Perinatal HIE/NE and the Contribution to Subsequent CP

Hypoxic–ischemic brain injury that occurs around the time of birth is the most recognized cause of the severe, long-term neurologic deficits in infants and children termed CP. This is often defined as a nonprogressive motor disorder of early onset affecting approximately 2–3 per 1,000 school-aged children. Brain injury resulting in CP may occur during the antepartum, peripartum, or postnatal period. However, overwhelming evidence suggests that the genesis of CP [Reference Aravamuthan, Fehlings and Novak10,Reference te Velde, Morgan, Novak, Tantsis and Badawi11,Reference McIntyre, Goldsmith and Webb12] in approximately 60–70% of cases is antenatal in onset [Reference Nelson and Ellenberg13,Reference Nelson and Leviton14,Reference Nelson15]. The timing of the antepartum insult is critical to the evolution of a specific lesion. Thus, cerebral ischemia prior to the 20th week of gestation results in a neuromigrational defect, for example, schizencephaly. Ischemia occurring between the 28th and the 34th weeks of gestation usually results in periventricular leukomalacia (PVL). When ischemia occurs between the 34th and 40th weeks of gestation, focal/multifactorial cerebral injury usually develops (Figure 1.1) [Reference Aravamuthan, Fehlings and Novak10]. By contrast, the proportion of CP attributed to perinatal events is small and has been estimated to represent approximately 15–20% of cases [Reference Aravamuthan, Fehlings and Novak10,Reference Hollung, Bakken and Vik16,Reference Novak, Hines, Goldsmith and Barclay17]. We will briefly review the evidence supporting the substantial antenatal contribution to CP.

Figure 1.1

Antenatal ischemic injury to the brain is dependent upon the timing of the insult. A, schizencephaly; B, periventricular leukomalacia; C, focal cerebral infarction.

Figure 1.1 Long description

Panel A shows a large, irregular anechoic cavity extending from the cortical surface toward the ventricular region. Panel B shows multiple small, rounded anechoic areas clustered near the central region. Panel C shows two asymmetric anechoic regions within the cerebral parenchyma.

Radiologic Evidence

Magnetic resonance imaging was used to retrospectively analyze the brains of 29 patients born at term suffering from CP. Three major patterns were noted. They included (1) gyral anomalies, suggestive of polymicrogyria, consistent with mid second trimester injury; (2) isolated PVL reflecting late second- or early third-trimester injury; and (3) watershed cortical or deep gray nuclear damage, consistent with late third-trimester, perinatal, or postnatal injury. Overall, the contribution of these abnormalities was noted in 55% of patients [Reference Påhlman, Gillberg and Himmelmann18]. In a second study, CT brain scans obtained from 83 children born at term were used to evaluate the most likely etiology of hemiplegic CP. The most common finding was periventricular atrophy considered to represent PVL (37%), often as a complication of preterm birth. As this lesion was found in children born at term it was considered to reflect an in-utero etiology of CP. Cerebral maldevelopment, due to a very early intrauterine lesion, was detected in 17%, whereas cortical/subcortical lesions were seen in 16% of the children. Altogether the CT findings indicated a prenatal origin of hemiplegia in approximately 70% of cases [Reference Whitney, Warschausky, Ng, Hurvitz, Kamdar and Peterson19].

Clinical Evidence

A study of children with spastic CP (n = 183) were compared to a matched group of control children (n = 549). There was a 2.8-fold increased risk for spastic CP in those children with clinical signs of BA as compared to controls. It was estimated that the possible cause of subsequent brain damage related to intrapartum “asphyxia” was approximately 8% of children [Reference Blair and Stanley20].

Contribution of Intrapartum versus Antenatal Factors in the Genesis of NE/HIE

A study involving 405 infants ≥35 weeks’ gestation with early-onset encephalopathy, meeting criteria for perinatal asphyxia, and with neuroimaging findings consistent with acute HIE, were compared with 239 neurologically normal infants. Case infants were fivefold more often delivered by emergency cesarean section (CS). On logistic regression analysis, only one antepartum factor (gestation ≥41 weeks) and several intrapartum factors – prolonged rupture of membranes (PROM), abnormal cardiotocography, thick meconium, and sentinel event – remained independently associated with HIE (area under the curve [AUC] 0.88; p < 0.001). Notably 3.7% of cases and 19.7% of controls had no identifiable risk factors for HIE [Reference Fahey, Maclennan, Kretzschmar, Gecz and Kruer21]. These findings strongly point to the intrapartum period as important in the evolution of HIE.

In another study of infants with HIE, 87% had intrapartum complications that increased risk, a sentinel event (36%, relative risk [RR] = 16.1), clinical CA (40%, RR = 5.2), or both (RR = 11). Urinary tract infection during pregnancy (RR = 2.6) was a potential antenatal risk factor. It was concluded that a significant proportion of HIE is preceded by intrapartum events [Reference Jin, Lewis and Bakhtiari22]. In another study involving 351 term infants >36 weeks who presented with NE (n = 262) or seizures (n = 90) without other signs of global encephalopathy, MRI images showed evidence of an acute insult in 80%, whereas evidence of established injury was noted in 2 infants (<1%). In the seizure group, acute focal damage was noted in the majority of infants. Two infants also had evidence of antenatal injury. These findings point to the immediate perinatal period as the most important precursor to NE [Reference Cooper, Fahey and Mackay23].

In another study of infants with moderate or severe NE, the incidence was found to be 1.64 per 1,000 term live births. The main cause was perinatal hypoxic ischemia (52%). The timing was intrapartum in 56%, antepartum in 13%, and ante-intrapartum in 10% of cases. In 19% of cases, no underlying cause for NE was identified [Reference Jöud, Sehlstedt, Källén, Westbom and Rylander24]. Another large study comprised singleton neonates of gestational age (GA) ≥36 weeks, who met criteria for NE within 3 days of birth [Reference Novak, Jackman and Finch-Edmondson25]. Clinically recognized seizures (60%) were the most prevalent condition; 49% had a 5-min Apgar score of ≤3 and 18% had a reduced level of consciousness. An abnormal maternal or fetal condition predated labor in 46%; maternal hypertension (16%) or small for gestational age (SGA) (16%) were the most frequent risk factors. The most prevalent perinatal event was elevated maternal temperature ≥37.5°C (27% of cases). Acute asphyxial indicators were reported in 35% and inflammatory indicators in 24%. Much of NE remained unexplained. This group had very heterogenous inclusion criteria and only 18% had evidence for reduced level of conscious, contributing to the unclear etiology of NE in many cases.

Perinatal Conditions Associated with NE on the First Postnatal Day

Neonatal encephalopathy that occurs around the time of birth may be related to interruption of placental blood flow (PBF), which may be acute (a sentinel event) or intermittent (fetal heart rate [FHR] abnormality), ICH, CA/maternal fever, intrapartum-acquired sepsis, placental abnormalities, and maternal medications (Figure 1.2). Factors that may affect the brain around delivery and manifest as NE during the first postnatal day. The most distinctive of these states is HIE secondary to interruption of PBF, reflective of asphyxia (see section on Perinatal Hypoxic–Ischemic Encephalopathy). Over the years it has become apparent that not all cases of “acute” brain injury present immediately at birth and are triaged directly to the neonatal intensive care unit (NICU). Rather, some infants are initially triaged to the well-baby nursery (WBN) and subsequently present with clinical signs of encephalopathy including seizures during the first 24 h. Two case series presented next highlight a group of infants with delayed clinical presentation of perinatal injury.

Figure 1.2

Factors that may affect the brain around delivery and manifest as neonatal encephalopathy during the first postnatal day.

Figure 1.2 Long description

The diagram shows a brain at the centre surrounded by the following factors: Interruption of Placental Blood Flow includes Acute and Intermittent. Sentinel Event includes Abruption, Cord prolapse, Uterine Rupture, and Cord Abnormalities. Placental Abnormalities include Maternal and/or Foetal Vascular Malperfusion, Foetal Thrombotic Vasculopathy, and Villitis. Maternal Medications include Selective Serotonin Uptake Inhibitor and Tricyclics. Sepsis is listed as a separate factor. Intracranial Haemorrhage includes Subdural, Subarachnoid, Epidural, Intraventricular, and Intraparenchymal. Chorioamnionitis includes Clinical with Maternal Fever, and Histologic with Evidence of Chorioamnionitis or Funisitis.

Second Case Series (2004–2009)

Between January 2004 and December 2009, 33 infants of 29,597 term infants delivered at New York Presbyterian Hospital, New York presented clinically with early signs consistent with intrapartum brain injury [Reference Whitney, Kamdar, Ng, Hurvitz and Peterson27]. The underlying conditions included HIE (n = 8, 0.27/1,000), or after neuroimaging, SDH (n = 10, 0.34/1,000), intraventricular hemorrhage (IVH) (n = 3), or intraparenchymal hemorrhage (IPH) (n = 2, 0.17/1,000), and NAIS (n = 4, 0.14/1,000 live term births), reflecting an overall rate of 1.12/1,000 term live births. Of the 33 infants, 13 (39%) were asymptomatic in the DR and triaged to the WBN. Delayed presentation included apnea (n = 6) and, upon neuroimaging, IVH (n = 3), IPH (n = 2), and sinus venous thrombosis (SVT) (n = 1) (Figure 1.2). Three infants presented with desaturation episodes and were noted to have SDH (n = 1) and NAIS (n = 2), and one infant with a visible fracture had an epidural hemorrhage on neuroimaging (Figure 1.3). Twenty (61%) infants were triaged directly to the NICU due to perinatal depression requiring respiratory support. Diagnoses included HIE with seizures (n = 4) presenting from 6 to 48 h post-delivery; one infant who was comatose from birth died within 6 h. One infant required prolonged respiratory support but had a normal neurologic course. In the remaining 14 infants, clinical signs included seizures (n = 12) and apnea (n = 2) with resolution within 24 h. On MRI the associated changes included SDH (n = 7), basal ganglia and bilateral parasagittal injury (n = 3), NAIS (n = 2), periventricular cysts (n = 1), and normal (n = 1). Electroencephalogram (EEG) seizures were noted in nine cases. Placental pathology revealed changes in 16/18 cases including CA (n = 7) and fetal thrombotic vasculopathy (FTV) (n = 5); the latter was associated with NAIS in 3/4 cases.

