Since the publication of the first edition of this text in 1989, a great deal has been written regarding the issues of neonatal asphyxia and hypoxic–ischemic encephalopathy (HIE) in term and near-term infants. These manuscripts have addressed the incidence, etiology, pathophysiology, treatment, and outcome of such patients, often relating outcomes to the development of cerebral palsy (CP) and/or mental retardation in survivors [1–29]. Much of the understanding of the pathophysiology has been the result of studies carried out in laboratory animals, which have been extrapolated to the human fetus and newborn. Additional studies of complications and outcome have been population-based, comparing the injured infant to carefully selected normal controls. These studies have added a great deal to our understanding of risk factors for brain injury, and have enhanced our ability to predict and to identify patients with increasing accuracy. This has become increasingly important, as newer modalities of treatment have evolved which require more precision in the early identification of these infants so that the validity of these therapies can be ascertained. As can be seen in Chapters 39, 41, and 42, early institution of treatment becomes of paramount importance if an improved outcome is to be achieved.
While some still believe that the major injuries in these patients occur in the intrapartum period, many studies suggest otherwise, and allude to the fact that many of the problems arise antenatally, and may be exacerbated in the intrapartum period. Clearly, several important publications have defined specific criteria that must be present in order to establish that intrapartum events are the primary causes of the infant's difficulties, but not everyone agrees that such rigorous definitions are valid in each and every case [7,10].
Unfortunately, the terms birth asphyxia and HIE have been and continue to be used interchangeably to identify the depressed infant. Stanley et al. noted that “Birth asphyxia is a theoretical concept, and its existence in a patient is not easy to recognize accurately by clinical observation” [1]. Similarly, the description HIE would indicate that the cause of the condition is clearly identified. Most now refer to such infants as having neonatal encephalopathy, a term used by Nelson and Leviton to describe “a clinically defined syndrome of disturbed neurological function in the earliest days of life in the term infant, manifested by difficulty in initiating and maintaining respiration, depression of tone and reflexes, subnormal level of consciousness and often seizures” [2].
This terminology has been adopted by the International Cerebral Palsy Task Force [7] as well as by the American College of Obstetricians and Gynecologists (ACOG) and the American Academy of Pediatrics (AAP) in their text entitled Neonatal Encephalopathy and Cerebral Palsy: Defining the Pathogenesis and Pathophysiology [10].
Neonatal encephalopathy is a purely clinical description that avoids identification of the etiology or pathogenesis of the infant's condition. Unfortunately, neonatal encephalopathy does not exist as a distinct diagnostic category in the International Classification of Diseases, 9th Revision (ICD-9). Table 1.1 lists the categories and numerical designations used to define these infants.
Nevertheless, it is crucial that the caretakers of the injured infant have a more rigorous understanding of the factors that could possibly contribute to the infant's condition, and that they are not unduly influenced by circumstantial evidence. Many of the events leading to the infant's presentation at birth occur long before the onset of labor. With the use of sophisticated imaging, more and more infants are being recognized as having abnormalities that have already led to significant damage prior to the intrapartum period. In addition, careful examination of the placenta has been of great value in identifying lesions that are associated with infections or other anomalies that have or can lead to fetal injury (see Chapter 20).
We recognize that these infants are often unable to tolerate the stress of labor well, may have fetal heart rate abnormalities either prior to delivery or during the early stages of labor, or have abnormal contraction stress tests or non-stress testing. These infants are often difficult to resuscitate, and show neurological features that seem excessive considering the problems that occurred during labor or the birthing process. In addition, some infants may have suffered a significant intrauterine catastrophe, recover, and may even be able to
Table 1.1. International Classification of Diseases 9th Revision (ICD-9) categories used to designate neonatal encephalopathyUnspecified birth asphyxia in newborn infants
768.9
Encephalopathy, not classified
348.3
Encephalopathy, other
348.39
Cerebral depression, coma, and other abnormal cerebral signs
779.2
Asphyxia, mild–moderate
768.6
Severe birth asphyxia
768.5
Asphyxia and hypoxemia
799.0
If seizures are present, add
779.0
Table 1.2. Acute causes of fetal brain injury (sentinel events)tolerate labor well enough not to have abnormalities noted on their fetal heart-rate tracings [23].Prolapsed umbilical cord
Uterine rupture
Abruptio placentae
Amniotic fluid embolism
Acute neonatal hemorrhage
Vasa previa
Acute blood loss from cord
Fetal–maternal hemorrhage
Acute maternal hemorrhage
Any condition causing an abrupt decrease in maternal cardiac output and/or blood flow to the fetus
We also recognize that events leading to difficulties in the prematurely born infant may be different from those in infants born near or at term. Similarly, the preterm infant may have many more and vastly different difficulties in the postpartum rather than the intrapartum period, and will have different clinical features from those seen in the full-term infant. In attempts to evaluate etiology, pathogenesis, intervention, and management, one must be aware that similar events may have different consequences depending upon the patient's capacity to respond to various insults, and some of these are determined by gestational age. In addition, infants with intrauterine growth restriction (IUGR) make up a disproportionate share of infants with neonatal brain injury, suggesting that the underlying cause or causes of the growth restriction may have started in utero and continued through the intrapartum and postpartum periods.
Asphyxia is defined as progressive hypoxemia and hypercapnia accompanied by the progressive development of metabolic acidosis. The definition has both clinical and biochemical components, and indicates that, unless the process is reversed, it will lead to cellular damage and ultimately to the death of the patient. We currently do not have the sophisticated technology of routinely measuring fetal cerebral activity or the neurocellular response to unfavorable conditions such as hypoxia, ischemia, or acidosis, the compensatory mechanisms that protect the brain cells, or, when such mechanisms are inadequate, the documentation of cell injury and cellular death.
