fosphenytoin is used in place of its parent drug (ie, phenytoin) because of the adverse reactions related to the diluent (ie, propylene glycol) used with phenytoin, especially those related to extravasation of the intravenous infusion. Lorazepam is an alternative for patients whose seizures do not respond to fosphenytoin and phenobarbital. Seizures can recur with lorazepam because of its 8- to 12-hour duration of effect.

An alternate strategy is the administration of repeated boluses of phenobarbital until the seizures stop. In this strategy, serum levels are raised to 100 mcg/mL or more, depending on the patient's cardiovascular tolerance. If this approach is used, the physician should be ready to provide ventilatory support because respiratory failure almost always ensues when phenobarbital levels are raised rapidly. One of the drawbacks to this approach is that after phenobarbital levels are >40 mcg/mL, large dosing increments are necessary to achieve additional small increases in the percentage of patients becoming seizure free.

Anticonvulsants

These agents may be used in the treatment of neonatal seizures.

Phenobarbital (Barbita, Luminal, Solfoton)

Volume of distribution in neonates 0.8-1 L/kg. Half-life 50-200 h; may be 150-200 h in neonates with HIE. Serum levels variable; higher levels correlated with higher percentage of seizure-free patients: 10-30 mcg/mL, 40%; 40 mcg/mL, 70%; 100 mcg/mL, 77%; and adding second antiepileptic drug (AED), 88%.
Respiratory failure usually occurs during acute loading, when serum levels >40-50 mg/dL but may occur earlier or later. Comedication with other sedating drugs, especially benzodiazepines, may increase respiratory depression. Levels >100 mcg/mL may cause cardiovascular instability in some patients.

Dosing

Adult

Pediatric

20 mg/kg IV
Infusion rate: 1 mg/kg/min IV
Repeat bolus: 10 mg/kg IV may be necessary
Maintenance: 2.5-5 mg/kg/d PO/IV

Interactions

May decrease effects of chloramphenicol, digitoxin, corticosteroids, carbamazepine, theophylline, verapamil, metronidazole, and anticoagulants (may need to adjust dose in patients whose coagulation parameters stabilized with anticoagulants); alcohol may produce additive CNS effects and death; chloramphenicol, valproic acid, and MAOIs may increase toxicity; rifampin may decrease effects; induction of microsomal enzymes may decrease effects of PO contraceptives in women (must use additional contraception to prevent unwanted pregnancy)

Contraindications

Documented hypersensitivity; severe respiratory disease; marked impairment of liver function; nephritis

Precautions

Pregnancy

D - Fetal risk shown in humans; use only if benefits outweigh risk to fetus

Precautions

Use cardiorespiratory monitoring during infusion; respiratory depression may occur with high doses or serum levels; be ready to provide respiratory support; menstrual irregularities

Fosphenytoin (Cerebyx)

Second-line drug after phenobarbital fails. Seizure control 31-41% with loading doses of 15-20 mg/kg. Full-term neonates may need levels 18-20 mg/dL; achieved only with high loading and maintenance doses. Phenytoin poorly bound (range, 6-41%; mean, 24%) to protein in critically ill neonates, especially those with hyperbilirubinemia. High free phenytoin levels may exacerbate seizures and cause cardiac arrhythmias. Volume of distribution of phenytoin in neonate is 0.89 ± 35 kg/L.

Half-life of phenytoin in newborns 6-140 h; tends to shorten with postnatal age (decreases by two thirds at age 1-4 wks). Serum concentration linearly correlated with half-life after age 8 d. Finding compatible with concentration-dependent kinetics that rely on saturable enzymatic metabolism for excretion; also seen in older patients.

Kinetics in first week of life may be unpredictable. In most cases, decreased, slowed PO absorption plus increasing clearance makes increasingly high PO maintenance doses (7-25 mg/kg/d) necessary near end of first month to keep therapeutic serum levels with PO maintenance.

Total urinary excretion in infancy vs older children is 30% vs 59%. In infants, PO absorption decreased further by fasting and increased (by 69%) with food. Therapeutic serum range 10-20 mg/dL.

Phenytoin prodrug. Popular recently because of relatively favorable toxicity profile. Dephosphorylates into phenytoin, but prodrug most tightly bound to serum protein, which rapidly increases serum levels after IV infusion.

Dosing

Adult

Pediatric

Loading doses: phenytoin equivalent 15-20 mg/kg IV (use creatinine clearance [CrCl] monitors); equivalents marked on vial
Bolus infusion rate: 1 mg/kg/min
Maintenance: phenytoin 7-25 mg/kg/d PO

Interactions

Amiodarone, benzodiazepines, chloramphenicol, cimetidine, disulfiram, ethanol (acute ingestion), omeprazole, phenacemide, phenylbutazone, Succinimides, fluconazole, isoniazid, metronidazole, miconazole, sulfonamides, trimethoprim, and valproic acid may increase toxicity; barbiturates, carbamazepine, theophylline, diazoxide, ethanol (chronic ingestion), rifampin, antacids, charcoal, and sucralfate may decrease effects; may decrease effects of acetaminophen, corticosteroids, dicumarol, disopyramide, doxycycline, estrogens, haloperidol, amiodarone, carbamazepine, cardiac glycosides, methadone, metyrapone, mexiletine, PO contraceptives, quinidine, theophylline, valproic acid

Contraindications

Documented hypersensitivity; sinoatrial block; second- and third-degree AV block; Adams-Stokes syndrome

Precautions

Pregnancy

D - Fetal risk shown in humans; use only if benefits outweigh risk to fetus

Precautions

Cardiorespiratory monitoring required for phenytoin and fosphenytoin; blood dyscrasias (order blood counts and urinalysis at start and q1mo thereafter); discontinue use if skin rash appears (if exfoliative, bullous, or purpuric, do not resume); caution in acute intermittent porphyria; may raise blood glucose levels; discontinue if hepatic dysfunction occurs

Lorazepam (Ativan)

Benzodiazepine to treat neonatal seizures. Main excretion urinary. No active metabolites known. Small volume of distribution for unbound form; partly why effective duration of action against seizures may be longer than that of diazepam.

Dosing

Adult

Pediatric

0.05 mg/kg IV over 3-5 min; repeat up to 3 times q15min

Interactions

Alcohol, phenothiazines, barbiturates, and MAOIs increase CNS toxicity

Contraindications

Documented hypersensitivity; preexisting CNS depression; hypotension; narrow-angle glaucoma

Precautions

Pregnancy

D - Fetal risk shown in humans; use only if benefits outweigh risk to fetus

Precautions

Caution in renal or hepatic impairment, myasthenia gravis, organic brain syndrome, or Parkinson disease; seizure recurrence possible because of 8- to 12-h duration of efficacy; attenuation of background or burst-suppression pattern may appear during and persist several hours after infusion

Follow-up

Prognosis

• Outcome measures and prognostic factors in the history of patients in whom HIE is suspected
  • One difficulty in assessing patients with suspected HIE-NE is finding reliable outcome measures that reflect the cerebral injury caused by cerebral hypoperfusion or ischemia. The rate of CP and mental retardation in the surviving cohorts is frequently used as an end point for comparison among the various factors and symptoms thought to be associated with perinatal hypoxemia and ischemia.
    • CP is defined as a chronic disability associated with an aberrant control of movement and posture appearing early in life but not associated with any known progressive disease.
    • In neuropathologic terms, CP reflects damage to the corticospinal tract that causes weakness or injury to the basal ganglia leading to a movement disorder (eg, dystonia, choreoathetosis).
    • One problem in using CP as an outcome measure is that only about 10% of the patients with CP have a history of an event in the perinatal period suggestive of hypoxic damage.
  • A few generalizations are possible regarding prediction of a patient's neurologic outcome after a hypoxic-ischemic event.
    • In full-term infants with birth weights greater than 2.5 kg, obstetric complications significantly increase the risk of subsequent neurologic morbidity in the infant only when followed by low Apgar scores at 5 minutes (or later) and signs of NE.
    • Both the duration and the severity of clinical NE are closely correlated with the outcome. Mild clinical NE suggests a normal neurologic outcome, whereas severe NE suggests motor deficits or death on follow-up. Prognostication in intermediate degrees of clinically defined NE depends on the duration of symptoms and on results of auxiliary testing, such as EEG, study of evoked potentials, and imaging (especially MRI). MRS may be helpful in these cases.
    • Many, if not most, clinical features of HIE-NE are nonspecific and can be due to other causes (see Differentials).
    Table 3. Factors Predictive of Adverse Neurologic Outcomes in HIE

    Data Source	Finding	Description
    Labor and delivery	General factors	Obstetric complications followed up 5-min Apgar scores <5 and signs of NE Umbilical-cord pH <7 (limited value) Birth weight <2.5 kg Microcephaly at birth Apgar score £ 5 at 5 min and after
    Neurologic examination	Neonatal neurologic syndrome	Decreased activity after first day of life Need for incubator >3 d in term infants Feeding problems, need for gavage in term infants Poor suck Respiratory difficulties Hypotonia or hypertonia (12- to 15-fold increase in CP rate) Weak cry (14- to 15-fold increase in CP rate) Facial palsy (18-fold increase in CP rate) Staging and severity Mild syndrome = Good prognosis Moderate syndrome = Prognosis difficult to predict (poor prognosis if >5 d) Severe syndrome with stupor or coma = Poor prognosis (high mortality and psychomotor disability rate on follow-up)
    	Neonatal seizures	Early onset (questioned) Persistent, difficult to treat (questioned) ICP Oliguria >36 h
    Ancillary tests	Imaging studies	Ultrasonography - Hemorrhage CT - Hemorrhage, hypoattenuation MRI - Hemorrhage, abnormal signal intensity MRS - NAA on proton study; ratio of inorganic phosphate to phosphocreatine and ATP on phosphorus study
    	Neurophysiology	EEG - Low voltage or burst suppression (poor outcome, normal and mild sequelae, intermediate patterns decrease predictability) SSEPs - Good prediction of normal outcome, bilateral lack of cortical potentials suggests poor outcome VEPs - If abnormal, good predictors of abnormal outcome BAEPs - Limited value alone Combined VEPs, SSEPs, BAEPs - Powerful predictors of outcome
    	Biochemical marker	Elevated CK-BB - CSF value more useful than blood value