Figure 1.3

Schema of infants who present with neurologic signs in the first 24 h and triaged initially to the well-baby nursery (n = 13) or to the neonatal intensive care unit (n = 33).

Figure 1.3 Long description

The flowchart begins with the heading number of infants n equals 33. The infants are divided into two groups. One group is labelled triaged to well baby nursery, n equals 13. Under this group, clinical signs include apnoea, n equals 6; desaturation episodes, n equals 3; seizures, n equals 3; and skull fracture, n equals 1. Also under this group, neuroimaging includes IVH, n equals 3; IPH, n equals 2; venous sinus thrombosis, n equals 1; bilateral parasagittal injury, n equals 1; subdural haemorrhage, n equals 3; NAIS, n equals 2; and epidural haemorrhage, n equals 1. The second group is labelled admitted to NICU, n equals 20. Under this group, clinical signs include HIE, n equals 5; seizures, n equals 12; apnoea, n equals 2; and respiratory distress, n equals 1. Also under this group, neuroimaging includes bilateral parasagittal injury, n equals 3; subdural haemorrhage, n equals 7; NAIS, n equals 2; periventricular cysts, n equals 1; frontal cysts, n equals 1; and normal, n equals 1.

Summary

1. 1. Clinical signs of NE including HIE may manifest in the DR or be delayed with onset of clinical signs in the first 24 h of life.

2. 2. Hypoxic–ischemic encephalopathy appears to be less frequent in occurrence (range 0.27–0.34/1,000 births), whereas ICH and SDH were relatively more common.

3. 3. With the exception of infants who progressed to HIE, no perinatal marker was able to prospectively identify the remaining infants destined for brain injury; such infants were only identified at the time of clinical signs/symptoms.

4. 4. Histologic CA was associated with both SDH and FTV with cerebral infarction (CI), consistent with prior studies [Reference Bourke-Taylor, Joyce, Morgan, Reddihough and Tirlea28,Reference De Felice, Toti, Laurini, Stumpo, Picciolini and Todros29].

Pathway to Brain Injury

Studies with term fetal or neonatal monkeys provide important insights into the relation of FHR changes to hypoxia ischemia (HI) and subsequent neuropathology [Reference McNamara, Morgan and Novak30,Reference Novak, Morgan and Adde31]. With severe “total” asphyxia the resultant neuropathology involved principally the thalamus and brainstem. Partial interference in respiratory gas exchange (partial asphyxia) resulted in slower evolving hypoxia and acidosis, followed by late FHR decelerations, a reduction in cardiac output, hypotension, and eventual evidence for cerebral ischemia. Depending on the duration and severity, brain injury became apparent from one to several hours after the insult. The subsequent injury was located within a watershed distribution, that is, parasagittal cerebral cortex and white matter as well as basal ganglia (BG) [Reference McNamara, Morgan and Novak30,Reference Novak, Morgan and Adde31,Reference Kinney, Volpe and Volpe32]. The topography of the lesions is similar to that described in humans [Reference Kinney, Volpe and Volpe32].

Definition of Asphyxia

Birth asphyxia develops during the first and second stages of labor secondary to interruption of PBF, with impaired gas exchange, which leads, if persistent, to hypoxemia and hypercapnia. The process is identified by fetal acidosis (measured in umbilical arterial blood), which broadly reflects the degree of anaerobic metabolism that evolves during periods of hypoxia or increased oxygen demand [Reference Rainaldi and Perlman33]. The umbilical arterial pH that best defines asphyxia remains unclear. Fetal acidemia had originally been defined as an umbilical arterial pH < 7.2. With this definition the rate of ‘‘asphyxia’’ ranged from 5% to 20% and the risk of neonatal neurologic sequelae was low [Reference Einspieler, Prayer and Marschik34]. Subsequently it was noted that severe fetal acidemia, that is, umbilical arterial pH < 7.00, reflects a degree of acidemia where the risk of adverse neurologic sequelae is increased [Reference Olsen, Brown and Eeles35,Reference Einspieler, Bos, Libertus and Marschik36,Reference Goodwin, Belai, Hernandez, Durand and Paul37]. A cord pH of <7.0 is present in from 0.3 to 3.9 per 1,000 of term newborns [Reference Prechtl, Einspieler, Cioni, Bos, Ferrari and Sontheimer38,Reference Kraus, Saucedo, Byrne, Chalak, Pruszynski and Spong39]. A systematic review evaluated the association of umbilical arterial pH <7.0 with the incidence of HIE and the proportion of cases of CP in term infants. The incidence of an arterial pH <7.0 at birth was 3.7 of 1,000 term newborns of which 17.2% survived with neonatal neurologic morbidity, 16.3% had seizures, and 5.9% died during the neonatal period. The incidence of HIE was 2.5 of 1,000 live births. The proportion of CP associated with intrapartum HI was 14.5%. This adds further support for the relatively small contribution of intrapartum HIE to CP [Reference Alexander, Amery and Salt40].

Circulatory Changes Accompanying Interruption of Placental Blood Flow (Asphyxia) to Preserve Cerebral Perfusion

When PBF is compromised, cardiac output redistributes to protect more vital organs (e.g., brain, myocardium, and adrenal glands) (Figure 1.5). This chemoreflex-mediated alteration of blood flow [Reference Kwong, Eeles, Olsen, Cheong, Doyle and Spittle41] is at the expense of decreased flow to “less vital” organs such as the kidney or intestine, as discussed in detail in Chapter 5 [Reference Cohn, Sacks, Heymann and Rudolph42,Reference Peeters, Sheldon, Jones, Makowski and Meschia43,Reference Jensen, Garnier and Berger44]. With time there is loss of cerebral vascular autoregulation resulting in a pressure passive circulation, eventual diminution in cardiac output with resultant hypotension, and ultimately a critical decrease in cerebral perfusion and oxygen delivery (Figure 1.4) [Reference Volpe45,Reference Dawes, Jacobson, Mott, Shelley and Stafford46,Reference Himmelmann, Horber and De La Cruz47]. The mechanisms involved in the blood flow redistribution include peripheral vasoconstriction, triggered by a carotid chemoreflex, and endocrine factors and then maintained or subsequently modified by endocrine and local components [Reference Himmelmann, Horber and De La Cruz47,Reference Morrison, Gardner, Fletcher, Bloomfield and Giussani48,Reference Jones, Sheldon, Peeters, Makowski and Meschia49]. The critical relationship between blood pressure and CBF has been categorized in the experimental model. Thus, with initial arterial hypoxemia, fetal vascular resistance can decrease by at least 50% to maintain CBF with no or a minimal decrease in cerebral oxygen delivery [Reference Koehler, Jones and Traystman50,Reference Haataja, Mercuri and Guzzetta51,Reference Johnson, Palahniwk and Tween52]. Critical to this state is a normal or elevated mean arterial blood pressure. However, persistent, severe hypoxemia leads to hypotension, and cerebral vascular resistance cannot decrease further, resulting in a marked reduction in CBF and a high likelihood of ischemic cellular injury [Reference Volpe45,Reference Romeo, Ricci and van Haastert53].

Figure 1.5

Fetal circulatory response to interruption of placental blood flow.

Figure 1.5 Long description

The flowchart begins with Cardiovascular Response to Interruption of Placental Blood Flow. This leads to Redistribution of Cardiac Output. Two outcomes are shown next. One path shows Increased Brain, Coronary, Adrenal Blood Flow. The other path shows Decreased Renal, Intestinal, Hepatic Blood Flow. These paths lead to Ongoing Progressive Asphyxia. The next step shows Decreased Cardiac Output and Loss of Autoregulation. The final step shows Decreased Cerebral Blood Flow.

Respiratory Breathing Patterns in Response to Asphyxia

In addition to the cardiovascular changes that occur with asphyxia, characteristic changes in breathing patterns occur. Critical to understanding the relationship between respiratory and circulatory changes is the work of Dawes et al. (Figure 1.6) [Reference Dawes, Jacobson, Mott, Shelley and Stafford46,Reference Dawes54]. In rhesus monkeys, asphyxia was induced by umbilical cord ligation and covering the head with a small bag of warm saline with a characteristic series of changes seen. Within 30 s of total asphyxia, a brief period of rapid rhythmic respiratory effort occurred. This culminated in apnea (primary) and bradycardia, which lasted for approximately 30–60 s (Figure 1.6). The animal then began to have gasping respirations, but spontaneous regular respiration could be induced via prompt physical stimulation and PPV after reoxygenation. If no intervention was performed, the gasping lasted for approximately 4 min. It gradually became weaker until a terminal “last gasp” occurred. This was deemed to be secondary apnea with a heart rate (HR) invariably <60 bpm and pH < 7.00. Unless resuscitation was initiated, death followed. These changes appeared to be pH related, that is, when the arterial pH was >7.17, asphyxiated monkeys were initially able to maintain rhythmic respirations followed by apnea and gasping respirations. By contrast, when the arterial pH was <7.14 there was no onset of rhythmic respiration; rather, the monkey progressed directly to gasping. Notably for every 1-min delay in initiating resuscitation at this stage, it took approximately 2 min before gasping was observed and 4 min until respirations became apparent [Reference Dawes54]. In this regard a nonbreathing newborn observed immediately following delivery is likely to be in primary apnea because in most cases they respond to stimulation and PPV with a rapid increase in HR. For those in secondary apnea with pH < 7.0, the likelihood of cardiopulmonary resuscitation (CPR) is markedly increased.

