Hypothermia Treatment and the Newborn

As this quotation shows, the use of cold water to stimulate the onset of respiration in depressed newborn infants has a long history. However, it was James Miller from New Orleans and Bjorn Westin from Stockholm who in the 1950s developed a scientific rationale for the use of hypothermia in “asphyxia neonatorum.” In a series of animal studies they demonstrated that rapid cooling in asphyxia prolonged survival by allowing the establishment of an equilibrium between cerebral oxygen demand and supply (reviewed in reference 1). They followed this in 1958 with a striking clinical study involving 10 severely depressed term infants.23 The infants all had Apgar scores of 1 at birth and they had failed to respond to conventional methods of resuscitation available at that time. The apneic infants were immersed in a specially constructed cold water bath with the aim of reducing core temperature rapidly. Cooling was stopped when spontaneous respirations commenced or when the rectal temperature approached 27°C. In 4 infants transfusions of oxygenated blood were also administered. The infants were then dried and allowed to rewarm spontaneously. One of the 10 infants died due to hyaline membrane disease but the remaining 9 survived, despite minimum core temperatures of between 23°C and 30°C, and periods of apnea ranging from 8 to 79 minutes! At follow-up 10 years later, none of the nine infants showed neuromotor impairment or cerebral palsy.4Subsequently, Miller and Westin1 treated all 65 infants who were unresponsive to routine resuscitative methods out of a total of 4400 births in their hospital in Stockholm. Remarkably for that period, 52 of 65 survived, (9 of those who died were of very low birth weight) and no surviving infant had evidence of cerebral palsy. After the introduction of positive pressure ventilation for apneic newborns, Miller and Westin continued to champion the role of hypothermia into the 1970s, as an additive therapy designed to minimize the risk of cerebral injury.45 Small-scale trials were performed in California, Switzerland, and Finland.5-8 Yet despite their positive experience, therapeutic hypothermia was never adopted on a significant scale.In the history of clinical medicine a single scientific paper may exert an influence on the thinking of a generation. The paper of Silverman and colleagues9 published in Pediatrics in 1958 was in this mold, and it sounded the death knell for hypothermia. Silverman reported the outcome of 182 premature infants, half of whom were nursed during the first 5 days of life in a “hypothermic” incubator (mean air temperature 28.9°C) and the remainder in a “normothermic” incubator (mean air temperature 31.7°C). Survival overall was 68% in the hypothermic group compared with 83% in the warmer incubators. More detailed analysis of their figures shows that the majority of the effect was in infants with birth weights <1000 g, with only a small difference in survival in infants with weights >1500 g. In addition the age at admission of the normothermic group was significantly greater than that of the other group, which may have contributed to the difference in survival.Silverman's paper provided a major incentive for basic research into thermoregulation in the newborn and led to the clinical practice of rigorously controlling core temperature in all newborn infants. Although the study was limited to preterm infants, the assumption gained ground that hypothermia was dangerous for term as well as preterm infants and to be avoided at all cost. Because of the axiomatic nature of this assumption, few subsequent studies have been performed on the association between hypothermia and mortality. A more recent study of Hazan and colleagues10 from Montreal suggested that, although there was an association between low temperature on admission to their neonatal intensive care unit and increased mortality, there was little evidence of a causal relationship. Many clinicians can provide anecdotal accounts of the intact survival of infants exposed inadvertently to significant hypothermia. In contrast to the very preterm baby, the normally-grown term infant has a repertoire of adaptive and protective responses to cold stress. Indeed there is evidence that when exposed to hypoxia, the normal thermogenic response to hypothermia is suppressed as part of an adaptive response to reduce metabolic demands.11 In 1958 Burnard and Cross12 demonstrated that infants who failed to establish spontaneous respiration within 3 minutes of birth had significantly lower rectal temperatures in the first 24 hours of life compared with normal infants.The rationale for the beneficial effects of hypothermia provided by Miller and Westin remains unproven. However, over the last decade evidence has accumulated from a range of neonatal and fetal animal models to show that small reductions in brain temperature after resuscitation from hypoxia-ischemia have a significant and long-lasting neuroprotective effect (reviewed in reference 13). In addition, hypothermia may prolong the “therapeutic window” after exposure to acute hypoxia-ischemia, and exert a synergistic action with other rescue therapies such as excitatory amino acid antagonists.14 Conversely, an elevated brain temperature increases susceptibility to ischemic injury.15 Of particular interest is the recent epidemiologic study from California showing a strong association between maternal infection and cerebral palsy in infants weighing >2500 g.16 Maternal fever exceeding 38°C was associated with a ninefold increase in the risk of “unexplained” spastic cerebral palsy. Although maternal fever may have acted merely as a marker for chorioamnionitis, it is plausible that elevated fetal temperature was causally associated with adverse outcome by increasing the vulnerability of neural tissue to various forms of cellular injury.Thus, brain temperature is an important biologic variable in infants at risk of ischemic injury. Inadvertent exposure of the fetal and neonatal brain to hyperthermia is not uncommon and may be an important contributor to perinatal brain injury. Conversely, deliberate reduction of brain temperature after birth has clear therapeutic potential. In our view mild hypothermia is currently the most promising option for a clinically feasible neural rescue therapy in encephalopathic newborn infants after acute perinatal asphyxia.13 However, several important questions remain unresolved and must be addressed before large-scale clinical trials are contemplated. The size of the therapeutic window after resuscitation is unknown, but it is likely to be no more than 2 to 6 hours, requiring the very early identification of suitable infants for treatment.17 The optimal duration for hypothermia treatment is also unknown. Although prolonging hypothermia may improve neuroprotection, it may also lead to an increase in adverse effects, including increased blood viscosity,18 pulmonary hemorrhage,19 and reduced cardiac contractility.20 Selective brain cooling using a “cooling cap” has been shown to be feasible in infant piglets21 and fetal sheep.22 If a satisfactory temperature gradient between the brain and the trunk can be established, this approach may allow the benefits of cerebral hypothermia to be obtained while minimizing the risk of adverse systemic side effects.The definitive testing of hypothermia and other promising rescue treatments will clearly demand large, multicenter randomized trials. However, preliminary feasibility studies are needed before a suitable trial protocol can be designed. Coordination and collaboration between researchers in this area is essential if the therapeutic promise of hypothermia, first demonstrated 40 years ago by Miller and Westin, is to be fulfilled.Human infants are born virtually hairless, covered in fluid, with a large head, and a scalp circulation that does not vasoconstrict in response to cold. The term infant is equipped with a range of mechanisms to minimize adverse effects from mild hypothermia and responds to a decrease in core temperature by initiating nonshivering thermogenesis, leading to gradual rewarming over the first 24 hours.23 Mild hypothermia is almost universal in deliveries outside hospital and in many developing countries.24 Perhaps infants are meant to get a little cold after birth.