Bakken Lecture: The brain, the heart, and therapeutic hypothermia

Author and Disclosure Information


Therapeutic hypothermia in acute resuscitation medicine has a long history, but its currently recommended use dates back to work in the mid-1960s by the late Dr. Peter Safar and colleagues. Compared with normothermia, mild therapeutic hypothermia, induced right after restoration of spontaneous circulation in comatose survivors of cardiac arrest, leads to 1 additional patient with intact neurological outcome for every 6 patients treated. Demonstrating benefit from therapeutic hypothermia in other acute neurological insults, such as traumatic brain injury, has been more difficult. Current research to optimize the benefits of mild therapeutic hypothermia in cardiac arrest is focused on hypothermia’s profound effects on drug metabolism, determining the best anesthetics and sedatives to use with cooling, and identifying compounds that may promote induction of hypothermia or create a poikilothermic state. Future applications of therapeutic hypothermia may include induction of emergency preservation and resuscitation to buy time for damage-control surgery in patients with exsanguination cardiac arrest.



Therapeutic hypothermia has been a central focus of research at the Safar Center for Resuscitation Research since the center was founded—as the International Resuscitation Research Center—at the University of Pittsburgh School of Medicine in 1979. In this article, which is based on my 2008 Bakken Lecture, I will discuss historical, contemporary, and futuristic applications of therapeutic hypothermia. Given that the key mission of the Safar Center is “to save hearts and brains too good to die,” the basis of my discussion will consist of how therapeutic hypothermia impacts both heart and brain—and the lessons that can be learned in each case.


Reprinted, with permission, from Journal of the Iowa Medical Society (Safar P. Community-wide cardiopulmonary resuscitation. J Iowa Med Soc 1964; 54:629–635).

Figure 1. Classic publication of the “first ABCs of resuscitation” by the late Dr. Peter Safar. As shown, therapeutic hypothermia was recommended for use in comatose patients who achieve restoration of spontaneous circulation, even in the 1960s.3

The late Dr. Peter Safar and his colleague, the late Dr. Hubert Rosomoff, played an instrumental role in the use of therapeutic hypothermia in the early 1960s in patients with acute neurological insults. Their classic 1965 publication, “Management of the comatose patient,”1 contained recommendations that in many ways outline the current use of mild therapeutic hypothermia as recommended by the American Heart Association and the International Liaison Committee on Resuscitation.2 In addition, in 1964, Dr. Safar recommended in his historic “first ABCs of resuscitation” that hypothermia be used in patients who remain comatose after successful restoration of spontaneous circulation (Figure 1).3 That recommendation holds true in today’s guidelines.2 However, therapeutic hypothermia in acute resuscitation medicine has a remarkably long history.

Baron Dominique Jean Larrey, surgeon-in-chief of the Napoleonic armies and the father of modern military medicine, observed in 1814 that the wounded “privileged” soldiers lying closer to the campfire died sooner than those in more remote, colder areas.4 Similarly, Dr. Charles Phelps, surgeon to the New York City Police Department, in 1897 recommended the use of the “ice cap” for traumatic brain injury.5

In the 1980s, however, therapeutic hypothermia began to fall out of favor. This resulted, in part, from overzealous application in some patients, who were treated for durations longer than a week and at temperatures in the moderate (28°C to 32°C) rather than mild (33°C to 35°C) range. This led to an increase in complications.6 Laboratory studies in a rat model of global cerebral ischemia by Busto et al 7 in 1987 and in a canine model of cardiac arrest by Leonov et al8 in 1990 demonstrated that benefit could be produced using mild cooling after the insult. This and parallel work in neonatology led to the ultimate breakthrough that translated into improved outcomes with the use of mild therapeutic hypothermia in adults with cardiac arrest9,10 and in newborns with hypoxic-ischemic encephalopathy.11

Clinicians and scientists familiar with hypothermia might suggest that its potential therapeutic benefit has been known for decades, given the use of hypothermic circulatory arrest for neuroprotection and cardioprotection in open heart surgery. However, one of the most interesting aspects of neuroprotection provided by mild therapeutic hypothermia is that it is not clearly linked to attenuation of energy failure.7 Unlike the setting of deep hypothermic circulatory arrest—where the induction of hypothermia occurs before the insult, and levels of hypothermia are such that energy failure is prevented—mild cooling, applied after cardiac arrest, appears to confer benefit by other mechanisms. Effects on cell signaling, oxidative and nitrative stress, apoptosis, excitotoxicity, and other mechanisms appear to mediate this benefit.12,13


Use in cardiac arrest

Compared with normothermia, mild therapeutic hypothermia, induced immediately after restoration of spontaneous circulation in comatose survivors of ventricular fibrillation cardiac arrest, leads to 1 additional patient with intact neurological outcome for every 6 patients treated.9 This is a remarkable effect given the extremely poor overall outcomes observed after out-of-hospital cardiac arrest. Studies in animal models, however, suggest that the therapeutic potential of mild hypothermia can be maximized with application either during or as early as possible after the insult.14 However, clinicians in the field of cardiology appropriately have cause for concern about the possibility that even mild cooling could reduce that potential for successful defibrillation or lead to re-arrest. In 2005, an important paper by Boddicker et al15 explored the impact of mild hypothermia on defibrillation success in experimental ventricular fibrillation in pigs and found, remarkably, that the success rate actually improved with mild or moderate hypothermia! That report opened the door for a number of studies that are now focused on rapid cooling during cardiopulmonary resuscitation (CPR) and on the rapid induction of mild hypothermia using intravenous cooling.16,17

Support for the use of intra-arrest cooling came initially from work in animal models of cardiac arrest—first from the work of Abella et al18 in a mouse model of potassium-induced cardiac arrest, and later from a canine model in work by Nozari et al.19 In the latter study, delaying the onset of mild hypothermia during advanced cardiac life support markedly worsened both multisystem organ failure and survival. Cardiovascular function in that model appeared to be substantively improved by early intra-arrest cooling.

The potential for the use of intravenous cooling in patients with a bolus of crystalloid to induce mild hypothermia was pioneered in a seminal paper by Bernard et al.16 In that report, an approximately 2°C reduction in core temperature could be achieved with infusion of about 30 mL/kg of fluid over 30 minutes. Mean arterial blood pressure increased mildly, and the intervention was well tolerated when applied early after restoration of spontaneous circulation. Kim and colleagues20 built upon that initial work and demonstrated the feasibility of the use of intravenous iced normal saline to induce mild hypothermia by paramedics in the prehospital setting. This approach, and its impact on neurological outcome and survival, is currently being evaluated in a randomized controlled trial. Combining intra-arrest cooling with the use of intravenous fluids is the obvious next step. This could facilitate rapid induction, which could then be maintained with commercially available surface cooling devices.21

Next Article: