In the past 12 years neonatal intensive care has been responsible for a dramatic decrease in neonatal mortality. Concomitantly, neonatal neurologic morbidity has also declined, as evidenced by many follow-up studies. Reports of "intact survival" of 50Og to 800g infants are beginning to emerge.1,2 The marked decrease in mortality at all birth weights, however, has resulted in an increase in the number of children who survive with varying degrees of brain damage. This has created problems for families and has placed stress on the health resources of society at large.3
This review focuses on the most common neurological problems in the newborn period, emphasizing the application of new techniques available for prognosis, diagnosis and treatment. Abnormal movements and behavioral patterns in the newborn as well as their most common alleged etiologies, intraventricular-periventricular hemorrhage (IVH-PVH) and hyppxic-isdhemic encephalopathy (HIE), will be discussed in detail.
ABNORMAL NEONATAL MOTOR AND BEHAVIORAL PATTERNS
True seizures in the neonatal period should be considered symptomatic of serious underlying neurological or systemic disease, and imply a guarded prognosis. The reported incidence of seizures varies between 3 to 15/10004 and has been claimed to be as high as 25% in premature infants.5 Even optimistic follow-up studies report a 17% mortality and 30% serious morbidity.6,7 The prognosis of affected babies depends mainly on the underlying cause of the "seizures,"6 and to an important but unknown extent on the rapid establishment of diagnosis and treatment. Therefore, the appropriate interpretation and investigation of adventitious movements that occur in the newborn is of utmost importance.
However, not all deviant motor or behavioral patterns seen in the newborn represent seizures. Hughlings Jackson defined a seizure as a clinical manifestation of an abnormal discharge of cortical neurons. Even in the premature neonate, it is very rare to have a clinical seizure without a coincident abnormal paroxysmal discharge on the electroencephalogram (EEG).4·8 There remains considerable controversy concerning which adventitious movements, postures, or behavioral patterns constitute true seizures. The term "subtle seizures," which has become common parlance in neonatal intensive care units, may at times oversimplify the situation and lead to an inappropriate designation.9 For this reason, the following patterns will be discussed in detail: 1) tonic postures; 2) abnormal eye movements; 3) respiratory patterns including apnea; and 4) other repetitive stereotyped movements (ie, buccal and lingual movements).
Tonic postures in the newborn may represent benign primitive reflexes, be an indication of brain stem damage as seen in transtentorial herniation, or be a manifestation of true seizure activity.
An example of a tonic posture of the benign variety is the tonic neck reflex initiated by changes in the position of the baby's head. Similarly, brief tonic patterns may be seen during rapid eye movement (REM) sleep.
In the premature infant less than 32 weeks gestational age (GA). common postures often associated with catastrophic intraventricular hemorrhage (IVH), are usually interpreted as seizures. In this circumstance, the infant usually exhibits dramatic downward and often convergent eye movements. At the same time, intermittent flexion or inward rotation of the arms together with extension of the legs is seen. These are decerebrate or decorticate postures, the result of transtentorial herniation. Severe apnea and bradycardia often accompany these postures. The EEG during these events shows nonspecific slowing or flattening, but does not show abnormal cortical paroxysmal activity.6,10
The tonic postures that represent true seizures are seen principally in infants older than 35 weeks gestational age. One or two extremities, usually on the same side, go into slow extension with or without simultaneous adverse phenomena (tonic deviation of head and eyes toward the extended extremities). An EEG obtained coincidentally with this activity will usually show epileptiform paroxysmal activity emanating from the contralateral frontal or temporal lobe. In the term infant following a hypoxic-ischemic insult, it may be difficult to distinguish decerebrate or decorticate postures associated with cerebral edema and intermittent herniation from true seizures. In this circumstance, the EEG is particularly helpful.4
Before concluding the discussion of tonic postures it should be mentioned that opisthotonic posturing (neck retroflexion) may be seen in infants with meningitis or subarachnoid hemorrhage presumably on the basis of meningeal irritation. Opisthotonus is also a pre-terminal event in currently rare cases of kernicterus.