In lieu of direct measurements, we have utilized indirect indicators that have been based on studies carried out in laboratory animals and extrapolated to be used in the human fetus. In a few instances direct measurements have been possible, but have not been convincingly linked to outcome. Indirect assessments include the biophysical profile, fetal heart rate measurements, evidence of severe metabolic acidosis, depressed Apgar scores, abnormal newborn neurological function, and development of seizures. As mentioned, the timing of the events is often unknown and difficult to ascertain as far as onset, duration, and severity are concerned. Based on studies in monkeys by Myers [30], and also substantiated to a great extent in fetal lambs by the group in New Zealand [31], two major types of intrauterine asphyxial conditions have been recognized. These include acute total asphyxial events and prolonged partial asphyxia.
The causes of the acute total events are listed in Table 1.2 and have been referred to as “sentinel events” by MacLennan and the International Cerebral Palsy Task Force [7]. In the acute type of asphyxia, there is a catastrophic event: the fetus is suddenly and rapidly deprived of his or her lifeline, and usually does not have the opportunity to protect the brain by “invoking the diving reflex.”
The conditions most commonly encountered include prolapse of the umbilical cord, placental abruption, fetal hemorrhage, and uterine rupture. For a period of time there was an increase in the use of vaginal birth after cesarean section (VBAC), but recently, because of the risk of uterine rupture, there has been a decrease in VBAC, with many physicians and hospitals reluctant to provide such care for patients (see Chapter 12).
Infants with acute asphyxia have damage to the deep gray matter of the brain involving the thalamus, basal ganglia, and brainstem, often with sparing of the cerebral cortex. If successfully resuscitated, these infants may not have evidence of multisystem or multiorgan dysfunction. Laboratory animals that were healthy prior to the onset of the acute asphyxial event develop evidence of neurological damage as early as 8 minutes after the acute event. Major irreversible lesions were found after 10–11 minutes, and the animals usually succumbed if not resuscitated within 18 minutes. After 20 minutes of asphyxia, some animals could be resuscitated, but usually died of cardiogenic shock within 24–48 hours even with intensive care [30].
Although data in humans are lacking, studies of infants following prolapsed cords or uterine rupture suggest similar time frames, and those infants who have occult prolapse often have a better outcome than those with overt prolapse. A study from Los Angeles County University of Southern California (LAC/USC) Medical Center noted that if it required greater than 18 minutes to deliver the fetus after spontaneous rupture of the uterus, neurological sequelae would ensue [32]. Unfortunately, the long-term follow-up of the surviving infants in this study is not available. Thus the 30-minute timing of “
Table 1.3. Factors that have been associated with decreased incidence and mortality due to neonatal encephalopathy in term and near-term infantsdecision to incision,” as recommended by ACOG, is not valid in these situations. The infants who have suffered this type of acute asphyxia will have varying degrees of neurological injury, often manifesting extrapyramidal types of CP and varying degrees of mental impairment depending upon the severity and extent of the injury.More stringent awareness and documentation of the appropriate diagnosis of neonatal encephalopathy
Early prenatal care and recognition of mothers who are at high risk of delivering an infant with neonatal encephalopathy
Pregnancy termination when severe congenital malformations are detected
Early recognition of infants with growth restriction as well as macrosomic infants and avoidance of intrapartum complications
Improved education and training of personnel who are responsible for resuscitation and stabilization of the depressed neonate
Appropriate treatment of mothers who are carriers of Group B streptococci
More liberal use of cesarean section for infants in the breech position
Improved recognition and treatment of mothers with chorioamnionitis
More appropriate induction and use of obstetrical anesthesia
Ready access to neonatalintensive care
The second group of infants are those that have been subjected to prolonged partial asphyxial episodes, and have involvement of the cerebral cortex in a watershed type of distribution. They often have multiorgan involvement and have pyramidal signs of CP. The incidence and severity of cognitive impairment also depend upon the extent and severity of the lesion.
An acute event may also occur in a fetus who has already been subjected to a partial prolonged asphyxial condition or a pre-existing neurological insult. That fetus may demonstrate complications of both processes and have both pyramidal and extrapyramidal neurological findings associated with varying degrees of auditory, visual, and/or cognitive abnormalities.
Haider and Bhutta noted that of the over 130 million babies born yearly worldwide, about 4 million expire in the neonatal period, primarily from complications arising during birthing [21]. Most of the infants are born in developing nations, and at least 50% of the deaths occur at home, where most of these infants are born.
In industrialized nations the incidence of neonatal encephalopathy is much lower, and it has continued to fall over the past three decades. Depending upon the criteria used to document the incidence as well as the severity of the encephalopathy, the incidence varies from 1 to 7/1000 live births. Using the Sarnat score or modifications of the score, infants are classified as having mild, moderate, or severe encephalopathy [33] (see Chapter 16). In the United Kingdom [3], France [18], and Australia [19], similar declines have been noted not only in the incidence of severe encephalopathy but also in the mortality rate of infants so affected.
In Sweden, where an Apgar score of less than 7 at 5 minutes was used to identify babies with this problem, the incidence increased from 5.7/1000 live births to 8.2/1000 over a 7-year period. The incidence of severe depression varied between 1.4 and 2.6/1000 live births. However, the incidence of stillborn infants had decreased significantly [26].
Wu et al. evaluated data from the state of California from 1991 to 2000, and included 5 364 663 live-born infants. Using ICD-9 classifications of 768.5, 768.6, and 768.9, the incidence of neonatal asphyxia fell from 14.8/1000 to 1.3/1000 live births, a 91% decrease during the study years [34]. Data from the 1996 annual summary of vital statistics also demonstrated that the infant mortality rate due to asphyxia fell 72% between 1979 and 1996 [35], and in the most recent surveys the mortality rates were 0.13/1000 in 2001 and 0.15/1000 in 2003.
