Hypothermia for newborns with hypoxic-ischemic encephalopathy

---

The Canadian Paediatric Society gives permission to print single copies of this document from our website. For permission to reprint or reproduce multiple copies, please see our copyright policy.

BACKGROUND

Neonatal encephalopathy is a clinical syndrome of disturbed neurological function that presents early in life, with an incidence of approximately 1/1000 to 6/1000 live births [1]. Hypoxic-ischemic encephalopathy (HIE) accounts for a significant proportion of encephalopathic newborns. Despite advances in perinatal care, moderate-to-severe acute perinatal HIE in late preterm and term infants remains an important cause of mortality, acute neurological injury and subsequent long-term neurodevelopmental disability [1]. Impaired cerebral blood flow, in the setting of hypoxia, is the main mechanism causing brain injury following intrapartum hypoxia-ischemia.

At the cellular level, hypoxia-ischemia results in two phases of energy failure. The primary phase follows the reduction in blood flow and oxygen supply, with a fall in adenosine triphosphate (ATP), failure of the sodium (Na+)/potassium (K+) pump, depolarization of cells, lactic acidosis, release of excitatory amino acids, calcium entry into the cell and, when severe, cell necrosis [2]. Following resuscitation and reperfusion, there is a latent period of 1 to 6 hours where the impairment of cerebral oxidative metabolism can at least partially recover, before irreversible failure of mitochondrial function [2]. This latent phase is the therapeutic window for neuroprotective interventions [2][3].

The risk for disability and impaired cognitive development correlates with the severity of HIE [4][5]. A mild reduction in brain temperature of 2°C to 4°C, initiated within 6 hours after birth, was the first therapy to demonstrate neuroprotection in newborn animals [6]. The neuroprotective mechanism of therapeutic hypothermia is multifactorial and attributable to broad inhibitory activity against a variety of harmful cell processes.

HIE is often unanticipated, and many infants are initially cared for in community hospitals. It is important, therefore, that professionals involved in caring for these infants—family physicians, obstetricians, midwives, nurses, community paediatricians and neonatologists—recognize when cooling may be beneficial. This position statement summarizes the current evidence pertaining to therapeutic hypothermia for the neuroprotection of neonates with HIE and provides guidance for clinicians regarding this therapy.

STATEMENT DEVELOPMENT

A search of MEDLINE, including ‘in-process and other non-indexed citations’ (1946 to May 6, 2016), Embase (1974 to May 6, 2016), and the Cochrane Central Register of Controlled Trials (CENTRAL, April 2016) was performed by a certified librarian, using the OVID interface [7]. Search terms included: ‘hypothermia, induced’, ‘hypother* or cooling’, ‘hypoxia-ischemia, brain’, ‘hypox*or ischem*’, ‘newborn infant’, and ‘infan* or newborn* or newborn* or neonat*’. Reference lists of publications were reviewed. A total of 1511 references were retrieved, of which full text articles and Cochrane reviews were reviewed.

The hierarchy of evidence from the Centre for Evidence-Based Medicine (www.cebm.net) was applied to the publications identified. Recommendations are based on the format developed by Shekelle et al. [8].

SHOULD HYPOTHERMIA BE ROUTINELY OFFERED TO INFANTS WITH HIE?

A systematic review and meta-analysis of 11 trials (totalling 1505 newborns) and including seven studies with follow-up to at least 18 months, concluded that hypothermia is effective for decreasing mortality or moderate-to-severe neurodevelopmental delay (NDD) at 18 to 24 months (RR 0.75; 95% CI 0.68 to 0.83), with a number needed to treat (NNT) of 7 (95% CI 5 to 10), and increasing survival without NDD (RR 1.63; 95% CI 1.36 to 1.95) [9]–[19]. A further analysis based on the degree of clinical encephalopathy at randomization (five trials) showed benefits of hypothermia for both infants with moderate encephalopathy (RR 0.68; 95% CI 0.56 to 0.84), an NNT of 6 (95% CI 4 to 13) and those with severe encephalopathy (RR 0.82; 95%CI0.72to0.93), and an NNT of 7 (95% CI 4 to17).

Therapeutic hypothermia is therefore considered to be the standard of care for infants with moderate-to-severe HIE who meet inclusion criteria. Level of evidence: 1a

WHICH INFANTS SHOULD BE TREATED WITH HYPOTHERMIA?

