Mild hypothermia via selective head cooling as neuroprotective therapy in term neonates with perinatal asphyxia: an experience from a single neonatal intensive care unit

Article metrics



The objective of this study was to determine the efficacy of mild hypothermia via selective head cooling as a neuroprotective therapy in term infants with perinatal asphyxia.

Study design:

Full-term newborns who had 5 min Apgar scores <6, first arterial blood gas pH<7.10 or BD>15 mEq/l, and with the clinical signs of encephalopathy were enrolled within 6 h after birth. Patients were randomized to receive mild hypothermia treatment via selective head cooling for a total of 72 h or receive routine treatment as a control. Brain hypoxic-ischemic injury was quantified based on the head computed tomographic scan (CT scan) at postnatal age 5–7 days and a Neonatal Behavioral Neurological Assessment (NBNA) score at 7–10 days of life.


A total of 58 patients (30 hypothermia, 28 control) completed the study. Hypothermia was well tolerated in this study and attenuated the hypoxic-ischemic brain injury due to perinatal asphyxia. Head CT scan demonstrated moderate to severe hypoxic-ischemic changes in only 4/30 cases from the hypothermic group. In contrast, 18/28 cases in the control group showed moderate to severe hypoxic-ischemic changes (χ2=15.97, P<0.01). Brain hypothermia also significantly improved the NBNA score (32±2 in the hypothermic group vs 28±3 in the control group, P<0.01).


s: Our results suggest that selective head cooling may be used as a neuroprotective therapy in term neonates with perinatal asphyxia. A long-term follow-up study is needed to further validate the results of this study.


Hypoxic-ischemic encephalopathy (HIE) due to perinatal asphyxia remains one of the major causes of neonatal death and later neurodevelopmental disability. There are limited therapeutic interventions available to rescue brain function during or immediately after perinatal asphyxia. The prognosis for an infant diagnosed with HIE has not changed during the past two decades. It is estimated that 25–30% of HIE survivors will have long-term neurodevelopmental disabilities that include cerebral palsy, seizure disorder and mental retardation.1 The magnitude of HIE due to perinatal asphyxia is much worse in developing countries such as China.2

Over the past two decades, our understanding of the mechanisms of neonatal HIE has increased dramatically from experiments both in vitro and in vivo. It is well accepted that depletion of cellular stores of high-energy phosphates, principally ATP and phosphocreatine, initiates the cascade of events leading to neuronal death after hypoxic ischemia.3, 4 On the basis of extensive animal studies, there is increasing evidence suggesting that mild or moderate brain hypothermia is an effective intervention to ameliorate the neuronal injury due to hypoxic ischemia.5, 6, 7 A few small clinical trials have demonstrated the safety of head cooling when used in neonates with HIE.8, 9 The purpose of this study was to use this simple method to treat full-term infants with perinatal asphyxia and examine its measurable efficacy compared to a control group that received routine treatment.

Patients and methods

This study was approved by the Ethics Committee of Yuying Children's Hospital of Wenzhou Medical College. After written consent from parents, newborn infants admitted to the Yuying Children's Hospital of Wenzhou Medical College from July 1, 2000 to June 30, 2003 were enrolled into the study when the following criteria were fulfilled: (1) gestational age 37 weeks; (2) Apgar scores <6 at 5 min with first postnatal arterial blood gas pH<7.10 or BD>15 mEq/l; and (3) the clinical signs of postpartum encephalopathy (decreased muscle tone, lethargy, coma, or seizures) starting within 6 h after birth. Consecutively admitted patients who met the inclusion criteria were randomized to the hypothermic group or the control group based on whether it was on an odd or even day of admission. Infants with major congenital anomalies and prolonged hypoxemia due to severe persistent fetal circulation were excluded from the study.

Once a patient was enrolled, the complete obstetric history was obtained, and the degree of HIE was determined with Sarnat's clinical staging system by our attending neonatologists.10 In the hypothermic group, selective head cooling was initiated as soon as the patient was admitted and enrolled. The selective head cooling was achieved by applying a cooling cap device (SDL-V) with circulating cold water at 10°C (10±1°C) (Tianyuan Scientific Development Inc. Changchun, China). The patients were kept under a radiant warmer with the targeted temperature set at 34–35°C for a total of 72 h. A shield was used to block the cooling cap from the radiant heat during head cooling. An indwelling rectal temperature (TR) probe was used to monitor the temperature constantly and used as the target temperature. Nasopharyngeal temperature (TNP) was also monitored during head cooling. Infants were then spontaneously rewarmed in room temperature to a normal temperature after the completion of 72 h of head cooling. The radiant warmer was added for rewarming if a patient's TR remained less than 36.0°C after 12 h from the completion of head cooling. The serum electrolytes were monitored daily and routine liver enzymes checked on day 3. The vital signs including TNP and TR were monitored continuously and recorded bihourly in our NICU during the study period. The control group received the same neonatal intensive care and monitoring except head cooling. Their TR was measured with a regular thermometer not by constant rectal probe. All patients in both groups were treated with a loading dose of 20 mg/kg of phenobarbital (regardless of having clinical seizures or not) as soon as they were enrolled and then 5 mg/kg per day for a minimum of 72 h as part of our NICU routine treatment protocol for HIE patients. An additional 5–10 mg/kg of phenobarbital was administered if clinical seizures continued and the phenobarbital levels were subtherapeutic. Furthermore, all patients were kept on maintenance intravenous fluids (10% dextrose water with or without sodium and potassium) and NPO for at least 72 h. Dopamine at 5 μg/kg/min was also routinely used in both groups to maintain a normal blood pressure or for renal blood perfusion during the 72-h study period.

A cranial head computed tomographic (CT) scan (GE 2000) was obtained on all enrolled infants at 5–7 postnatal days. All CT scans were reviewed by a neuroradiologist without the knowledge of group assignment. The extent and location of hypodensity areas were used to quantify the brain injury into three grades as previously described by Fitzhardinge et al.11: (1) Mild: scattered, minor areas of hypodensity usually localized in the periventricular area with some extension into the frontal or parieto-occipital areas of the cortex; (2) Moderate: widespread involvement extending from white matter well into the gray matter giving a mottled appearance; and (3) Severe: homogeneous decreased brain tissue density throughout most of the supratentorial compartment with only the basal ganglia and cerebellum presenting as normal density and the lateral ventricles small and compressed. The presence of intraparenchymal or intraventricular hemorrhage was graded as severe.

A well-described 20-item Neonatal Behavioral Neurological Assessment (NBNA) score was also used in the study. The NBNA score system, which was modified from the Brazelton Neonatal Behavioral Assessment Scale (BNBAS),12 has been well described before in detail and used as an assessment tool for neurological impairment in HIE patients.13, 14 NBNA contains five clusters: behavior (six items); passive tone (four items); active tone (four items); primary reflexes (three items); and general assessment (three items). Each item of NBNA can be scored as 0, 1 or 2, and has a maximum score of 40 for normal full-term infants.13 The scoring was done in all patients at 7–10 days of life by two specially trained individual investigators. If the scores were different between the two investigators, the mean value of the two scores was used. All statistical analysis was performed by using the SPSS 8.0 software and data presented as mean and s.d. when appropriate. Analysis of variance (ANOVA), Student's t-test and χ2 were used to compare the difference between two groups, and P<0.05 was considered to be statistically significant.


There were a total of 62 patients enrolled into the study during the study period, among them 32 in the hypothermic group and 30 in the control group. There were only nine cases (four from hypothermic group and five from control group) who were intubated upon admission for ventilatory support. There were four deaths (two in each group) after enrollment into the study due to multiorgan failure secondary to severe asphyxia and parental wish to withdraw therapy. Therefore, a total of 58 patients completed the study. The demographic data and characteristics of the two groups are presented in Table 1. There are no significant differences between the two groups for all the factors listed.

Table 1 The demographic data and initial clinical characteristics of the study population

All enrolled patients were born in other community hospitals or clinics and were immediately transferred to the Yuying Children's Hospital if the patient had an Apgar score <6 at 5 min. Most of the enrollments were within 4 h after birth. Both groups were relatively hypothermic at admission due to the lack of a neonatal transport incubator for transporting patients. Once enrolled into the hypothermic group, it took an average of 55±20 min to cool the temperature down to the target level. After 72 h of head cooling, it took an average of 300±50 min to rewarm the patients to a normal temperature.

The changes of temperature, heart rate and mean blood pressure during the 72 h trial period are presented in Table 2. As expected, the heart rates in the hypothermic group were significantly lower than the control group. However, the mean blood pressures were maintained by the routine use of dopamine. The hypothermia was well tolerated by all the patients in the hypothermic group. There was no difference between the two groups for electrolytes, liver enzymes, blood urea nitrogen (BUN) and creatinine levels during the 72-h period (data not shown).