Abnormal Eye Movements
Nursery personnel often fail to observe significant deviant eye movements either because the neonate's eyes are covered for protection or because the examiner does not open the eyes as part of routine observation. Again, it should be emphasized that not all deviant eye movements represent seizure phenomena. Horizontal repetitive nystagmoid movements of both eyes are frequently observed as part of clonic seizures, either unifocal or multifocal. These may accompany the clonic jerking of an extremity or the face. These conjugate jerky eye movements occur toward the same side as the extremity involved. They may also occur in isolation either preceding or following a focal clonic seizure of an extremity. After a prolonged seizure, one may observe tonic conjugate deviation of the eyes toward the side of the electrical seizure focus in the brain. This most likely represents a post-ictal phenomenon due to exhaustion of the neurons in the seizure focus. Tonic conjugate deviation of the eyes may also be seen occasionally in the context of severe brain stem damage due to hypoxia or hemorrhage. This type of tonic deviation is usually sustained and cannot be overcome by oculocephalic maneuver as is usually the case with cortical lesions described above.
Brief horizontal, or occasionally vertical, jerking eye movements, usually associated with minimal but perceptible lid activity, are frequently seen during REM or light sleep in the newborn period.4 These may be associated with periodic breathing and less commonly brief jerks of the extremities. This represents a normal physiologic phenomenon.
Intermittent conjugate downward movement of the eyes of variable duration with the sclera being seen above the eyes represents the well-known phenomenon of sunsetting. This reflects increased intracranial pressure on midbrain structures or less commonly, primary midbrain injury. A similar but distinct phenomenon of rapid downward movement of the eyes followed by a slow upward drift is termed ocular bobbing. This is usually associated with damage to the pons and can be seen in the context of severe IVH, HIE, or less commonly in metabolic disorders.9
Rapid chaotic conjugated eye movements often in association with lid movements are termed opsoclonus and have been seen rarefy in neonates. They are thought to be associated with cerebellar or brain stem disorders. A search for infection or central nervous system (CNS) hemorrhage is indicated when they are seen. Opsoclonus has been found to be associated in some cases with neuroblastoma in older children but this has not to date been observed in the neonatal period.
Abnormal respiratory patterns, particularly recurrent apnea, may be a manifestation of seizure activity in the more mature newborn. It should be remembered, however, that in light (REM) sleep, breathing is characteristically periodic and that this physiologic state constitutes 50% to 80% of the infant's day. Moreover, recurrent apneic attacks, particularly in the premature infant, usually suggest non-convulsive apnea indicative of brain stem malfunction or immaturity. This "central apnea" frequently accompanies severe respiratory distress and periventricular hemorrhage but can be associated with sepsis, hypoglycemia, or any number of serious disorders in the newborn. Bradycardia usually occurs early in this type of apneic spell, often within eight seconds of onset." There is no EEG evidence of abnormal cortical discharge. The apnea may respond to increases in PO2, theophylline, or frequent stimulation.
In more mature neonates, apnea with late bradycardia, occurring 15 or 20 seconds into the episode, may prove to be true seizure activity. When this type of apnea occurs accompanied by other overt clinical seizures, the diagnosis is not difficult. In doubtful cases, an EEG should be done. Occasionally, a trial of anticonvulsant therapy may be necessary.
Other abnormal respiratory patterns may be seen in the newborn and usually are linked to severe central injury secondary to structural or metabolic disease. These include central neurogenic hyperventilation, or irregular ataxic breathing. These represent severe brain stem injuries. These abnormal respiratory patterns do not suggest seizure activity. Occasionally, intrinsic lung disease or systemic etiologies may be responsible for abnormal respiratory patterns which superficially resemble the central patterns described above.