Figure 1.6

Respiratory and circulatory responses to asphyxia. Note with primary apnea the arterial pH is >7.0 and recovery in most cases will respond to stimulation and rarely positive pressure ventilation, whereas at the point of secondary apnea the heart rate is often <60 bpm and the arterial pH < 7.00. In this situation resuscitation is more intensive and often will require intubation ± CPR.

Figure 1.6 Long description

The three graphs have a shared horizontal axis labelled Time From Onset of Asphyxia in minutes, scaled from 0 to 20 in increments of 5. At the top of the figure, text lists P subscript C O 2 values of 45, 100, 150, 200, and 40, with corresponding pH values of 7.3, 7.0, 6.8, 6.75, and 7.1. The upper graph is labelled Gasps per minute. The vertical axis is scaled from 0 to 6 in increments of 2. The tracing begins with a sharp decrease labelled Primary Apnoea. This is followed by a moderate increase in gasping activity labelled Last Gasp. The tracing then reaches zero during a phase labelled Secondary or Terminal Apnoea. The tracing increases again at a point labelled Onset of Gasping. The middle graph is labelled Beats per minute and Heart Rate. The vertical axis is scaled from 50 to 200 in increments of 50. The tracing begins near 200, decreases, and continues at lower values. A point along the tracing is labelled Resuscitation, after which the tracing rises toward higher values. The lower graph is labelled Blood Pressure, with units shown as m m H g. The vertical axis is scaled from 20 to 60 in increments of 20. The tracing initially increases, then decreases to lower values, and later increases again following resuscitation. Along the horizontal axis beneath this graph, a segment is labelled Brain Damage, extending approximately from 10 to 15 minutes. Values are estimated.

Noncirculatory Responses to Asphyxia

Several biologic factors aid in preserving critical organ viability during and following asphyxia. Cerebral metabolic rate is lower in the fetus than in the term infant or adult, creating a more favorable ratio of energy supply and demand. Additionally, the neonatal brain has the capacity to use alternate energy sources when needed [Reference Romeo, Cioni, Scoto, Pizzardi, Romeo and Guzzetta55]. During relative oxygen and glucose depletion, energy substrates such as lactate and ketones become critical for cerebral metabolism [Reference Yager, Heitjan, Towfighi and Vannucci56,Reference Dawes, Mott and Shelley57]. The fetal and neonatal myocardium is more resistant to HI than the adult myocardium [Reference Thompson58]. In addition to the brain and heart, fetal hemoglobin may also allow for a greater tolerance to a hypoxic environment [Reference Wimberley59], because at low oxygen tensions (i.e., below a “crossover pO₂”), a left-shifted fetal hemoglobin–oxygen dissociation curve may be advantageous in delivering more oxygen to tissues [Reference Aberman60,Reference Romeo, Cowan and Haataja61].

Pathophysiology of Acute/Subacute Hypoxic–Ischemic Brain Injury

When the compensatory mechanisms described above are overwhelmed and CBF can no longer meet demand, a cascade of biochemical events ensues. These events are complex, interrelated, and will ultimately lead to cell death without intervention. With compromised CBF, oxygen delivery is reduced, there is a switch to anaerobic glycolysis with a reduction in high-energy phosphate compounds [i.e., adenosine triphosphate (ATP) and phosphocreatinine]. Lactic acid accumulates and membrane ion pumps fail (Na⁺/K⁺ ATPase and Na+/Ca₂⁺ exchanger). With membrane pump failure, sodium and water influx into cells, leading to cell swelling. Calcium also flows into cells, which initiates the release of excitatory amino acids such as glutamate into the extracellular space. This overexcitation leads to more calcium influx, fostering an excitotoxic cycle [Reference Romeo, Cowan and Haataja62,Reference Maitre, Burton and Duncan63]. Further consequences include formation of free radicals, production of nitric oxide, and lipid peroxidation of cell membranes. The endpoint of cell death is classically described to occur via necrosis or apoptosis. Necrosis represents a rapid and severe breakdown of cellular function that occurs with the primary HI insult [Reference Romeo, Cowan and Haataja62,Reference Maitre, Burton and Duncan63,Reference te Velde, Tantsis and Novak64]. Following resuscitation, cerebral perfusion and oxygenation is restored, along with partial restoration of energy sources over 30 min, followed by a latent phase in which the patient has near-normal oxidative metabolism. However, there is a subsequent progressive decrease in high-energy phosphates from 24 to 48 h later, that is, secondary energy failure [Reference Romeo, Cowan and Haataja62,Reference Vannucci, Towfighi and Vannucci65,Reference McNamara, Scott, Boyd, Webb, Taifalos and Novak66]. During this state, reperfusion injury occurs due to extended reactions from the primary insult, characterized by inflammation, generation of reactive oxygen species and free radicals, and eventual apoptotic cell death. Apoptosis may be induced through caspase-dependent or gene transcription (caspase-independent) processes [Reference te Velde, Tantsis and Novak64]. Because of its delayed nature, preventing apoptosis has become an enticing target for therapeutic hypothermia (TH) and other therapies [Reference Romeo, Cowan and Haataja62,Reference Maitre, Byrne and Duncan67].

Pathologic Brain Injury following Perinatal Asphyxia in Term Infants

Brain injury following interruption of PBF (asphyxia) is HI in nature and is seen in characteristic locations at autopsy or upon MRI. The major neuropathologic types of injury in the term infant include selective neuronal necrosis (SNN) and parasagittal cerebral injury. Multi-/focal injury is a rare form attributed to HI. Three basic patterns of SNN are described. First, diffuse neuronal injury occurs with very severe and prolonged insults with involvement of hippocampus, perirolandic cortex, deep nuclear structures (caudate, putamen, globus pallidus, and thalamus), and brainstem (cranial nerve nuclei, pons, and inferior olivary nuclei). Second, a cerebral cortical – deep nuclear neuronal – predominance occurs with moderate to severe insults. Deep nuclear involvement includes BG, especially the putamen and thalamus. Third, deep nuclei–brainstem neuronal predominance occurs with severe, relatively abrupt insults (e.g., abruption placentae). Parasagittal cerebral injury refers to a lesion of cerebral cortex and subcortical white matter that occurs involving bilateral parasagittal and superomedial aspects of the cerebral convexities. In the more severe injury this extends to the lateral cerebral convexity, particularly in the parieto-occipital regions [Reference Rainaldi and Perlman33,Reference Romeo, Cowan and Haataja62].

MRI

MRI imaging has proved invaluable in diagnosing brain injury in infants with HIE following intrapartum HI [Reference Morgan, Romeo and Chorna68,Reference Ryll, Wagenaar, Verhage, Blennow, de Vries and Eliasson69,Reference Ouwehand, Smidt and Dudink70]. The major types of brain injury attributed to HI are illustrated in Figures 1.7–1.9. Heterogeneous patterns are common [Reference Steinman, Gorno-Tempini and Glidden71], as elements of more than one of these patterns are frequently appreciated. Partial lesions may also be found, as described in a recent study where 10 infants that underwent TH had isolated bilateral hippocampal injury [Reference Guttmann, Flibotte and DeMauro72]. MRI findings associated with poor outcome include involvement of the BG and thalamus, posterior limb of the internal capsule, and loss of gray–white matter differentiation [Reference Morgan, Romeo and Chorna68,Reference Ryll, Wagenaar, Verhage, Blennow, de Vries and Eliasson69,Reference Ouwehand, Smidt and Dudink70].

Figure 1.7

There is extensive confluent restricted water diffusion on the diffusion-weighted imaging (DWI) sequences involving the basal ganglia with corresponding hypointensity on the apparent diffusion coefficient (ADC) maps.

Figure 1.7 Long description

Panel A shows an axial brain scan with increased signal intensity in the bilateral basal ganglia regions relative to surrounding brain tissue. The ventricles are visible centrally, and cortical structures are visible peripherally. The basal ganglia appear as symmetric areas of higher signal. Panel B shows an axial brain scan at the same level. The bilateral basal ganglia regions appear darker relative to the surrounding brain tissue. The ventricular system and cortical outlines are visible, with symmetric low-signal areas in the basal ganglia compared with panel A.

Figure 1.8

Diffuse cerebral edema involving the entire bilateral cerebral cortex and deep gray matter that demonstrate profound confluent diffusion hyperintensity and corresponding ADC hypointensity. Mass effect on the brain is noted with swelling of the cortical gyri and diffuse sulcal effacement.

Figure 1.9

There is restricted diffusion in the bilateral medial temporal lobes/hippocampus.