The reasons for the decrease in neonatal encephalopathy have not as yet been clearly elucidated, but several factors have been suggested (Table 1.3). Perhaps one of the most important reasons is close adherence to a more specific diagnosis of encephalopathy. Other factors playing a significant role include early prenatal care and recognition of women with high-risk pregnancies, increased recognition and appropriate treatment of mothers who are carriers of Group B Streptococcus, early dating of pregnancy, and avoiding post-term deliveries as well as recognizing the fetus who is over- or undergrown. There has also been early termination of pregnancy where infants with significant congenital malformation are detected. Lastly, education programs have been developed to insure that depressed infants are given appropriate resuscitation and stabilization and ready access to intensive care [34] (see Chapters 39 and 40).
Most of the data regarding risk factors have been derived from data accumulated from the Western Australia case–control studies [5,19]. The infants were born at or near term and had moderate to severe neonatal encephalopathy as defined by strict criteria. The criteria included seizures alone or associated with abnormal consciousness, difficulty in maintaining respiration, difficulty in feeding, and abnormal tone and/or reflexes.
These are listed in Table 1.4 and include poor socioeconomic status and advanced maternal age. The findings of a family history of seizures and/or neurological disorders are similar to those previously described by Nelson and Ellenberg [36]. The increased incidence associated with in vitro fertilization is discussed in Chapter 6.
These are delineated in Table 1.5 and include maternal and fetal characteristics. Mothers with thyroid disease were nine times more likely to have infants with neonatal encephalopathy compared to mothers who were euthyroid. This association has also been noted in infants with CP born to mothers with various
Table 1.4. Risk factors prior to conceptionSocioeconomic factors
Increased maternal age
Unemployment
Women without health insurance
Medical conditions
Family history of recurrent non-febrile seizures
Family history of other neurological disorders
Infertility treatment
Poorly controlled chronic illnesses
Table 1.5. Risk factors in the antepartum periodtypes of thyroid diseases. Similarly, severe pre-eclampsia, moderate to severe vaginal bleeding, and severe viral illness were other prepartum risk factors. Growth restriction and those who were post-dates were at increased risk as well.Maternal conditions
Thyroid disease
Severe pre-eclampsia
Moderate to severe vaginal bleeding
Viral infection requiring medical attention
Late or no prenatal care
Poorly controlled diabetes
Systemic lupus erythematosus (SLE)
Infant complications
Post-datism
Intrauterine growth restriction (IUGR)
Abnormal placenta
Congenital malformations
The Australian study deliberately excluded infants with birth defects and abnormal antepartum fetal birth rate tracings, but both of these findings would indicate an at-risk infant [5].
Congenital malformations involving systems other than the nervous system were found more frequently in infants with encephalopathy, suggesting these are antepartum risk factors as well [37]. Women with chronic illnesses such as systemic lupus erythematosus (SLE) and diabetes have an increased risk of having neonates with encephalopathy.
Intrapartum risk factors are often a continuum of those factors that placed the infant at risk in the antepartum period, such as growth restriction and pre-existing congenital abnormalities. Additional factors include maternal fever, a tight nucchal cord, a persistent occiput posterior position, and a persistent non-reassuring fetal heart rate pattern that develops during the intrapartum period after being normal initially [5,6]. Chorioamnionitis, a diagnosis made clinically because of maternal pyrexia, leukocytosis, and malodorous amniotic fluid, is associated with a marked increase in the incidence and severity of neonatal encephalopathy and CP [38]. In laboratory animals, the presence of various cytokines, especially interleukin 6 (IL-6), causes an increased sensitivity of the fetus to ischemia and hypoxia, as well as having a direct deleterious effect on the brain. Measurements of various cytokines in blood and cerebrospinal fluid (CSF) have been found to be significantly higher in infants with encephalopathy who were later found to have abnormal neurodevelopmental outcome [15,39]. The issue of chorioamnionitis is discussed in greater detail in Chapter 12.
Factors that are associated with sudden changes in fetal heart rate patterns leading to bradycardia that does not resolve readily have been described as sentinel events; these are listed in Table 1.2 and are highly correlative with neonatal encephalopathy.
Several findings have been correlated to some extent with the severity of encephalopathy occurring in the intrapartum period. These include a persistently low Apgar score, presence of meconium in the amniotic fluid, evidence of significant metabolic acidosis, the onset of seizures within the first 72 hours of life, the need for cardiopulmonary resuscitation, abnormal electroencephalography, evidence of multiorgan damage, corroborative findings on imaging studies, and corroborative laboratory findings.
The Apgar score was designed to identify infants who were depressed at birth and who required resuscitative efforts [40]. The scoring system required an “advocate” in order to evaluate the infant and provide a numerical score of the infant's condition. Dr. Virginia Apgar did not design this scoring system to evaluate neurological damage or outcome. However, a score of less than 7 at 5 minutes has been used in numerous studies to identify an infant who has suffered from intrapartum events. Unfortunately, there are many factors that influence the Apgar score including immaturity, maternal anesthesia and/or analgesia, fetal and neonatal sepsis, and neuromuscular abnormalities. Using the Apgar score as an isolated finding by itself is inappropriate to define neonatal encephalopathy.
However, the persistence of a low score for greater than 5 minutes despite intensive and appropriate resuscitation has been associated with an increase in morbidity and mortality [27]. Perlman and Risser found that an Apgar score of 5 or less at 5 minutes in combination with significant fetal acidosis and the need for cardiopulmonary resuscitation increased the risk significantly (340-fold) for the infants to develop seizures, a marker of moderate to severe encephalopathy [4,13].