Table 2 Changes in temperatures, heart rate and mean arterial blood pressure starting from enrollment in both hypothermia treatment and control group infants

Hypothermia treatment significantly attenuated the acute hypoxic-ischemic brain injury based on the head CT scan done at 5–7 days of life and NBNA score at 7–10 days of life. In the hypothermia treatment group, there were four out of 30 total cases with moderate to severe hypoxic-ischemic changes seen on head CT scan. In contrast, 18 of 28 cases in control group demonstrated moderate to severe hypoxic-ischemic changes (P<0.01, Table 3). As seen in the Figure 1, brain hypothermia treatment also significantly improved the NBNA score performed at 7–10 days of life (32±2 vs 28±3, P<0.01).

Table 3 The number of cases with different degree of HIE changes on head CT scan in hypothermia treatment or control patients at 5–7 days of age
Figure 1

Neonatal Behavioral Neurological Assessment (NBNA) scores in hypothermia treatment and control group. Mild hypothermia via selective head cooling significantly improves the NBNA score in term infants with perinatal asphyxia at 7–10 days of life.


There is a long history of using hypothermia for trauma patients to prevent further tissue injury. In the care of neonates, some clinicians used hypothermia to treat asphyxiated neonates as early as in 1950s.5, 15 After extensive experiments in animal models, several pilot or small clinical trials using hypothermia for neuroprotection in asphyxiated full-term infants have been performed.8, 9, 16, 17, 18, 19 From those pilot human studies it appears that mild hypothermia is safe when used in term infants with perinatal asphyxia as long as close monitoring can be implemented in the NICU setting.

With our particular setting and patient population, we are able to demonstrate in this randomized intent-to-treat clinical trial that mild hypothermia for 72 h via selective head cooling started within 6 h of perinatal asphyxia can significantly attenuate the brain injury based on the head CT scan and NBNA score changes at about 1 week of life. Areas in brain tissue with decreased density are a common finding in term infants with HIE, the degree of which has been correlated with subsequent neurodevelopmental defect.11 Further studies using diffusion-weighted magnetic resonance imaging and quantitative magnetic resonance spectroscopy may be needed,20, 21 but these modalities are not readily available to a developing country such as ours. The NBNA scoring system, which was modified from the BNBAS by Bao et al.13, 14 in China, has been used to predict neurodevelopmental outcome. A similar scoring system has also been used by other investigators in other parts of the world for predicting neurodevelopmental outcome in HIE patients.22 We understand the limitation of these clinical tools for evaluating the acute neurological outcome of patients. A long-term neurological outcome follow-up study of our enrolled patients will significantly enhance the value of this study.

There are several other limitations to our study. This is an intent-to-treat clinical trial carried out in a single children's medical center and the number of patient enrolled was relative small. The study assignment was not blinded to investigators. The randomization method based on the odd or even day, which was specifically designed to fit our staff scheduling, was not considered to be a strict randomization. Further, all enrolled patients were born in other hospitals or clinics and the acute perinatal care was inadequate in most of those community hospitals. Almost all of the babies enrolled in this study were born vaginally (only two babies in hypothermia group and one baby in control group were born by cesarian section). During the study period, most of the local perinatal care providers had not been adequately trained in the neonatal resuscitation program, and most of the asphyxiated babies might not have been properly resuscitated at delivery room. Some severe HIE cases probably died soon after birth in local community hospitals before they had a chance to be admitted to the only level three NICU in the region. Therefore, only a few severe HIE cases were enrolled in this study and patient population might be different from those seen in other developed countries where a multicenter trial of using head cooling for HIE was conducted recently.23 However, the beneficial effects of using selective head cooling in our particular patient population are consistent with the report of the international multicenter trial. In that report, hypothermia was only effective in patients with moderate HIE based on the results of initial amplitude-integrated electroencephalography.23

In our study, therapeutic dose of phenobarbital was routinely used in every enrolled patient. Thoresen et al.24 has shown that hypothermia alone without sedation was not effective in protecting the brain tissues of newborn piglets after a severe global hypoxic-ischemic insult. We speculate that the use of phenobarbital may help to reduce cerebral metabolism and oxygen consumption25 and facilitate the hypothermia process, which might enhance the neuroprotection of head cooling for infants with perinatal asphyxia. Although it remains controversial whether using high-dose phenobarbital alone is neuroprotective in asphyxiated infants,4, 26 prophylactic use of a conventional dose of phenobarbital for a few days in asphyxiated infants remains a common practice in China. On the other hand, routine use of dopamine for asphyxiated term infants is not a common practice. Based on our experience with asphyxiated infants, as well as reported hemodynamic and renal perfusion changes in other studies,8, 9, 27 we decided to use dopamine at 5 μg/kg/min for all enrolled patients to prevent hypothermia-related adverse hemodynamic changes, and as an attempt to improve renal blood flow.