• Labor-and-delivery complications and Apgar scores as predictors of outcome
  • NCPP researchers analyzed the obstetric and neonatal features associated with increased morbidity and mortality rates in the first year of life.
  • Placenta previa, abruptio placentae, breech delivery, face-brow presentation, and cord prolapse are associated with a higher rate of low birth weight (<2.5 kg) and low 5-minute Apgar scores. Breech delivery is associated with a higher rate of CP, as high as 5% among low-birth-weight infants.
  • Chorioamnionitis and rupture of membranes more than 24 hours before the delivery are associated with increased rates of infant mortality and CP on follow-up among infants with birth weight less than 2.5 kg.
  • An umbilical cord <40 cm is associated with increased mortality rate in low-birth-weight infants.
  • Among infants with birth weight greater than 2.5 kg, obstetric complications are associated with an increased rate of CP only if 5-minute Apgar scores are 5 or less or if signs of NE are seen. In these infants, 5-minute Apgar scores of 3 or less are associated with even more morbidity; however, those with Apgar scores of 7 or more have no increase in the risk of CP.
  • Among obstetric complications, only breech delivery and cord prolapse are associated with an increased risk of death in the first year in infants with birth weight greater than 2.5 kg. Although meconium-stained amniotic fluid is often mentioned as a risk factor for motor disability, this association is valid only for newborns with 5-minute Apgar scores of 3 or less.
  • In infants with HIE-NE, need of a respirator for >3 days, beginning of full feedings delayed by >6 days, and reappearance of normal activity >12 days are also associated with increased rates of CP. Infants with these findings also have an increased incidence of neonatal seizures.
• Apgar scores and NE
  • In the NCPP, features of NE, including decreased activity after the first day of life, need for incubator for >3 days, feeding problems (need for gavage), poor suck, and respiratory difficulties were associated with increased morbidity on follow-up. The combination of signs of NE, neonatal seizures, and 5-minute Apgar scores of 5 or less was associated with a 55% risk of CP on follow-up and a 33% risk of death in the first year. Neonatal seizures increase the risk of morbidity and mortality in patients with signs of NE or low Apgar scores. Table 4. Risk Factors in Neonates with Birth Weight >2.5 kg
    

    >Risk Factor	>CP Risk, %	>Death at Age <1 Year, %
    NE + neonatal seizure + 5-min Apgar score >5	10	29
    NE + neonatal seizure + 5-min Apgar score £ 5	55	33
    5-min Apgar score £ 3	4.7	No data
    10-min Apgar score £ 3	17	69*
    15-min Apgar score £ 3	36	69*
    20-min Apgar score £ 3	57	69*
    All patients with neonatal seizure	11.6	30
    Resuscitation >5 min + neonatal seizure	46	56

    Source.–NCPP.
    *A total of 270 (69%) of 390 patients with an Apgar score of £ 3 at 10 minutes or later died before the age of 1 year.
• Neurologic examination in the neonatal period
  • In the NCPP, limb or axial (neck or trunk) hypotonia or hypertonia in the neonatal period was associated with a 12- to 15-fold increase in the risk of CP (as high as 4%). Weak suck or cry was associated with a 14- to 15-fold increase in the rate of CP at follow-up examinations. Facial palsy was increased in CP at a rate of 18 times from baseline.
  • Several methods based on the neurologic examination were used to predict outcome after HIE. The Sarnat and Sarnat staging system is described in Grading systems for HIE.
    • Stage 1 is almost always associated with a good outcome, and stage 3 is strongly correlated with severe neurologic disabilities or neonatal death.
    • If signs and symptoms are reversed within 5 days, stage 2 is associated with recovery in all cases in the original study. Many patients progress to stage 2; in these patients, the outcome cannot be predicted in a timely fashion in the neonatal period, especially in the first 2-3 days of life.
    • The combination of stupor, flaccid hypotonia, and absence of primitive reflexes (clinical component of Sarnat and Sarnat stage 3) has been validated in large studies and is associated with death before discharge in 50% of patients. Neurologic sequelae are found in all survivors.
    • Studies in 79 patients with hyperalertness and hyperexcitability (similar to the clinical findings in Sarnat and Sarnat stage 1) showed good outcomes in 69 (10 were lost to follow-up).
    • Unimpaired survivors of moderately severe NE are most likely to have reading, spelling, and arithmetic scores more than 1 grade below their expected level when tested at the age of 8 years. Patients with mild NE do not differ from the peer groups in any of the test scores.
  • A scoring system of the neurologic examination known as the Lipper postasphyxial index was designed to predict outcome in the first 48 hours of life. The index is correlated with neurodevelopmental outcome at 1 year, especially when used in conjunction with a low attenuation index on the CT scan.
• Neonatal seizures and prognosis of HIE-NE
  • Neonatal seizures are useful markers of HIE. Large population-based studies have shown a substantial effect of neonatal seizures on the long-term morbidity and mortality rates after HIE. Table 5. Duration of Neonatal Seizures and Risks of Cerebral Palsy and Epilepsy
    

    Neonatal Seizures, d	CP, %	Epilepsy,%
    1	7	11
    2	15	22
    3	25	25
    >3	46	40

    Source.–NCPP follow-up data.
    Table 6. Neonatal Seizure Onset and Mortality Risk
    
    

    Neonatal Seizure Onset, d	Mortality Rate, %
    <0.5	53
    0.5-1	39
    1-2	37
    2-7	19
    7-28	31*