The presence of meconium in the amniotic fluid has long been thought to indicate fetal stress. Meconium is found in 8–20% of all deliveries, being uncommonly encountered in preterm gestations and more frequently in the post-term baby [41]. If meconium is recognized in amniotic fluids of infants at 34 weeks' gestation or younger, significant intrauterine stress or intrauterine infection must be suspected. In term and post-term infants, meconium staining is usually light, and the fetus and newborn are essentially symptom-free. However, heavy, thick meconium passed early in labor tends to have a more ominous significance than when it is passed more proximate to delivery. But even this finding has not been substantiated.
The presence of meconium per se in term infants is not predictive of neurological sequelae; in fact, Nelson and Ellenberg noted that fewer than 0.5% of the infants weighing more than 2500 g with meconium staining had neurological sequelae [36]. In other studies, the presence of meconium-stained amniotic fluid had no predictive value in regard to outcome, the development of neurologic symptoms in the newborn period, or acidosis measured by the pH of cord blood. Even when the presence of meconium was ascertained and used in conjunction with either Apgar scores or cord pH values or both, the finding did not alter the incidence of subsequent neurological abnormalities (see Chapters 20 and 36).
Fetal heart-rate monitoring has now been used for over 40 years, having been developed to decrease the rates of neonatal encephalopathy and intrauterine deaths. While it has been successful in decreasing the rate of stillbirths, its use has not significantly decreased the incidence of neurological sequelae. In one carefully controlled study comparing continuous fetal heart rate monitoring with intermittent auscultation of the fetal heart rate, the incidence of seizures was reduced in the group that was continually monitored electronically. However, the long-term neurological outcomes were the same in both groups, and most of the infants who developed CP did not have seizures in the newborn period [42]. A more complete discussion of intrapartum monitoring is found in Chapter 15.
Since the mid-1960s, obstetricians have utilized fetal acid–base measurements as adjuncts to fetal heart-rate monitoring to evaluate the well-being of the fetus and to identify those who were at risk to develop intrapartum difficulties. Fetal scalp blood sampling during labor has been abandoned to a great extent, and fetal heart-rate monitoring has been used exclusively. The acid–base status of the fetus has been monitored at the time of delivery by assaying the umbilical arterial and venous blood gases immediately after birth.
Initially, an arterial pH below 7.20 was considered to be abnormal, but few such infants were found to have any neonatal or subsequent neurological abnormalities. Correlative data were noted when the umbilical arterial pH was less than 7.0, and especially when it was associated with newborns who required various forms of resuscitation. The vast majority of infants with low pH and no other findings almost always have benign neonatal courses.
Chauhan et al. [43] evaluated their own data as well as several large previously published studies totaling over 43 000 infants born at term who had umbilical arterial pH levels of 7.0 or less. The prevalence of this low pH ranged from 0.2% to 1.6% of live births, with a mean of 0.6%. The incidence of neurologic injury in these infants ranged from 4.3% to 30.9%, and the mortality rate ranged from 0% to 8%.
Low [17] has noted that the threshold for significant metabolic acidosis was a base deficit between 12 and 16 mmol/L. As the degree of acidosis increased, the number of neurological abnormalities increased. In addition, the longer the acidosis persisted, the greater was the risk of neurological defects.
The finding of a low pH in itself is of little consequence unless other abnormalities are found. Perlman [23] reported on a total of 115 infants who were found to have an umbilical cord arterial pH of less that 7.0; 68 of the infants were cared for in the well-baby nursery and discharged home following an uneventful neonatal course. Over 80% of those who were admitted to the intensive care nursery also had benign courses.
Is there an arterial pH level that would predict an abnormal outcome in an infant who is depressed at birth? Goodwin et al. [44] evaluated over 120 infants born at term with a cord pH of less than 7.0. Approximately 4% died, 8% had major neurological abnormalities, 4% were suspected of having neurological problems, 6% were lost to follow-up, and 78% were normal. The same investigators, in a follow-up study of these same infants, noted that if the arteriovenous difference in PCO2 was greater than 25 mmHg, the infants had an increased incidence of seizures, encephalopathy, and cardiac, pulmonary, and renal dysfunction, as well as abnormal neurological outcomes. The arteriovenous difference in PO2 correlated to a much lesser extent [45].
Not all infants with neonatal encephalopathy will have abnormal umbilical arterial blood gases. Those infants with normal gases behave as if the umbilical cord had been clamped at the onset of the asphyxial episode, with little blood flow taking place from the placenta to the fetus. This would be most likely a result of cord prolapse, cord impingement, or even an asystolic event. These infants are depressed, often pale, and poorly responsive. After appropriate resuscitation has been instituted and cardiopulmonary function restored, an arterial sample of the infant's blood will be found to be markedly acidemic.
The onset of seizures within the first 2–3 days of life has been equated with the incidence and severity of neonatal encephalopathy, as well as with the quality of intrapartum care. The incidence of seizures varies from less than 1 to 3.5/1000 live births, and has decreased significantly over the past 20 years from an incidence as high as 14/1000 live births. In a recent comprehensive study of 89 infants by Tekgul et al. [46] the major etiological factor was global encephalopathy, found in 40% of the patients. These investigators also felt that 60% of the seizures were the result of intrapartum factors, and only 25% were secondary to antepartum factors.