Clinical and Biomarkers Used to Identify Infants at High Risk for Evolving HIE

CPR in the DR

A study of 39 infants who received CPR in the DR demonstrated the importance of an associated cord arterial pH < 7.00 (Figure 1.10). All term infants who received CPR with a cord pH > 7.0 had a normal neurologic course, whereas infants who received CPR with a cord pH < 7.0 all had an abnormal neurological examination including HIE and seizures [Reference Perlman and Risser74]. In another study, 47 infants with a cord arterial pH < 7.00 were admitted to intensive care of whom 37 (95%) exhibited a benign neurologic course and two had mild hypotonia [Reference Toldo, Varishthananda and Einspieler75]. Eight infants progressed to severe brain injury with signs of HIE and seizures (Figure 1.11). These infants were 234-fold more likely to have received CPR in the DR as compared to infants with a benign neonatal course (Figure 1.11) [Reference Toldo, Varishthananda and Einspieler75].

Figure 1.10

Intensive resuscitation in the delivery room. Note that 39 infants were administered cardiopulmonary resuscitation [26].

Figure 1.10 Long description

The chart begins with Live Births (n equals 30898). This leads to Chest Compressions or Epinephrine, with n equals 39 (0.12 percent). From this point, three paths are shown. One path is labelled Term n equals 15 and ends at Severe Foetal Acidemia with n equals 10. A second path ends at Improper E T T Placement with n equals 5. A third path ends at Preterm n equals 24 and leads to Ineffective or Improper Ventilator support with n equals 24.

Figure 1.11

Infants born in the presence of fetal acidence (cord arterial pH < 7.00) and admiitted to intensive care. Note that 8 presented with seizures. Note that 6/8 of these infants required CPR. The odds ratio estimate for developing seizures with CPR was 234.

Figure 1.11 Long description

The classification chart begins with Severe Foetal Acidemia at the top. Two paths are shown below. One path is labelled Seizures asterisk with n equals 8. From this group, three outcome categories are shown: Death with n equals 5, Abnormal with n equals 2, and Normal with n equals 1. The second path is labelled No Seizures with n equals 39. From this group, two outcome categories are shown: Normal with n equals 37 and Abnormal with n equals 2. At the bottom of the figure, a line of text reads 6 of 8 infants required C P R in the D R. The O R estimate for developing seizures following.

Subacute HI Injury Presenting at Birth

From the preceding section infants at high risk for acute HIE can be identified based on evidence of a sentinel event and/or FHR abnormalities during labor plus a combination of postnatal markers, that is, cord arterial pH < 7.00, prolonged ± CPR, and a low Apgar score. However, a recent study a group of infants delivered with severe encephalopathy were considered to have suffered a subacute insult based on the following characteristics: absence of an acute sentinel event, often with decreased fetal movements for more than 6 h before delivery and with severe encephalopathy immediately following birth without the need for CPR and often with a cord arterial pH > 7.00 [Reference Kasdorf, Grunebaum and Perlman81]. Compared to infants who presented with acute interruption of PBF (placental abruption), infants with subacute injury were significantly less likely to deliver via emergent CS, required less CPR in the DR, and exhibited a higher cord arterial pH and BD compared with the acute group. Infants with subacute injury were more likely to present with severe rather than moderate encephalopathy (Table 1.2). An abnormal outcome was universal with subacute injury as opposed to 27% in those infants with an acute insult (p = 0.0001).

Table 1.2Clinical characteristics of infants with subacute versus acute injury

Table 1.2 Long description

The table has four columns: Characteristic, Subacute (n equals 7), Acute (n equals 26), and p-Value. It reads as follows. Row 1. Emergent C S: 3 (43 percent); 23 (88 percent); 0.02. Row 2. C P R or epinephrine: 1 (14 percent); 18 (69 percent); 0.02. Row 3. Cord arterial p H: 6.99 plus minus 0.19; 6.74 plus minus 0.15; 0.002. Under Sarnat staging: Row 4. Moderate: 1 (14 percent); 15 (58 percent); 0.0001. Row 5. Severe: 6 (86 percent); 11 (42 percent); 0.08. Row 6. Time to start T H (hours): 4.05 plus minus 0.89; 4.25 plus minus 1.060; less than 0.0001. Row 7. Abnormal outcome: 7 (100 percent); 7 (27 percent); less than 0.0001.

CS, cesarean section; TH, therapeutic hypothermia.

Clinical Examination

The Sarnat staging is frequently used for assessing the extent of encephalopathy (Table 1.3) [Reference Sarnat and Sarnat91]. In the original description the maximum staging of encephalopathy was determined in the first 2–3 days of life. This has been compressed to within the first 6 h for the purposes of administering TH. Thus, this represents a sicker group of infants than originally described. We consider the level of consciousness critical to the assessment of encephalopathy. The presence of lethargy is consistent with moderate encephalopathy and is invariably accompanied by other signs, for instance, decreased tone or weak suck. The infant becomes a strong candidate for TH if the other criteria mentioned previously are present. A stuporous/comatose infant should be cooled expeditiously in the absence of any contraindication. Infants with mild encephalopathy should be monitored repeatedly within the 6-h time frame for evolution to moderate to severe encephalopathy.

Amplitude Electroencephalograph

An amplitude electroencephalograph (aEEG) was used as an entry criterion for infants who received selective head cooling. aEEG recordings are characterized as high voltage (normal) (15–25 µV), moderate suppression (upper > 10 and lower ~5 µV), severe suppression (entire band width <5 µV) (Figure 1.12) [Reference al Naqeeb, Edwards, Cowan and Azzopardi91].

Figure 1.12

Representative aEEG tracings that are characterized as high voltage (normal) (15–25 µV), moderate suppression (upper > 10 and lower ~5 µV), severe suppression (entire band width <5 µV) [Reference Shankaran, Laptook and Ehrenkranz94].

TH as a Neuroprotective Strategy to Reduce Evolving HI Injury

TH has been shown to be effective therapy for NE secondary to HI. Multiple randomized trials have demonstrated that reductions in core temperature either alone or in combination with reduced head temperature [Reference Gluckman, Wyatt and Azzopardi93] consistently reduce death or disability at 18 months [Reference Shankaran, Laptook and Ehrenkranz94,Reference Azzopardi, Strohm and Edwards95,Reference Jacobs, Morley and Inder96]. A Cochrane meta-analysis indicated that TH reduced death or major neurodevelopmental disability in survivors from 61% in noncooled infants versus 46% in infants treated with hypothermia, yielding a risk ratio of 0.75 (95% CI 0.68 to 0.83),with number needed to treat (NNT) for an additional beneficial outcome of 7 (95% CI 5 to 10) [Reference Shevell7]. A second similar meta-analysis showed a beneficial effect in infants that extended to those with moderate encephalopathy (RR, 0.67; 95% CI 0.56 to 0.81) and those with severe encephalopathy (RR 0.83; 95% CI 0.74 to 0.92). The NNT to prevent 1 is 6 for infants with moderate encephalopathy and 7 for those with severe encephalopathy [Reference Tagin, Woolcott, Vincer, Whyte and Stinson97]. The neuroprotective effects of TH persisted at 6–7 years (RR 0.77; 95% CI 0.63 to 0.94) [Reference Shankaran, Pappas and McDonald98,Reference Azzopardi, Strohm and Marlow99]. Given the beneficial effects of hypothermia, the Committee on the Fetus and Newborn of the American Academy of Pediatrics provided an overview of the available data and expectations for centers that provide this therapy [Reference Papile, Baley and Benitz100].

Ventilation

Changes in carbon dioxide (CO₂) are important to monitor carefully as hypercarbia increases and hypocarbia decreases CBF [Reference Kasdorf, Perlman, Perlman and Inter101,Reference Rosenberg, Jones, Traystman, Simmons and Molteni102]. In term infants, several studies have demonstrated an association of hypocarbia (typically <35 mmHg) and adverse ND outcome especially in the setting of HIE [Reference Klinger, Beyene, Shah and Perlman103,Reference Pappas, Shankaran and Laptook104,Reference Lingappan, Kaiser, Srinivasan and Gunn105]. A retrospective cohort study of 198 term infants with moderate to severe HIE treated with TH showed an association of lowest PaCO₂ averaged over days 1–4 with identification of brain injury on MRI [OR (95% CI) 1.07 (1.00 to 1.14); p = 0.04] [Reference Lopez Laporte, Wang and Sanon106]. It is recommended that CO₂ be maintained in the normal range in mechanically ventilated infants at risk for HIE. This goal may be clinically difficult to achieve as infants with HIE often demonstrate hypocapnia, which may reflect the underlying CNS disease[Reference Nadeem, Murray, Boylan, Dempsey and Ryan107].

Fluid Status

Infants who have suffered HI often progress to a fluid overload state. This may reflect renal failure secondary to acute tubular necrosis or to the syndrome of inappropriate antidiuretic hormone release (Figure 1.5). This often presents with increase in weight, low urine output, and hyponatremia. Although fluid restriction is common practice, a randomized controlled trial of neonates with HIE treated with TH showed no difference in death or ND disability at 6 months between infants randomized to normal fluid versus restricted fluid intake (defined as 2/3 of normal intake) in the first 4 days of life [Reference Tanigasalam, Plakkal, Vishnu Bhat and Chinnakali108]. Thus, fluid and electrolyte administration should be individualized.

