Therapeutic Hypothermia for Perinatal Asphyxia

Hypoxic-ischemic encephalopathy (HIE) results when the newborn brain is deprived of adequate oxygen and blood flow around the time of birth. It affects roughly 2 per 1,000 live births and, before effective treatment existed, carried a grim prognosis — about 60% of affected infants either died or survived with significant neurological disability, including cerebral palsy, intellectual disability, and epilepsy. For decades, neonatologists could only offer supportive care: maintain temperature, treat seizures, support breathing and circulation. The idea that something could actually protect the brain seemed out of reach.

The intellectual breakthrough came from a seemingly simple observation made using a technology called phosphorus magnetic resonance spectroscopy (MRS) in asphyxiated newborns at University College London in the late 1980s and early 1990s. Researchers noticed that brain energy metabolism often appeared near-normal immediately after an asphyxia event and resuscitation, only to crash again 6–24 hours later — a pattern called secondary energy failure. Animal experiments confirmed this, most influentially in a 1994 newborn piglet study by Lorek and colleagues. The crucial implication was that there is a window of opportunity between the primary insult and this secondary deterioration — roughly 6 hours — during which the brain’s still-living cells might be rescued if the right intervention could be delivered in time.

The scientists who turned this insight into a treatment strategy were primarily based in Auckland, New Zealand. Peter Gluckman and Alistair Gunn used a chronically instrumented fetal sheep model to test whether cooling the brain after an ischemic insult could prevent the secondary deterioration. In a landmark 1997 paper in the Journal of Clinical Investigation, they showed that selective head cooling initiated 90 minutes after 30 minutes of cerebral ischemia produced dramatic neuronal rescue compared to untreated controls. Follow-up sheep studies refined the concept further, establishing two critical principles: cooling must begin during the latent phase before seizure onset to be effective, and 72 hours of sustained cooling was needed to cover the period of secondary injury. Simultaneously, Marianne Thoresen and colleagues at UCL were showing in piglet models that even mild systemic hypothermia after hypoxia-ischemia substantially reduced the severity of secondary energy failure — converging evidence from two continents that this was a real and reproducible protective mechanism.

With animal data in hand, the Auckland group moved cautiously to humans. A small safety study published in Pediatrics in 1998 showed that selective head cooling could be achieved in asphyxiated newborns without dangerous side effects. In the United States, Seetha Shankaran at Wayne State University and the NICHD Neonatal Research Network ran their own animal studies and a pilot human trial, providing the foundation for a large multicenter randomized controlled trial. These were not simple studies to design: HIE requires rapid enrollment within 6 hours of birth, standardized encephalopathy grading, sophisticated temperature management, and long-term neurodevelopmental follow-up — logistical challenges that required coordinated multicenter networks.

The pivotal year was 2005, when two independent large randomized trials were published simultaneously. The CoolCap trial, led by Gluckman, Wyatt, and Azzopardi, enrolled 243 infants across multiple centers and used selective head cooling with a water-circulating cap. The NICHD whole-body cooling trial, led by Shankaran and the Neonatal Research Network, enrolled 208 infants and cooled them using a whole-body blanket to 33.5°C for 72 hours. Both trials showed a significant reduction in the primary outcome of death or severe neurodevelopmental disability at 18 months of age. The effect wasn’t dramatic enough to eliminate disability — roughly one in nine treated infants benefited — but in a condition where there had previously been no treatment at all, it was a landmark achievement.

A third major trial, TOBY (Total Body Hypothermia), led by Denis Azzopardi and A. David Edwards in the UK, confirmed the findings in 2009, again using whole-body cooling. The Australian/New Zealand ICE trial added further confirmatory evidence. By the time meta-analyses pooled all the data, the picture was consistent: hypothermia reduced the risk of death or major neurodevelopmental disability with a relative risk around 0.75–0.81 and a number needed to treat of approximately 6–9. Multiple guideline bodies — including NICE in the UK and ILCOR internationally — endorsed therapeutic hypothermia as standard of care, and the AAP issued formal clinical guidance. By the late 2000s, most major neonatal centers in high-income countries had implemented cooling programs.

An important and initially uncertain question was whether the benefit seen at 18–24 months would hold up at school age — early improvements sometimes reflect delayed rather than prevented injury. Long-term follow-up answered this reassuringly. The TOBY trial’s school-age data showed that 52% of cooled children survived with an IQ ≥85, compared to 39% of controls, with significantly less cerebral palsy and severe disability. The NICHD trial showed a strong trend in the same direction at 6–7 years. These findings confirmed that therapeutic hypothermia genuinely alters the trajectory of brain injury rather than simply deferring it.

Subsequent trials have refined the protocol rather than changed it fundamentally. The NICHD’s Optimizing Cooling trial tested whether deeper cooling (to 32°C) or longer duration (120 hours instead of 72) would improve outcomes further. Neither did — suggesting that the original parameters were already close to optimal, and that more is not always better when it comes to temperature management. A sobering lesson came from the HELIX trial, which tested hypothermia in India, Sri Lanka, and Bangladesh, where HIE is far more common. Unexpectedly, cooling provided no benefit — and may have caused harm — likely because many of those infants had been injured well before birth, beyond the 6-hour window where treatment can still help. This highlighted that therapeutic hypothermia is not a universal remedy but a time-sensitive intervention that depends critically on early recognition and rapid treatment initiation.

Today, the standard protocol — cooling to 33.5–34.5°C within 6 hours of birth, continued for 72 hours, followed by gradual rewarming — is established practice for term and late preterm infants (≥36 weeks) with moderate to severe HIE, and has been considered “best practice” worldwide since 2009-2012. Treatment requires careful monitoring, because hypothermia has its own physiological effects: altered drug metabolism, coagulopathy, bradycardia, and increased pulmonary vascular resistance can all complicate management. The treatment is only partially effective — many cooled infants still die or survive with disability — and research is ongoing into adjunctive therapies such as erythropoietin, melatonin, and stem cells that might augment the neuroprotection hypothermia provides. The story of therapeutic hypothermia is, however, one of neonatology’s genuine success stories: a treatment developed through rigorous translational science, tested in well-designed multicenter trials, and confirmed over decades of follow-up to meaningfully change the lives of affected newborns and their families.

Zanelli SA, Wusthoff CJ, Lucke AM, Kaufman DA, Committee on Fetus and Newborn, Section on Neurology: Therapeutic Hypothermia for Neonatal Hypoxic-Ischemic Encephalopathy: Clinical Report. American Academy of Pediatrics, January 26, 2026.

Last Updated on 04/03/26