Neonatal Hypoxic-Ischemic Encephalopathy and Hypothermia Treatment

Abstract

Neonatal hypoxic-ischemic encephalopathy (HIE) is an important clinical entity since it is associated with death and long-term disability including cognitive impairment, cerebral palsy, seizures and neurosensory deficits. Over the past 40 years there has been an intensive search to identify therapies to improve the prognosis of neonates with HIE. Hypothermia treatment represents the culmination of laboratory investigations including small and large animal studies, followed by pilot human studies, and finally randomized controlled trials to establish efficacy and safety. Clinical trials have demonstrated that hypothermia treatment reduces mortality and improves early childhood outcome among survivors. HIE is a multi-system disease process which requires intensive medical support for brain monitoring and non-central nervous system organ dysfunction; it must be conducted in a level III or IV Neonatal Intensive Care Unit with infrastructure for an integrated approach to care for critically ill infants. Hypothermia treatment is the first and currently the only therapy to improve outcomes for neonates with HIE and indicates that HIE is modifiable. However, outcomes can likely be further improved. Hypothermia treatment has accelerated investigation of other therapies to combine with hypothermia. It has also stimulated a more intensive approach to brain monitoring which allows earlier intervention for complications. Finally, HIE and hypothermia treatment negatively influences the psychological state of affected families, and there is growing recognition of the importance of trauma informed principles to guide medical professionals.

Precis:

Hypothermia treatment for neonatal hypoxic-ischemic encephalopathy of infants ≥ 36 weeks gestation improves outcome and represents a major advance in neonatal care.

Introduction

Neonatal hypoxic-ischemic encephalopathy (HIE) is a serious neurological disorder that presents most frequently at or shortly after birth among late preterm and term neonates. Worldwide HIE is one of the leading causes of neonatal mortality along with preterm births. Among survivors, HIE is associated with a spectrum of neurodevelopmental problems from mild to extreme disability in childhood and later in life. Neonatal HIE is a condition that is not anticipated and maybe encountered in community hospitals as well as referral centers equipped with Neonatal Intensive Care Units. There rarely is time to prepare a family for the possibility of this condition. In the past, treatment was limited to intensive supportive care without a brain-oriented therapy. A critical observation of the pathogenesis of brain injury associated with HIE is that there is a sequence of events which unfolds over the first hours to days following hypoxia-ischemia. This sequence provides a time window when hypothermia therapy can be initiated and prevent or attenuate downstream events, and ultimately the extent of brain injury.

This article is intended for the obstetric community to provide an overview of neonatal encephalopathy and more specifically HIE and the use of hypothermia treatment. Areas that will be reviewed include the pathogenesis of HIE, neuroprotection provided by hypothermia, clinical evidence supporting the use of hypothermia treatment, what type of facility should provide hypothermia treatment, and the downstream effects of HIE and hypothermia on the family.

Newborn Encephalopathy and Hypoxic-Ischemic Encephalopathy

Neonatal encephalopathy is a clinical syndrome characterized by altered neurologic function with a subnormal level of consciousness, difficulty initiating and maintaining respirations at and following birth, depressed tone and reflexes, disturbances of cranial nerve function, and possibly seizure activity.¹ Neonatal encephalopathy is an overarching term reflecting many causal pathways manifesting with this syndrome including brain malformations, intracranial hemorrhage, stroke, metabolic disorders, genetic disorders, congenital infections, toxin exposure as well as HIE. Neonatal encephalopathy occurs in approximately 3/1000 live births and approximately 50% are attributable to HIE.² Neonatal encephalopathy is an important clinical entity since the underlying etiologies are often associated with mortality and serious neurodevelopmental deficits (cognitive impairment, cerebral palsy, seizures, and neurosensory deficits).³ The phenotype of encephalopathy is not specific to an etiology although some diagnoses may have characteristic neurological findings.

Some have suggested that the term HIE should not be used since a single cause attribution may be difficult to prove and a general descriptor, neonatal encephalopathy, may be better.⁴ However, a cogent, comprehensive rationale has been provided for using HIE.⁵ Much of this reflects the observed relationships between known, controlled hypoxic-ischemic insults in multiple preclinical studies spanning rodents to primates and the resultant type and topography of neuropathological lesions. The latter is mirrored by the neuropathology among neonates with autopsies who have clinical events indicative of severe hypoxic-ischemic insults, often with intrapartum sentinel events.⁶ Among survivors, magnetic resonance imaging (MRI) provides brain tissue assessment, and multiple scoring systems have been developed to capture location, extent, and patterns of injury.^7–10 The topography of brain MRI signal abnormalities is generally consistent with the distribution of animal neuropathology. Abnormalities detected by multiple MRI scoring systems for HIE have been associated with an increased risk of neurodevelopmental disability including cognitive delays, motor impairments including cerebral palsy, and visual and auditory deficits. Based on these considerations, the term HIE will be used in this article.

