Shifting Outlooks after Neonatal Encephalopathy in the Era of Therapeutic Hypothermia

Abstract

In the era of therapeutic hypothermia (TH), more infants are surviving moderate to severe hypoxic-ischemic encephalopathy (HIE) with no or less apparent injury on neonatal brain magnetic resonance imaging (MRI). Despite a reduction in death and severe neurodisability for infants with neonatal encephalopathy from HIE, children remain at risk for challenges in learning, language, coordination, behavior, and socioemotional development. Neither neonatal MRI nor early developmental testing is completely predictive of outcomes at school age. This review summarizes current data on long-term outcomes of infants who have received TH for HIE, most in the absence of cerebral palsy or significant neonatal brain injury. Many children with a history of neonatal HIE face challenges in their motor skills, emotion regulation and behavior, language and communication, cognition and learning, and academic achievement. All children with a history of neonatal HIE can benefit from close neurodevelopmental surveillance into adolescence. It is important for providers to counsel families about the spectrum of long-term outcomes and potential effects on later stages of neurodevelopment even when brain MRI is reassuring. Understanding the broader spectrum of neurodevelopmental impairment at school-age in children with neonatal HIE can inform new therapies, early intervention strategies, and pre-school readiness to optimize outcomes.

Introduction

Hypoxic ischemic encephalopathy (HIE) is a leading cause of brain injury in neonates worldwide, affecting between 0.5 to 3 per 1000 births per year in developed countries.^1–4 HIE is a form of neonatal encephalopathy caused by a transient disruption of cerebral blood flow and oxygenation during or near the time of birth.⁵ The terms neonatal encephalopathy and HIE can at times be used interchangeably causing some confusion in the literature, as not all presentations of neonatal encephalopathy are caused by exposure to hypoxia-ischemia.^{3, 6} For the hypoxia-ischemia (HIE) type of neonatal encephalopathy, brain injury may result from the primary ischemic event or secondary energy failure driven by oxidative stress, excitotoxicity, and inflammation.⁷ Therapeutic hypothermia (TH) involves inducing targeted hypothermia in the neonate for 72 hours, with the goal of minimizing this secondary phase of brain injury and preventing the associated long-term neurologic sequelae of neonatal HIE.⁸ Since its approval by the United States Food and Drug Administration in 2007, TH has become the standard of care for the treatment of moderate to severe HIE in the United States and throughout many high-income countries.^{9, 10}

The era of TH has seen a reduction in death and disability associated with HIE.^11–16 Infants with HIE have lower risks of cerebral palsy and other adverse neurologic outcomes after TH,^17–19 demonstrate less severe brain injury on post-cooling MRI, and are more likely to have normal post-cooling brain MRI than those infants with HIE who are not cooled.^20–22 Despite the neuroprotective benefits of TH, brain injury and long-term neurologic sequelae may still occur, which has prompted important additional neuroprotection studies to try and reduce neurologic morbidity.^23–27

When significant brain injury results from HIE, neurologic sequelae are often more global and can present with early motor disability and cerebral palsy, intellectual disability, epilepsy, and visual impairment.²⁸ However, even with mild or no apparent injury on neonatal brain MRI, there may be long-term sequelae that are not inconsequential to the child and their family. Many of these sequelae may not appear until later stages of child neurodevelopment such as at school-age, which is often after dismissal from ongoing evaluations in neonatal neurodevelopmental follow-up clinics. This review aims to characterize the broader spectrum of long-term outcomes reported for neonatal HIE in the era of TH, particularly with a focus on child outcomes for those who do not have significant brain injury or cerebral palsy. Given the protracted period of child neurodevelopment, as pediatric providers we must counsel about the spectrum of outcomes that may emerge at older ages of childhood following neonatal HIE. In the era of TH, the outlook after neonatal HIE is shifting from a high concern for cerebral palsy and severe neurodisability to recognizing the other long-term sequelae affecting learning, language, coordination, behavior, and socioemotional development. These “milder” outcomes require better characterization through further research, and unique long-term follow-up and support of children and their families.

Role of neuroimaging in prognostication

Brain MRI remains one of the most useful tools to define brain injury in neonatal HIE.²⁹ Risks for cerebral palsy and neurologic disability correlate well with specific patterns of injury on conventional neonatal neuroimaging.^30–32 Several scoring systems have been developed that correlate motor and cognitive outcomes with a higher burden of injury.^33–35 The pattern of injury is also a strong indicator of the type of neurologic deficits that may be expected and which are used to guide prognostic counseling in the neonatal intensive care unit (NICU).^{36, 37} For example, abnormal signal in the basal ganglia or the posterior limb of the internal capsule is often associated with significant motor impairment.^38–40 A watershed distribution pattern of injury and injury to cortical gray matter have been associated with abnormal neurocognitive outcomes at school age,^{34, 41, 42} and moderate to severe cortical and even subcortical gray matter injury has been associated with epilepsy.^{43, 44} Studies have confirmed these lesion-specific relationships in the era of TH.⁴⁵

However, in the absence of severe brain injury post completion of TH, predictions of neurodevelopmental outcomes from neonatal brain MRI are more challenging and less accurate.⁴⁶ Physicians tend to present MRI with greater certainty than actually exists.⁴⁷ While normal neonatal MRI has been associated with more favorable outcomes in early childhood,¹⁸ it does not predict entirely normal outcomes at school age.⁴⁸ Using advanced imaging techniques, infants with HIE who are cooled and without cerebral palsy demonstrate altered structural connectivity of cerebral white matter, with reduced fractional anisotropy in the fornix, corpus callosum, anterior and posterior limbs of the internal capsule bilaterally and cingulum bilaterally at 6–8 years.⁴⁹ This suggests that even in the absence of apparent major brain injury on neonatal MRI, that subsequent microstructural brain development may be altered in these children following the exposure to HIE.