Maintenance of Adequate Perfusion

Given the high likelihood of a pressure-passive cerebral circulation, management should aim to maintain the arterial blood pressure within a normal range for age and gestation. Hypotension is not uncommonly related to myocardial dysfunction, endothelial cell damage, or rarely to volume loss. Treatment should be directed toward the cause, that is to say: inotropic support for myocardial dysfunction, and volume replacement for intravascular depletion [Reference Wyckoff, Perlman and Laptook109].

Renal Function

Acute renal injury is common in infants in the setting of HIE (see above) and appears to be modulated by TH. A 2021 meta-analysis showed a significant difference between the incidence of acute renal injury in infants treated with TH versus controls (RR = 0.81, p = 0.03) [Reference van Wincoop, de Bijl-Marcus, Lilien, van den Hoogen and Groenendaal110]. In a randomized controlled trial of infants treated with TH versus standard treatment, the incidence of acute renal injury was approximately 50% less with TH versus standard treatment (p < 0.05) [Reference Tanigasalam, Bhat, Adhisivam and Sridhar111]. The mechanism of this potential protective effect is not entirely understood.

Control of Blood Glucose Concentration

Experimental studies suggest that both hyperglycemia and hypoglycemia may exacerbate brain damage. In adult experimental models of HI, hyperglycemia accentuates brain damage, whereas in immature rodents significant hyperglycemia (blood glucose 600 mg/dL) entirely prevented HI brain damage [Reference Vannucci and Mujsce112]. The effects of hypoglycemia in experimental neonatal models vary, as do the mechanisms of inducing hypoglycemia. Thus, insulin-induced hypoglycemia is detrimental to the immature rat brain subjected to HI. However, with fasting-induced hypoglycemia, a high degree of protection was noted, and may be secondary to increased concentrations of ketone bodies, presumably serving as alternative substrate to the immature brain [Reference Yager113]. Clinically, hypoglycemia when associated with HI appears detrimental to brain. Thus, term infants delivered in the presence of severe fetal acidemia (umbilical arterial pH < 7.0) who presented with an initial blood glucose concentration <40 mg/dL versus >40 mg/dL were 18 times more likely to progress to moderate to severe encephalopathy [Reference Salhab, Wyckoff, Laptook and Perlman114]. Post-hoc analysis of the Cool Cap Study noted an unfavorable outcome at 18 months more often in infants with hypoglycemia (≤40 mg/dL) and hyperglycemia (≥150 mg/dL) within the first 12 h following randomization [Reference Basu, Kaiser and Guffey115]. After adjusting for Sarnat stage and 5-min Apgar score, hyperglycemic infants randomized to TH had a significant reduced risk of death and/or severe disability at 18 months (adjusted risk ratio 0.8; 95% CI 0.66 to 0.99); hypoglycemia or normoglycemia infants did not show any benefit. Derangements in glucose levels have also been associated with varying patterns of brain injury on MRI including when treated with TH. Thus, infants with hypoglycemia had higher adjusted odds of watershed or focal–multifocal injury, whereas infants with labile glucose levels had predominantly basal ganglia or global injury on MRI compared to infants with normal glucose values [Reference Basu, Ottolini and Govindan116].

Seizure Management

HI cerebral injury is a common cause of early-onset neonatal seizures. Experimental evidence strongly suggests that repetitive seizures disturb brain growth and development as well as increase the risk for subsequent epilepsy [Reference Dzhala, Ben-Ari and Khazipov117,Reference Holmes, Gairsa, Chevassus-Au-Louis and Ben-Ari118]. Human studies show conflicting evidence. Glass et al. found that clinical seizures in the setting of HIE pre-TH were associated with worse cognitive and motor outcome at age 4 years [Reference Glass, Glidden, Jeremy, Barkovich, Ferriero and Miller119]. The Cool Cap study also demonstrated that aEEG seizures at time of enrollment were independently associated with an unfavorable outcome [Reference Gluckman, Wyatt and Azzopardi120]. By contrast, the presence of clinical seizures at any time during hospitalization was not associated with death or moderate to severe disability at 18 months in a secondary analysis of the Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) whole-body cooling trial [Reference Kwon, Guillet and Shankaran121]. The optimal management of neonatal seizures in the setting of HIE treated with TH remains unclear [Reference van Rooij, Hellstrom-Westas and de Vries122,Reference DeLaGarza-Pineda, Mailo and Boylan123]. Clinical and/or EEG seizures are invariably treated with an anticonvulsant, commonly phenobarbital as a first-line agent, and treatment is continued if seizures persist until the anticonvulsant therapy has been optimized

Adjuvant Neuroprotective Agents

Multiple alternative and adjuvant agents that target various mechanisms in the cascade of events that lead to neuronal necrosis and apoptosis have been evaluated without much success. Melatonin has been evaluated in experimental and human studies [Reference Robertson, Faulkner and Fleiss124]. A recent meta-analysis concluded that clinical data supporting the neuroprotective effects in neonates are limited [Reference Ahmed, Pullattayil, Robertson and More125]. Erythropoietin has been shown in experimental rodent models to reduce the extent of brain injury [Reference Juul, McPherson and Bammler126]. The drug was subsequently evaluated in a large multicenter, double-blind, randomized, placebo-controlled trial in conjunction with TH versus a placebo group where the incidence of death or neurodevelopmental impairment was comparable between treatment groups. Moreover, the number of serious adverse events per child was significantly higher in the erythropoietin versus the placebo group. Currently there are no adjunctive therapies that enhance the neuroprotection of TH.

Chorioamnionitis

Case Presentation

The infant was a 3,155 g 40 1/7-week female delivered via CS secondary to arrest of descent. There was PROM of 36 h duration. The infant emerged meconium stained, depressed, hypotonic with a HR at 1 min < 100 bpm. PPV was initiated with improvement in HR >100 bpm. The infant was intubated for lack of respiratory effort. Apgar scores were 1 and 8 at 1 and 5 min, respectively. Initial temperature was 38.2°C. No cord blood gas was available. Initial blood gas revealed a pCO₂ 40, pH 7.23, BD –10 mEq/L. She was extubated within 12 h. No blood pressure issues were noted. She was started on antibiotics. Seizures developed at approximately 36 h of life. An EEG revealed seizures emanating from the right frontocentral region with continuous diffuse slowing. She was treated with anticonvulsants. An EEG 3 days later showed improved continuity. A spinal tap was unremarkable. An MRI showed multiple infarcts (Figure 1.13). Magnetic resonance venography (MRV) was normal. Placental pathology was consistent with acute CA and funisitis. She was discharged home on phenobarbital on day of life 13.

Figures 1.13

Multifocal areas of cortical/subcortical T2 hyperintensity (A and B) with corresponding ADC-restricted diffusion (C and D) are identified throughout the supratentorial region in multiple vascular distributions most prominent within the posterior perisylvian frontoparietal junctions, right greater than left. These are most compatible with multiple acute infarctions.

Summary

1. 1. Several important issues are raised. The infant had an elevated temperature at birth, and exhibited BD, was hypotonic, developed seizures at 36 h of life, and had bilateral symmetric lesions on MRI in a distribution distinct from that noted with HI injury. There was placental evidence of CA and funisitis.

In the sections below, the relationship between clinical CA and maternal fever and neonatal neurologic signs including BD, NE, and seizures will be explored, and the potential link of maternal fever and cytokine release to subsequent NE will be reviewed.

Hyperthermia and Neonatal Brain Injury

Hyperthermia in an infant is defined as a rectal or axillary temperature >37.5°C, usually secondary to an elevated maternal temperature. Importantly the fetus must maintain a temperature gradient of approximately 0.5°C higher than the mother to dissipate heat across the placenta [Reference Kasdorf and Perlman131]. Hyperthermia may induce a relative hypermetabolic state [Reference Kasdorf and Perlman131,Reference Perlman132] that can be deleterious to neonatal brain [Reference Baumgart133,Reference Sedin, Martin and Walsh MC134,Reference Slinko, Caspersen and Ratner135]. The increase in metabolic demand is associated with a simultaneous increase in CBF, which may exacerbate the extent of reperfusion injury following prior hypoxia–ischemia [Reference Busija, Leffler and Pourcyrous136,Reference Inder, Volpe and Volpe137]. Moreover, an elevated temperature may alter blood–brain barrier permeability, promoting cerebral edema [Reference Natah, Srinivasan, Pittman, Zhao and Dunn138]. The association of hyperthermia and brain injury may be related to several mechanisms including free radical production, glutamate and glycine release, and increased metabolic demands with depletion and subsequent slow recovery of phosphocreatine and adenosine triphosphate levels (Figure 1.14) [Reference Kasdorf and Perlman131]. In addition, there is a clear link between hyperthermia and cytokine release as well as hemostatic changes including thrombosis, which may lead to adverse neurologic outcome [Reference Bouchama, Roberts and Al Mohanna139]. In a baboon experimental model subjected to heatstroke, a significant increase in prothrombin time, activated partial thromboplastin time, and decreases in D-dimer, fibrinogen, and platelets was noted [Reference Leon, Blaha and DuBose140]. Additionally cytokine release with a marked elevation of interleukin (IL)-6 was demonstrated compared to controls [Reference Leon, Blaha and DuBose140], and in mice exposed to heat strain significant elevations in IL-6, IL-10, and IL-1b were noted [Reference Leon and Helwig141].

Figure 1.14

Potential pathways to neuronal injury with hyperthermia (temperature > 37.5°C).