The severity of HIE is often characterized by Sarnat stages of mild, moderate or severe involvement (Table 1).¹¹ In the era before hypothermia, a cohort from Western Canada meeting clinical criteria of hypoxia-ischemia was prospectively identified at birth and followed through early school age to delineate neurodevelopmental outcomes.^{12, 13} Neonates were characterized by the worst Sarnat stage manifested in the first week after birth and brain injury was deemed present by any cognitive delay, cerebral palsy, visual or auditory loss, or the presence of seizures.¹⁴ At 3.5 years of age, if the worst Sarnat stage was mild, outcome was uniformly good with no associated death or brain injury. At the other end of the spectrum, if the worst stage was severe, outcome was uniformly poor with either death or brain injury among all. Infants with a moderate stage had a mixed outcome; some died, and some survived with or without brain injury. The same cohort was followed until 8 years of age and compared with a peer group.¹⁵ Assessments of intelligence quotients (IQ) among neonates with mild, moderate and severe Sarnat stages showed lower IQ scores with increasing severity of Sarnat stages in the first week after birth.

Table 1:

Principal characteristics of the Sarnat stages of encephalopathy

Category	Mild (Stage 1)	Moderate (Stage 2)	Severe (Stage 3)
Level of consciousness	Hyper-alert	Lethargic	Stupor or coma
Spontaneous activity	Normal or decreased	Decreased	None
Posture	Mild distal flexion	Distal flexion/complete extension	Decerebrate
Tone	Normal or increased	Hypotonia or hypertonia	Flaccid or rigid
Primitive reflexes
Suck	Weak or incomplete	Weak or incomplete	Absent
Moro	Low threshold to illicit	Incomplete	Absent
Autonomic function
Pupils	Mydriasis	Myosis	Variable or non-reactive
Heart rate	Tachycardia	Bradycardia	Variable
Respirations	Hyperventilation	Periodic breathing	Apnea or on ventilator support

Stages were modified from the original description¹¹ to facilitate rapid assessment shortly after birth for enrollment in clinical trials. Neonates were evaluated for 6 categories of neurologic function. Within each category, signs were used to determine findings consistent with mild, moderate or severe encephalopathy. There is overlap between mild and moderate encephalopathy, which can make it challenging to distinguish the stage for some infants.

Pathogenesis of HIE and Modification by Hypothermia Treatment

Hypoxia-ischemia of a severity to injure the brain must have critical decrements in the cerebral delivery of oxygen and glucose to trigger a sequence of early and subsequent downfield events.¹⁶ The initial effects of hypoxia-ischemia are reductions in high energy phosphorylated compounds, adenosine triphosphate (ATP) and phosphocreatine (PCr), the hallmark of primary energy failure (Figure 1). Multiple energy dependent cellular events are triggered by primary energy failure.¹⁷ If resuscitation is timely and successful, ATP and PCr recovers and marks the start of a latent phase, which is thought to extend for approximately 6 hours. The latent phase may be marked by the development of mitochondrial dysfunction which may act as a focal point in determining presence or absence of subsequent injury.¹⁸ Aberrant mitochondrial function may provide triggers for a second interval of energy failure between 6–12 hours after hypoxia-ischemia.^{19, 20} Secondary energy failure is accompanied by pathogenic processes that differ from primary energy failure (inflammation, accelerated apoptosis, deafferentation) along with ongoing excitotoxic and free radical injury (Figure 1). Secondary energy failure may last hours to days and is followed by a tertiary phase characterized by evolution of brain injury including removal of necrotic and apoptotic tissue, scarring or gliosis, and loss of brain volume (global or regional). Simultaneously there is evidence for initiation of tissue repair with stem cell proliferation, angiogenesis, and reestablishment of connectivity which may continue for weeks to months.^21–23

Figure 1:

The sequence of critical phases of altered brain metabolism after neonatal hypoxia-ischemia is depicted. Brain injury is a not a single event but rather a process which unfolds and extends over days. The interval between primary and secondary energy failure, the latent phase, represents a potential therapeutic window where interventions such as hypothermia can be initiated and modify outcome. ATP, adenosine triphosphate; PCr, phosphocreatine.

Neuroprotection from hypothermia includes a wide range of actions rather than a single pathway and favorably alters multiple processes associated with primary and secondary energy failure and the latent phase.²⁴ Specifically, hypothermia reduces energy consumption by the brain, promotes coupling of blood flow and metabolism, reduces the release of excitatory neurotransmitters, suppresses production of oxygen free radicals, inhibits inflammation and attenuates the acceleration of apoptotic programmed cell death. Hypothermia in effect slows or stops the cascade of pathologic processes but does not promote tissue repair. Pre-clinical fetal sheep studies have demonstrated that hypothermia is most effective when initiated within 6 hours of hypoxia-ischemia.^25–27 Based upon this work, the latent phase is considered to represent a therapeutic window when treatment can be initiated with the potential to attenuate or modify the extent of brain injury.

