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

Cardiac arrest vs traumatic brain injury

One of the interesting aspects of the beneficial effects of mild therapeutic hypothermia in the setting of cardiac arrest relates to the following question: Why is hypothermia effective in improving neurological outcome after cardiac arrest while it has been more difficult to demonstrate benefit in other acute neurological insults, such as traumatic brain injury?²²

Application of hypothermia in cardiac arrest may represent something of a “perfect storm.” First, a recent study by Berger et al²³ provides some insight into the time course of neuronal death after cardiac arrest versus traumatic brain injury. In that study of children who suffered either cardiac arrest or severe traumatic brain injury requiring management in the intensive care unit, peak levels of the serum biomarker of neuronal death, neuron-specific enolase (NSE), occurred days after cardiac arrest, whereas they occurred generally within a few hours of traumatic brain injury. This suggests a broader therapeutic window for the application of mild hypothermia in cardiac arrest as opposed to traumatic brain injury. In addition, the only neuroprotective therapy used in cardiac arrest is mild hypothermia. In contrast, in traumatic brain injury, myriad therapies are part of standard of care, including intracranial pressure monitoring and cerebrospinal fluid drainage, mannitol, hypertonic saline, barbiturates, and surgical interventions such as decompressive craniectomy.²⁴ These intracranial pressure–directed therapies in traumatic brain injury may confer a variety of neuroprotective actions, thus raising the bar for hypothermia to show benefit. A similar case could be made regarding the surgical treatment of subarachnoid hemorrhage, where hypothermia has been ineffective.²⁵

Efforts to optimize hypothermia

Given the benefit of mild therapeutic hypothermia in cardiac arrest, we and other investigative teams are actively pursuing ways to further optimize its effects beyond the use of a more rapid induction, as discussed above.

One of the most overlooked areas of study relates to hypothermia’s profound effects on drug metabolism; despite the need for many drugs in critically ill patients after cardiac arrest, knowledge of how hypothermia alters drug metabolism and how best to adjust drug doses is limited. Therapeutic hypothermia has recently been shown, during cooling, to directly inhibit binding of drugs to the active site of the key drug-metabolizing enzyme, cytochrome P450.²⁶ In contrast, in the setting of experimental cardiac arrest and resuscitation, mild hypothermia also protects against induction of cytokines such as interleukin-6, which downregulates cytochrome P450. Thus, mild hypothermia reduces drug metabolism during cooling but leads to a better recovery of drug metabolism after rewarming. This dichotomous effect will need to be studied at the bedside. Hypothermia can also reduce drug effects.²⁶ Thus, until we know how to optimally dose various therapies in patients treated with hypothermia, it is probably best to carefully monitor levels (when possible) and also drug effects. The best example of this at the bedside is the use of monitoring neuromuscular blockade in patients treated with vecuronium or pancuronium during mild hypothermia.

Another interesting area of study involves defining the best anesthetics or sedatives to use with cooling. For example, a recent report by Statler et al²⁷ showed that hypothermia was much less effective as a neuroprotectant after experimental traumatic brain injury in rats anesthetized with fentanyl than with isoflurane. In that study, fentanyl was unable to blunt the stress response to cooling. Given the variety of sedatives and analgesics used at the beside in both neurointensive care units and coronary care units, understanding which agents work best with hypothermia could further enhance hypothermia’s therapeutic benefit.

There is also a search for agents that may promote induction of hypothermia or create a poikilothermic state, thereby facilitating tolerance of the hypothermic state without a stress response. One agent that has shown some promise in the setting of experimental cardiac arrest is the endogenous peptide neurotensin, which has direct effects on temperature regulation at the hypothalamic level. In an experimental model of asphyxial cardiac arrest in rats, Katz et al²⁸ reported that the neurotensin analog NT69L facilitated induction of hypothermia and improved outcome. Another agent that has been touted to induce a state of “hibernation on demand” is hydrogen sulfide gas. A recent experiment by Blackstone et al²⁹ demonstrated induction of deep hypothermia and a hibernation-like state in mice allowed to breathe hydrogen sulfide gas at 80 parts per million. This state was completely reversible upon discontinuation of the agent. Unfortunately, studies in large animal models have not been able to demonstrate induction of hypothermia with this approach.³⁰ Nevertheless, these drugs represent prototypes for future exploration; if the right agent is found, it could lead, in theory, to markedly enhanced efficacy of cooling.