    Source.–NCPP follow-up data.
    *Possibly due to a high incidence of meningitis.
  • In the NCPP, the duration (days) of neonatal seizures was analyzed as an independent predictor of motor deficits (ie, CP) and epilepsy (ie, recurrent, unprovoked postneonatal seizures) on follow-up. Patients who had neonatal seizures for 1 day had a 7% CP rate, and 11% of them had epilepsy on follow-up. Approximately 46% of the patients who had seizures for > 3 days had CP, and 40% had epilepsy on follow-up.
  • Timing of the onset of neonatal seizures is somewhat correlated with early death. The death rate is maximal with onset in the first 12 hours and decreases in the first week of life. Between the end of the first week and the end of the neonatal period, the rate increases, most likely because neonatal meningitis causes seizures in this age group. In a Canadian (Edmonton, Alberta) cohort prospectively followed up to the age of 3.5 years, neonatal seizures in moderately severe NE (hypotonia and suppressed primitive reflexes) or severe NE (stupor, absent primitive reflexes) were associated with a 5- to 6-fold increase in sequelae on follow-up. Timing of neonatal seizures was not correlated with subsequent morbidity.
  • Neonatal seizures >30 minutes are associated with increased rates of mortality and mild-to-moderate mental retardation.
    • The question of whether improvements in obstetric and neonatal care in the past 20 years have improved outcomes in HIE-NE remains to be answered.
    • Recent data from a select population of infants with neonatal seizures (probably skewed toward severe cases) showed a mortality rate (33%) equivalent to that in the NCPP but with high rates of mental retardation (67%), CP (63%), and subsequent epilepsy (56%). Gestational age, birth weight, 1- and 5-minute Apgar scores, and age at onset of neonatal seizures had no effect on the incidence of subsequent epilepsy; however, neurologic and EEG abnormalities in the neonatal period did.
• Auxiliary testing in the prognostication of HIE-NE
  • Imaging has become useful in the prognostication of HIE. Follow-up of patients with perinatal focal cerebral infarction in the 1980s and 1990s showed that long-term sequelae included hemiparesis (55%), cognitive deficits (32%), and chronic seizures (30%). Language dysfunction after focal cerebral infarction had little to do with lateralization of the lesion.
  • As discussed, EEG and evoked potentials are helpful for prognostication in HIE. EEGs and evoked potentials are particularly helpful in neonates with moderately severe NE. SSEP is most helpful because of its negative predictive value (ie, predicting normal outcome), and VEP offers positive predictive value (ie, predicting abnormal outcome) in full-term infants. Bilaterally absent SSEPs are good indicators of disability on follow-up. Combined SSEP, VEP, and BAEP results are powerful predictors of the outcome after HIE.
  • CSF CK-BB and neuron-specific enolase may become useful prognostic markers of HIE in the future.
• Prophylaxis
  • Several attempts have been made to reduce the burden of the brain damage related to HIE-NE. Strategies to prevent neuroexcitatoxicity, inhibit the formation of nitrous oxide, prevent epileptogenicity, and others have been attempted. For the most part, the results have been disappointing.
  • Mild head cooling has recently been attempted. Early results showed a mild but significant improvement in outcomes in patients with mild-moderate aEEG changes, but further studies are necessary to verify this finding. Investigators observed no improvement in outcomes in patients with severe aEEG changes.

Miscellaneous

Medicolegal Pitfalls

• One of the main medicolegal implications of a diagnosis of HIE-NE is the misperception that poor obstetric management has resulted in the HIE. Although obstetric complications are poor predictors of long-term neonatal outcome, the diagnosis of HIE-NE may lead to claims of medical malpractice.
• Neonatologists, pediatricians, pediatric neurologists, and other health professionals always should refrain from using the statement "hypoxic-ischemic encephalopathy due to intrapartum asphyxia or intrapartum ischemia" unless unmistakable, documented proof is present. Such proof is rarely present. The mere presence of the clinical syndrome in the neonate does not prove anything because it may have multiple etiologies, as already noted (see Differentials).
• Motor and cognitive deficits on follow-up examination also do not prove anything because most cases of CP and/or mental retardation do not have a clear-cut, established etiology. Only situations such as clinically significant maternal hypotension during delivery or severe dystocia followed by severe neonatal depression qualify as evidence that intrapartum factors may have contributed to the abnormal neurologic outcome.
• Nonetheless, data from current epidemiologic studies indicate that most fetuses who undergo true hypoxic-ischemic events die or survive with minimal morbidity. Patients who survive with sequelae are in the minority.

References

1. Aida N, Nishimura G, Hachiya Y, et al. MR imaging of perinatal brain damage: comparison of clinical outcome with initial and follow-up MR findings. AJNR Am J Neuroradiol. Nov-Dec 1998;19(10):1909-21. [Medline].

2. Albani M, Wernicke I. Oral phenytoin in infancy: dose requirement, absorption, and elimination. Pediatr Pharmacol (New York). 1983;3(3-4):229-36. [Medline].

3. Amiel-Tison C, Ellison P. Birth asphyxia in the fullterm newborn: early assessment and outcome. Dev Med Child Neurol. Oct 1986;28(5):671-82. [Medline].