Impaired cardiac, liver and renal functions are common clinical presentations in asphyxiated patients.28 As expected, on average our enrolled patients in both group had elevated BUN, creatinine and liver enzymes compared to the normal values, however, there were no statistical difference for any of those parameters between our hypothermic treatment group and control group. Further, there were no obvious abnormal bleeding symptoms recognized in any of those 58 patients, although clotting factors were not routinely measured in all patients. Mild hypothermia was therefore well tolerated by our enrolled patients and no adverse effect related to hypothermia was recorded in our study, which is consistent with other reported clinical trials.8, 17, 23

In conclusion, our study adds data suggesting that mild hypothermia via selective head cooling in term neonates with perinatal asphyxia is relatively safe and can be used in a regular NICU setting. Mild hypothermia does not aggravate cardiac, hepatic and renal dysfunction if close monitoring can be implemented. More importantly, our data suggest that the mild hypothermia via selective head cooling has a potential to be used as neuroprotective therapy in term neonates with perinatal asphyxia. A long-term follow-up study is needed to further validate the results of this study.



hypoxic-ischemic encephalopathy

T NP :

nasopharyngeal temperature

T R :

rectal temperature

CT scan:

computed tomographic scan


Neonatal Behavioral Neurological Assessment


  1. 1

    Vannucci RC, Perlman JM . Interventions for perinatal hypoxic-ischemic encephalopathy. Pediatrics 1997; 100: 1004–1014.

  2. 2

    Xu JL, Gao XX, Wang SX . Retrospective analysis of neonatal hypoxic-ischemic encephalopathy in last 10 years. Chin J Perinatal Med 2001; 4: 236–239 (Chinese).

  3. 3

    Whitelaw A . Systemic review of therapy after hypoxic-ischemic brain injury in the perinatal period. Semin Neonatol 2000; 5: 33–40.

  4. 4

    Grow J, Barks JDE . Pathogenesis of hypoxic-ischemic cerebral injury in the term infant: current concepts. Clin Perinatol 2002; 29: 585–602.

  5. 5

    Miller JA, Miller FS, Westin B . Hypothermia in the treatment of asphyxia neonatorum. Biol Neonate 1964; 6: 148–163.

  6. 6

    Gunn AJ, Gunn TR, Haan HH, Williams CE, Gluckman PD . Dramatic neuronal rescue with prolonged selective head cooling after ischemia in fetal lambs. J Clin Invest 1997; 99: 248–256.

  7. 7

    Thoresen M, Simmonds M, Satas S, Tooley J, Silver IA . Effective selective head cooling during posthypoxic hypothermia in newborn piglets. Pediatr Res 2001; 49: 594–599.

  8. 8

    Gunn AJ, Gluckman PD, Gunn TR . Selective head cooling in newborn infants after perinatal asphyxia: a safety study. Pediatrics 1998; 102: 885–892.

  9. 9

    Akisu M, Huseyinov A, Yalaz M, Cetin H, Kultursay N . Selective head cooling with hypothermia suppresses the generation of platelet-activating factor in cerebrospinal fluid of newborn infants with perinatal asphyxia. Prostaglandins Leukot Essent Fatty Acids 2003; 69: 45–50.

  10. 10

    Sarnat HB, Sarnat MS . Neonatal encephalopathy following fetal distress. A clinical and electroencephalographic study. Arch Neurol 1976; 33: 696–705.

  11. 11

    Fitzhardinge PM, Flodmark O, Fitz CR, Ashby S . The prognostic value of computed tomography as an adjunct to assessment of the term infant with postasphyxial encephalopathy. J Pediatr 1981; 99: 777–781.

  12. 12

    Als H, Tronick E, Lester BM, Brazelton TB . The brazelton neonatal behavioral assessment scale (BNBAS). J Abnorm Child Psychol 1977; 5: 215–231.

  13. 13

    Bao XL, Yu RJ, Li ZS, Zhang BL . Twenty-item behavioral neurological assessment for normal newborns in 12 cities of China. Chin Med J 1991; 104: 742–746.