Other Adventitious Movements
Early in the course of an encephalopathy, usually HIE, repetitive and stereotyped buccal and lingual movements may be observed. These include opening and closing the mouth, lip smacking or chewing. When these are associated with other clinical seizure phenomena, particularly of the extremities, they may well represent additional seizure activity with EEG concomitants. l2 Later in the encephalopathy, when other forms of seizures have subsided, often some lip smacking or mouthing movements will persist. This activity is usually not accompanied by abnormal paroxysmal discharges on the EEG, but may show a non-specific slowing of the background activity. These persistent mouthing movements appear to be sequelae of the encephalopathy, perhaps implicating some frontal lobe release phenomenon.
Similar mouthing movements are also perceived in normal newborns without other suggestions of a CNS insult. These may occur immediately before vomiting and are particularly frequent in intubated babies and those with feeding tubes, almost certainly reflecting a gagging phenomenon.
Paradoxically, though many movements are inappropriately construed as seizures, the behavioral state of total or near total unresponsiveness is seldom thought of in this light. The lack of responsiveness may be the end state of a severe encephalopathy, or the result of heavy sedation needed to control seizures. In some infants, however, this state may reflect continuing underlying seizure activity, status epilepticus. The baby may lie motionless with a dull fixed stare. The EEG in these infants will usually show continuous seizure activity. In this circumstance, further anticonvulsant therapy is essential.
APPLICATIONS OF THE NEONATAL EEG
The neonatal EEG is helpful both in distinguishing true seizure activity from other adventitious motor or behavioral patterns and, in addition, provides information concerning the prognosis in neonates with seizures.
When it is not clear whether a repetitive behavior pattern actually represents a seizure, the EEG is most helpful if obtained during the suspect behavior.4 The electroencephalographer or technician should clearly mark during the recording, the timing and description of any suspect clinical activity observed. A polygraph recording with simultaneous EEG, electro-oculogram (EOG), electromyogram (EMG), electrocardiogram (EKG), and respiratory monitoring is most helpful. Split screen video techniques often applied to older children would be of immense value in the neonatal setting.
The judiciously interpreted EEG may also be helpful in determining prognosis. The proper time for an EEG is within 24 to 48 hours after the last seizure. Unlike the older child, where a longer post-ictal time lapse is recommended, in the newborn, EEG often normalizes after a few days and may lose its prognostic significance.
Five interictal EEG patterns are observed in infants 36 to 42 weeks gestation. ,3 These are: 1) normal; 2) unifocal spikes with normal background activity; 3) multifocal spikes; 4) burst suppression patterns; and 5) inactive and flat. The first two types of tracing suggest a good outcome. Multifocal spikes, especially with abnormal background activity, have only a 20% chance of a good prognosis. The burst suppression pattern, not to be confused with a normal infant deep sleep pattern (trace-alternant), almost invariably points to a poor prognosis usually resulting in a severe static encephalopathy if the patient survives. The inactive or flat EEG has the same poor prognosis although it may be hard to interpret in the face of high anticonvulsant drug levels. It is not synonymous with brain death.
TREATMENT OF NEONATAL SEIZURES
The symptomatic treatment of neonatal seizures, where an underlying treatable metabolic cause cannot be found, can be conveniently divided into the management of a single seizure or a few widely separated seizures and that of continuous seizures, "status epilepticus." There is little controversy about treating a single seizure with a loading dose of 20 mg per kilo of intravenous phenobarbital. This achieves levels in the 15 to 25 µg/ml range. Maintenance doses of 3 to 4 mg/ kilo/day are then used to maintain levels in this range for the first week.14
It is in the area of status epilepticus where controversy exists as to the therapeutic level of phenobarbital and when to add a second anticonvulsant. Phenobarbital alone, as given above, will control 36% to 70% of neonatal seizures. If unsuccessful, most authors generally recommend adding Phenytoin slowly, IV, 20 mg/kg, to reach levels of 15 to 20 (ig/ml.4 Our experience suggests that an additional 20 mg per kilo of phenobarbital , given slowly in separate doses of 10 mg per kilo can be safely administered before a second drug is tried. This will bring levels into the 35 to 60 µg/ml range, often causing sedation, but no serious cardiorespiratory problems.15 At this higher blood level, 80% to 90% of seizures will be controlled with phénobarbital alone. One is reminded that the usual therapeutic range is but a maintenance level to be used as a general guide once control of seizures is attained, but often may not be sufficient to combat the seizures of a severe encephalopathy. If one accepts the fact that ongoing seizures per se are harmful to the brain, then any level sufficient to stop convulsions, short of cardiovascular collapse, is acceptable. The consequences of hypoxicischemic encephalopathy are high and might possibly be related to adequate seizure control. Short-term decrease in responsiveness, spontaneous activity, and feeding is a small price to pay if seizures can be adequately and quickly controlled.