Further, timing of the brain MRI may affect the yield of neonatal images. A recent study of 94 infants with HIE treated with TH who underwent both an “early” brain MRI at a median of 4 days, and a “late” brain MRI at a median of 16 days found some cases to have findings on the later scan that were unrecognized on early brain MRI. Of the 24 infants with normal early MRI scans, 13% had abnormalities shown on their late scan, including abnormal T1 and T2 signal and restricted diffusion in the deep gray matter. Seventy infants had abnormalities on the early scan, with 16% having complete resolution by the time of the late scan, and 6% of infants demonstrated worsening or new findings.⁵⁰ Current standard practice at most centers involves an early brain MRI within the first week of age. In many institutions a late MRI is not routinely acquired unless there is discrepancy in the early imaging and clinical condition of the infant.⁵¹ Thus, for some infants, brain injury may be missed on an early scan as the injury may still be evolving or undergoing repair, leading to potential underestimation of brain injury. In the era of TH where there are increasingly more infants with normal or mild injury only, providers should use caution with counseling parents about prognosis based on MRI,⁵² and be able to discuss the broader array of effects HIE can have on long-term learning, communication, emotional and behavioral regulation, and motor function, as well as other unknown impacts.

Long-term neurodevelopmental outcomes

Studies in the pre-cooling era demonstrated that even in the absence of clinically evident neurologic disability, children with a history of HIE performed more poorly on scores of cognitive, neuropsychological, and educational assessments compared to peers at school age.^{53, 54} These differences include skills that are foundational in school such as arithmetic, reading, and spelling.⁵⁵ In addition, attention-deficit hyperactivity disorder (ADHD), developmental coordination disorder (DCD), anxiety and depression scores,⁵⁶ behavioral and social problems,^{57, 58} and autism spectrum disorder (ASD)⁵⁹ have been reported at higher frequency in these children.

Follow up analyses from randomized controlled trials of TH for neonatal HIE lend insight into long-term outcomes in the era of TH. Childhood outcomes from the NICHD trial compared outcomes of cooled and non-cooled infants, demonstrating lower rates of death or severe disability at 6 to 7 years of age.²⁸ Outcomes of the TOBY trial demonstrated that children undergoing TH were more likely to be without neurologic abnormalities at age 6 to 7 years, have significantly lower rates of cerebral palsy, and lower risk of moderate or severe disability.¹⁹ However, even with the proven benefits of TH children with perinatal HIE continue to have higher rates of cognitive and learning impairment than peers.^{21, 60} These risks are present even in the absence of cerebral palsy and warrant ongoing assessments through school age.⁶¹ The impacts of HIE on long-term neurodevelopmental impairments including higher-order cognitive functions in children without cerebral palsy is now only starting to be understood.

Cognitive development

Long-term outcome studies of school-age children with a history of neonatal encephalopathy from the pre-cooling era indicate a range of cognitive outcomes, even in the absence of motor disability or cerebral palsy.^{53, 62–64} The same holds true in the era of therapeutic hypothermia. It is known that cognitive functioning is critical for academic success, achieving adult independence, and for occupational opportunities. The currently available studies in the post-cooling era are limited by small sample size, lack of a control cohort, and/or duration of child follow-up but overall find that there is a range of cognitive sequelae for children with neonatal HIE in the absence of significant neuromotor disability (Table 1). These sequelae include deficits in global intelligence (IQ), executive function, memory, language and nonverbal reasoning skills which are all important components of cognitive functioning.^{65, 66}

Table 1.

Summary of studies reporting cognitive outcomes in school-age children with history of perinatal HIE who underwent TH.