Figure 1.14 Long description

The flowchart begins with Hyperthermia (Infant Temperature greater than 37.5 degrees Celsius). From this, multiple parallel pathways are shown. One pathway shows Vascular Endothelial Injury. A second pathway shows Increased Blood Brain Permeability, which proceeds to Loss of Autoregulation and then to Cerebral Edema. A third pathway shows Increased Metabolic Demand, which proceeds to Increased Cerebral Blood Flow and then to Reperfusion Injury. A fourth pathway shows Cytokine Free Radical Glutamate Release. All pathways converge at the bottom of the diagram at Neuronal Injury.

The Association of Maternal Fever and BD, NE, and Seizures

BD

A clinical study demonstrated a significant relationship between maternal fever (temperature > 38°C), BD (5-min Apgar < 5) and need for assisted ventilation ± chest compressions [Reference Perlman142]. In a second study term infants born to mothers with clinical CA whose infants had a higher rectal temperature at 30 min, were 3.9-fold more likely to have BD (requiring PPV for >2 min, and/or intubation) and were 2.8-fold more likely to be admitted to the NICU [Reference Shalak, Perlman, Jackson and Laptook143].

Maternal fever and NE

Numerous studies have assessed maternal fever and subsequent NE. The risk for NE ranged from 3.1- to 8-fold among studies. The risk was highest with associated acidosis (see Table 1.4) [Reference Impey, Greenwood, MacQuillan, Reynolds and Sheil144,Reference Blume, Li, Loch and Koepsell145,Reference Impey, Greenwood, Black, Yeh, Sheil and Doyle146].

Table 1.4Association of maternal fever with the risk for neonatal encephalopathy

Table 1.4 Long description

The table has three main columns: Study, Maternal Fever greater than 37.5 degrees Celsius, and Risk for N E. It reads as follows. Row 1. Study 144: yes; increased 4.7-fold. Row 2. Study 145: yes; increased 5.4-fold. Row 3. Study 146 asterisk: yes; increased 8.7-fold.

*Associated acidosis.

Maternal Fever and Neonatal Seizures

Several studies have described seizures in association with intrapartum maternal fever [Reference Lieberman, Lang, Richardson, Frigoletto, Heffner and Cohen147,Reference Lieberman, Eichenwald, Mathur, Richardson, Heffner and Cohen148,Reference Glass, Pham, Danielsen, Towner, Glidden and Wu149]. In one study, infants with seizures were threefold more likely when born to mothers with fever >100.4°C during labor even after controlling for numerous factors, such as epidural use [Reference Lieberman, Eichenwald, Mathur, Richardson, Heffner and Cohen148]. In another study, maternal fever was associated with a significant 2.9-fold adjusted increased risk for seizures [Reference Glass, Pham, Danielsen, Towner, Glidden and Wu149].

Cytokine Responses in Infants with Clinical CA and Associated HIE/NE

The association between the inflammatory cytokine response [IL-6, IL-8, and regulated on activation, normal T-cell expressed and secreted (RANTES) in cord blood, at 6 and 30 h] in term infants born to mothers with clinical CA and BD, abnormal neurologic examination, and HIE was assessed [Reference Shalak, Laptook, Jafri, Ramilo and Perlman130]. Cord cytokine concentrations were higher in infants with CA than in controls. IL-6 increased at 6 h, followed by a fivefold decline by 30 h. In contrast, IL-8 and RANTES progressively decreased over time. There was no relationship between cytokines and BD, but the neurologic examination correlated significantly with IL-6 at 6 h (r < 0.5). All cytokine concentrations were significantly higher in infants with HIE/seizures versus none at 6 h of age. There were no differences in cord or postnatal concentrations for any of the cytokines evaluated between infants with and without HCA. In a second study, serum neuronal glial fibrillary acidic protein (GFAP), ubiquitin carboxyl-terminal hydrolase L1(UBCH), and inflammatory biomarkers (cytokines measured in serum from umbilical artery at 6–24, 48, 72 h) in term newborns with mild, moderate, or severe HIE who received TH was undertaken to determine whether measurements could stratify severity of HIE. At birth, serum GFAP and UBCH increased significantly with the severity of HIE, and serial GFAP remained elevated in neonates with MS HIE. IL-6, IL-8, and vascular endothelial growth factor (VEGF) were significantly greater at 6–24 h with MS HIE. Serial values were unaffected by TH or rewarming. Elevated GFAP, IL-1, IL-6, IL-8, tumor necrosis factor (TNF), interferon (IFN), and VGEF at 6–24 h were associated with abnormal neurological outcomes at 15–18 months [Reference Chalak, Sánchez, Adams-Huet, Laptook, Heyne and Rosenfeld150]. It was concluded that both neuronal and inflammatory biomarkers were elevated in infants with MS encephalopathy. A third prospective study evaluated measures of 10 pro- and anti-inflammatory cytokines including IL-1α, IL-1β, IL-6, IL-8, IL-10, TNF-α, IFN-γ, VEGF, and granulocyte macrophage/colony-stimulating factor (GM-CSF) from term infants requiring resuscitation at birth. Infants with NE and abnormal neuroimaging had significantly elevated GM-CSF at 0–24 h and IL-8, IL-6, and IL-10 at 24–48 h. Elevated IL-10 levels were associated with mortality [Reference Hare, Watson and O’Neill151]. In another study from the same investigator, infants with NE who received TH had significantly higher levels of erythropoietin (Epo), IL-6, and IL-1ra versus controls at baseline. Following lipopolysaccharide (LPS) stimulation an excessive Epo and IL-1ra response was noted. By contrast, IL-8, IL-2, IL-6, TNF-α, GM-CSF, VEGF, and IFN-γ were significantly lower than controls following LPS stimulation. GM-CSF, IFN-γ, IL-1β, IL-1ra, and VEGF were higher on days 1–2 in NE infants with abnormal neuroimaging [Reference O’Dea, Kelly and McKenna152].

Summary

1. 1. A host of cytokines are released to varying degree in infants born to mothers with clinical CA and play an important role in the evolution of MS encephalopathy.

Cytokine Storm

The above observations raise the possibility of whether the increased release of cytokines in infants born to mothers with CA and subsequent NE may reflect a “cytokine storm release,” akin to that described in adults [Reference Fajgenbaum and June153]. It is an umbrella term characterized by constitutional symptoms, systemic inflammation, and multiorgan dysfunction that can lead to multiorgan failure if inadequately treated. Laboratory findings include nonspecific markers of inflammation such as C-reactive protein and prominent elevations in serum inflammatory cytokine levels, as described above [Reference Fajgenbaum and June153].

Effects of the Inflammatory Response on CBF

SDH

A recent report indicates that SDH has become the most common lesion identified in newborn infants undergoing MRI with an occurrence of 0.34/1,000 term infants [Reference Whitney, Kamdar, Ng, Hurvitz and Peterson27]. The clinical expression varies depending on the size and location of the blood. The commonest presentation is unexplained respiratory depression in the DR, apnea, or seizures. Other signs include lethargy, hypotonia, or rarely a full fontanel [Reference Inder, Perlman, Volpe and Volpe167,Reference Chamnanvanakij, Rollins and Perlman170]. However, depending on location, in particular within the posterior fossa, even small amounts of blood sometimes result in obstructive hydrocephalus or striking neurological deficits [Reference Inder, Perlman, Volpe and Volpe167]. Chronic SDH over the convexity usually presents with macrocephaly, seizures, obstructive hydrocephalus, or striking neurological deficit.

SAH

SAH can be categorized as either primary or secondary in origin. Primary SAH refers to hemorrhage in the subarachnoid space most frequently from rupture of small veins bridging the leptomeninges [Reference Inder, Perlman, Volpe and Volpe167,Reference Perlman171]. Secondary SAH is due to extension of subdural, intraventricular, or intraparenchymal hemorrhage and occurs less often. Coagulation disorders and rupture of an intracranial aneurysm or arteriovenous malformation are occasionally responsible [Reference Inder, Perlman, Volpe and Volpe167]. The most common clinical presentation is seizures, with onset usually on the second postnatal day. In the interictal period, the infants appear remarkably well. A second presentation is that of recurrent apnea, with onset may be in the first 24 h. Rarely, a third presentation is that of more extensive bleeding into the CSF causing seizures, a bulging anterior fontanel or diminished response, or intraparenchymal hemorrhages and occurs less often [Reference Inder, Perlman, Volpe and Volpe167].

EPH

EPH is a rare lesion accounting for approximately 2% of all cases of neonatal ICH observed at autopsy [Reference Perlman171]. Bleeding is derived from branches of the middle meningeal artery or from major veins or venous sinuses. Often, there is overlying fracture. The importance of an EPH is that it can rapidly expand, resulting in acute progression of encephalopathy in the hours following delivery. Such infants need emergent surgical exploration to relieve pressure and stop bleeding.

ICH

ICH in the term infant is uncommon. Blood may be found within the germinal matrix, ventricles, or brain parenchyma (Figure 1.15). The thalamus appears to be a common site of hemorrhage in the term infant and frequently also includes IVH [Reference Trounce, Dodd, Fawer, Fielder, Punt and Levene172]. Typically the hemorrhage develops later, that is, during the second week of life, but it may occur earlier. Predisposing factors include prior HI cerebral injury, sepsis, coagulopathy [Reference Roland, Flodmark and Hill173,Reference Bray and Luban174], and arteriovenous malformations [Reference Heck, Gailloud and Cohen175], or may be observed in association with subarachnoid or SDH [Reference Perlman171]. Recent data point to an important role of SVT in the pathway to ICH. Thus, in a study by Wu et al. one-third of infants with IVH had evidence of SVT [Reference Wu, Hamrick and Miller176]. Primary IVH is rare in the term infant. The clinical presentation of ICH is characterized by sudden onset of marked neurologic abnormalities, including seizures, evidence of increased intracranial pressure, and downward eye deviation. A bulging fontanel or blood in the CSF may be present. Hemiparesis or other focal neurological deficits are uncommon. The prognosis is surprisingly good with primary thalamic hemorrhage [Reference Inder, Perlman, Volpe and Volpe167].