Short-acting Valium, which may be valuable in stopping a single seizure, should not be given in combination or in close approximation with IV phenobarbital therapy. Respiratorycardiac arrests have been frequently reported.9 In combination with phenytoin, the use of IV Valium is safe, but sustained control of seizures is difficult. Thus, its use in the neonate is not recommended.
Intracranial hemorrhage appears to have replaced asphyxia as the most common identifiable form of neonatal cerebral insult. New instrumentation, particularly fetal heart monitoring, has led to improved obstetrical care with the delivery of fewer asphyxiated infants. Improved respiratory care in the nursery has decreased post-partum asphyxia and led to the survival of very small prematures. At the same time, CAT scan and ultrasound have permitted the classification and identification of subtle and silent hemorrhages, particularly in these immature survivors. These devices have also allowed the clinician to follow the evolution of these lesions.16
In the newborn, certainly an infant's gestational age is the best statistical indicator of the probable site of bleeding. Thus supratentorial subdural bleeding occurs almost exclusively in large mature babies with difficult deliveries. On the other hand, parenchymal supratentorial cerebral bleeding, originating from the periventricular area, is very common in infants 34 weeks gestation or less, while it is quite unusual in infants more mature than 36 weeks gestation. A clinical presentation of primary subarachnoid bleeding of venous origin is commonly recognized in full-term newborns presenting with focal seizures and a benign clinical course.6 Neither the incidence nor the symptomatology of primary or secondary subarachnoid bleeding is established in prematures. Posterior fossa hemorrhages are often terminal events in prematures, but in term infants may be compatible with infant survival.17 Surgical evacuation of the hematoma is indicated in some cases.18
Periventricular-intraventricular hemorrhage (PVH-IVH), so renamed to denote the actual site of bleeding, occurs in as high as 40% to 50% of infants weighing less than 1500 g. Most infants are less than 35 weeks gestational age. The frequency increases with decreasing gestational age.19 Only this type of hemorrhage is discussed in detail.
Pathology and Pathophysiology
In most cases, the site of origin of periventricular hemorrhage in prematures is the germinal matrix between the head of the caudate and the thalamus near the foramen of Monro. This area contains thin-walled vessels. A periventricular hemorrhage may be confined to this friable matrix area, Type I bleed in the terminology of Papile. Approximately 60% to 80% of the time, the bleeding extends into the lateral ventricles (Type II), expanding them (Type HI), and extending into the cerebral parenchyma, (Type IV).20 It is not established whether the initial bleed is arterial, principally from the recurrent artery of Huebner, off the anterior cerebral artery, or venous from the terminal vein which runs along the lateral border of the ventricle entering the internal cerebral vein at a very sharp angle. Proponents of the arterial theory cite hypotension followed by hypertension in the asphyxiated brain which has lost its autoregulation (debatable) as the cause of the bleed into this vulnerable site. Those favoring a venous hypothesis point to pathologic evidence that suggests that stasis, thrombosis, rupture, and bleed occur in venules secondary to increased venous pressure.21 Whatever the mechanism, a temporal association with low gestation, hyaline membrane disease, and pneumothorax is wellestablished. Hypercapnia, acidosis, hypo- or hypertension and increased venous pressure can all be encountered in severe respiratory distress.22 It is still not clear which factor constitutes the main cause, or whether they all act in concert on a susceptible area. Bleeding usually occurs in the first 48 hours after birth, but it has been reported in utero as well as after several days of life.