Study	Age	Number of children	CP	Cooled?	Key findings	Outcome measure	Category
Cainelli, E., et al., 2021	5–7 years	40	No CP	Yes	Children with HIE scored lower than controls on measures of IQ, visuomotor skills, executive functions, and attention. IQ test effectively differentiated between the two groups (HIE vs control group).	WPPSI-IV, WISC-IV, NEPSYII, Semantic verbal fluency test, Tower of London task, Corsi test, Visual-Motor Integration Test, CBCL, CRS-R	attention, cognition, emotion-behavior
Edmonds, C.J., et al., 2021	5–7 years	31	Excluded children with CP > level 2 on GMFCS	Yes	Children with HIE were not as school-ready as controls - general cognitive abilities within normal range, but composite scores mark specific underlying difficulties identified by more focused assessments. Teachers identified significantly more subjective concerns with learning in multiple areas of executive functioning including initiation, working memory, planning and organizing, and behavioral regulation. Scored lower than controls on auditory attention, memory, language.	WPPSI-IV, WIAT-II, CELF-2, NEPSY-II, BRIEF, RBMT, Touwen Neurological Exam, MABC-2	cognition, motor, emotion-behavior, language
Jary, S., et al., 2019	6–8 years	29	No CP	Yes	MABC-2 scores at or below 15% were associated with lower full-scale IQ, Working Memory, Perceptual Reasoning scores, and 64% needed greater support in school.	MABC-2, WISC-IV, BSID-III	motor, cognitive
Lee-Kelland, R., et al., 2020	6–8 years	29	No CP	Yes	Children with HIE had lower mean full-scale IQ than controls, more emotional and behavioral difficulties, and approximately one third needed extra support in school.	MABC-2, WISC-IV, SDQ	motor, emotion-behavior, cognition
Pappas, A., et al., 2015	18–22 months and 6–7 years	63	CP and no CP	Yes	9% of children without CP had IQ <70, 31% had IQ 70–84, similar between TH and normothermia. Mental developmental index <70 at 18 months had lower IQ at 6 to 7 years. 20% of children with normal IQ and 28% with IQ 70–84 received special education services or were held back a grade level.	BSID-II, WISC, NEPSY	attention, cognition, language
Robertsson Grossmann K et al., 2023	6–8 years, 10–12 years	57	9 had CP	Yes	In the absence of major neuromotor impairment, children had normal intelligence. Longitudinal evaluation with assessment in early adolescence demonstrated emerging deficits in executive function in about a quarter of children with previously favorable prognosis.	MABC-2, WISC IV-V	cognition

BSID-II or III: Bayley Scales of Infant Development-II or Bayley Scales of Infant Development-III

BRIEF: Behavior Rating Inventory of Executive Functions

CBCL: Child Behavior Checklist

CELF-2: Clinical Evaluation of Language Fundamentals-2

CHQ: Child Health Questionnaire

CRS-R: Connors Rating Scales- Revised

GMFCS: Gross Motor Function Classification System

MABC-2: Movement Assessment Battery for Children-2

NEPSY-II: Neuropsychologic Assessment-2

RBMT: Rivermead Behavioral Memory Test

SDQ: Strengths and Difficulties Questionnaire

WIAT-II: Wechsler Individual Achievement Test

WPPSI-IV: Wechsler Preschool and Primary Scale of Intelligence

WISC-IV or WISC-V: Wechsler Intelligence Scales for Children 4 or Wechsler Intelligence Scales for Children 5

It is helpful to identify cognitive and learning disabilities early, ahead of school entry. For children with more significant cognitive impairments, they can often be recognized early by standard evaluations. However, especially for children with milder impairments, early childhood evaluations during preschool years of cognitive function may not be able to fully appreciate higher-order cognitive dysfunction that may emerge and be identified only at older ages when more comprehensive cognitive testing can be performed.⁶⁷ For example, the Bayley Scales of Infant and Toddler Development-III/-IV which is a widely used evaluation for outcomes in early childhood has poor concurrent validity with cognitive scores at school-age especially among children with more mild impairments.⁶⁸ Thus, a Bayley cognitive score in the normal range at age 2 or 3 years does not provide assurance that cognitive function at school-age will be normal. Additionally, longitudinal evaluations with assessment in early adolescence identify emerging deficits in executive function in about a quarter of children with previously favorable prognosis.⁶⁹ Thus, it is important to monitor cognitive development through school-age for all children with a history of neonatal encephalopathy or HIE.

In a systematic review by Schreglmann and colleagues, including articles from the pre- and post-cooling for HIE eras of studies reporting cognitive and/or behavioral outcomes of children 4 years or greater without cerebral palsy, a substantial number of children from these studies (25–63%) have an impairment in general cognitive abilities.⁷⁰ In the systematic review, the age-appropriate Wechsler was the most commonly used test to measure intellectual abilities, and overall showed a greater number of children to have full-scale IQ scores in the −1 standard deviation and −2 standard deviation ranges compared to normative cohorts suggesting that children with a history of neonatal HIE without cerebral palsy have an increased risk for cognitive impairment at school-age.⁷⁰ In a case-control study of 29 children aged 6–8 years with a history of HIE who received TH and do not have cerebral palsy, mean full-scale IQ scores on WISC-IV were significantly lower in the cases than controls (mean difference: −13.62, 95% CI −20.53 to −6.71).⁶⁵ About a third of the children were receiving additional academic support at school,⁶⁵ indicating a recognized need for modifications to help support their learning and school performance. In a case-control study assessing school readiness in children with a history of HIE treated with TH, children with HIE (cases) had lower memory performance scores including working memory compared to non-HIE controls.⁷¹ These differences contribute to early difficulties with learning and academic achievements. Given the important differences recognized by the available literature on long-term childhood outcomes and their limitations with many having a small sample size, a long-term outcome study is underway in Canada to better describe neurodevelopmental and cognitive consequences of neonatal encephalopathy.⁷² Large case-control studies of school-age children with HIE treated with TH are needed to better understand the higher-order cognitive challenges that these children may face as they enter adolescence and beyond.