The second most common cause of seizures was stroke, with 18% of patients having this abnormality. Intracranial
Table 1.6. Effect of asphyxia on various organs in the newbornhemorrhage accounted for 17% of patients, with most due to extra-parenchymal rather than intra-parenchymal bleeding. Only 5% of the patients had developmental abnormalities and 3% had meningitis or encephalitis. Metabolic disturbances, such as hypoglycemia or hypocalcemia, were encountered infrequently, although these could accompany a severe pre- or intrapartum episode as well [46].Central nervous system injury
Hypoxic–ischemic encephalopathy (HIE)
Cerebral necrosis
Cerebral edema
Seizures
Hemorrhage
Spinal cord injury
Renal injury
Oliguria
Hematuria
Proteinuria
Acute renal failure
Pulmonary injury
Respiratory failure
Pulmonary hemorrhage
Persistent pulmonary hypertension of the newborn (PPHN)
Pulmonary edema
Meconium aspiration syndrome
Cardiovascular injury
Decreased ventricular function
Abnormalities of rate and rhythm
Tricuspid regurgitation
Papillary muscle necrosis
Hypotension
Cardiovascular shock
Gastrointestinal injury
Gastrointestinal hemorrhage
Sloughing of mucosa
Necrotizing enterocolitis (NEC)
Hepatic injury
Hyperammonemia
Elevated liver enzymes
Coagulopathies
Hematological abnormalities
Elevated nucleated red cell count
Neutropenia or neutrophilia
Thrombocytopenia
Coagulopathy
Metabolic abnormalities
Hypoglycemia
Hypocalcemia
Sodium and potassium abnormalities
Hypo- or hypermagnesemia
Source: Modified from Carter et al. [51].
Neonatal stroke is being encountered more frequently as the use of imaging techniques has increased. It is found as frequently as 1/4000 live births, is probably the second most common cause of neonatal seizures, and may be associated with various types of genetic hypercoagulable states. This topic is discussed in Chapter 25.
The mortality rate associated with neonatal seizures has also fallen dramatically over the past 20 years, and has been reported to be as low as 7%. Unfortunately, the prevalence of adverse long-term outcome was 28%, with a 20% rate of later seizure recurrence. With advances and improvements in perinatal care, the incidence and mortality rate associated with seizure have decreased significantly, but the long-term damage in survivors remains unchanged [47]. The seizures themselves may contribute to the already existing brain injury by impairing energy utilization and integrity of the neurons. An aggressive approach to the treatment of neonatal seizures is warranted in order to mitigate further damage to an already compromised central nervous system (CNS) [48] (see also Chapters 17 and 43).
The fetal response to an asphyxial episode is to preserve perfusion and oxygenation of the heart, brain, and adrenal gland at the expense of the other “non-vital” organs, such as the kidney, lungs, gastrointestinal tract, and musculoskeletal system. The incidence of single or multiple organ injury in association with neonatal encephalopathy has varied from 40% to almost 100%, and seems to correlate with the severity of the CNS injury [7,23,25,49–51]. Most often, the renal system is involved, and it is the easiest to evaluate. Findings range from mild oliguria (less than 1 mL/kg/h), proteinuria, and hematuria to renal tubular necrosis and acute renal failure.
Cardiac manifestations vary from minor arrhythmias, ST segment changes on EKG, and tricuspid insufficiency to papillary muscle necrosis, poor ventricular contractions, and cardiogenic shock. Patients with moderately severe or severe asphyxia may have a fixed, non-varying rapid heart rate of 140–160 beats/minute, which may be a prelude to impending failure and cardiogenic shock.
Pulmonary manifestations of asphyxia vary from increased pulmonary vascular resistance that responds readily to correction of acidosis and hypoxia, to persistent pulmonary hypertension of the newborn, severe pulmonary insufficiency, or pulmonary hemorrhage, all of which are difficult to manage.
Other organs that are involved, and the manifestations of their involvement, are listed in Table 1.6. One area often
Table 1.7. Laboratory studies used to support the diagnosis and severity of neonatal encephalopathyoverlooked in the patient with severe asphyxia is damage to the spinal cord. Clancy et al. [52] described 18 severely asphyxiated newborns, 12 or whom expired. On autopsy, five of the 12 demonstrated severe ischemic necrosis in the spinal cord gray matter. Electromyographic studies in the six survivors were abnormal and consistent with recent injury to the lower motor neurons above the level of the dorsal root ganglion. It is often difficult to distinguish clinically between damage to the cortical motor area and damage to the spinal cord.Study
Body fluid
Ammonia
Serum
Lactate
Serum
Creatine kinase BB (CK-BB)
Serum, CSF
Erythropoietin
Serum, CSF
Neuron-specific enolase
CSF
Myelin basic protein
CSF
Glutamate
CSF
Troponin T
Serum
S100B protein
Serum
Interleukin 6
Serum, CSF
CSF, cerebrospinal fluid.Source: Modified from Volpe [29].
Phelan et al. [53] described 57 infants with HIE, of whom 14 had no evidence of multisystem problems. Six infants were delivered following uterine rupture, one had fetal exsanguination, one had a cord prolapse, and one was delivered following maternal cardiopulmonary arrest. Five fetuses had sudden and prolonged fetal heart rate decelerations, which persisted until delivery. All of these infants would be classified as having an acute asphyxial or a sentinel episode, and would not have had the opportunity to develop the “diving reflex” necessary to protect the brain and heart at the expense of other organs.
If an infant with intrapartum asphyxia demonstrates only CNS involvement without other organ abnormalities, it may be that there was an acute hypoxic event, that the CNS damage did not occur in the intrapartum period, or that it was due to a cerebrovascular event that did not cause profound hypoxia or hypotension to affect other organs.
In addition to the abnormalities in acid–base determinations that have been described previously, various metabolic parameters have also been found to be abnormal in these patients. Some, but not all, have correlated to a certain degree with severe encephalopathy, but they often do not differentiate those infants from infants with mild to moderate encephalopathic states [29].