4. Amiel-Tison C, Grenier A. Neurologic Evaluation of the Newborn and the Infant. New York, NY: Masson Publishing;1983.

5. Ashwal S. Brain death in the newborn. Current perspectives. Clin Perinatol. Dec 1997;24(4):859-82. [Medline].

6. Azzopardi D, Wyatt JS, Cady EB, et al. Prognosis of newborn infants with hypoxic-ischemic brain injury assessed by phosphorus magnetic resonance spectroscopy. Pediatr Res. May 1989;25(5):445-51. [Medline].

7. Barkovich AJ, Westmark K, Partridge C, et al. Perinatal asphyxia: MR findings in the first 10 days. AJNR Am J Neuroradiol. Mar 1995;16(3):427-38. [Medline].

8. Barkovich AJ, Sargent SK. Profound asphyxia in the premature infant: imaging findings. AJNR Am J Neuroradiol. Oct 1995;16(9):1837-46. [Medline].

9. Barkovich AJ, Hajnal BL, Vigneron D, et al. Prediction of neuromotor outcome in perinatal asphyxia: evaluation of MR scoring systems. AJNR Am J Neuroradiol. Jan 1998;19(1):143-9. [Medline].

10. Barkovich AJ, Westmark KD, Bedi HS, et al. Proton spectroscopy and diffusion imaging on the first day of life after perinatal asphyxia: preliminary report. AJNR Am J Neuroradiol. Oct 2001;22(9):1786-94. [Medline].

11. Battin MR, Dezoete JA, Gunn TR, et al. Neurodevelopmental outcome of infants treated with head cooling and mild hypothermia after perinatal asphyxia. Pediatrics. Mar 2001;107(3):480-4. [Medline].

12. Bloom RS, Cropley C, Neonatal Resuscitation Steering Committee. The American Heart Association and American Academy of Pediatrics Textbook of Neonatal Resuscitation. Elk Grove, VA: American Academy of Pediatrics;1994.

13. Bourgeois BF, Dodson WE. Phenytoin elimination in newborns. Neurology. Feb 1983;33(2):173-8. [Medline].

14. Boyle RJ, Kattwinkel J. Ethical issues surrounding resuscitation. Clin Perinatol. Sep 1999;26(3):779-92. [Medline].

15. Bydder GM, Rutherford MA. Diffusion-weighted imaging of the brain in neonates and infants. Magn Reson Imaging Clin N Am. Feb 2001;9(1):83-98, viii. [Medline].

16. Clancy RR, Legido A. Postnatal epilepsy after EEG-confirmed neonatal seizures. Epilepsia. Jan-Feb 1991;32(1):69-76. [Medline].

17. Cowan FM, Pennock JM, Hanrahan JD, et al. Early detection of cerebral infarction and hypoxic ischemic encephalopathy in neonates using diffusion-weighted magnetic resonance imaging. Neuropediatrics. Aug 1994;25(4):172-5. [Medline].

18. Davis DJ. Neonatal subgaleal hemorrhage: diagnosis and management. CMAJ. May 15 2001;164(10):1452-3. [Medline].

19. De Praeter C, Vanhaesebrouck P, Govaert P, et al. Creatine kinase isoenzyme BB concentrations in the cerebrospinal fluid of newborns: relationship to short-term outcome. Pediatrics. Dec 1991;88(6):1204-10. [Medline].

20. Dodson WE. Phenytoin elimination in childhood: effect of concentration-dependent kinetics. Neurology. Feb 1980;30(2):196-9. [Medline].

21. Eken P, Toet MC, Groenendaal F, de Vries LS. Predictive value of early neuroimaging, pulsed Doppler and neurophysiology in full term infants with hypoxic-ischaemic encephalopathy. Arch Dis Child Fetal Neonatal Ed. Sep 1995;73(2):F75-80. [Medline].

22. Ekert P, Perlman M, Steinlin M, Hao Y. Predicting the outcome of postasphyxial hypoxic-ischemic encephalopathy within 4 hours of birth. J Pediatr. Oct 1997;131(4):613-7. [Medline].

23. Ellenberg JH, Nelson KB. Cluster of perinatal events identifying infants at high risk for death or disability. J Pediatr. Sep 1988;113(3):546-52. [Medline].

24. Flodmark O, Becker LE, Harwood-Nash DC, et al. Correlation between computed tomography and autopsy in premature and full-term neonates that have suffered perinatal asphyxia. Radiology. Oct 1980;137(1 Pt 1):93-103. [Medline].

25. Gluckman PD, Wyatt JS, Azzopardi D, et al. Selective head cooling with mild systemic hypothermia after neonatal encephalopathy: multicentre randomised trial. Lancet. Feb 19-25 2005;365(9460):663-70. [Medline].

26. Goodman SI. Inherited metabolic disease in the newborn: approach to diagnosis and treatment. Adv Pediatr. 1986;33:197-223. [Medline].

27. Gunn AJ, Gluckman PD, Gunn TR. Selective head cooling in newborn infants after perinatal asphyxia: a safety study. Pediatrics. Oct 1998;102(4 Pt 1):885-92. [Medline].

28. Hammerman C, Kaplan M. Ischemia and reperfusion injury. The ultimate pathophysiologic paradox. Clin Perinatol. Sep 1998;25(3):757-77. [Medline].