Motor outcomes

While significant motor impairments like cerebral palsy tend to correlate strongly with abnormal imaging patterns on neonatal MRI, the milder spectrum of motor impairments do not correlate well with neonatal MRI findings,⁷³ and can occur even in the absence of brain injury on MRI.⁷⁴ Furthermore, since the standardization of TH, risks of abnormal motor outcomes in childhood after neonatal HIE seem to evolve over time and may be missed by early developmental screenings. In one study in the cooling era, 38% of children with HIE had impairments in manual dexterity, aiming and catching, and balance at 6–8 years.⁷⁵ Importantly, this poor motor performance at school age was not predicted by 18-month Bayley Scales of Infant and Toddler Development-III (Bayley-III) scores.⁷⁵

Motor impairment in HIE post-cooling is a spectrum that may manifest in more subtle ways than cerebral palsy (Table 2). Developmental coordination disorder (DCD) is a neurodevelopmental disorder of motor coordination that affects balance, coordination, dexterity, fine motor and gross motor skills.⁷⁶ DCD can impact quality of life, activities of daily living, and occupational opportunities and may not become evident until school age. In a study of long-term motor outcomes of cooled infants with HIE without moderate to severe cerebral palsy, 14 of 45 children (31%) had significant motor difficulties at 11 years of age on the Movement Assessment Battery for Children-2 (MABC-2).⁷⁷ Of these children, five were diagnosed with DCD. These children had scores that were within the lower range of normal on their 2-year Bayley-III evaluation,⁷⁷ and thus were not identified as being at risk for a complex DCD or other developmental motor concerns. Without early recognition of motor impairments, early intervention therapies may not be initiated at the developmental time windows when there is the greatest opportunity for developmental neuroplasticity.

Table 2.

Summary of studies reporting motor outcomes in school-age children with history of perinatal HIE who underwent TH and do not have a major motor disability.

Study	Age	Number of children	Cerebral palsy	Cooled?	Key findings	Outcome measure	Category
Edmonds, C.J., et al., 2021	5–7 years	31	Excluded children with CP > level 2 on GMFCS	Yes	Scored lower than controls on fine motor skills (even when excluding those with CP).	WPPSI-IV, WIAT-II, CELF-2, NEPSY-II, BRIEF, RBMT, Touwen Neurological Exam, MABC-2	cognition, motor, emotion-behavior
Erdi-Krausz, G., et al., 2021	5–7 years	27	No CP	Yes	Compared to controls, HIE cases were more likely to have minor neurologic dysfunction and motor difficulties on MABC-2 (p=.006). Parents were more concerned about motor skills. Neither neonatal MRI findings nor Bayley scores at 2 years correlated with school age motor outcomes.	MABC-2, Simplified Touwen Neurological Exam, WIPPSI-IV, ADHD-RS, PedsQL	motor, attention
Eriksson Westblad, M., et al., 2023	11-years	45	No moderate to severe CP	Yes	14 children with significant motor difficulties in early adolescence, not predicted by 2-year Bayley-III scores. 5 children diagnosed with Developmental Coordination Disorder.	MABC-2, WISC-V, BSID-III, AIMS, MABC-2 Checklist (parent-survey)	motor
Jary, S., et al., 2019	6–8 years	29	No CP	Yes	38% demonstrated poor motor performance (at or below 15th percentile), 24% were at or below the 5th%. Poor motor performance not predicted by 18-month Bayley-III scores. MABC-2 scores at or below 15% were associated with lower full-scale IQ, working memory, and perceptual reasoning scores.	MABC-2, WISC-IV, BSID-III	motor, cognitive
Lee-Kelland, R., et al., 2020	6–7 years	29	No CP	Yes	Children with HIE had lower motor performance in manual dexterity, aiming and catching compared to controls.	MABC-2, WISC-IV, SDQ	cognitive, emotional-behavioral, motor

ADHD-RS: Dupaul ADHD Rating Scale

AIMS: Alberta Infant Motor Scale

BRIEF: Behavior Rating Inventory of Executive Functions

BSID-III: Bayley Scales of Infant Development-III

CELF-2: Clinical Evaluation of Language Fundamentals-2

MABC-2: Movement Assessment Battery for Children-2

NEPSY-II: Neuropsychologic Assessment-II

PedsQL: Pediatric Quality of Life Inventory

RBMT: Rivermead Behavioral Memory Test

SDQ: Strengths and Difficulties Questionnaire

WIAT-II: Wechsler Individual Achievement Test

WISC-IV or WISC-V: Wechsler Intelligence Scales for Children 4 or Wechsler Intelligence Scales for Children 5

WPPSI IV: Wechsler Preschool and Primary Scale of Intelligence 4

These more subtle motor difficulties, especially affecting manual dexterity and coordination, have also been demonstrated in multiple cohort studies of children with a history of HIE who were cooled.^{65, 71, 73} Table two summarizes relevant clinical studies on school age motor function in the era of TH. Some authors have proposed that these types of findings reflect a problem with cognitive processing impairments rather than pure motor skill dysfunction, due to the timed nature of the study tasks.⁵⁸ However, when surveyed, parents reported being subjectively more concerned about their children’s motor function.⁷³ The day-to-day observations of parents are important context for interpreting standardized motor developmental testing and suggest these may be true motor skill impairments.

