Therapeutic Hypothermia and Targeted Temperature Management With or Without the “Cold Stress” Response

In this issue of Therapeutic Hypothermia and Temperature Management, Choudhary and Jia (2017) present a review of studies that have explored approaches to reduce and regulate temperature in a manner that circumvents the need for “forced” temperature manipulation. The approaches reviewed include activation or inhibition of specific hypothalamic nuclei via pharmacologic strategies and/or using electrical stimulation. The pharmacologic approaches described range from direct neurotransmitter manipulation such as of α₁ or α₂ receptors or manipulation of nitric oxide synthetic pathways by microinjection of either noradrenaline or N-monomethyl-L-arginine (L-NMMA), respectively, into the hypothalamus (Mallick et al., 2002; Osaka, 2010) or local infusion of vasopressin 1A-receptor agonists to modulate “warm sensitive” (WS) or “cold sensitive” (CS) neurons (Tang et al., 2012). More clinically relevant approaches include systemic administration of drugs such as neurotensin (Katz et al., 2015), cannabinoids (Ma et al., 2014), thyroxine derivatives (Doyle et al., 2007), hydrogen sulfide (Lefer, 2007), adenosine, or many others have been suggested (also reviewed in Zhang et al., 2013; Liu et al., 2016), as has electrical stimulation of selected hypothalamic nuclei (Choudhary and Jia, 2017). However, not all approaches have been successful or easy to replicate across various models of acute brain injury and resuscitation—where they might be the most useful, and importantly work in large animal models—where cooling is more challenging than in rodents is more limited and more variable in success (Drabek et al., 2011; Katz et al., 2012).

Notably, the majority of evidence supporting the concept that manipulation of hypothalamic circuits is a promising strategy to induce neuroprotection is derived from experiments in which animals were uninjured. More work is needed to test if hypothalamic pathways can be similarly manipulated after damage in different types of brain injury. Hypothalamic insufficiency is well known to occur after a traumatic brain injury (TBI) (Schneider et al., 2017), and the extent of hypothalamic damage after other insults relevant to therapeutic hypothermia (TH) and targeted temperature management (TTM) such as cardiac arrest need to be further clarified. Elucidation of the mechanisms that mediate hypothalamic/pituitary dysfunction postinjury is an ongoing area of investigation. Clinical translation of hypothalamic modulating strategies, for cooling brain-injured patients, will require additional research to define whether or not synaptic functions of target brain nuclei (e.g., containing WS or CS neurons) remain relatively spared/normal and have predictable pharmacological responsivity—compared with the spectrum of abnormal states manifested in vulnerable brain regions damaged by injury and which might benefit most from TH. Also, this could differ greatly across the various CNS insults.

Another practical question, to be addressed, is the extent to which pharmacological modulation of hypothalamic circuits is amendable to controlling the rate of rewarming, which would be important to address if these types of therapies were used for a sustained period, for example, to prevent shivering—rather than to induce hypothermia. The 2015 American Heart Association guidelines recommend that patients treated with TH be rewarmed at 0.3–0.5°C/h. Rewarming at faster rates is also known to worsen outcome in TBI patients (Hutchison et al., 2008). It would be interesting to test if drug-induced hypothermia avoids challenges associated with rewarming compared with standard cooling devices; might hypothalamic modulation result in less hemodynamic instability or changes in sympathetic tone during rewarming (Tokunaga et al., 1996)? However, it is unclear in the setting of drug-induced cooling—with resetting of hypothalamic regulation, if residual drug was present, a situation could develop where a return to normothermia might be perceived as induction of “forced” hyperthermia.

Preventing some of the adverse consequences of “forced cooling” or allowing more prompt temperature reductions by blunting endogenous temperature regulation could be highly desirable and beneficial in the application of TH or TTM, as pointed out by the authors. Mitigating complications that result from cold stress, most notably, shivering, could importantly reduce metabolic demands during acute neurologic crises and prevent the cascade of secondary consequences that ensue from preventing or treating this side effect, notably the use of sedatives, narcotics, intubation, mechanical ventilation, and/or neuromuscular blockade, if other more benign strategies to prevent shivering such as surface warming are insufficient or impractical. Clearly, this review confirms that interventions that might facilitate the induction and maintenance of TH and/or TTM deserve additional exploration. Furthermore, a pharmacological method to rapidly initiate endogenous cooling and/or augment intravenous or surface cooling could have tremendous utility for first responders during the transport of patients, or to augment cooling in the emergency department and/or intensive care unit.