Signs and Symptoms
Type I hemorrhage is usually asymptomatic or presents with signs and symptoms so subtle that they are usually missed. Some recent studies suggest that these babies exhibit difficulty in tracking and fixation, but most observers cannot be sure these functions are demonstrable in normal pre-term infants. When the hemorrhage breaks into the ventricles, Type II, non-specific symptoms and signs such as lethargy and retrocolis may be present. In Type III and Type IV bleeds, one of three clinical courses is likely to ensue. A dramatic deterioration may occur soon after the bleed associated with a 50% mortality. The infant develops severe apnea and bradycardia. They exhibit decerebrate and decorticate postures described under abnormal movements. Then pupils become fixed and the infants become flaccid and unresponsive, dying within minutes to hours. A less dramatic deterioration may occur over a few days to a week with only some of the above symptoms observed. Lastly, a rather severe ventricular dilatation can occur without any symptomatology or change in head size. This is designated as silent hydrocephalus.23
The laboratory features that point to the more severe types of PVH-IVH are a drop in hematocrit of at least 10%, hypercapnia, acidosis, hypocalcemia, hypo- or hyperglycemia, and changes in the cerebrospinal fluid. A lumbar puncture (LP) performed early after a bleed should reveal many thousand red blood cells (RBCs) and a cerebrospinal fluid (CSF) protein in the 250 to 1500 mg range. There is some statistical correlation with elevation of the protein and severity of bleed. The LP is not always conclusive for two reasons. First, the incidence of traumatic taps is common in small pre-term infants and RBCs do not necessarily clear in successive samples of CSF. Second, a small number of infants may exhibit normal CSF at the time of the first LP, possibly because the blood has not yet descended to the lumbar level or because of an infrequent aqueductal stenosis caused by the bleed. A few days after the original intraventricular bleeding, the fluid becomes xanthochromic and may exhibit high white blood cell (WBC) counts and hypoglycorrhachia in addition to RBCs and elevated protein. The findings are suggestive of meningitis which is excluded by negative cultures.
Computed tomography was the original cornerstone for the diagnosis and classification of intracerebral hemorrhage. In all but the smallest hemorrhages, it visualizes the sites of parenchymal blood, the presence of blood in the ventricles and subarachnoid space, ventricular size and shift of major structures. However, in high-risk infants most prone to hemorrhage, transport to a CT scanner can be hazardous. Thus, portable ultrasonography has become a very important tool in both the diagnosis and management of intracranial hemorrhage. Small periventricular bleeds can be visualized as well as intraventricular blood. The ventricular size can be noted and followed without apparent risk to the infant. The technique is not as suitable for subarachnoid or subdural blood as a CT scan, but early periventricular Ieukomalacia, when it is sizable, can sometimes be noted by this method.
Prognosis and Therapy
In Type I and Type II hemorrhages, prognosis appears quite good, with 80% to 90% intact survival. The morbidity in these groups would appear to correlate with ischemic damage presumably due to the degree of birth asphyxia and respiratory complications. In Type III and Type IV, mortality reaches a 50% to 75% level and subsequent morbidity is high. Although most studies have lumped grade III and grade IV hemorrhages, this may be inappropriate. A significant percentage of infants with ventricular hemorrhage resolve their hydrocephalus spontaneously, while others develop progressive hydrocephalus. A significant percentage will die from parenchymal extensions of their lesions from complications of their concurrent cardiopulmonary disease.