Urinalysis will usually detect proteinuria and hematuria. Elevations of serum and hepatic enzymes document renal and liver involvement, but may not correlate well with the degree of CNS damage. Elevation of serum ammonia is usually found when severe neurological damage has occurred. Elevations of creatine kinase (CK) in serum and CSF are often found, and both resolve quickly over a one- to two-day period. Erythropoietin, both in serum and in CSF, are excellent markers of severe encephalopathy. The level of urinary lactate/creatine ratio, measured within the first 6 hours of life, has been correlative with the severity of encephalopathy and adverse neurological sequelae at 1 year of age. Follow-up data on this measurement have not been forthcoming, and many encephalopathic infants fail to pass urine within the first 6 hours of life.
Recently, increased levels of troponin T [54] and S100B protein [55] in serum, and IL-6 in serum and CSF, have been found to correlate with the severity of the encephalopathy [15]. Troponin T has been used as a marker of myocardial damage, and has been correlative with the degree of encephalopathy as well. Unfortunately, most of the reports have been based on a small number of patients, and their findings must be evaluated in a much larger group of encephalopathic infants.
If abnormalities are encountered, they should be re-evaluated frequently in order to assess the evolution of the injury. Table 1.7 lists the various laboratory determinations that have been used to evaluate the extent of CNS injury in affected infants. Some of these measurements are not readily available in many clinical laboratories, and samples of sera of CSF have to be sent to specialized laboratories for assay.
It is important to do a lumbar puncture in the encephalopathic individuals, in order to obtain fluid for the assays as well as to rule out meningitis and encephalitis as causes of the infant's encephalopathy [29].
The measurements and interpretations of white blood cells and nucleated red blood cell counts [56] in these infants are discussed in Chapter 21.
If significant damage has occurred to the CNS, the infant should demonstrate neurological abnormalities in the neonatal period. It is often difficult to appreciate such abnormalities in preterm infants, especially those who have cardiopulmonary abnormalities and who are being treated with assisted ventilation. Often these infants cannot be distinguished from other prematurely born infants with similar cardiopulmonary abnormalities. However, in the term or near-term infant, signs of encephalopathy are readily discernible. Sarnat and Sarnat [33] developed an infant scoring system that categorizes the patients into three stages of “postasphyxial encephalopathy,” identifying mild, moderate, and severe. Although they correlated many of the findings with electroencephalographic changes, one can use their classification even if the electroencephalographic changes are not evaluated. The clinical manifestation and the
Table 1.8. Criteria to define an acute intrapartum event sufficient to cause cerebral palsygradation of severity of the encephalopathy are discussed in Chapter 16. Similarly, the electroencephalographic abnormalities found in these infants are discussed in Chapter 17.Essential criteria (must meet all four)
(1) Evidence of metabolic acidosis in fetal umbilical cord arterial blood obtained at delivery (pH < 7.00 and base deficit ≥ 12 mmol/L)
(2) Early onset of severe or moderate neonatal encephalopathy in infants born at 34 weeks or more of gestation
(3) Cerebral palsy of the spastic quadriplegic or dyskinetic type
(4) Exclusion of other identifiable etiologies such as trauma, coagulation disorders, infectious conditions, or genetic disorders
Criteria that collectively suggest an intrapartum timing (within close proximity to labor and delivery, e.g., 0 to 48 h), but are non-specific to asphyxial insults
(1) A sentinel (signal) hypoxic event occurring immediately before or during labor
(2) A sudden and sustained fetal bradycardia or absence of fetal heart rate variability in the presence of persistent, late, or variable decelerations, usually after a hypoxic sentinel event when the pattern was previously normal
(3) Apgar scores of 0–3 beyond 5 min
(4) Onset of multisystem involvement within 72 h of birth
(5) Early imaging study showing evidence of a non-focal cerebral abnormality
Source: Reproduced with permission of BMJ and ACOG.
With the increased use of neuroimaging in the encephalopathic infant, investigators have been able to more clearly define the extent of the injury, and have, to some degree, determined the timing of the insult. While not being able to pinpoint timing in minutes or even hours, by following the changes that occur on imaging studies the neuroradiologist has been able to help in determining the time frames of injury in many infants. Magnetic resonance spectroscopy is also of help in evaluating the extent of damage, as well as the timing of the insult. These aspects are discussed in detail in Chapter 18.
In evaluating all of the factors that have been associated with neonatal encephalopathy, it is obvious that no one factor taken by itself can identify the infant who will have neurological injury. Using only the Apgar score, unless it is very low for a protracted period, is not a very good marker, nor is it predictive of long-term outcome.
In his excellent review of intrapartum asphyxia and its relationship to CP, Perlman noted that “a single marker of in utero stress provided little useful information regarding the asphyxial process or the fetal adaptive responses, and thus the relationship to neonatal brain injury or subsequent cerebral palsy.” He noted that there had to be a “constellation of markers” in linking intrapartum events to neonatal encephalopathy and then to CP. The infants with severe encephalopathy, including seizures, could be identified by using a 5-minute Apgar score of 5 or less, the need for intubation or CPR, and an umbilical cord arterial pH of 7.0 or less [23].
For years, it was postulated that a depressed infant who developed seizures within the first 72 hours of life had suffered from an adverse intrapartum event. Currently, there is disagreement as to the correlation of intrapartum events with the development of neonatal encephalopathy. The case–control studies from Australia found that 69% of the infants with encephalopathy had only antepartum risk factors, 25% had both antepartum and evidence of intrapartum markers, 4% had evidence of intrapartum issues only, and 2% had no recognizable causes [5].