29. Hanrahan JD, Cox IJ, Azzopardi D, et al. Relation between proton magnetic resonance spectroscopy within 18 hours of birth asphyxia and neurodevelopment at 1 year of age. Dev Med Child Neurol. Feb 1999;41(2):76-82. [Medline].

30. Hillman LS, Hillman RE, Dodson WE. Diagnosis, treatment, and follow-up of neonatal mepivacaine intoxication secondary to paracervical and pudendal blocks during labor. J Pediatr. Sep 1979;95(3):472-7. [Medline].

31. Huddleston JF. Intrapartum fetal assessment. A review. Clin Perinatol. Sep 1999;26(3):549-68, v. [Medline].

32. Johnson AJ, Lee BC, Lin W. Echoplanar diffusion-weighted imaging in neonates and infants with suspected hypoxic-ischemic injury: correlation with patient outcome. AJR Am J Roentgenol. Jan 1999;172(1):219-26. [Medline].

33. Kendall G, Peebles D. Acute fetal hypoxia: the modulating effect of infection. Early Hum Dev. Jan 2005;81(1):27-34. [Medline].

34. Khong PL, Lam BC, Tung HK, et al. MRI of neonatal encephalopathy. Clin Radiol. Nov 2003;58(11):833-44. [Medline].

35. Klesh KW, Murphy TF, Scher MS, et al. Cerebral infarction in persistent pulmonary hypertension of the newborn. Am J Dis Child. Aug 1987;141(8):852-7. [Medline].

36. Klipstein CA, McBride MC. Predictors of Cerebral Palsy in Perinatal Hypoxic Ischemic Encephalopathy. Ann Neurol. 1992;32:478 (abstract).

37. Leviton A, Nelson KB. Problems with definitions and classifications of newborn encephalopathy. Pediatr Neurol. Mar-Apr 1992;8(2):85-90. [Medline].

38. Levy S. Somatosensory evoked potentials in pediatrics. In: Chiappa K, ed. Evoked Potentials in Clinical Medicine. New York, NY: Lippincott-Raven;. 1997: 453-69.

39. Lipp-Zwahlen AE, Deonna T, Micheli JL, et al. Prognostic value of neonatal CT scans in asphyxiated term babies: low density score compared with neonatal neurological signs. Neuropediatrics. Nov 1985;16(4):209-17. [Medline].

40. Lipper EG, Voorhies TM, Ross G, et al. Early predictors of one-year outcome for infants asphyxiated at birth. Dev Med Child Neurol. Jun 1986;28(3):303-9. [Medline].

41. Majnemer A, Mazer B. Neurologic evaluation of the newborn infant: definition and psychometric properties. Dev Med Child Neurol. Oct 1998;40(10):708-15. [Medline].

42. Majnemer A, Rosenblatt B, Riley PS. Prognostic significance of multimodality evoked response testing in high-risk newborns. Pediatr Neurol. Nov-Dec 1990;6(6):367-74. [Medline].

43. Maytal J, Novak GP, King KC. Lorazepam in the treatment of refractory neonatal seizures. J Child Neurol. Oct 1991;6(4):319-23. [Medline].

44. Mercuri E, Barnett AL. Neonatal brain MRI and motor outcome at school age in children with neonatal encephalopathy: a review of personal experience. Neural Plast. 2003;10(1-2):51-7. [Medline].

45. [Best Evidence] Miller SP, Ramaswamy V, Michelson D, et al. Patterns of brain injury in term neonatal encephalopathy. J Pediatr. Apr 2005;146(4):453-60. [Medline].

46. Mizrahi EM, Kellaway P. Characterization and classification of neonatal seizures. Neurology. Dec 1987;37(12):1837-44. [Medline].

47. Myers RE. Two patterns of perinatal brain damage and their conditions of occurrence. Am J Obstet Gynecol. Jan 15 1972;112(2):246-76. [Medline].

48. Natsume J, Watanabe K, Kuno K, et al. Clinical, neurophysiologic, and neuropathological features of an infant with brain damage of total asphyxia type (Myers). Pediatr Neurol. Jul 1995;13(1):61-4. [Medline].

49. Nelson KB, Ellenberg JH. Neonatal signs as predictors of cerebral palsy. Pediatrics. Aug 1979;64(2):225-32. [Medline].

50. Nelson KB, Ellenberg JH. Apgar scores as predictors of chronic neurologic disability. Pediatrics. Jul 1981;68(1):36-44. [Medline].

51. Nelson KB, Ellenberg JH. Obstetric complications as risk factors for cerebral palsy or seizure disorders. JAMA. Apr 13 1984;251(14):1843-8. [Medline].

52. Nelson KB, Ellenberg JH. The asymptomatic newborn and risk of cerebral palsy. Am J Dis Child. Dec 1987;141(12):1333-5. [Medline].

53. Nelson KB, Leviton A. How much of neonatal encephalopathy is due to birth asphyxia?. Am J Dis Child. Nov 1991;145(11):1325-31. [Medline].

54. Nelson KB, Emery ES 3rd. Birth asphyxia and the neonatal brain: what do we know and when do we know it?. Clin Perinatol. Jun 1993;20(2):327-44. [Medline].