Criteria to Initiate Hypothermia Treatment

Consistent maternal antepartum risk factors (e.g., chronic hypertension, diabetes, antepartum hemorrhage, or thyroid disease) have not been identified among neonates who develop HIE. In contrast, the frequency of intrapartum complications is very high²⁸ and as indicated in Table 2, these events are part of proximate pathways which culminate in acute fetal hypoxia-ischemia. Most of the intrapartum complications are unanticipated and some may be viewed as sentinel events.

Table 2:

Selected maternal and neonatal characteristics derived from trials of hypothermia for HIE

Maternal Characteristics
Morbidities during Pregnancy	Hypothermia Treatment n/N (%)	Non-Cooled (Usual care) n/N (%)
Chronic Hypertension	82 / 449 (18)30, 43	14 / 106 (13)30
Antepartum hemorrhage	48 / 449 (11)30, 43	20 /0106 (19)30
Thyroid Disease	15 / 449 (3.3)30, 43	1 / 106 (0.9)30
Diabetes	53 / 449 (11)30, 43	9 / 106 (8)30
Intrapartum Complication	Hypothermia Treatment	Control
Fetal Heart Rate Decelerations	411 / 559 (73)30, 43, 34	148 / 217 (68)30, 34
Cord Prolapse	86 / 621 (14)30, 43, 34, 33	31 / 280 (11)30, 34, 33
Nuchal Cord	4 / 62 (6)33	4 / 63 (6)33
Uterine Rupture	41 / 186 (22)30, 43, 33	21 / 169 (12)30, 33
Placental Abruption	6 / 62 (10)33	12 / 63 (19)33
Maternal Hemorrhage	79 / 559 (14)30, 43, 34	28 / 217 (13)30, 34
Shoulder Dystocia	47 / 559 (8)30, 43, 34	24 / 217 (11)30, 34
Maternal Pyrexia	70 / 784 (9)30, 43, 34, 33, 32	30 / 442 (7)30, 34,33, 32
Meconium-Stained Fluid	15 / 62 (24)33	18 / 63 (28)33
Emergency cesarean delivery	375 / 621 (60)30, 43, 34, 33	180 / 280 (64)30, 34, 33
Neonatal Characteristics
Characteristic Prior to Intervention	Hypothermia Treatment	Non-Cooled (Usual Care)
Range of Mean Gestational Age (weeks)	38.5 – 39.343, 34, 33
Range of Mean Birth Weight (grams)	3300 – 343130, 43, 34, 33
Sex (Male)	446 / 784 (57)30, 43, 34, 33, 32	254 / 442 (57)30, 34, 33, 32
Seizures	274 / 722 (38)30, 43, 34, 32	172 / 379 (45)30, 34, 32
Moderate Encephalopathy	419 / 621 (67)30, 43, 34, 33	137 / 280 (50)30, 34, 33
Severe Encephalopathy	184 / 621 (29)30, 43, 34, 33	115 / 280 (41)30, 34, 33

Maternal and neonatal characteristics of infants with HIE dichotomized by randomization to infant treatment with hypothermia or usual care. Data were compiled from 5 randomized trials using all data available.

Unless otherwise indicated, data are presented as n/N (%) where n represents the number with the characteristic and N represents the total number evaluated.

Based on preclinical data supporting a putative latent phase of approximately 6 hours following a hypoxic-ischemic event, randomized clinical trials were designed to enroll neonates by 5.5–6 hours of age. However, as summarized by a Report of the American College of Obstetricians and Gynecologists Task Force on Neonatal Encephalopathy,²⁹ a definitive test for HIE does not exist. The Task Force outlined multiple variables that need to be assessed to determine if acute peripartum or intrapartum hypoxia-ischemia contributed to a neonate born with encephalopathy. These include a sentinel event immediately before or during labor, a fetal heart rate pattern indicative of an acute event (e.g., a change from a category 1 to category 3 tracing), neonatal signs (depressed Apgar scores, fetal acidemia, encephalopathy and seizures), electrographic evidence of encephalopathy, multi-system organ dysfunction, changes on MRI imaging, placental abnormalities, and the absence of other etiologic risk factors. The full manifestation of a hypoxic-ischemic injury (e.g., neurological examination, electrographic findings, non-central nervous system dysfunction, MRI findings) evolves over days following birth consistent with the pathogenesis as outlined earlier. The certainty of a diagnosis of HIE increases when more of these variables are present and provides a stronger link between peripartum events and early childhood neurodevelopmental outcomes.

Neonatal practitioners confronted with a neonate at risk for acute HIE at birth do not have the time to acquire all the data outlined by the Task Force if hypothermia treatment is to be initiated by 6 hours. At risk neonates are typically critically ill when born and require intensive stabilization in the first hours after birth while being assessed for hypothermia treatment. Clinical trials of hypothermia therefore used more limited data for inclusion criteria. In broad terms, inclusion criteria sought evidence of impaired placental gas exchange (clinical or biochemical) followed by evidence of moderate or severe encephalopathy using the modified Sarnat stages (Table 1). The presence of encephalopathy serves as important verification of the biological effects of altered placental gas exchange on the neonate. An example of the inclusion criteria from one trial, the National Institute of Child Health and Human Development (NICHD) Neonatal Research Network (NRN) Whole Body Cooling trial,³⁰ is depicted in Figure 2.