The timing of injury is crucial. Therapeutic hypothermia is effective in acute perinatal insults (e.g., placental abruption, cord prolapse), rather than for antenatal or chronic insults. Current evidence shows that the infants who benefit from hypothermia are term and late preterm infants ≥36 weeks GA with HIE who are ≤6 hours old and who meet either treatment criteria A or treatment criteria B, and also meet criteria C [20]:

A. Cord pH ≤7.0 or base deficit ≥−16, OR

B.
 pH 7.01 to 7.15 or base deficit −10 to −15.9 on cord gas or 
blood gas within 1 h AND

1. 
History of acute perinatal event (such as but not limited to cord prolapse, placental abruption or uterine rupture) AND
2. Apgar score ≤5 at 10 minutes or at least 10 minutes of 
positive-pressure ventilation 


C. Evidence of moderate-to-severe encephalopathy, demonstrated by the presence of seizures OR at least one sign in three or more of the six categories shown in Table 1. 


When available, assessment with an amplitude-integrated electroencephalogram (aEEG) for at least 20 minutes to document abnormal tracings or seizures, particularly for infants with moderate encephalopathy, may be useful for determining eligibility. A normal aEEG does not predict a normal MRI or favourable acute outcome [21]. The use of continuous EEG (when available) or aEEG within the first 48 hours postbirth may be useful for detecting electrical seizures or mild HIE [22]. Recent retrospective studies suggest that infants with mild HIE may have sequelae and perhaps should be considered for therapeutic hypothermia [23]–[25]. However, at this point, there are insufficient data to make this recommendation.

Table 1. Criteria for defining moderate and severe encephalopathy
Category	Moderate encephalopathy	Severe encephalopathy
1. Level of consciousness	Lethargy	Stupor/coma
2. Spontaneous activity	Decreased activity	No activity
3. Posture	Distal flexion, full extension	Decerebrate (arms extended and internally rotated, legs extended with feet in forced plantar flexion)
4. Tone	Hypotonia (focal, general)	Flaccid
5. Primitive reflexes
Suck Moro	Weak Incomplete	Absent Absent
6. Autonomic system
Pupils Heart rate (HR) Respirations	Constricted Bradycardia Periodic breathing	Skew deviation/dilated/ nonreactive to light Variable HR Apnea

All infants who are depressed at birth should be assessed to determine whether they fulfill criteria A or B. Infants who fulfill criteria A or B should then undergo a careful neurological examination to determine whether they fulfill criteria C. Infants who meet criteria A and C or B and C should be offered hypothermia. Some cases are difficult to categorize and require discussion with a tertiary care neonatologist. As preparations for cooling are made, if an infant’s neurological status has fully returned to normal (within 30 to 45 minutes after birth), cooling may be deferred with careful ongoing observation for signs of returning encephalopathy. Level of evidence: 1a

WHICH INFANTS SHOULD NOT RECEIVE HYPOTHERMIA?

In most randomized controlled trials (RCTs) of therapeutic hypothermia, exclusion criteria included: moribund infants or infants with major congenital or genetic abnormalities for whom no further aggressive treatment is planned; infants with severe intrauterine growth restriction; infants with clinically significant coagulopathy; and infants with evidence of severe head trauma or intracranial bleeding. Isolated intraventricular hemorrhage may not constitute an absolute contraindication; however, this guidance is based on very limited evidence [26]. Despite a paucity of data showing that initiating cooling after 6 hours of age may be beneficial, therapeutic hypothermia may still be considered after discussion with parents, always recognizing the uncertainty with regards to its benefits and considering possible side effects (e.g., bleeding diathesis, hypotension, pulmonary hypertension and cardiac arrhythmias) [19][27]. Indications and contra-indications should be discussed as soon as possible with a tertiary care centre. Level of evidence: 1a

WHAT ARE THE SIDE EFFECTS OF HYPOTHERMIA?

In infants who received hypothermia, clinical trials have reported sinus bradycardia (a heart rate of 80 to 100 beats per minute), hypotension with possible need for inotropes, mild thrombocytopenia, and persistent pulmonary hypertension with impaired oxygenation. Hypothermia can also prolong bleeding time. Infants with HIE, whether they receive therapeutic hypothermia or not, are more prone to arrhythmias, anemia, leukopenia, hypoglycemia, hypokalemia, urinary retention and coagulopathy [11]. Clinicians caring for infants with HIE who receive therapeutic hypothermia should anticipate and closely monitor possible complications. Indications to stop hypothermia and rewarm the infant include: hypotension despite inotropic support; persistent pulmonary hypertension with hypoxemia, despite adequate treatment; and clinically significant coagulopathy, despite treatment. Subcutaneous fat necrosis, with or without hypercalcemia, has been reported as a potential rare complication [28]. Cooling is stopped uncommonly (<10% of cases) due to complications. Level of evidence: 2a

WHICH METHOD SHOULD BE USED TO PROVIDE HYPOTHERMIA?