Figure 1.15

CT scan that demonstrates both intraventricular hemorrhage and intraparenchymal hemorrhage. The infant was initially admitted to the WBN and presented at 10 h with seizures.

Case Presentation

AA is a 3,340 g 38 6/7-week female infant. Delivery was via CS. The baby emerged with a spontaneous cry and was given oxygen and CPAP for 6 min. Apgar scores were 7 and 8 at 1 and 5 min, respectively. The cord arterial blood gas revealed a pH 7.32 pCO₂ 35 and BD 8 mEq/L. The infant was transferred to the WBN. On DOL 2 the mother noted that the infant’s cry appeared different, with poor feeding. Over a 2-h period, the infant was observed to have four episodes of generalized tonic–clonic movements of legs and arms each lasting approximately 1–2 min. She was transferred to the NICU and treated with anticonvulsants. A video EEG showed slowing but no EEG seizures. MRI revealed extensive restriction diffusion within the right parasagittal frontal lobe and right parietotemporal lobes, with correlating decreased signal on the ADC maps compatible with an acute infarction (Figure 1.16). Magnetic resonance angiography (MRA) and MRV were both negative. A thrombophilia work up was negative. No placental pathology was available.

Figure 1.16

Infant with a stroke with presentation on DOL 2 with seizures. The MRI revealed extensive diffusion restriction within the right parasagittal frontal lobe and right parietotemporal lobes, with correlating decreased signal on the ADC maps compatible with an acute infarction.

Pathogenesis

The pathogenesis of NAIS is not completely understood; however, several risk factors have been identified including maternal CA and PROM (threefold increased risk for both) [Reference Lee, Croen and Backstrand179], maternal hypertension (twofold increased risk, 180), intrapartum fever (threefold increased risk, early-onset sepsis/meningitis OR 5.8) [Reference Harteman, Groenendaal, Kwee, Welsing, Benders and de Vries181], emergency CS, male sex [Reference Srivastava, Dunbar and Shevell182], and SGA (threefold increase) [Reference Sorg, von Kries and Klemme183]. A meta-analysis of 11 studies found additional factors including preeclampsia, oligohydramnios, FHR abnormalities, and resuscitation at birth [Reference Li, Miao and Xu184]. A recent review noted that perinatal inflammation often induced by infection was the only consistent independent risk factor for NAIS [Reference Giraud, Guiraut, Chevin, Chabrier and Sébire185]. The mechanism(s) of injury remain unclear and may include elevated temperature, cytokine release associated with infection, alterations in coagulation, and a subsequent prothrombotic state associated with hyperthermia as discussed earlier [Reference Bouchama, Roberts and Al Mohanna186,Reference Leon and Helwig187]. Additional possible causes include inherited or acquired coagulopathies, trauma, and hypoxia–ischemia. However, a specific cause often cannot be identified.

Clinical Studies Evaluating Infants with NAIS

The clinical course of eight term infants with NAIS was evaluated using Doppler assessment of CBFV from the MCA and ACA during the acute phase of the disease [Reference Perlman, Rollins and Evans188]. Seven of eight patients required no DR resuscitation; one infant required brief PPV. All infants were initially admitted to the WBN. Infants were identified on the basis of abnormal clinical findings, that is, seizures (n = 6), hypotonia and apnea (n = 2) observed during the first 48 h. Neuroimaging revealed nonhemorrhagic left focal (n = 6) and right focal MCA infarction (n = 2). Transient ipsilateral decreases in CBFV as compared to the contralateral unaffected side was noted upon clinical presentation in four infants. These differences were not present at subsequent CBFV measurements. In a second study risk factors for HIE were compared to those infants with NAIS. All infants with NAIS had a negative thrombophilia work up. Infants with NAIS had a significantly higher incidence of seizures whereas infants with HIE were notable for a high occurrence of CS delivery, often secondary to a nonreassuring FHR tracing, suggesting that NAIS was not associated with perinatal risk factors that allow for early identification [Reference Adami, Grundy, Poretti, Felling, Lemmon and Graham189].

Placental Findings with NAIS

Placental examination provides a plausible argument that many cases of NAIS have associated placental changes. In a study of 46 children with NAIS, FVM occurred 4.8-fold and amniotic fluid inflammation 2.6-fold more often in NAIS versus controls. In addition, evidence of a “stress response” (meconium plus elevated nucleated red blood cells) was 31-fold more common in the NAIS compared to a control group. These placental findings reflect subacute-to-chronic intrauterine stressors and were associated with both arterial and venous stroke [Reference Bernson-Leung, Boyd and Meserve190]. A second report in 12 infants with symptomatic NAIS (n = 5) or SVT (n = 7) noted the following placental changes: thrombo-inflammatory processes (n = 6), sudden catastrophic events (n = 5), decreased placental reserve (n = 3), and a stressful intrauterine environment (n = 2) [Reference Elbers, Viero, MacGregor, DeVeber and Moore191].

Background

Glucose is an essential metabolic fuel for the brain. Central nervous function is rapidly perturbed by hypoglycemia [Reference Perlman, Volpe and Volpe199]. Under conditions where glucose is limited, alternative substates such as lactate, pyruvate free fatty acids, and a variety of keto acids, such as ketone bodies, may help spare brain function and structure [Reference Hawdon, Ward Platt and Ainsley-Green200,Reference Harris, Weston and Harding201].

Case Presentation

GG is a 2,460 g 39-week gestation SGA male infant born via emergent CS secondary to fetal decelerations. The infant emerged limp with a weak cry with HR > 100 bpm. Initial resuscitation included CPAP for about 1 min for weak respiratory effort. Apgar scores were 7 and 8 at 1 and 5 min, respectively. Cord arterial pH was 6.95, pCO₂ 77, and BD 15.2 mEq/L. He was admitted to the NICU due to mild respiratory distress requiring nasal cannula and room air. Initial lactate was 11, which normalized over 24 h. He soon developed hypoglycemia with a blood glucose < 20 mg/dL, which required four glucose boluses for correction. Within 6 h of birth, leg movements concerning for seizures were noted. He was placed on a vEEG within 12 h, which detected seizure activity, without a clinical correlate and was indicative of severe diffuse cerebral dysfunction. Anticonvulsants were initiated. The vEEG improved with eventual resolution of seizures. A spinal tap was unremarkable. He was maintained and discharged on phenobarbital. The placental pathology revealed maternal and fetal vascular malperfusion, intervillous thrombus, mild chronic multifocal villitis, and fetal membranes filled with meconium-filled macrophages. MRI brain revealed bilateral cortical restricted diffusion involving the temporal, parietal, and occipital lobes (Figure 1.17). Follow-up MRI at 3 months demonstrated evolution of cystic encephalomalacia in the temporal, parietal, and occipital lobes.

Figure 1.17

MRI brain of an infant with hypoglycemia showing bilateral cortical restricted diffusion involving the temporal, parietal, and occipital lobes (A). Follow-up evolution of cystic encephalomalacia in the temporal, parietal, and occipital lobes was noted by 3 months (B).

Relationship between Hypo-/Hyperglycemia and Brain Injury in Infants with HIE/NE

A retrospective review was undertaken to determine the potential contribution of initial hypoglycemia (defined as blood glucose ≤ 40 mg/dL) to the development of neonatal brain injury in 185 term infants with severe fetal acidemia (umbilical arterial pH < 7.00) [Reference Salhab, Wyckoff, Laptook and Perlman114]. Infants with a blood sugar ≤40 mg/dL versus >40 mg/dL were 6.3-fold more likely to have an abnormal neurologic outcome. By multivariate logistic analysis, four variables significantly associated with abnormal outcome, including: initial blood glucose (OR 18.5), cord arterial pH ≤6.90 versus >6.90 (OR 9.8), a 5-min Apgar score ≤5 versus >5 (OR 6.4), and requirement for intubation ± CPR versus neither (OR 4.7).

A second prospective cohort of term neonates evaluated the impact of hypoglycemia (glucose < 46 mg/dL) in the first 24 h on MRI changes and early neurodevelopmental outcome at one year [Reference Tam, Haeusslein and Bonifacio202]. Neonatal hypoglycemia was associated with a 3.72-fold increased odds of corticospinal tract injury and 4.82-fold increased odds of a one-point worsened neuromotor score and a 15-point lower cognitive and language score on the Bayley Scales of Infant Development. A third study investigated the association of neonatal hypoglycemia (defined as ≤40 mg/dL) and hyperglycemia (defined as >150 mg/dL) with neurocognitive outcomes in 234 infants with moderate–severe HIE who received TH. By univariate analysis unfavorable outcome was more common with hypoglycemia (81%, p = 0.004), hyperglycemia (67%, p = 0.01), and any glucose derangement within the first 12 h (67%, p = 0.002) versus normoglycemic infants [Reference Basu, Kaiser and Guffey115]. A study from the same group of investigators determined whether the early glycemic profiles in infants with HIE [defined as follows: normoglycemia (glucose concentrations of >47 to ≤150 mg/dL); hypoglycemia (≥1 glucose measurement ≤47 mg/dL); hyperglycemia (≥1 concentration >150 mg/dL)] and labile glucose (both hypoglycemia and hyperglycemia) were associated with distinct patterns of brain injury on MRI. Infants with normal versus abnormal blood glucose values had a normal brain MRI in 60% versus 32%, respectively (p = 0.02) [Reference Basu, Ottolini and Govindan116]. In particular, the odds of predominant watershed or focal–multifocal injury was sixfold higher in association with hypoglycemia and labile glucose, whereas the odds of predominant BG or global injury was 5.6-fold higher with labile glucose levels versus infants with normal blood glucose values [Reference Basu, Ottolini and Govindan116].