Treatment is clearly not necessary for Type I and Type II hemorrhages. The treatment for progressive hydrocephalus has included serial lumbar taps, external ventricular drainage, ventriculo-peritoneal shunting, and medical treatment designed to decrease CSF production or increase diuresis. The results of frequent lumbar punctures are equivocal. They do not appear to affect the course of a severe bleed when attempted during the first three weeks of therapy.23 Later, they may be useful in preventing progressive hydrocephalus by giving the process a chance to arrest spontaneously. This may postpone or avoid the need for ventriculo-peritoneal shunt. The lumbar punctures are best performed while measuring pressures with external devices applied to the fontanelle or with a CSF manometer attached to the LP needle. Enough fluid should be withdrawn to bring pressure below 80 to 100 cm of water. A very helpful refinement is to perform ultrasonography before and after the tap. There should be a reduction in ventricular size together with a pressure drop if the taps are to be effective. The failure to obtain enough fluid suggests aqueductal stenosis when the procedure is properly performed. These measures have resulted in a decrease in shunting procedures. When serial LPs do not seem to work, a 5 to 14-day trial of external ventricular drainage is recommended. Beyond five days, the chances of infection and ventriculitis become high.24 This procedure may arrest an ongoing process, again, avoiding definitive neurosurgical procedures. The use of acetazolamide in high doses has allowed one group to postpone and sometimes avoid shunting.25 The placement of ventriculo-peritoneal shunts carries a higher complication risk with smaller infants, and should be postponed until it is clearly inevitable.
In spite of decreasing incidence, asphyxia neonatorum and its clinical correlate, hypoxic-ischemic encephalopathy, is still among the leading causes of static encephalopathy. Although anoxic brain injury occurs with equal frequency in pre-term and full-term babies, overt clinical patterns are seen mainly in infants above 36 weeks GA.
Pathology and Pathophysiology
Anoxic-ischemic damage usually occurs preceding or during labor.26 Postnatal hypoxic injury can be superimposed when there is cardiopulmonary dysfunction. Brain pathology depends on the maturity of the brain at the time of insult as well as the mechanism, timing, and duration of the hypoxia or ischemia.
Pre-term infants develop periventricular Ieukomalacia, punctate to large islands of ischemia in the white matter.9,27 Their preferential location in the frontal and occipital regions probably accounts for the increased incidence of spastic diplegia and eye problems in survivors.
Newborns greater than 36 weeks GA with partial subacute asphyxia evidence mainly grey matter supratentorial lesions. The pathological picture varies from generalized cerebral edema and hemorrhage to neuronal necrosis in the depths of the cerebral sulci. The lesions may have a parasagittal watershed distribution (the area where lateral and medial cerebral circulations overlap).28 This occurs most frequently with fetal hypotension. These subacute variants of asphyxia are the ones most often clinically recognized in the neonatal period.
Signs and Symptoms
Asphyxia occurring in the neonatal period results in Apgar scores of less than 6 at one and five minutes. The clinical picture usually translates chemically into hypercapnia and hypoxia with acidosis. Clinical patterns in infants less than 34 weeks GA differs from that in more mature infants. The findings are very non-specific. Unless accompanied by grade III or grade IV PVH-IVH, there may be no symptoms except for poor arousal and frequent apnea and bradycardia.
The three most common clinical pictures of hypoxicischemic encephalopathy29,30 in infants greater than 36 weeks GA are: a) minimal lethargy and hypotonia; b) hyperalertness often associated with hypertonia; and c) depression and severe hypotonia. All three may have similar birth histories with evidence of fetal distress by fetal monitoring and meconium staining.
In the first category, the history of asphyxia may be as severe as in the latter two types, but for unknown reasons, these infants exhibit at most minimally reduced arousal and tone.
The second group appears hyperalert with normal or increased muscle tone. They are often tremulous with a low threshold Moro response. They exhibit tachycardia and dilated pupils, as if there were sympathetic overflow. Many of these infants become normal within a few hours. Some, however, progress to obtundation and seizures as in the next group.
The third presentation of HIE, the most ominous, consists of those infants depressed from the beginning, very hypotonic with minimal spontaneous movements, depressed or absent Moro and suck reflexes, and hypoactive deep tendon reflexes. In some, proximal hip and shoulder weakness predominates (watershed lesions). Abnormal movements and behavior, often seizures in this setting, begin between 6 and 24 hours. Their appropriate recognition has been discussed.
The clinical course of these infants is variable. Some show gradual improvement from the beginning, others will have a few clinical seizures before they begin to improve. Those infants with severe encephalopathy will become more stuporous over the first 24 hours. They develop status epilepticus and begin to show respiratory abnormalities. They exhibit brain stem dysfunction along with intermittent decerebration. Their fontanelle may bulge due to cerebral edema. Without vigorous supportive and anticonvulsant therapy 20% to 30% will die. If the infants survive the first 48 to 72 hours, seizures will gradually abate.