Two methods of cooling have been used in clinical trials— selective head cooling and whole body cooling. One small single-centre trial compared both methods of cooling and found no difference in outcomes at 12 months of age in infants enrolled [29]. Larger clinical trials found similar effects on death and disability outcomes when either method was compared with placebo [30].

Selective head cooling can be achieved with cooling caps fitted around an infant’s head, with the aim of maintaining fontanelle temperature below 30°C. A rectal temperature of 34°C ± 0.5°C is maintained using a servo-controlled radiant heater. This system is expensive and labour-intensive, and may produce scalp edema or skin breakdown. It is also more difficult to maintain rectal temperature and there is limited access for EEG leads. Whole body cooling to a rectal temperature of 33.5°C ± 0.5°C can be achieved with passive cooling, cool packs and/or commercially available cooling blankets. Servo-controlled cooling blankets produce more consistent target temperatures [31].

Whole body cooling is recommended preferentially because it is easier to set up and use, less expensive, provides better access to EEGs and is more readily available. Level of evidence: 1b

WHERE SHOULD HYPOTHERMIA BE PROVIDED?

Hypothermia should be provided in centres that have both the personnel with expertise and competency in the intricacies of treatment—including the recognition of complications—and the resources to treat not only the multiorgan failure sometimes associated with HIE but other potentially serious complications of hypothermia. Centres providing hypothermia must have access to ultrasound, computed tomography and magnetic resonance imaging and be able to perform electroencephalograms (EEGs). Level 3 and 4 neonatal units are able to meet these requirements. Level of evidence: 5

WHEN SHOULD COOLING BE INITIATED?

In newborn rat pups, a 12-hour delay in starting hypothermia increased brain injury after severe hypoxia-ischemia, which is of clinical concern if the same findings are applicable to human newborns [32]. While awaiting transport to a tertiary care unit, initiation of passive cooling (e.g., removing the infant’s hat, blanket and turning off an overhead warmer) should be strongly considered in community hospitals, following consultation with a receiving neonatologist. Rectal temperature or, if not possible, axillary temperature should be monitored every 15 minutes to ensure the infant’s temperature does not decrease below 33°C.

The earlier passive cooling is initiated, the earlier target temperature will be reached [33]. The use of ice packs or cool gels by untrained personnel, which can lead to severe hypothermia, is not recommended. Level of evidence: 2b

WHAT IS THE TARGET TEMPERATURE TO BE REACHED?

The optimal rectal or esophageal temperature appears to be 33.5°C ± 0.5°C for whole body cooling and 34.5°C ± 0.5°C for selective head cooling. Cooling to 32.0°C in one study was not found to yield superior outcomes [31]. Level of evidence: 1a

HOW LONG SHOULD HYPOTHERMIA BE CONTINUED?

Studies derived from animal models suggest that hypothermia is most effective when instituted in the latent phase, before secondary failure of oxidative mechanisms, and continued throughout the secondary phase [2]. Most clinical trials used 72 hours of cooling in their treatment group. A recent study did not find cooling longer (i.e., 120 hours) superior to cooling for 72 hours [34]. Level of evidence: 1a

HOW SHOULD THE INFANT BE REWARMED?

The speed of rewarming remains controversial. In randomized clinical trials, infants were rewarmed over 6 to 12 hours (0.5°C every 1 to 2 hours). Most centres rewarm infants by 0.5°C every 1 to 2 hours. Seizures and worsening of clinical encephalopathy upon rewarming have been reported [35]. In such circumstances, experts suggest recooling for 24 hours and resuming rewarming [36]. Level of evidence: 5

CAN HYPOTHERMIA BE PROVIDED TO PRETERM INFANTS?

Hypothermia is associated with increased mortality in the preterm infant. There is currently no evidence that therapeutic hypothermia offers any benefits to infants younger than 35 weeks GA. Data in infants <35 weeks is limited to case reports, small cohort studies or studies evaluating cooling for conditions other than perinatal asphyxia, such as necrotizing enterocolitis [26][37][38]. A few infants as young as 35 weeks GA were included in some clinical trials [12][19] and some Canadian centres offer this therapy to infants at ≥35 weeks GA. Level of evidence: 4

SHOULD ADJUNCTIVE THERAPIES BE PROVIDED?