Figure 2:

The inclusion criteria of one clinical trial of hypothermia, National Institute of Child Health and Human Development Neonatal Research Network Whole Body Cooling is depicted.³⁰ Neonates who meet screening criteria are evaluated in a stepwise process. The initial step is to assess if biochemical or clinical markers of impaired placental gas exchange are present. If these criteria are fulfilled, neonates undergo a neurological examination to determine the presence of moderate or severe encephalopathy. If encephalopathy is present, neonates meet criteria for hypothermia treatment.

Five randomized trials were published between 2005–2011 which were remarkably homogenous in terms of inclusion and exclusion criteria.^30–34 Neonates considered for inclusion were all ≥ 36 weeks gestation (except for one trial which enrolled infants ≥ 35 weeks gestation³⁴), birthweight > 1800 grams, and < 6 hours of age. Similar to the NICHD NRN trial,³⁰ the other trials used markers of impaired placental gas exchange (low Apgar scores, need for resuscitation or fetal acidemia) followed by a neurological examination for moderate or severe encephalopathy using the Sarnat stages.^31–34 Three of the initial trials used either amplitude integrated EEG (aEEG) or a full Montage EEG to provide electrophysiologic confirmation of encephalopathy.^31–33 The average age of randomization of neonates among these 5 trials ranged from 4–5 hours following birth. Although earlier initiation of hypothermia in the 6-hour window after birth is supported by preclinical data,^25–27 there are no prospective clinical studies to verify improved outcome with initiation earlier within the first 6 hours of life. Mild encephalopathy was not an inclusion criterion in any of the 5 trials since the outcome of neonates with mild encephalopathy was historically better than moderate or severe encephalopathy.^13–15

Since neonatal encephalopathy has a broad differential diagnosis,³⁵ there is the potential to initiate hypothermia treatment for neonates who do not have HIE. After completion of the NICHD NRN whole body cooling trial, enrolled neonates were reviewed and none had a brain malformation, metabolic or neuromuscular disorder as a cause of the encephalopathy.³⁶ One neonate in the hypothermia group and two neonates in the control group had positive blood cultures before initiation of the study intervention. These observations provide reassurance that the inclusion criteria from randomized trials identifies at risk neonates with HIE.

Efficacy of Hypothermia for HIE

The initial trials of hypothermia^30–34 have been combined with an early smaller trial³⁷ and a larger trial from China³⁸ for a meta-analysis to estimate the efficacy of hypothermia for HIE.³⁹ All trials used a stepwise process for inclusion, randomization to cooling or standard care for moderate or severe encephalopathy, a 72-hour duration of the intervention, a rewarming rate of 0.5°C/hour, and evaluated death or disability as the primary outcome most commonly at 18–22 months of age. Hypothermia was achieved with either whole body cooling (target core temperature 33.5°C) or head cooling combined with more modest body cooling (target core temperature 34.5°C). Major disability included any of the following: cerebral palsy, developmental delay greater than 2 standard deviations below the mean, blindness in both eyes or sensorineural hearing loss requiring amplification. Selected results of the meta-analysis are listed in Table 3. ​ Outcomes were similar irrespective of the mode of cooling (whole body vs head cooling with more modest body cooling).

Table 3:

Neurodevelopmental outcomes of children at 18–22 months of age treated with hypothermia or usual care at birth for HIE

Outcome*	Hypothermia	Non-Cooled (Usual Care)	Risk Ratio (95% Confidence Interval)
Death or major disability	46.0†	61.4	0.75 (0.68–0.85)
Death	25.3	34.2	0.75 (0.64–0.88)
Major disability	26.2	39.3	0.67 (0.55–0.80)
Cerebral palsy	22.9	35.2	0.66 (0.54–0.82)
Blindness	5.9	10.0	0.62 (0.38–1.01)
Hearing loss	3.8	5.9	0.66 (0.35–1.26)
Death or major disability among neonates with:
Moderate encephalopathy‡	36.6	53.7	0.68 (0.56–0.84)
Severe encephalopathy‡	69.9	85.7	0.82 (0.72–0.93)

Results of a meta-analysis to provide the primary outcome and components of the primary outcome at 18–22 months among neonates enrolled in trials of hypothermia³⁹

^*Outcomes among survivors are based on the total number of neonates evaluated.

^†All results for hypothermia and non-cooled neonates are percentages.

^‡Among infants with moderate or severe encephalopathy at randomization.