The patients will remain hypotonic and stuporous for a variable period. Recovery proceeds at a variable pace, poor suck, slow feeding and head lag being particularly notable in this period.
The most useful test performed in perinatal asphyxia is one actually performed before birth. Fetal heart-tone records indicating late and variable decelerations, as well as a fixed intrauterine heartbeat are strong evidence of placental insufficiency and fetal distress which lead to postnatal anoxic encephalopathy. There is a good correlation between fetal heart rate (FHR) abnormalities and prognosis.31
After birth, serum chemistries will show low p02 high p COt and low pH. Serum sodium may be decreased suggesting inappropriate antidiuretic hormone secretion secondary to severe encephalopathy. Lumbar puncture will usually be normal in terms of cells and protein. However, serum CPK brain isoenzymes have been correlated with severity of asphyxia and outcome.
Among electrical studies, the appropriate use and value of EEG has been discussed. Visual evoked responses, and more recently brain stem auditory evoked responses, when abnormal, indicate a guarded prognosis.32,33
Radioisotope studies have shown an increased uptake in some cases in the parasagittal watershed areas of the cortices which have correlated with the findings of hip and shoulder weakness in some newborns. Recent CAT findings have been of two varieties. Some may show areas of infarction as do the isotope findings. The long-term significance is not yet clear, but is probably not encouraging. Other babies show extensive hypodense areas of both grey and white matter. The infants with these latter findings have done very poorly at follow-up.
Treatment and Prognosis
Measures which include respiratory support, maintenance of normal blood pressure and normal renal output are to be emphasized while specific treatment for seizures is given.
The status of anti-edema agents is still not known. Steroids have at best given equivocal results, probably because they are usually ineffective in conditions causing generalized cytotoxic edema. Diuretics and osmotic agents have not been adequately evaluated in controlled studies.
The prognosis of HIE is always guarded and relates mainly to clinical course. The more rapid the rate of recovery from initial depression, the better the outlook. Those who develop seizures have a less favorable prognosis. Some early predictive statements can be cautiously made by the judicial interpretation of the aforementioned laboratory studies in conjunction with the clinical course.28-30.35
1. Britten SB. Rtzhardinge PM. Ashley S: Is intensive care justified for infants weighing less than 801 gms at birth? J Pediutr 1981: 99:937.
2. Driscoll JM. Driscoll YT. Steir ME. el al: Mortality and morbidity in infants less than 1001 grams at birth. Pediatrics 1982: 69.21.
3. Ross G: Mortality and morbidity in very low birthweight infants. PediatrAnn 1983; 12:32.
4. Bergman I. Painter MJ. Crumrine PK: Neonatal seizures. Semiti Permutai 1982; 6:54.
5. Seay AR. Bray PF: Significance of seizures in infants weighing less than 2500 grams. Arch Neurol 1977: 34:381.
6. Hill A. Volpe JJ: Seizures, hypoxic-ischemic brain injury, and intraventricular hemorrhage in the newborn. Ann Neurol 1981; 10:109.
7. Dennis J: Neonatal convulsions aetiology, late neonatal status, and long-term outcome. Dev Med Child Neurol 1978: 29:143.
8. Dreyfus-Brisac C. Monod N: Electroclinical studies, in Kelloway P. Petersen I (eds): Neurological and Electroencephalographs Correlative Studies in lnfiincy. New York. Grune & Stratton. 1964. ? 273.
9. Volpe JJ: Neurology tfthe Newborn. Philadelphia. WB Saunders Co. 1981.
10. Engel R: Abnormal electroencephalograms in the neonatal period. Thomas CC (ed): Springfield. 111. 1975.
11. Fenichel GM. Olson BJ. Fitzpatrick HJE: Heart rate changes in convulsive and non-convulsive neonatal apnea. Ann Neurol 1980: 2:577.