55. Painter MJ, Gaus LM. Neonatal seizures: diagnosis and treatment. J Child Neurol. Apr 1991;6(2):101-8. [Medline].

56. Painter MJ, Pippenger C, Wasterlain C, et al. Phenobarbital and phenytoin in neonatal seizures: metabolism and tissue distribution. Neurology. Sep 1981;31(9):1107-12. [Medline].

57. Painter MJ, Pippenger C, MacDonald H, Pitlick W. Phenobarbital and diphenylhydantoin levels in neonates with seizures. J Pediatr. Feb 1978;92(2):315-9. [Medline].

58. Painter MJ, Minnig B, Mollica L. Binding profile of anticonvulsants in neonates with seizures [abstract]. Neurology. 1997;22:413.

59. Pasternak JF, Predey TA, Mikhael MA. Neonatal asphyxia: vulnerability of basal ganglia, thalamus, and brainstem. Pediatr Neurol. Mar-Apr 1991;7(2):147-9. [Medline].

60. Perlman JM, Tack ED. Renal injury in the asphyxiated newborn infant: relationship to neurologic outcome. J Pediatr. Nov 1988;113(5):875-9. [Medline].

61. Piazza AJ. Postasphyxial management of the newborn. Clin Perinatol. Sep 1999;26(3):749-65, ix. [Medline].

62. Pourcyrous M. Cerebral hemodynamic measurements in acute versus chronic asphyxia. Clin Perinatol. Dec 1999;26(4):811-28. [Medline].

63. Rivkin MJ. Hypoxic-ischemic brain injury in the term newborn. Neuropathology, clinical aspects, and neuroimaging. Clin Perinatol. Sep 1997;24(3):607-25. [Medline].

64. Robertson C, Finer N. Term infants with hypoxic-ischemic encephalopathy: outcome at 3.5 years. Dev Med Child Neurol. Aug 1985;27(4):473-84. [Medline].

65. Robertson CM, Finer NN. Long-term follow-up of term neonates with perinatal asphyxia. Clin Perinatol. Jun 1993;20(2):483-500. [Medline].

66. Robertson CM, Finer NN, Grace MG. School performance of survivors of neonatal encephalopathy associated with birth asphyxia at term. J Pediatr. May 1989;114(5):753-60. [Medline].

67. Robertson RL, Ben-Sira L, Barnes PD, et al. MR line-scan diffusion-weighted imaging of term neonates with perinatal brain ischemia. AJNR Am J Neuroradiol. Oct 1999;20(9):1658-70. [Medline].

68. Rose AL, Lombroso CT. A study of clinical, pathological, and electroencephalographic features in 137 full-term babies with a long-term follow-up. Pediatrics. Mar 1970;45(3):404-25. [Medline].

69. Rothman SM, Olney JW. Glutamate and the pathophysiology of hypoxic--ischemic brain damage. Ann Neurol. Feb 1986;19(2):105-11. [Medline].

70. Ruth V, Virkola K, Paetau R, Raivio KO. Early high-dose phenobarbital treatment for prevention of hypoxic-ischemic brain damage in very low birth weight infants. J Pediatr. Jan 1988;112(1):81-6. [Medline].

71. Rutherford M, Ward P, Allsop J, et al. Magnetic resonance imaging in neonatal encephalopathy. Early Hum Dev. Jan 2005;81(1):13-25. [Medline].

72. Sarnat HB, Sarnat MS. Neonatal encephalopathy following fetal distress. A clinical and electroencephalographic study. Arch Neurol. Oct 1976;33(10):696-705. [Medline].

73. Saugstad OD, Rootwelt T, Aalen O. Resuscitation of asphyxiated newborn infants with room air or oxygen: an international controlled trial: the Resair 2 study. Pediatrics. Jul 1998;102(1):e1. [Medline].

74. Scher MS. Seizures in the newborn infant. Diagnosis, treatment, and outcome. Clin Perinatol. Dec 1997;24(4):735-72. [Medline].

75. Shankaran S, Kottamasu SR, Kuhns L. Brain sonography, computed tomography, and single-photon emission computed tomography in term neonates with perinatal asphyxia. Clin Perinatol. Jun 1993;20(2):379-94. [Medline].

76. Sie LT, van der Knaap MS, Oosting J, et al. MR patterns of hypoxic-ischemic brain damage after prenatal, perinatal or postnatal asphyxia. Neuropediatrics. Jun 2000;31(3):128-36. [Medline].

77. Simon NP. Long-term neurodevelopmental outcome of asphyxiated newborns. Clin Perinatol. Sep 1999;26(3):767-78. [Medline].

78. Steinberg A. Decision-making and the role of surrogacy in withdrawal or withholding of therapy in neonates. Clin Perinatol. Sep 1998;25(3):779-90, xii. [Medline].

79. Taylor MJ, Murphy WJ, Whyte HE. Prognostic reliability of somatosensory and visual evoked potentials of asphyxiated term infants. Dev Med Child Neurol. Jun 1992;34(6):507-15. [Medline].

80. Thornberg E, Thiringer K, Odeback A, Milsom I. Birth asphyxia: incidence, clinical course and outcome in a Swedish population. Acta Paediatr. Aug 1995;84(8):927-32. [Medline].

81. Triulzi F, Baldoli C, Parazzini C. Neonatal MR imaging. Magn Reson Imaging Clin N Am. Feb 2001;9(1):57-82, viii. [Medline].