Based on these initial clinical trials, hypothermia treatment for HIE is efficacious. The number of neonates needed to treat to avoid one neonate with death or disability was seven. Hypothermia treatment also has an acceptable safety profile. Hypothermia was associated with more frequent sinus bradycardia, thrombocytopenia, and trends towards higher rates of pulmonary artery hypertension. However, use of pressor support and coagulopathy resulting in thrombosis or hemorrhage did not differ between cooled and non-cooled groups.³⁹

A limited number of trials evaluated infants at 6–7 years to determine if the benefit of hypothermia observed at 18–22 months persisted. In the NICHD NRN Whole Body Cooling trial, death or an IQ < 70 (2 standard deviations below the mean) at 6–7 years was present in 47% of cooled infants and 62% of infants receiving standard care (aRR 0.78, 95% CI, 0.61–1.01).⁴⁰ There was a reduction in mortality, but no differences in CP, IQ<70, or moderate/severe disability. In the TOtal Body HYpothermia Trial (TOBY) investigators assessed a positive outcome of hypothermia, survival with an IQ ≥ 85 (within 1 standard deviation of the mean).⁴¹ Hypothermia treatment was associated with a greater percentage of survivors with an IQ ≥ 85 compared with standard care (52% vs 39%, respectively, aRR 1.31, 95% CI, 1.01–1.71). There was no difference in mortality, but CP was reduced. Despite different results at 6–7 years between these trials, the results are reassuring that outcomes are improved, and hypothermia did not simply salvage infants with profound injury destined to die without cooling treatment. Functional status has been examined among 6–7-year-old children in the NICHD NRN trial.⁴² In the standard care arm, 47% of children were receiving special education, 43% were receiving speech therapy, and 9% were diagnosed with behavioral problems. Among neonates receiving hypothermia, these outcomes were present in 32%, 30% and 7%, respectively. These observations indicate that ongoing educational needs is an important concern among infants with moderate and severe HIE even if they received hypothermia treatment.

A subsequent clinical trial attempted to further reduce death or disability by cooling to a deeper temperature (32.5°C) and/or cooling for a longer interval (120 hours) among neonates born and enrolled between 2010–2013.⁴³ Deeper and/or longer cooling did not improve outcomes compared to cooling to 33.5°C for 72 hours. However, in this trial the rate of death or disability following cooling at 33.5°C for 72 hours for moderate or severe HIE was 29%, a marked reduction from earlier trials. The improvement in death or disability compared with the initial trials reflects multiple factors including earlier recognition of at-risk neonates, improved stabilization, earlier initiation of hypothermia treatment, and more intensive neuro-monitoring.⁴⁴ Randomization by 6 hours after birth was an inclusion criterion for hypothermia trials based on preclinical studies of a well-timed ischemic insult with initiation of cooling at different post-ischemia intervals.^25–27 However, there are neonates in whom encephalopathy progresses after 6 hours of age or who were born remote from a center that provides hypothermia treatment preventing initiation in the desired time interval. Furthermore, hypothermia treatment is based on a hypoxic-ischemic event occurring near or at the time of delivery; the frequency of well-timed sentinel events near delivery (e.g., acute abruption, ruptured uterus, prolapsed umbilical cord, etc.) was approximately 60% of cases enrolled in one of the first clinical trials.^,45

It seems plausible that some neonates enrolled in hypothermia trials may experience in utero events potentially hours before birth, and the interval between hypoxia-ischemia to initiating hypothermia treatment may actually be beyond 6 hours. The use of hypothermia beyond 6 hours from birth therefore represents an important issue. The NICHD NRN performed a randomized trial of initiation of hypothermia compared with targeted normothermia for neonates presenting between 6–24 hours after birth.⁴⁶ A preplanned Bayesian analysis was used since the sample size was anticipated to be small (168 neonates enrolled over 6 years). The mean age (± standard deviation) of randomization was 16±5 h and 15±5 h for the hypothermia and non-cooled groups, respectively. Approximately 32% of neonates enrolled in both groups were ≥ 6 h to ≤ 12 h after birth. Based on the observed results there was a 76% probability of less death or disability with hypothermia initiated after 6 hours of age compared with normothermic care. The results are not conclusive but suggest possible benefit. Some centers may choose not to use hypothermia after 6 hours of age due to the uncertainty of the results. Others may choose to use it given potential benefit of hypothermia, the seriousness of the outcome (death or disability), the absence of obvious harm, and lack of an alternative therapy other than supportive care.

In contrast to trials conducted in high income countries, a recent randomized controlled trial compared hypothermia with standard care among neonates born in India, Sri Lanka, and Bangladesh with moderate or severe encephalopathy.⁴⁷ This was a rigorous trial of whole body cooling initiated within 6 hours of birth, using a target core temperature of 33.5°C, and continued for 72 hours before rewarming compared with non-cooled neonates receiving supportive intensive care. The composite outcome of death or disability at 18–22 months was not reduced with hypothermia treatment, and death was significantly increased. The absence of hypothermic neuroprotection may reflect an earlier onset of brain injury in utero placing the neonate outside of the therapeutic window when hypothermia was initiated. Earlier injury is suggested by seizures observed within hours of birth and a high percentage of infants with white matter injury on MRI possibly reflecting a partial prolonged event.⁴⁸ Another consideration is an obstetric practice environment that differs from the U.S. trials with less intensive intrapartum monitoring and lower rates of obstetric interventions given that emergency cesarean deliveries were performed in 23% and 17% of hypothermic and standard care groups, respectively.