However, it has become newly apparent that there are also novel and unique aspects of the endogenous reaction to TH, namely, newly discovered components of the “cold stress” response that may confer powerful neuroprotective effects. In a provocative report in the journal Nature by Peretti et al. (2015), a brief (1–2 hours) period of deep hypothermia induced by injection of 5-AMP with cold exposure produced remarkable neuroprotection against long-term neurodegeneration in mice with either an Alzheimer's disease knock-in phenotype or prion disease (Peretti et al., 2015). In that study, induction of the cold-shock protein RNA binding motif-3 (RBM-3) appeared to represent the key facilitator of this long-term benefit—which included histological and behavioral benefit and prolonged survival.

That report, although extremely exciting, raised many questions about the clinical translatability of these findings, namely neuroprotection by hypothermia-induced RBM-3, particularly, with regards to the use of TH and TTM in acute brain injury and neurocritical care. Our group has provided some insight into that important question. We reported that temperature reductions as small as 1°C induced robust expression of RBM-3 in isolated neurons (Jackson et al., 2015), suggesting this mechanism might play a role in mild TH and/or even TTM. The response was particularly robust in young neurons, mimicking newborn neurons, but largely absent in “adult” neurons. In addition, data suggest that RBM-3 mediates its benefit, at least in part, through potent induction of various facets of global protein synthesis upon rewarming.

Also, because RBM-3 protein levels are robustly increased in large and small hibernating animals (Fedorov et al., 2011; Schwartz et al., 2013), drug-inducing cooling strategies, which activate hibernation regulatory pathways (e.g., adenosine receptor agonists) (Choudhary and Jia, 2017), may similarly increase or perhaps further augment this important neuroprotective mechanism vs. traditional cooling techniques. Also, relevant to the anatomical focus of the review, RBM-3 expression in hibernating ground squirrels is abundant, specifically, in the hypothalamus as well as in cortex during prehibernation, torpor, and during interbout arousal episodes (Schwartz et al., 2013). This finding is logical given that the hypothalamus is a major regulator of mammalian hibernation and because RBM-3 helps cells to function normally under conditions of cold stress (i.e., RBM-3 preserves global protein synthesis—which is a prerequisite for protein cell signaling and thus presumably for the hypothalamus to exert its biological actions).

Taken together, these studies clearly raise important questions, such as (1) how much of the acute or long-term benefit of TH or TTM is modulated by RBM-3 or other cold stress proteins—versus other more traditional and well-known benefits of TH or (2) might we be able to prevent unwanted side effects of “cold stress” such as shivering while maintaining its neuroprotective effects during TH or TTM? It also may be more than coincidence that the RBM-3 is robust in “young” neurons in vitro, but minimal in “adult” neurons (Jackson et al., 2015). Clinically, TH is well known to exhibit its greatest degree of neuroprotection in the newborn brain versus older children or adults (Jacobs et al., 2013; Moler et al., 2016; Tasker et al., 2017). Our work on the mechanism of action of RBM-3 and efforts to develop pharmacological inducers or mimics of it also suggest that drugs such as fibroblast growth factor-21 can induce RBM-3 in vitro, so there may even be a hope for a pharmacological avenue to induce this pathway without requiring “cold stress.”

Certainly, TH has long been known to those using it clinically and in preclinical models to represent a complex intervention, with multiple effects inside and outside of the CNS. It not only produces clear potent neuroprotection but also challenging side effects—well beyond shivering, such as impairment of drug metabolism, cold diuresis, alterations in systemic and cerebral hemodynamics, electrolyte abnormalities, alterations in coagulation, and increased infection risk (Tortorici et al., 2007; Hutchison et al., 2008; Fink et al., 2010; Empey et al., 2013; Anderson et al., 2016). This review by Choudhary and Jia (2017), taken along with other exciting molecular developments in the field, thus is timely and serves to increase our awareness that new approaches to inducing and/or augmenting TH and TTM may be on the horizon to take better advantage of both its potent neuroprotective effects while minimizing its complications. It may take just the right dose of cold stress.