Volpe, drawing from his vast experience in evaluating encephalopathic infants, noted that 20% had insults related primarily to antepartum events, 35% had intrapartum disturbances, 35% had both intra- and antepartum events, and 10% had issues in the postpartum period. The latter was encountered primarily in prematurely born infants [29].
The International Cerebral Palsy Task Force developed a template enumerating the criteria used to define an intrapartum event sufficient to cause CP [7]. This template was modified by the ACOG and the AAP, and published in 2003 [10]. These criteria, which are depicted in Table 1.8, include both essential criteria and criteria that collectively denote intrapartum timing. Not all investigators have agreed with these criteria, especially in regard to the issues of timing that have been advocated by the Task Force or the ACOG/AAP publications.
Cowan et al. [11] evaluated 351 infants with neonatal encephalopathy and/or seizures who were referred to two large intensive care units, and who were evaluated with MRIs and/or postmortem examinations. The infants were divided into two groups: 261 infants with neonatal encephalopathy and 90 who had seizures without encephalopathy. Imaging of the brain showed an acute insult in 80% of the encephalopathic infants and did not show evidence of prior injuries or atrophy. In the group with seizures only, focal damage (stroke) was found in 69% while 2% had evidence of antenatal injury. Their follow-up was disconcerting, as 66 infants in the encephalopathic group died and 85 had neurological sequelae. This study was not population-based or case-controlled, as were the studies from Australia, but was based on findings from tertiary referral centers that treated the most severely affected infants.
As noted previously, if the fetus has suffered from an acute intrapartum event, and there is not enough time to invoke the “diving reflex,” the infant may not show multiorgan damage. Similarly, if the umbilical cord is acutely impinged, prolapsed, or tightly wound around the fetus's neck, the cord blood gas may be normal and may not accurately reflect the fetal condition at the time of birth.
Table 1.9. Conditions causing neonatal depression and/or neonatal encephalopathy that mimic intrapartum asphyxiaNeonatal sepsis
Chorioamnionitis without documented neonatal sepsis
Congenital infections
Viral
Toxoplasmosis
Neuronal migration disorders
Congenital myotonic disorders, including congenital and transient myasthenia gravis
Metabolic conditions causing lactic acidosis
Genetic disorders associated with thrombotic or thrombophilic abnormalities, including
Protein C and protein S deficiencies
Factor V Leiden deficiency
Anticardiolipin antibodies
In other situations, the fetus may have suffered an acute or subacute intrauterine asphyxial event prior to the onset of labor, followed by a period of recovery, and at the time of delivery may show no overt signs of injury. Such infants were often sent to the well-baby nursery, but within hours they developed signs of encephalopathy, often with seizures, and had the neuroimaging findings that have been associated with an acute intrapartum asphyxial episode. Perlman noted that in his experience this group of infants makes up 50% of patients with encephalopathy [23]. In addition, these same infants often may not tolerate labor well, may have poor cardiopulmonary reserve, and may develop non-reassuring fetal heart rate tracings. Even though they are delivered expeditiously, they demonstrate all of the findings that have been associated with “intrapartum asphyxia” and have neuroimaging that is compatible with an intrapartum injury.
As techniques such as hypothermia are being developed and utilized to treat encephalopathic infants, it behooves the physician to be as precise as possible in identifying the infants with intrapartum injury, in order to provide the most appropriate types of care. Treatment with hypothermia must be initiated within 6 hours after birth in order to be effective. Many infants enrolled in these trials may not have had a “true” intrapartum event, and may not respond well to treatment. These infants must be identified in a retrospective manner, if possible, and evaluated as a separate group of patients from those who had a true intrapartum injury.
Nelson and Leviton were among the first to question whether all infants with neonatal encephalopathy had their insults secondary to intrapartum asphyxia [2]. One of the more common problems that can present in this fashion is the infant with neonatal sepsis. Currently, Group B Streptococcus is the most common organism involved. In many instances, the mother had been pretreated with antibiotics, and an organism was not able to be cultured from the newborn's blood or CSF. Indirect evidence of infection may be present, including an abnormally low or elevated white blood cell count, an elevated C-reactive protein, and/or evidence of severe chorioamnionitis. These infants have severe lactic acidosis, may have pulmonary hypertension or hemorrhage, and are very difficult to manage in the neonatal period. Even with the use of nitric oxide, high-frequency ventilation, and extracorporeal membrane oxygenation (ECMO), the mortality and morbidity rates are high.
Similarly, the infant born of a mother with chorioamnionitis may also behave like the infant with intrapartum asphyxia. Placental perfusion has been shown to be decreased in such pregnancies, further subjecting the fetus to increased risk of damage (see Chapters 12 and 20).
Although most infants with congenital infections such as cytomegalovirus, herpes, or toxoplasmosis are asymptomatic at birth, and later develop clinical manifestations of their disease, a few will be symptomatic in the neonatal period and behave as if they had suffered from intrapartum asphyxia. Infants with congenital parvoviral infection are often born with generalized edema, are difficult to resuscitate, and have significant rates of morbidity and mortality. Newborns with neuronal migration disorders and those with early-onset myotonic disease have also been mislabeled as infants suffering from intrapartum encephalopathy. The infant with an intrauterine stroke may also be depressed in addition to having seizures. Too often, without substantiating evidence, it is assumed that the stroke has been caused by an adverse intrapartum hypoxic event. Lastly, infants with metabolic disorders can also present in the immediate newborn period with signs suggesting intrapartum asphyxia (see Chapters 23 and 34).
Table 1.9 lists some of these conditions, and clinicians must be aware that not all patients with encephalopathy have their insult due to an intrapartum asphyxial event.