How is Hypothermia Treatment Achieved?

There are two approaches to cool the brain. One employs a cooling cap through which cool water is circulated and is combined with a reduction in core temperature to 34.5°C. The cooling cap minimizes the reduction in core temperature needed to cool the deep brain since systemic hypothermia may be harmful to critically ill neonates. The alternative is to cool the body to cool the brain. The cooling cap has largely fallen out of favor and whole-body cooling is the most common method currently used. As indicated in Figure 3, the neonate is placed on a cooling blanket to reduce the body temperature to a core target of 33.5°C (acceptable range of 33.0–34.0°C) and in turn reduce the brain temperature.

Figure 3:

Depiction of a mock-up simulation of the equipment used to care for neonates undergoing hypothermia treatment for hypoxic-ischemic encephalopathy. The neonate is positioned on a cooling blanket within an incubator set to not provide any exogenous heat. The cooling blanket is attached to a hyper-hypothermia device which regulates the neonate’s temperature to a core body temperature of 33.5°C. Other equipment around the bedside includes devices to monitor and support pulmonary and cardiovascular function and assess brain electrical activity. An amplitude integrated electroencephalogram (EEG) is a simplified EEG which can provide a limited number of channels. Alternatively, centers use a standard EEG machine with multiple channels. IV, intravenous.

Core body temperature is typically monitored with either an indwelling probe in the lower third of the esophagus or the rectum; animal studies have indicated that these sites are good measures of core body temperature.⁴⁹ There are multiple commercial systems that regulate core body temperature by a servo controller. These devices compare the desired temperature to the neonate’s actual core temperature and automatically adjust the temperature of water circulating through the blanket to keep the infant core temperature close to 33.5°C.

What Type of Neonatal Unit Should Provide Hypothermia Treatment?

Hypothermia treatment should only be performed in a level III or IV NICU with the infrastructure to care for infants with multi-organ system dysfunction including coordination of multiple pediatric subspecialty services. Hypothermia is a time sensitive treatment and referring nurseries/NICUs need well developed relationships with regional centers that provide hypothermia treatment to facilitate timely transfers.

Hypothermia needs to be monitored with continuous measurement of core temperature to ensure that target temperature is achieved for optimal neuroprotection. Hemodynamic instability during cooling and device malfunction can contribute to undesirable low or high core temperature.⁵⁰ The dissemination of hypothermia therapy has accelerated a more proactive approach to HIE management beyond solely hypothermia treatment.⁵¹ This has resulted in more intensive brain monitoring, most often by use of a full montage continuous EEG (cEEG) or amplitude integrated EEG (aEEG) if a cEEG is not available. EEG allows earlier seizure detection and treatment; however, it remains unclear if childhood outcomes are improved with the use of more intensive brain monitoring in the absence of a suitable comparison group. EEG also provides valuable prognostic information to complement other data in order to give the best estimate of early childhood outcome.⁵²

MRI is the optimal imaging technique for HIE and is typically performed following completion of hypothermia between days 4–7 after birth. Brain MRI provides the greatest discrimination to differentiate presence or absence of injury and is critical for providing prognostic information for families. Given these considerations, active collaborations between NICU providers, neurologists or epileptologists, and neuroradiologists is part of the approach to HIE.

In addition to encephalopathy and seizures, HIE is a multi-system organ disease process as a result of a redistribution of cardiac output triggered by hypoxia-ischemia that shifts blood flow from lesser to more critical parts of the body.⁵³ If hypoxia-ischemia is severe, organ system dysfunction may even affect essential organs such as the heart. Some of the more common non-central nervous system specific organ dysfunctions are reviewed below and further substantiate the need for intensive care in a level III or IV NICU.

Pulmonary:

Poor respiratory drive at birth secondary to hypoxia-ischemia results in the need for positive pressure ventilation and potentially intubation shortly after birth. Meconium aspiration can be associated with hypoxia-ischemia with a resultant pneumonitis. Persistent pulmonary artery hypertension occurs in greater than 20% of infants with HIE and is associated with meconium aspiration and severe HIE.⁵⁴ Hypoxic-ischemia pulmonary dysfunction can necessitate extra-corporeal membrane oxygenation in a small percent of infants. Intensive monitoring with continuous measurement of oxygen saturation, transcutaneous pCO₂ and intermittent arterial blood gas sampling is essential since persistent hypocapnia and hyperoxia are associated with an increased risk of neurodevelopmental disability.^{55, 56}

Cardiac:

Cardiovascular dysfunction occurs frequently among neonates with moderate or severe HIE and therapies to augment myocardial function, treat hypotension and stabilize peripheral vaso-paralysis are often used. Continuous blood pressure monitoring with indwelling arterial catheters in conjunction with serial echocardiographic assessments are core components of maintaining hemodynamic stability and preserving end organ perfusion to avoid further injury.⁵⁷

Renal:

Oliguria, weight gain and a positive fluid balance need to be anticipated among neonates with HIE. Elevation of serum creatinine and blood urea nitrogen are common and reflect effects of hypoxia-ischemia on the kidney (either pre-renal or renal in origin). Fluid restriction and meticulous attention to water balance and serum electrolytes is critical to avoid the effects of fluid overload especially in the setting of compromised cardio-pulmonary function. Most renal dysfunction is transient but rarely there can be progression to dialysis.