  14. 14

    Bao XL, Yu RJ, Li ZS . 20-item neonatal behavioral neurological assessment used in predicting prognosis of asphyxiated newborn. Chin Med J 1993; 106: 211–215.

  15. 15

    Westin B, Miller J, Nyberg R, Wedenberg E . Neonatal asphyxia pallida treated with hypothermia alone or with hypothermia and transfusion of oxygenated blood. Surgery 1959; 45: 868–879.

  16. 16

    Thoresen M, Whitlaw A . Cardiovascular changes during mild therapeutic hypothermia and rewarming in infants with hypoxic-ischemic encephalopathy. Pediatrics 2000; 106: 92–99.

  17. 17

    Battin MR, Dezoete JA, Gunn TR, Gluckman PD, Gunn AJ . Neurodevelopmental outcome of infants treated with head cooling and mild hypothermia after perinatal asphyxia. Pediatrics 2001; 107: 480–484.

  18. 18

    Shankaran S, Laptook A, Wright LL, Ehrenkranz RA, Donovan EF, Fanaroff AA et al. Whole-body hypothermia for neonatal encephalopathy: animal observations as a basis for a randomized, controlled pilot study in term infants. Pediatrics 2002; 110: 377–385.

  19. 19

    Battin MR, Penrice J, Gunn TR, Gunn AJ . Treatment of term infants with head cooling and mild systemic hypothermia (35.0 and 34.5°C) after perinatal asphyxia. Pediatrics 2003; 111: 244–251.

  20. 20

    Martin E, Buchli R, Ritter S, Regula S, Remo LH, Eugen B et al. Diagnostic and prognostic value of cerebral 31P magnetic resonance spectroscopy in neonates with perinatal asphyxia. Pediatr Res 1996; 40: 749–758.

  21. 21

    Krishnamoorthy KS, Soman TB, Takeoka M, Schaefer PW . Diffusion-weighted imaging in neonatal cerebral infarction: clinical utility and follow-up. J Child Neurol 2000; 15: 592–602.

  22. 22

    Thompson CM, Puterman AS, Linley LL, Hann FM, van der Elst CW, Molteno CD et al. The value of a scoring system for hypoxic ischemic encephalopathy in predicting neurodevelopmental outcome. Acta Paediatr 1997; 86: 757–761.

  23. 23

    Gluckman PD, Wyatt JS, Azzopardi D, Ballard R, Edwards A, Ferriero D et al. Selective head cooling with mild systemic hypothermia to improve neurodevelopmental outcome following neonatal encephalopathy: the coolcap study. Lancet 2005; 365: 663–670.

  24. 24

    Thoresen M, Satas S, Loberg EM, Whitelaw A, Acolet D, Lindgren C et al. Twenty-four hours of mild hypothermia in unsedated newborn pigs starting after a severe global hypoxic-ischemic insult is not neuroprotective. Pediatr Res 2001; 50: 405–411.

  25. 25

    Nilsson L . The influence of barbiturate anaesthesia upon the energy state and upon acid–base parameters of the brain in arterial hypotension and in asphyxia. Acta Neurol Scand 1971; 47: 233–253.

  26. 26

    Hall RT, Hall FK, Daily DK . High-dose phenobarbital therapy in term newborn infants with severe perinatal asphyxia: a randomized, prospective study with three-year follow-up. J Pediatr 1998; 132: 345–348.

  27. 27

    Guignard JP, Gillieron P . Effect of modest hypothermia on the immature kidney. Acta Paediatr 1997; 86: 1040–1041.

  28. 28

    Piazza AJ . Postasphyxial management of the newborn. Clin Perinatol 1999; 26: 749–765.

Download references


We wish to thank the entire staff members of the NICU of Yuying Children's Hospital of Wenzhou Medical College for helping to complete the study and Drs Sidhartha Tan and Ian Holzman for the critical review and comments of the manuscript.

Author information

Correspondence to Z-L Lin.

Rights and permissions

Reprints and Permissions

About this article

Cite this article

Lin, Z., Yu, H., Lin, J. et al. Mild hypothermia via selective head cooling as neuroprotective therapy in term neonates with perinatal asphyxia: an experience from a single neonatal intensive care unit. J Perinatol 26, 180–184 (2006) doi:10.1038/

Download citation


  • perinatal, asphyxia
  • hypoxic-ischemic encephalopathy
  • newborn, infant
  • induced, hypothermia

Further reading