12. Volpe. JJ: Neonatal seizures. Clin Pennato! 1972; 4:43.
13. Rose AL. Lombroso C: Neonatal seizure states: A study of clinical . pathological . and electroenccphalographic features in 137 full-term babies with long-term follow up Pediatrics 1970: 45:404.
14. Painter MJ. Pippenger C. Macdonald H. et al: Phénobarbital and phcnytoin levels in neonates with seizures. J Pediatr 1978; 92:315.
15. Gal P. Pharm D. Toback J. et al: Efficacy of phénobarbital monotherapy of neonatal seizures - Relationship to blood levels. Neurology (NY) 1982: 32:140.
16. Pape KE. Bennett-Britton S. Szymonowicz W. et al: Diagnostic accuracy of neonatal brain imaging: A postmortem correlation of computed tomography and ultrasound scans. J Pediatr 1983: 102:275.
17. RomS.SerfonteinGl. Humphreys RP: Intracerebellar hematoma in the neonate. J Pediatr Í978; 93:486.
18. Fishman MA. Percy AK. Cheek WR. et al: Successful conservative management of cerebellar hematomas in preterm infants. J Pediatr 1981: 98:466.
19. Ahman PA. Lazarra A. Dykes F. et al: Intraventricular hemorrhage in the high risk pretenu-infant: Incidence and outcome. Ann Neurol 1980: 7:118.
20. PapileLA. BursteinJ. Burstein R.etal: Incidence and evolution of subependymal and intraventricular hemorrhage: A study of infants with birthweight less than 1500 grams. J Pediatr 1978: 92:529.
21. Pape KE. Wigglesworth JS: Hemorrhage. Ischemia, and the Perinatal Brain. Philadelphia. JB Lippincott Co. 1979.
22. Hcllman J. Vannucci RC: Intraventricular hemorrhage in premature infants. Semin Perinatal 1982: 6:42.
23. Tasby TY. Volpe JJ: Intraventricular hemorrhage in the premature infant. Pediatr Clin North Am 1982: 29:1077.
24. Harbaurger E. Saunders RL. Edwards EH: External ventricular drainage for control of posthemorrhagic hydrocephalus in premature infants. J Neuwsurg 1981; 55:766.
25. Shinar S. Molteni RA. Gammon K. et al: Intraventricular hemorrhage in the premature infant: A changing outlook. N Engl J Med 1982: 306:1464.
26. Brown JK. Purvis RJ. Forfar JO. et al: Neurologic aspects of perinatal asphyxia. Dev Med Child Neurol 1974: 16:567.
27. Banker BG. Larroche JC: Periventricular Ieukomalacia in infancy. Arch Neurol 1962: 7:386.
28. Hill A. Volpe JJ: Hypoxic-ischemic brain injury in the newborn. Semin Perinatal 1982: 6:25.
29. Sarnat HS. Sarnat MS: Neonatal encephalopathy following fetal distress. Arch Neurol 1976: 33:696.
30. Koenigsberger MR: Acute encephalopathies of infancy, in Rudolph AM (ed): Pediatrics, ed 17. Norwalk. Appleton-Century-Crofts. 1982. ? 1598.
31. Painter MJ. Depp R. 0*Donoghue PD: Fetal heart rate patterns and development in the first year of life. Am J Obstel Gynecol 1978: 132:271.
32. Hrbek A. Karberg R Kellmer I. et al: Clinical application of evoked elecroencephalographic responses in newborn infants. I. Perinatal asphyxia. Dev Med Child Neurol 1977: 19:33.
33. Hecox KE. Cone B: Prognostic importance of brainstem auditory evoked responses after asphyxia. Neurology (NY) 1981: 31:1429.
34. Fitzhardinge PM. Flood mark O. Fitz CR. et al: The prognostic value of computed tomography as an adjunct to assessment of the term infant with post asphyxia! encephalopathy. J Pediatr 1981: 99:777.
35. Fenichel GM: Hypoxic-ischemic encephalopathy in the newborn. Arch Neurol 1983: 40:261.