Hematologic:

The most frequent hematological abnormality associated with HIE is thrombocytopenia. Platelet counts need to be serially followed since some neonates require replacement therapy. Disseminated intravascular coagulopathy is uncommon but can occur with severe injury.

Metabolic:

Glucose is the primary fuel for the neonatal brain and dysregulation of glucose metabolism is common. Hypoglycemia following hypoxia-ischemia reflects the shift from aerobic to anerobic metabolism and depletion of endogenous stores. Hyperglycemia may also occur and reflects reduced tissue uptake and utilization.⁵⁸ Prompt treatment is considered critical since both hypo- and hyperglycemia have been associated with an increased risk of neurodevelopmental disability.^58–60

Finally, infants with HIE are at increased risk of neurodevelopmental problems in childhood. Arrangements are needed to track and insure follow-up for developmental assessments, and screens for neuromotor and neural sensory disorders over the first 2–5 years after birth.

Application of Hypothermia Treatment to Infants Not Included in Clinical Trials

Like other therapies that demonstrate benefit, there has been enthusiasm to apply hypothermia to cohorts or clinical scenarios not addressed in randomized trials of hypothermia.

Mild HIE:

The most controversial issue is the use of hypothermia to treat neonates with mild HIE. Mild HIE was not included in the first series of hypothermia trials since the outcome of neonates classified as mild encephalopathy was not as severe as neonates with moderate or severe encephalopathy.¹⁴ However, the outcome of neonates with mild HIE was based on neonates classified by the worst stage of encephalopathy in the first week after birth. Follow-up of neonates with mild encephalopathy identified within 6 hours of birth indicates worse outcomes compared to mild encephalopathy classified by the maximal stage in the first week after birth.^{61, 62} In addition, there is overlap in the clinical findings among mild and moderate encephalopathy (Table 1) posing challenges in identifying infants with mild HIE. Based on these observations, there has been progressive therapeutic drift with many clinicians providing hypothermia for mild HIE in the US and the United Kingdom. Specifically, using data from the California Perinatal Quality Care Collaborative, the proportion of neonates with mild HIE receiving hypothermia treatment increased from 46% in 2010 to 79% in 2018.⁶³ Using the National Neonatal Research database derived from NICUs in England and Wales, hypothermia treatment among neonates with mild HIE increased from 24.9% during 2011–2013 to 35.8% in 2014–2016.⁶⁴

The progressive increase in the use of hypothermia for neonates with mild HIE has occurred despite the lack of evidence-based studies. Some earlier randomized and quasi-randomized trials of hypothermia inadvertently enrolled a small number of neonates with mild HIE while intending to enroll those with moderate or severe HIE. A meta-analysis of these data indicated that death or disability did not differ between neonates receiving hypothermia (19.6%) and those receiving usual care (19.7%, risk ratio 1.11, 95% confidence interval 0.55–2.25).⁶⁵ Since the confidence intervals are wide and the sample size was limited (n=117), benefits or harm of hypothermia treatment in the setting of mild HIE cannot be excluded. A number of clinical trials are planned or have been initiated to study hypothermia for neonates with mild HIE and their success will be dependent on physician equipoise.

Preterm Infants:

Based on the efficacy of hypothermia for moderate or severe HIE (Table 3), the Committee on the Fetus and Newborn of the American Academy of Pediatrics recommended use of this treatment by the neonatal community.⁶⁶ The recommendation stated that cooling neonates born at less than 35 weeks’ gestation should only be undertaken as part of a research initiative. However, most clinical trials that established the efficacy and safety of hypothermia enrolled infants with a gestational age of ≥ 36 weeks at birth.^30–33 Only two trials enrolled infants at 35 weeks gestation^{34, 37} and one of these trials was a smaller pilot with outcomes assessed at 12 months.³⁷ Many centers have used the inclusion criteria of these two trials and the recommendation by COFN to provide hypothermia to neonates born at or beyond 35 weeks’ gestation. Furthermore, some centers have extended the gestational age of eligibility for cooling to birth after 34 weeks’ gestation.⁶⁴ Interestingly, in the two trials that enrolled infants born between 35⁰ and 35⁶ weeks, there were only 7 infants randomized in this gestational age window.^{34, 37} Irrespective of the recommendation and ongoing clinical practice, there is a knowledge gap regarding the efficacy of hypothermia treatment for moderate or severe HIE at 35 weeks’ gestation. The NICHD NRN has addressed this knowledge gap by performing a clinical trial among 168 neonates born between 33⁰ and 35⁶ weeks with HIE who were randomized to hypothermia (esophageal temperature of 33.5°C for 72 hours) or targeted normothermia (esophageal temperature of 37°C for 72 hours, NCT 01793129). Enrollment has been completed and results are pending.