(see Chapter 48)
Although the incidence and mortality rates of infants with encephalopathy have decreased markedly, the complication rates found in the survivors have not changed appreciably in the past 20 years [11]. The infants with mild encephalopathy usually have benign courses, and have few, if any, neonatal sequelae. Those with severe encephalopathy (grade III Sarnat score) have a very high mortality rate, ranging from 25% to 100% [11]. Major handicaps are reported in as few as 42% to as high as 100% of survivors, with most studies showing that more than 80% are handicapped to a significant degree. The infants with moderate encephalopathic changes (grade II Sarnat score) have a much lower mortality rate of 5% or less, and fewer than 25% have major handicaps, with 75% or more having no discernible sequelae. Although careful follow-up of these patients to school age has shown an increased incidence of learning disabilities, similar results have been found by evaluating MRI findings in these infants and documenting the extent of the injury [57].
It is anticipated that, with the development and use of techniques such as hypothermia, growth factors, and oxygen-free-radical inhibitors, the outcome of the encephalopathic infants will improve. Early data suggest that these therapies are proving to be beneficial, especially in the moderately severe group of patients (see Chapter 42).
(see Chapter 46)
The relationship of neonatal encephalopathy and CP with and without cognitive impairment continues to be elusive and often difficult to ascertain. The incidence of CP varies, and is dependent on the type of injury present and the cause, if it can be identified. In most developed countries the incidence is remarkably similar, varying between 1.0 and 2.5/1000 live births, and it has not changed to any significant degree over the past two decades. The rates of CP rose in the early 1970s and tended to remain constant during the early 1980s, primarily due to increased numbers of very-low-birthweight infants surviving, some of whom developed CP. Himmermann et al. [58], however, noted that the subsequent reduction in neonatal mortality has resulted in far more healthy children surviving without CP than with CP. Despite this, prematurely born infants still make up at least 25–50% of the total number of infants so afflicted, and the risk of CP increases with decreasing gestational age and birthweight. Neuroimaging has been helpful in identifying not only the area of the brain involved, but also the timing of the insult. Malformations of the brain tend to occur primarily during the first and into the second trimester of pregnancy, while lesions in the white matter occur between the 20th and the 34th weeks of gestation, and gray matter lesions and injuries to the striatum occur after the 34th week of pregnancy [59,60].
Spastic diplegia is the most frequent type of CP found in the preterm infant, and there has been an increasing incidence of hemiplegia due primarily to cerebral infarcts. In the latest study from Sweden, 30% of patients born at or near term had a prenatal cause for their CP, and 35% had perinatal causes. However, in this latter group were infants who had evidence of HIE, intracerebral hemorrhage, or infections involving the CNS. It was also noted that the maternal risk factors increased from 4.8% in term infants with CP and 8.5% in preterm infants in the period 1969–74, to 17% and 35% respectively in the years 1995–8. The most frequently encountered risk factors were maternal fever at the time of delivery and maternal diabetes. A disconcerting factor in the Swedish studies has been the increased incidence of dyskinetic CP, which is often associated with intrapartum difficulties [61].
Neonatal encephalopathy was found in 24% of the term infants with CP reported from Australia [19], 22% of those from Canada [61], and 31% of those from Norway [61].
In the large Dublin randomized trial of electronic versus intermittent auscultation, six infants who had seizures in the neonatal period were found to have CP at 4 years of age. Three were from each group of monitored patients. Interestingly, 15 additional patients with CP who were diagnosed at 4 years of age did not have neonatal seizures and were not in the high-risk group. Of the total number of patients with CP at 4 years of age, only 29% had intrapartum difficulties [42]. Thus most reports have indicated that the incidence of CP due to neonatal encephalopathy ranges between 10% and 30% in term infants.
CP found in children following neonatal encephalopathy is primarily of the spastic quadriplegic or dyskinetic types, and is associated more frequently with severe cognitive impairment and epilepsy when compared to those children with CP who were not encephalopathic. In addition, the mortality rate is almost four times greater (19% vs. 5%) in the encephalopathic patients [11,19].
Investigators have also reported an association of CP with non-cerebral birth defects, particularly cardiac defects, which has added further evidence that many of the antecedents of CP occur in the antepartum period [37].
The most important approach would be to decrease the incidence of preterm births, because this group contributes at least 50% of patients to the CP population [62]. This would be a formidable task and would require a multipronged attack if it were to be successful. There has been a significant increase in the number of multiple births, which has increased the number of preterm infants. In addition to being born early, these infants have an increased risk of in utero complications such as twin-to-twin transfusion and the in utero death of one of the infants. Multiple births are frequently the result of fertility-enhancing techniques, which often are used in older patients. Hopefully, improved techniques, careful counseling, and the implantation of a single rather than multiple fertilized eggs will mitigate this problem to some extent.
Early recognition and treatment of women who have or are at risk of having chorioamnionitis could be a factor in decreasing the incidence of prematurity, and it may be an important factor in decreasing the incidence of CP in the term infant as well. Similarly, careful monitoring of pregnant women with chronic illnesses, especially SLE, diabetes, and thyroid disease, would be another approach [63,64].
Electronic fetal heart rate monitoring was hailed as a method to decrease the incidence of CP. This technique has helped reduce the number of stillbirths, but has had little, if any, effect on the incidence of CP. While the rate of cesarean sections in response to abnormalities noted in fetal heart rate tracings has increased almost fivefold, the incidence of CP has remained unchanged [65].
Infants who are growth-restricted in utero contribute significantly to the number of patients with neonatal encephalopathy, seizures, and CP. Improving the early recognition and earlier intervention in these pregnancies could potentially enhance the outcome for these infants.