Emotional and behavioral outcomes

Although emotion regulation and behavioral problems are common sequelae of other neonatal neurologic brain injuries, such as prematurity-related brain injury,⁷⁸ very few studies in the era of TH have focused on these types of long-term outcomes after HIE (Table 3). While few studies focus on the risks of mood disorders in neonatal HIE after TH, one small cohort study of children ages 3 to 6 years without cerebral palsy, demonstrated higher scores than controls for anxious and depressive symptoms on the Child Behavior Checklist (CBCL).⁷⁹ These children also scored higher, indicating more symptoms, on the Preschool Symptom Self-Report, a pictorial instrument that assesses presence of depressive symptoms in preschool aged children.⁷⁹ The authors stress the importance of continued investigations of mental health as a sequelae of perinatal HIE, noting that in adult stroke, major mood disorders follow in approximately a third of affected patients.⁸⁰ In younger HIE cohorts increased internalizing problems are associated with parent mental health concerns,⁸¹ underscoring the need for exploring the role of these potentially modifiable environmental factors in school age.

Table 3.

Summary of studies reporting emotional and behavioral outcomes in school-age children with history of perinatal HIE who underwent TH.

Study	Age	Number of children	CP	Cooled?	Key findings	Outcome measure	Category
Álvarez-García, M., et al., 2022	3–6 years	14	No CP	Yes	Children with HIE scored higher than controls for anxious and depressive symptoms, as well as aggressive behaviors on CBCL and more depressive symptoms on PRESS. Increased risk for mood disturbances.	CBCL, PRESS, ASQ-3	emotion-behavior
Cainelli, E., et al., 2021	5–7 years	40	No CP	Yes	Incidence of psychopathology was significantly higher in children with HIE. Clinically relevant psychopathological symptoms in 35% of children with HIE. Most frequent symptoms reported: oppositional behavior, attention problems, hyperactivity, somatic complaints, anxiety.	WPPSI-IV, WISC-IV, NEPSY-II, Semantic verbal fluency test, Tower of London task, Corsi test, Visual-Motor Integration Test, CBCL, CRS-R	attention, cognition, emotion-behavior
Cainelli, E., et al., 2022	6–9 years	19	No CP	Yes	Compared to controls, HIE group was more likely to have impaired affect recognition, higher rates of ADHD, separation anxiety, depressive mood, and behavioral concerns.	Social evaluation domain of the NEPSY-II, CBCL, K-SADS-PL	attention, emotion-behavior
Edmonds, C.J., et al., 2021	5–7 years	31	Excluded children with CP > level 2 on GMFCS	Yes	Children with HIE are not as school-ready as controls. Teachers identified significantly more subjective concerns about behavioral regulation.	WPPSI-IV, WIAT-II, CELF-2, NEPSY-II, BRIEF, RBMT, Touwen Neurological Exam, MABC-2	cognition, motor, emotion-behavior
Lee-Kelland, R., et al., 2020	6–7 years	29	No CP	Yes	Children with HIE had more emotional and behavioral difficulties compared to controls.	MABC-2, WISC-IV, SDQ	Cognitive, motor, emotion-behavior
Zareen, Z., et al., 2021	4–6 years	30 with moderate to severe HIE	7 with CP	24 cooled	Children with HIE had increased bedtime resistance and sleep anxiety compared to controls. Children with HIE had lower quality of life scores than controls.	CHSQ, PedsQL	emotion-behavior

ASQ-3: Ages and Stages Questionnaire-3

BRIEF: Behavior Rating Inventory of Executive Functions

CBCL: Child Behavior Checklist

CELF-2: Clinical Evaluation of Language Fundamentals-2

CHSQ: Childhood Sleep Habits Questionnaire

CRS-R: Connors Rating Scales- Revised

K-SADS-PL: Kiddie Schedule of Affective Disorders and Schizophrenia for School-Age Children-Present and Lifetime Version

MABC-2: Movement Assessment Battery for Children-2

NEPSY or NEPSY-II: Neuropsychologic Assessment or Neuropsychologic Assessment-2

PedsQL: Pediatric Quality of Life Inventory

PRESS: Preschool Symptom Self-Report

RBMT: Rivermead Behavioral Memory Test

SDQ: Strengths and Difficulties Questionnaire

WIAT-II: Wechsler Individual Achievement Test

WPPSI IV: Wechsler Preschool and Primary Scale of Intelligence 4

WISC-IV: Wechsler Intelligence Scales for Children 4

Aggressive behaviors,⁷⁹ sleep anxiety and bedtime resistance,⁸² separation anxiety,⁸³ and emotional and behavioral regulation difficulties,^{65, 71} have also been reported in HIE outcome studies of cooled children. In another sample of 40 children with HIE who underwent TH, psychopathology was significantly higher (35%) compared to control peers (12%) at school age.⁸⁵ The most frequent symptoms reported were oppositional behavior, attention problems, hyperactivity, somatic complaints, and anxiety.⁸⁵ More research is needed to better define long-term emotional and behavioral outcomes in children in this population, and additional risk factors, as the emotional well-being of a child with HIE is a core hope of most parents as well as a key component of health-related quality of life.⁸⁶

Autism spectrum disorder

Autism spectrum disorder (ASD) is a complex neurodevelopmental disorder comprising differences in communication and social skills recognized in early childhood that has been identified to have multiple etiologies and risk factors including gene-environment interactions.⁸⁷ For some children, the diagnosis of ASD can be made due to structural brain injury in locations that affect communication and socialization. In other children, ASD may be diagnosed in the absence of structural brain injury with a history of prenatal, perinatal, and postnatal exposures. Prenatal exposures to maternal infection and inflammation may also be risks for abnormal fetal-neonatal transition and HIE and thus may contribute to the etiology for both the diagnosis of neonatal encephalopathy and subsequent diagnosis of autism following early neurodevelopment.⁸⁸ For children with communication and language delay, the diagnosis of autism can be helpful to providing more therapy opportunities, (i.e. Applied Behavior Analysis therapy). Overall, there is a lack of studies in the post-cooling era on the specific outcome of autism following HIE, which represents an important knowledge gap to be addressed. Understanding the risk for this outcome is important for families and prognostic counseling as many families know about autism as a neurodevelopmental disorder due to its media attention and ask about this risk for their child.