Further, a prospective cohort of neonates ≥36 weeks’ gestational age with NE underwent continuous glucose monitoring for 72 h, with MRI at day 4 of life, and follow-up at 18 months. Maximum glucose during the first 48 h best predicted BG AUC = 0.81) and watershed injury (AUC = 0.85) [Reference Kamino, Widjaja and Brant203]. Minimum glucose was not predictive of brain injury. At ND follow-up, a glucose threshold of >180 mg/dL during the first 48 h was associated with significantly higher child behavior problem checklist, worse neuromotor score, and higher odds for CP; this association was lost after adjusting for brain injury severity [Reference Kamino, Widjaja and Brant203]. A retrospective analysis of newborn infants with HIE found that hyperglycemia on the first day was associated with poor outcome in newborn infants undergoing TH [Reference Chouthai, Sobczak, Khan, Subramanian, Raman and Raghavendra204]. A systematic analysis evaluated the association between neonatal hypoglycemia and hyperglycemia with adverse outcome and found that children exposed to neonatal hypoglycemia had twofold higher odds of ND impairment or death and exposure to hyperglycemia was associated with a threefold higher risk of death or neurodisability at 18 months [Reference Puzone, Diplomatico and Caredda205].

Case Presentation

IB was a 2,800 g 37-week male infant born to a mother via CS secondary to maternal fever and persistent fetal tachycardia with a nonreassuring FHR. The mother was GBS negative. Rupture of membranes occurred <5 h prior to delivery. The infant emerged limp, apneic, and cyanotic with poor color and low HR. He was intubated and administered CPR for 3 min. The Apgar scores were 2, 6, and 8 at 1, 5, and 10 min, respectively, with a cord arterial pH of 7.30. In the NICU, he soon became jittery with posturing. He had no gag or suck but was breathing spontaneously. Initial postnatal pH was 7.21, with pCO₂ 35 and a BD –13 mEq/L. A complete blood count revealed a white blood count of 10,000 with 39 bands and immature to mature ratio of 0.81. The platelets were 239,000. C-reactive protein was 8.4. Antibiotics were started. He was transferred to a tertiary NICU. Examination there revealed a spontaneously breathing intubated infant, lethargic but responsive to stimulation, a weak suck and absent gag, central hypotonia with intermittent sensory-induced clonus. The assessment was moderate encephalopathy. Selective head cooling was initiated close to 6 h. Within 24 h the blood cultures were positive for Listeria monocytogenes. The placental pathology was significant for acute CA with focal acute funisitis and acute villitis with associated necrosis. A spinal tap was unremarkable. A vEEG was abnormal due to the presence of left-sided rhythmic runs of predominantly sharp activity consistent with seizures persisting over 2 days. MRI revealed multifocal areas of cortical/subcortical T2 hyperintensity with corresponding ADC-restricted diffusion identified throughout the supratentorial region in multiple vascular distributions, and most prominent within the posterior perisylvian frontoparietal junctions, right greater than left, strongly suggestive of multiple acute infarctions (Figure 1.18). MRA and MRV were normal.

Figure 1.18

MRI from an infant with acute CA who underwent TH. Multifocal areas of cortical/subcortical T2 hyperintensity (A) with corresponding ADC-restricted diffusion (B) are identified throughout the supratentorial region in multiple vascular distributions most prominent within the posterior perisylvian frontoparietal junctions, left greater than right. These are most compatible with multiple acute infarctions.

Hyperbilirubinemia

Physiologic jaundice is noted in every newborn with a serum bilirubin peak of approximately 6 mg/dL by 48–72 h of life with subsequent decline [Reference Perlman, Volpe and Volpe210,Reference Gartner, Lee, Vaisman, Lane and Zarafu211]. The rate of rise of bilirubin is more rapid with hemolysis (ABO incompatibility) [Reference Bahr, Christensen, Agarwal, George and Bhutani212,Reference Bhutani, Zipursky and Blencowe213]. Neurons are selectively susceptible to bilirubin toxicity. The most commonly affected regions include the BG (particularly the globus pallidus and subthalamic nucleus), the hippocampus, various cranial nerve nuclei (particularly the oculomotor, vestibular, and auditory), various other brainstem nuclei, and certain cerebellar nuclei (particularly the dentate) [Reference Perlman, Volpe and Volpe210]. The clinical presentation is usually on the second day of life and beyond. Most infants exhibit definite neurologic findings with initial features of lethargy, sleepiness, a paucity of movements, hypotonia, poor suck, and a slightly high-pitched cry. During an intermediate phase the infant appears more stuporosed coupled with irritability and increased extensor tone with arching of the neck (retrocollis) and back (opisthotonos). Without intervention the encephalopathy becomes more pronounced with deep stupor and coma, absent suck, and a shrill cry [Reference Perlman, Volpe and Volpe210]. Although the acute neurologic syndrome has become increasingly rare, acute bilirubin encephalopathy is still estimated to occur in 1.6 per 100,000 live births [Reference Bahr, Christensen, Agarwal, George and Bhutani212,Reference Bhutani, Zipursky and Blencowe213]. Importantly, early neurologic signs are nonspecific for hyperbilirubinemia and may be seen with the other causes of NE discussed in this review.

Nonketotic Hyperglycinemia (NKH)

NKH is an inborn error of metabolism in which large amounts of glycine accumulate in body fluids. Glycine functions as a neurotransmitter, and during early development the glycine receptor is excitatory while the “classic” glycine receptor is inhibitory [Reference Perlman, Volpe and Volpe214]. Clinical features typically present in the first 2 days of life with ineffective suck, impaired ventilatory effort (apnea), hypotonia, seizures, myoclonus, and frequent hiccups. Metabolic features include hyperglycinemia.

Hyperammonemia

Hyperammonemia results from enzyme defects of the first four of five steps of the Krebs–Henseleit cycle, which include carbamoyl phosphate synthetase, ornithine transcarbamylase, argininosuccinic acid synthetase, and argininosuccinase [Reference Perlman, Volpe and Volpe214,Reference Enns215]. Ammonia and the conversion product glutamine may have a variety of toxic effects on the brain. The clinical syndrome invariably has its onset between 24 and 72 h with features that include poor feeding, disturbed level of consciousness (stupor, coma), and hyperventilation with concomitant respiratory alkalosis presumably reflecting a central effect of hyperammonemia.

Organic Acid Disorders

The most common of the organic acid disorders are propionic and methylmalonic acidurias, which result in serious neurological disturbances. Propionate and methylmalonate are vital intermediates in the catabolism of protein and fats. The accumulation of these acids results in a disturbance of myelin and involvement of the BG [Reference Perlman, Volpe and Volpe214,Reference Enns215]. Onset of symptoms is in the first days of life with an abrupt onset of vomiting, tachypnea, stupor, and seizures. Metabolic features include profound acidosis, propionic ± methylmalonic acidemia, hyperglycinemia, and hyperammonemia.

Disorders of Pyruvate and Mitochondrial Metabolism

Pyruvate dehydrogenase deficiency is the most common defect in oxidative metabolism [Reference Robinson, MacKay, Chun and Ling216]. The neonatal form is characterized by lactic acidosis with ketosis often present, with occasional hyperammonemia and severe neurologic compromise with predominant hypotonia [Reference Byrd, Krohn and Winkler217].

Mitochondrial Disease

Mitochondrial disorders can cause dysfunction of any organ system, either alone or in any combination. Neuromuscular signs and symptoms occur most commonly in the first days and include hypotonia, seizures, lethargy, feeding and respiratory difficulties, deafness, dystonia, or seizures [Reference Perlman, Volpe and Volpe214,Reference Sue, Hirano, DiMauro and De Vivo218].

Sulfite Oxidase Deficiency/Molybdenum Cofactor Deficiency

Sulfite oxidase deficiency may occur in isolation or combined with xanthine oxidase deficiency (molybdenum cofactor deficiency). Neonates typically present with feeding difficulties, vomiting, and drug-resistant seizures in the first few days after birth. A diagnosis may be made from a sulfite strip test on a fresh urine sample [Reference Sass, Gunduz and Araujo Rodrigues Funayama219].

Isolated Seizures

Metabolic disorders that may present with isolated neonatal seizures, without obvious clinical or laboratory markers of an underlying inborn error of metabolism, include NKH, sulfite oxidase deficiency, pyridoxine-responsive seizures, and folinic acid-responsive seizures. Seizures may be a prominent component in peroxisomal disorders (e.g., Zellweger syndrome), mitochondrial disorders (Leigh’s syndrome), organic acidemias, and urea cycle defects. These disorders are characterized by routine laboratory abnormalities (e.g., prominent hyperammonemia, lactic acidemia, or metabolic acidosis) that should establish an underlying diagnosis [Reference Perlman, Volpe and Volpe214,Reference Enns215].