82. Van Orman CB, Darwish HZ. Efficacy of phenobarbital in neonatal seizures. Can J Neurol Sci. May 1985;12(2):95-9. [Medline].

83. Vannucci RC. Experimental biology of cerebral hypoxia-ischemia: relation to perinatal brain damage. Pediatr Res. Apr 1990;27(4 Pt 1):317-26. [Medline].

84. Volpe JJ, Pasternak JF. Parasagittal cerebral injury in neonatal hypoxic-ischemic encephalopathy: clinical and neuroradiologic features. J Pediatr. Sep 1977;91(3):472-6. [Medline].

85. Volpe JJ. The neurological examination: Normal and abnormal features. In: Volpe JJ, ed. Neurology of the Newborn. 3rd ed. Philadelphia, Pa: WB Saunders;. 1995: 95-124.

86. Volpe JJ. Hypoxic-Ischemic Encephalopathy: Neuropathology and Pathogenesis. In: Volpe JJ, ed. Neurology of the Newborn. 3rd ed. Philadelphia, Pa: WB Saunders;. 1995: 279-313.

87. Volpe JJ. Hypoxic-Ischemic Encephalopathy: Clinical Aspects. In: Volpe JJ, ed. Neurology of the Newborn. 3rd ed. Philadelphia: WB Saunders;1995:314-369.

88. Williams CE, Mallard C, Tan W, Gluckman PD. Pathophysiology of perinatal asphyxia. Clin Perinatol. Jun 1993;20(2):305-25. [Medline].

89. Zarifi MK, Astrakas LG, Poussaint TY, et al. Prediction of adverse outcome with cerebral lactate level and apparent diffusion coefficient in infants with perinatal asphyxia. Radiology. Dec 2002;225(3):859-70. [Medline].

90. al Naqeeb N, Edwards AD, Cowan FM, Azzopardi D. Assessment of neonatal encephalopathy by amplitude-integrated electroencephalography. Pediatrics. Jun 1999;103(6 Pt 1):1263-71. [Medline].

91. de Vries LS. Somatosensory-evoked potentials in term neonates with postasphyxial encephalopathy. Clin Perinatol. Jun 1993;20(2):463-82. [Medline].

92. du Plessis AJ, Johnston MV. Hypoxic-ischemic brain injury in the newborn. Cellular mechanisms and potential strategies for neuroprotection. Clin Perinatol. Sep 1997;24(3):627-54. [Medline].

Keywords

HIE, ischemia/hypoxemia, hypoxemia/ischemia, perinatal asphyxia, newborn encephalopathy, neonatal encephalopathy, NE, HIE-NE, excitotoxicity, periventricular leukomalacia, cerebral ischemia, cerebral hypoxia, birth asphyxiation, hypoxic-ischemic brain injury in the newborn

Contributor Information and Disclosures

Author

Marcio Sotero de Menezes, MD, Associate Professor, Department of Neurology, Division of Pediatric Neurology, Children's Hospital of Seattle, University of Washington
Marcio Sotero de Menezes, MD is a member of the following medical societies: American Academy of Neurology and American Epilepsy Society
Disclosure: Nothing to disclose.

Coauthor(s)

Dennis WW Shaw, MD, Professor, Department of Radiology, Department of Radiology, University of Washington School of Medicine; Consulting Staff, Children's Hospital and Regional Medical Center of Seattle
Dennis WW Shaw, MD is a member of the following medical societies: American College of Radiology, American Roentgen Ray Society, American Society of Neuroradiology, American Society of Pediatric Neuroradiology, International Society for Magnetic Resonance in Medicine, Radiological Society of North America, Society for Pediatric Radiology, and Society of Cardiovascular and Interventional Radiology
Disclosure: Nothing to disclose.

Medical Editor

Ann M Neumeyer, MD, Clinic Director, Instructor, Departments of Neurology and Pediatrics, Massachusetts General Hospital, Harvard Medical School
Ann M Neumeyer, MD is a member of the following medical societies: American Academy of Neurology, Child Neurology Society, and Massachusetts Medical Society
Disclosure: Nothing to disclose.

Pharmacy Editor

Francisco Talavera, PharmD, PhD, Senior Pharmacy Editor, eMedicine
Disclosure: Nothing to disclose.

Managing Editor

Kenneth J Mack, MD, PhD, Senior Associate Consultant, Department of Child and Adolescent Neurology, Mayo Clinic
Kenneth J Mack, MD, PhD is a member of the following medical societies: American Academy of Neurology, Child Neurology Society, Phi Beta Kappa, and Society for Neuroscience
Disclosure: Nothing to disclose.

CME Editor

Matthew J Baker, MD, Consulting Staff, Collier Neurologic Specialists, Naples Community Hospital
Matthew J Baker, MD is a member of the following medical societies: American Academy of Neurology
Disclosure: Nothing to disclose.

Chief Editor

Amy Kao, MD, Assistant Professor, Department of Neurology, Department of Pediatrics, Division of Pediatrics, Oregon Health and Science University; Consulting Staff, Shriners Hospital
Amy Kao, MD is a member of the following medical societies: American Academy of Neurology, American Academy of Pediatrics, American Epilepsy Society, and Child Neurology Society
Disclosure: Nothing to disclose.

© 1994- by Medscape.
All Rights Reserved
(http://www.medscape.com/public/copyright)