Data from the pre-cooling era indicates an association between perinatal hypoxia-ischemia and risk for autism.⁸⁹ It may be that a combination of factors in addition to exposure to perinatal hypoxia-ischemia, perhaps also in those with a genetic risk, increases the risk of the outcome of autism as opposed to a single factor. In a population-based study in Western Australia in the pre-cooling era of newborns with moderate to severe neonatal encephalopathy and term control peers with follow up to five years of age, children with a history of neonatal encephalopathy were 5.9 times (95% CI 2.0–16.9) more likely to have an autism diagnosis compared to children without neonatal encephalopathy.⁵⁹ Overall, in this study, 12 of 239 (5%) children with neonatal encephalopathy had a diagnosis of ASD at age 5 years.⁵⁹ This large population-based study provides some evidence of the association between perinatal encephalopathy and autism risk, although in the pre-cooling era and should be replicated in a modern population-based study of children HIE who received TH.

Studies including autism in the era of TH for HIE are limited. In a single center study of 107 children with HIE treated with TH with evaluation at 2 years of age, a higher rate of autism was not observed by the Quantitative-Checklist for Autism in Toddlers (Q-CHAT).⁸⁴ In a systematic review of studies reporting the social emotional and psychological sequelae of HIE in the pre- and post-cooling eras, 21 of 43 articles were identified in which children received TH. Overall studies reported a wide range of behavioral, emotional, and psychological outcomes including autism and ADHD.⁹⁰ Thus, given the multifactorial nature of autism and the findings of available literature, it is important to ensure appropriate developmental screening for autism and related diagnoses for all children with a history of neonatal encephalopathy.

Attention and hyperactivity

In the pre-cooling era, a study comparing 28 children with HIE and without cerebral palsy to their siblings without HIE found a higher risk of ADHD on the Connors scale (p < 0.003) and the Inattention subscale of the ADHD Rating Scale IV (p < 0.006).⁵⁶ Marlow et al. found higher hyperactivity subscale scores (p < .01) in a similar cohort of children with HIE compared to controls without HIE.⁵³ While these two pre-cooling studies indicate a higher risk for perceived attention problems in this population, the literature on attention and associated disorders in the era of TH is sparse (Table 4). Two small studies of school age children without CP identified higher rates of ADHD⁸³ and poorer attention⁸⁵ in HIE compared to controls. Neither Pappas et al. nor Shankaran et al. found specific differences in attention between children with HIE who received TH and those who did not at 6–7 years,^{21, 60} which may suggest that TH is not protective against attention deficit disorders and that additional studies are needed in modern cohorts.

Table 4.

Summary of studies reporting language and attention-related outcomes in school-age children with history of perinatal HIE who underwent TH.

Study	Age	Number of children	CP	Cooled?	Key findings	Outcome measure	Category
Cainelli, E., et al., 2021	5–7 years	40	No CP	Yes	Children with HIE scored lower than controls on measures of attention. Clinically relevant psychopathological symptoms in 35% of children with HIE. Most frequent symptoms reported included attention problems, hyperactivity.	WPPSI-IV, WISC-IV, NEPSYII, Semantic verbal fluency test, Tower of London task, Corsi test, Visual-Motor Integration Test, CBCL, CRS-R	attention, cognition, emotion-behavior
Cainelli, E., et al., 2022	6–9 years	19	No CP	Yes	Compared to controls, HIE group with higher rates of ADHD.	Social evaluation domain of the NEPSY-II, CBCL, K-SADS-PL	attention, emotion-behavior
Edmonds, C.J., et al., 2021	5–7 years	31	Excluded children with CP > level 2 on GMFCS	Yes	Scored lower than controls on language assessments, particularly word structure.	WPPSI-IV, WIAT-II, CELF-2, NEPSY-II, BRIEF, RBMT, Touwen Neurological Exam, MABC-2	cognition, motor, emotion-behavior, language
Erdi-Krausz, G., et al., 2021	5–7 years	27	No CP	Yes	Higher inattention scores in HIE vs controls. Parents did not rate inattention as worse in HIE group vs controls, but teachers did.	MABC-2, Simplified Touwen Neurologic Exam, WIPPSI-IV, ADHD-RS, PedsQL	motor, attention
Pappas, A., et al., 2015	18–22 months and 6–7 years	63	CP and no CP	Yes	Attention scores are similar between normothermia and hypothermia group (although, 17% of hypothermia and 28% of normothermia deemed untestable due to attention or executive function).	BSID-II, WISC, NEPSY	attention, cognition
Robb, T.J., Tonks, J., Spencer, A.P.C. et al. 2022	6–8 years	48	No CP	Yes	Significantly lower general communication composite scores as well as structural and pragmatic language scores compared to controls, greater language impairments and need for more support in school. 18-month Bayley-III language composite scores correlated with early school age scores. Language scores correlated with FA of widespread white matter.	CCC-2	language
Shankaran, S., et al., 2012	6–7 years	97	12 in TH group	Yes	Similar proportion of children with attention and executive function problems (4% TH vs 13% normothermia).	WPPSI-III, WISC-IV, NEPSY, CHQ	attention, cognition