In animal models, induced hypothermia causes significant physiological stress and is associated with prolonged elevation of circulating cortisol levels after asphyxia, which could increase neuronal loss [39][40]. Infusion of an analgesic, such as morphine, significantly reduces plasma cortisol and noradrenaline concentrations in ventilated newborns, compared with placebo treatment [41]. However, in human newborns the effects of sedative and analgesic therapy during hypothermia on short- and long-term outcomes are unclear; exercising caution is essential [3]. Hypothermia leads to longer serum clearance of morphine, fentanyl and midazolam [42]. Infusions of morphine at rates higher than 10 mcg/kg/hour have been associated with toxic serum levels in a subset of infants enrolled in one study [43]. A low infusion of morphine (≤10 mcg/kg/h) or equivalent opioid is recommended as the initial approach for easing discomfort. Level of evidence: 3b

Similarly, antiepileptic drugs should be used with caution in newborns with HIE, due to their known neurotoxicity [44]. Despite this proviso, experts recommend to treat neonatal seizures, which are common in HIE and suspected to be an independent cause of brain injury [22][45][46]. Obtaining serum levels of antiepileptics, particularly in the first 72 hours if redosing is needed, should be strongly considered. Level of evidence: 4

Early minimal enteral feeding (10 mL/kg/day to 20 mL/kg/ day) during hypothermia, initiated during the first few days of life, is safe and feasible for newborns with HIE [47]. In fact, whole body hypothermia may even have beneficial effects on gastrointestinal morbidity and feeding tolerance [48]. However, more than minimal feeds is not as safe because gut perfusion may be decreased during cooling [49].

Minimizing fluctuations in blood carbon dioxide levels, avoiding hyperoxia, ensuring adequate tissue perfusion with appropriate use of vasopressors or inotropic agents, maintaining normal serum glucose, treating hyperbilirubinemia, and minimizing unnecessary stimulation or handling are additional management strategies to optimize outcome [50]–[53].

There has been considerable interest and ongoing research in evaluating the neuroprotective efficacy of different agents (allopurinol, xenon, melatonin, erythropoietin, neural stem cells and magnesium sulphate) in combination with hypothermia [54]–[60], but there is insufficient evidence to recommend their use at this time.

WHAT IS THE OPTIMAL TIMING FOR BRAIN IMAGING?

MRI is the preferred technique for imaging the brains of infants with neonatal encephalopathy [61]. Research exploring the predictive role of ultrasound and near-infrared spectroscopy is ongoing. In the precooling era, days 3 to 5 of life provided the best time window for MRI with diffusion-weighted imaging, for prognostic purposes and the possibility of redirecting clinical care [62]. A few cohort studies examining the correlation between MRI findings at various ages and later outcome now suggest that, in infants who receive therapeutic hypothermia, an MRI performed between days 2 and 4 correctly identifies lesions on the DWI sequence that are seen after day 10 on T1 and T2 sequences [63][64].

Performing an MRI in an infant undergoing active cooling is a challenge, due to the needs to maintain a steady temperature and compatibility of thermoregulation equipment with MRI. In the absence of an MRI-compatible isolette and other specialized equipment, it is recommended to obtain an MRI once rewarming has taken place, on day of life 4 or 5. Imaging can usually be done with the infant swaddled and after a feed, as opposed to under general anaesthesia. Centres should attempt to perform such MRIs on the same day of life for all patients, to increase local expertise in reading and interpreting findings. Consider a repeat MRI between days 10 and 14 of life when the imaging and clinical features are discordant or when diagnostic ambiguity persists [65]. Level of evidence: 3

WHAT FOLLOW-UP SHOULD BE ORGANIZED FOR INFANTS WHO RECEIVE HYPOTHERMIA?

Cerebral palsy or severe disability occurs in more than 30% of HIE-affected newborns, and is most common in infants with severe encephalopathy [66]. There is increasing recognition that cognitive deficits may be prominent, even in the absence of cerebral palsy [67]. Severe visual impairment or blindness occurs in up to 25% of children after moderate or severe encephalopathy, especially in the setting of hypoglycemia. Decreased visual acuity, visual fields or stereopsis are also described [68]. Sensorineural hearing loss, likely secondary to brainstem injury, affects up to 18% of survivors of moderate encephalopathy without cerebral palsy [69]. Cognitive deficits, particularly difficulties with reading, spelling and arithmetic, are seen in 30% to 50% of childhood survivors of moderate HIE [70]. Behavioural difficulties, such as hyperactivity and emotional problems, should also be considered even in survivors who do not experience motor disability [71]. In neonates with moderate-to-severe HIE treated with therapeutic hypothermia, childhood epilepsy is identified in 13% of survivors, with 7% necessitating an antiepileptic medication in school age [72].