Long-Term Effects of HIE and Hypothermia on the Family

Hypothermia for newborns with HIE is one of the most stressful medical problems that a family can experience in the NICU. In a very short period, families move from expectations of a healthy, term delivery to observing their baby receive varying extents of stabilization and possibly cardio-pulmonary resuscitation with intubation, chest compressions and administration of cardiotonic medications following delivery. The events at birth are followed in the first hours by decisions to initiate hypothermia or transfer neonates from community facilities or smaller NICUs to a regional center for hypothermia. Mothers experience long-lasting deleterious effects of this experience on themselves and their families, even after their babies are discharged from the NICU.⁶⁷ Mothers whose neonates require hypothermia are at increased risk for postpartum depression, anxiety, and post-traumatic stress disorder.^68–70

The adverse effects of HIE and hypothermia on family well-being starts with the loss of the bonding process at birth and includes immediate separation of mother and baby, inability to provide skin to skin care, and lack of initiation of breastfeeding. Although parents of any acutely ill neonate at birth experience a similar separation, the effects of HIE on the mother-baby dyad grows over the initial hospitalization due to the inherent prognostic uncertainty.

Given the complex and profound psychosocial effects of HIE on parental well-being, there is consensus that families of neonates with HIE experience trauma. A framework to conceptualize parental trauma associated with HIE has been documented and centers on the three E’s: the event, the experience, and the effect. Targeted interventions utilizing trauma-informed care principles are recommended at each stage and focus on communication, parental integration into NICU care, parental mental health and resource support.^{71, 72} Psychosocial support systems (social workers, psychologists, family support coordinators) are increasingly recognized as essential for family stability. The care of neonates with HIE is provided by a neonatologist who typically has had no prior interactions with the parents. Frequent updates are important to build a bond between the medical team and the family. The neonatologist also bears the responsibility of maintaining communication with obstetric/maternal providers especially if the mother is still hospitalized in another facility.

The physical separation of the mother and her neonate has a profound effect. Promoting families to be with their baby during treatment for HIE is an important initial step to decrease the chance of a second trauma and to restore a sense of family unity. Since HIE is a largely unanticipated event, many affected neonates are born in community hospitals necessitating transfer to a regional NICU. Unfortunately, health insurance coverage may pose limits on the ability to transport postpartum mothers. A challenging aspect of hypothermia treatment has been the inability of parents to hold their baby. There is increasing interest across the neonatal community to facilitate parents holding their babies while being cooled. Limited studies support the feasibility of this practice in specific circumstances.^{73, 74} There can be limited space at the bedside (Figure 3) but NICU personnel are resourceful in prioritizing the needs of the family. Acuity of illness and instability of the neonate are the primary considerations that constrain holding them while undergoing hypothermia treatment.

Conveying Information to Families about HIE and Hypothermia

Communicating information to families about a complex subject like HIE during a perinatal crisis is extremely challenging. Parents have a multitude of questions, but some parents verbalize that they do not know what question to ask and feel totally helpless. Qualitative research using semi-structured interviews with parents whose babies had HIE and underwent hypothermia indicated three important themes.⁷⁵ The first is that communication is often fragmented with many different clinicians conveying information regarding HIE. Second, information explaining HIE and hypothermia is often provided at a comprehension level above what is recommended for non-medical personnel. Third is the uncertainty of the prognosis for many neonates after HIE. These themes helped to formulate trauma-informed principles and guide clinicians regarding communication, integrating the family into the NICU environment, supporting parental health, and connecting families with resources.⁷¹

Conclusions

Prior to 2005, care of neonates with moderate or severe HIE was limited to intensive supportive care. Following birth, abnormalities of hemodynamics, acid-base status, oxygenation, ventilation, and metabolism were assessed and corrected with subsequent surveillance for organ system dysfunction. Most non-central nervous system dysfunction recovers with this paradigm but not so for the brain, where injury may progress over time. Hypothermia treatment for HIE represents a landmark therapy since it is the first and only therapy to date that has been demonstrated to reduce mortality, and improve neurodevelopmental outcomes. Based on the randomized trials reviewed in this publication, the Committee on Fetus and Newborn of the American Academy of Pediatrics provided a framework for the implementation of this therapy, and use of hypothermia has been widely disseminated.⁶⁶ Hypothermia treatment has unequivocally demonstrated that HIE is a modifiable condition and opened many doors for further investigation since the first trials were published in 2005. The observed success with hypothermia has quickened the pace of laboratory investigations to identify viable pharmacologic therapies to justify randomized clinical trials and ultimately combine with hypothermia.⁷⁶ Implementation of hypothermia has led to identification of components of supportive care which may affect early childhood outcomes (hypo- and hyperglycemia, hypocarbia, and blood pressure support) and require prospective investigation. With the dissemination of hypothermia, comprehensive organizations such as HOPE for HIE (https://www.hopeforhie.org, accessed 05/05/2023) have emerged to provide support and resources for parents of affected neonates. Such organizations are essential partners in ongoing and developing research to insure inclusion of the family perspective.