BRIEF: Behavior Rating Inventory of Executive Functions

CBCL: Child Behavior Checklist

CCC-2: Children’s Communication Checklist

CELF-2: Clinical Evaluation of Language Fundamentals-2

CHQ: Child Health Questionnaire

K-SADS-PL: Kiddie Schedule of Affective Disorders and Schizophrenia for School-Age Children-Present and Lifetime Version

MABC-2: Movement Assessment Battery for Children-2

NEPSY or NEPSY-II: Neuropsychologic Assessment or Neuropsychologic Assessment-2

RBMT: Rivermead Behavioral Memory Test

WIAT-II: Wechsler Individual Achievement Test

WPPSI-III or IV: Wechsler Preschool and Primary Scale of Intelligence 3 or Wechsler Preschool and Primary Scale of Intelligence 4

WISC-IV: Wechsler Intelligence Scales for Children 4

Language and communication

Early language development was studied by Chin et al. in a cohort of 45 preschool aged children who underwent TH for HIE.⁹¹ At a mean age of 26 months, the cohort demonstrated poor expressive language, decreased vocabulary, and shorter utterances. While overall receptive language scores were within the normal range, the presence of injury on neonatal brain MRI predicted poorer receptive language. Few studies have focused on school age language outcomes in these children (Table 4). One study of children with HIE and TH age 5–7 years identified particular difficulty with word structure rules compared to peers,⁷¹ which involves challenges with word morphology, inflections, derivations, and pronoun use. In another cohort of 6- to 8-year-old children with a history of TH for HIE, 48 children without CP demonstrated poorer communication and pragmatic language, and had a higher degree of language impairments than age-matched controls.⁹² Unlike cognition and motor skills,^{68, 75, 77} school-age language scores in this cohort were predicted by 18-month Bayley-III scores.⁹² Additionally, in this study, the general composite language score was found to correlate with the fractional anisotropy of the corpus callosum, the cingulum, and the superior longitudinal fasciculus at 6–8 years.⁹² Additional quantitative imaging markers have been shown to predict early childhood language outcomes (18 to 30 months) in children cooled for HIE,^{93, 94} suggesting that advanced imaging methods have the potential to supplement conventional MRI, particularly in cases without overt brain injury. Further exploration of these tools may lead to important prognostic information about language function after neonatal HIE.

Socioeconomic status (SES) also has a striking influence on language outcomes in HIE. In the Chin study, estimated household income predicted deficits in expressive and receptive language.⁹¹ In another study, increased social risk was associated with lower 3-year cognitive and language scores, adjusting for neonatal brain imaging findings. When adjusting for social risk, associations between neonatal brain imaging and 3-year cognitive and language outcomes were not significant.⁸¹ These findings highlight the fallibility of a single imaging study and the concerning reality that children of lower SES may be negatively, disproportionately impacted by HIE.

Counseling parents about outcomes

Discussing prognosis for neonates with HIE is both challenging for clinicians and emotional for families. Parents at this uncertain time are often experiencing a sense of loss and grief,⁹⁵ and a provider communicating in a clear and empathic manner can promote parental well-being.⁹⁶ When surveyed, parents of children with HIE treated with TH indicate that communication from the clinician and receiving information as quickly as possible are of greatest importance.⁹⁷ Throughout the course of the NICU hospitalization for HIE and TH, providers interpret and synthesize neuromonitoring data in the form of neurological examinations, electrophysiology, and available imaging to share an impression with parents about the degree of the injury as it evolves.⁹⁸ Early and frequent conversations with parents may help to set expectations, as parents are often anxiously awaiting MRI results for several days until TH is completed.⁹⁹ While MRI is a crucial factor in determining severity of brain injury after TH, its utility and its limitations in the absence of overt injury should be shared with families. When brain MRI is reassuring against severe forms of injury, clinicians should still discuss the wider range of possible long-term outcomes and the need for ongoing follow-up and support.

In the case of a relatively normal brain MRI, providers should acknowledge the feelings of relief and joy that news of an MRI without significant injury brings to many parents. Families should be given reassurance that severe outcomes like cerebral palsy, global developmental delays, and major disability are unlikely (unless there are other indicators on EEG or neurological examinations that indicate otherwise). Rather, the risk for subtler impacts to learning, behavior, language, and motor coordination that are not as predictable from MRI should be discussed. While caution is recommended, the goal of these conversations is not to burden families with vigilant anxiety about the future, but lace optimism with awareness that milder impairments may yet emerge during childhood. We simultaneously provide assurance that these impacts are often mild and manageable when identified early, and that the quality of life for these children remains high.⁸⁶ However, because milder impairments often may not become evident until school age, serial neurodevelopmental assessments are recommended. Establishing a therapeutic alliance with the follow-up team early on and maintaining their follow-up through school age is critical to ensuring that the child is getting the evaluations and support that they need at the time different types of impairments may arise.