Follow-up at 18 to 24 months has been the standard of care in hypothermia trials. However, given the broad spectrum of neurodevelopmental impairments following hypoxic-ischemic brain encephalopathy and individual heterogeneity, following affected newborns closely through infancy and into later childhood is important. Multidisciplinary follow-up could involve a neonatologist or paediatrician, nurse, physiotherapy, occupational therapy, psychology, an infant development program, neurology, developmental paediatrician, ophthalmology and audiology. Specialists working together to assess long-term motor, psycho-educational, auditory and cognitive outcomes is an important care component for infants who have received therapeutic hypothermia. Level of evidence: 2b

CONCLUSION

Mild therapeutic hypothermia to a core rectal temperature of 33.5°C ± 0.5°C, initiated as soon as possible within the first 6 hours of life, decreases mortality and or severe long-term neurodevelopmental disabilities in infants with moderate-to-severe HIE, without increasing the incidence of disability in survivors. Therapeutic hypothermia should be provided in tertiary neonatal units and initiated in consultation with a level 3 neonatologist before transport. Close surveillance of infants during the cooling process is needed, given the risk for complications of both HIE and cooling. Long-term, multidisciplinary follow-up of survivors to assess and address neurocognitive function is essential to quality care.

A SUMMARY OF RECOMMENDATIONS

• Infants ≥36 weeks GA with moderate-to-severe HIE who meet inclusion criteria should receive therapeutic hypothermia. Recommendation grade A.
• Infants ≥35 weeks GA with moderate-to-severe HIE who meet other inclusion criteria should be considered for therapeutic hypothermia. Recommendation grade B.
• Therapeutic hypothermia should not be offered for the following patients: moribund infants or infants with major congenital or genetic abnormalities for whom no further aggressive treatment is planned, infants with severe intra-uterine growth restriction or clinically significant coagulopathy, or infants with evidence of severe head trauma or intracranial bleeding. Recommendation grade A.
• Community physicians caring for infants with suspected HIE for whom therapeutic hypothermia is considered, should consult a neonatologist regarding initiation of passive cooling as soon as possible after birth. All infants requiring therapeutic hypothermia should be transferred to a tertiary care Neonatal Intensive Care Unit with appropriate expertise and resources. Recommendation grade D.
• Both selective head cooling and whole body cooling are effective. Whole body cooling is easier to set up and use, less expensive and provides access to EEG; it is therefore recommended for centres not currently using selective head cooling. Recommendation grade B.
• Therapeutic hypothermia should be continued for 72 hours, with a target rectal (or esophageal) temperature of 33°C to 34°C for whole body cooling, or 34°C to 35°C for selective head cooling. Rewarming should occur over 6 to 12 hours (0.5°C every 1 to 2 hours). Recommendation grade A. 

• Therapeutic hypothermia in infants younger than 35 weeks is not recommended. Recommendation grade C. 

• Treatment of pain or discomfort during cooling with a low-dose opioid is recommended. Recommendation grade B. 

• Treatment of seizures, despite limitations of knowledge 
about the side effects of antiepileptic medications, is recommended, because benefits likely outweigh the risks. Recommendation grade D. 

• Follow-up of infants who received hypothermia to a minimum of 2 years, but ideally until school age, in a neonatal follow-up clinic, is recommended. Follow-up should be in conjunction with care by a community paediatrician. Recommendation grade B. 


Acknowledgements

This position statement was reviewed by the Community Paediatrics and Acute Care Committees of the Canadian Paediatric Society.

---

CPS FETUS AND NEWBORN COMMITTEE

Members: Mireille Guillot MD, Leonora Hendson MD, Ann Jefferies MD (past Chair), Thierry Lacaze-Masmonteil MD (Chair), Brigitte Lemyre MD, Michael Narvey MD, Leigh Anne Allwood Newhook MD (Board Representative), Vibhuti Shah MD


Liaisons: Radha Chari MD, The Society of Obstetricians and Gynaecologists of Canada; William Ehman MD, College of Family Physicians of Canada; Roxanne Laforge RN, Canadian Perinatal Programs Coalition; Chantal Nelson PhD, Public Health Agency of Canada; Eugene H Ng MD, CPS Neonatal-Perinatal Medicine Section; Doris Sawatzky-Dickson RN, Canadian Association of Neonatal Nurses; Kristi Watterberg MD, Committee on Fetus and Newborn, American Academy of Pediatrics

Principal authors: Brigitte Lemyre MD, Vann Chau MD

---

References

---

Disclaimer: The recommendations in this position statement do not indicate an exclusive course of treatment or procedure to be followed. Variations, taking into account individual circumstances, may be appropriate. Internet addresses are current at time of publication.