Developmental follow-up timelines are likely to vary widely depending on local providers and resources. Ideally, neonatal neurologists or child neurologists would be part of multidisciplinary neurodevelopmental evaluations from early on, as they may be best able to synthesize all available neurologic data from the NICU stay and during serial follow-up evaluations. Our current practice involves an outpatient follow-up visit at 2–3 months, 6–9 months, 15–18 months, 2-years, and 3-years with a neonatal neurologist and a physical therapist. Developmental child psychologists evaluate these children as well at staggered time points depending on needs, and at 4–6 years of age, a school readiness neuropsychology evaluation is offered. Neuropsychological testing should continue to be offered periodically throughout school age, for example, at 8–10 years, and at the transition to middle school. More focused neuropsychological assessments at these ages may identify specific underlying challenges, that are often missed or masked as low-normal scores in broader categories or composite scores.⁷¹ While this type of follow-up program may look different at different institutions, the most important aspect of the program is to have a serial follow-up schedule with providers trained in monitoring child development to enable early identification of differences in child development and learning. Through this approach we can help to optimize the long-term outcomes for these children and support them and their families as they grow.

Socio-economic influences on outcomes of HIE

Socio-economic status (SES) is an often overlooked but crucial factor in child health, well-being, and long-term outcomes. Despite TH, neurodevelopmental outcomes are impacted by SES in a myriad of ways. It is well-established that SES impacts neurodevelopmental outcomes of healthy children¹⁰⁰ and preterm infants.¹⁰¹ In the aforementioned Robb et al. study on language outcomes, SES correlated with language outcomes in children with HIE, but not in controls,⁹² suggesting a vulnerable neonatal brain may be more susceptible to inequalities of SES than a healthy one. Recent research suggests that parental education level and home literacy environment, both proxies for SES, positively impact cognitive outcome in cooled infants with HIE.¹⁰²

Poorer parental mental health and social risks can negatively impact child psychosocial and language outcomes.⁸¹ The parent experience of HIE is often traumatic regardless of the long-term outcome due to separation from the infant, lack of physical bonding, touch, basic caretaking roles in the NICU, and lack of peer support.^{95, 103} Given these risks, mental health and social workers are a benefit in the acute NICU setting,¹⁰⁴ and a trauma informed approach to care and communication is recommended through NICU admission and afterwards.¹⁰⁵ Emotional burdens, financial strains, and fear of the unknown continue to burden parents after discharge.¹⁰⁶ Clinicians should be aware of community resources for parental mental health, food insecurity, early intervention services, and screen for these concerns at follow-up visits to help identify families who may need more support. A strong relationship with a child’s provider in longitudinal follow-up can help parents cope and process their experiences.¹⁰⁵ For added layers of social support, parents should be directed to relevant peer support groups specific to families affected by HIE. Additionally, parents of children with HIE should be empowered with developmentally appropriate activity ideas and encouraged to read to their children early and often.

Parallels with evolution of care in premature neonatal brain injury

A parallel for long-term outcomes of HIE in the era of TH may be drawn from another neonatal neurologic pathology, preterm brain injury. With advances in care for complications of preterm birth, the rates of major prematurity-related structural brain injury have fortunately decreased.¹¹⁴ While the risks of spastic cerebral palsy in these vulnerable infants has declined,^{115, 116} behavioral and emotional issues comprising the “preterm behavioral phenotype”^116–119 and DCD have become more prevalent.¹²⁰ The preterm behavioral phenotype is a pattern of behavioral and emotional difficulties, ADHD, ASD, and anxiety disorders that is highly prevalent in preterm born children.¹¹⁸ While a more favorable outcome than cerebral palsy and severe developmental disability, these sequelae nevertheless can have a range of impacts on everyday social interactions, play, learning, and long-term outcomes. As neuroprotective care for infants with HIE continues to improve and more long-term data emerges, attention to long-term outcomes of the whole child will be necessary to characterize the milder phenotypes of survivors of HIE.

Conclusion

In the era of TH, while children with a history of moderate to severe HIE are less likely to have severe brain injury and disability, they remain at risk for many challenges that impact learning, emotions and behavior, language, and coordination that emerge at older ages of child development. Risks for these more subtle and insidious challenges are not well-correlated with neonatal imaging findings, making early prognostication and counseling families in the NICU period and during neonatal follow-up clinics challenging. Additionally, early developmental testing in childhood does not always predict these school-age challenges. Thus, all children with a history of perinatal HIE in the cooling era should have long-term follow up and support through school age. Similar to studies in preterm neonates,¹²¹ children with socioeconomic deprivation should be considered especially vulnerable to the long-term sequelae of HIE, even in the era of TH.

Impact Statement.

• Children with a history of moderate to severe hypoxic ischemic encephalopathy (HIE) without evidence of significant brain injury on post-cooling MRI are still at risk for a range of neurodevelopmental challenges at school-age and beyond.

• Neonatal neuroimaging and early neurodevelopmental evaluations do not reliably indicate outcomes at school-age, making early prognostication and counseling families in the NICU period and during neonatal follow-up clinics challenging.

• All children with perinatal HIE, regardless of MRI findings, should have long-term follow up and support through school age.