If there is a “time to target temperature paradox” in post-cardiac arrest care, would we know?
If there is a “time to target temperature paradox” in post-cardiac arrest care, would we know?
Kelly N. Sawyer – Michael C. Kurz
In the application of post-cardiac arrest care and targeted temperature management (TTM), the relationship between “time to target temperature”, or induction time, and neurologic outcome is unclear. Contemporary scientific literature presents conflicting results. Mirroring animal models, of cardiac arrest, some researchers have reported an association between delays to achieving target temperature and poor outcome in humans., , Others, however, have identified a seeming paradox: longer time to target temperature may herald a good neurologic outcome.
In this issue of Resuscitation, Dr. Perman and colleagues suggest a shorter induction time may differentiate those cardiac arrest victims with more substantial brain injury. The authors hypothesize: patients with steeper induction curves (i.e. rapid achievement of target temperature) are more likely to have impaired thermoregulatory control while shallower induction curves (i.e. slower achievement of target temperature) indicate less injury, as evidenced by an ability to resist TTM. While theories of hypothalamic dysfunction post-arrest abound, few, if any, data directly measuring this injury, or measuring derangements to energy metabolism, are available.
Perman et al. retrospectively examined the relationships between initiation of cooling (pre-induction) and achievement of target temperature (induction) times with neurologic outcome in 321 post-arrest patients at two hospitals. The population was heterogeneous, including 52.7% out of hospital cardiac arrests and 23.3% transfers from outside facilities. Standard Utstein template data were collected and factors associated with poor outcome included older age, non-shockable arrest rhythm, and longer times to return of spontaneous circulation (ROSC). Though no significant relationship between pre-induction time and outcome was identified, the authors report that patients with an induction time <2?h were half as likely to have a favorable neurologic outcome (15.6% vs. 32.5%; p?=?0.003). Those whose induction time was >5?h were twice as likely to have a good outcome (35.6% vs. 18.9%; p?=?0.002). For the nearly 50% of patients in the study whose induction time fell between 2 and 5?h, there was no difference in outcome. In light of these findings, the authors suggest that induction time may be useful as another tool for prognostication in the immediate post-arrest period – a concept that would be welcomed by those caring daily for these patients.
Some critics of the investigation by Perman et al. may suggest that without a baseline assessment of illness severity the heterogeneity of the population studied makes controlling for potential confounders related to temperature regulation and brain injury difficult at best. Initial core temperature pre-induction was not available for a majority of this cohort, although so-called “auto-cooling” and rebound hyperthermia have been described elsewhere as prognosticators for survival., Furthermore, those patients with a longer induction time actually are hypothermic longer, even if mildly (induction time plus 24?h at goal temperature), the effect of which may be beneficial., The authors note that post-arrest care, including the method of TTM induction, rapid induction goal, shivering suppression, and control of rewarming rate, was protocoled to control for differences in individuals treated with TTM. While these authors and others have also demonstrated that BMI is not associated with outcome, longer induction time has been directly associated with increasing BMI. Without data describing how much of a heat sink the TTM induction devices are providing, such as heat index, it is impossible to assess whether the shape of a patient’s induction curve is due to intact thermoregulatory reflexes, BMI, or some other, yet unmeasured factors.
Human thermoregulation is fundamentally a balance of energy transfer. Maintaining a given temperature requires the amount of heat generated by a patient to equal the amount of heat lost to the environment. When unequal, the patient’s temperature changes in the direction of the inequality. The application of TTM manipulates equilibrium by drastically increasing heat loss so that a new, lower body temperature can be reached and maintained. A patient with intact thermoregulatory reflexes may increase his metabolic rate and oxygen demand via shivering or other compensatory metabolic mechanisms. However, similar to other critical illnesses, the effect of the post-cardiac arrest syndrome upon energy expenditure is varied and unpredictable. For example, shivering may not occur in a “normal window” or at all.While we have some understanding of normal metabolism and the protective hibernation phenotype, our current ability to reliably measure mitochondrial injury and internal heat production in the clinical post-arrest period is only a crude estimation and not routinely employed.,
Most TTM protocols call for induction with up to 2?L cold intravenous saline and surface cold packs while advanced temperature management devices are being applied. Typical devices include cooling blankets or either surface or intravascular feedback-looped systems. Non-traditional mechanisms of cooling include using peritoneal lavage, total liquid ventilation, or extracorporeal membrane oxygenation (ECMO). Each induction adjunct has a unique efficiency to alter body temperature and create heat transfer. Studies mixing methods of cooling, without a standardized measure of energy transfer, may lead to confounded and perhaps the contemporary conflicting results. We cannot assume one patient’s response to these modalities is identical to the next any more than we can expect their clinical courses to be similar. Indeed, what we do to actively induce hypothermia and affect metabolic hibernation represents only one piece of the thermodynamic equation.
To give a different context, using induction time as a prognostic measure is analogous to evaluating an automobile’s performance by measuring acceleration without taking into account the terrain upon which it was measured. The utility of acceleration as a performance measure comes from the flat, straight track upon which it and other comparable vehicles are tested. In contrast, the clinical course of a post-arrest patient is rarely flat or straight, with wide variations in duration of and tissue tolerance to ischemia (i.e. time to ROSC or “downtime” while pulseless). These additional factors are simply unaccounted for even by the rigor of the Utstein template. Standardization of post-arrest TTM care, as the authors suggest, may help to reduce the number of variables in the equation. However, saying we have “the pedal to the metal” does little to quantify the impact of those unmeasured or uncontrolled variables remaining.
While induction time may ultimately become a useful surrogate for neurologic injury, it currently has only been studied retrospectively, in isolation. As applied, TTM treatment intervals remain uniform irrespective of methods used to manage temperature, measures of metabolic insult, and individual patient-level thermodynamics. However, the optimal temperature target, duration of treatment, and rates of both cooling and rewarming for post-arrest TTM are unknown. The complex interplay between impaired versus irreconcilably damaged internal thermoregulation, active internal or external temperature modification, and ultimately neurologic outcome remains poorly delineated. Before abandoning TTM or prognosticating based on rate of body temperature change, perhaps we should begin to discuss a standard method of describing the post-arrest care bundle, inclusive of the metabolic demands of TTM, while we continue the search for a measurable, therapeutic target. Identifying a potential paradox in clinical medicine is only useful if we explore how to take advantage of it to improve patient outcome. Until we can reliably quantify the “dose” of TTM beyond the interval in which it is applied, the utility of changes in duration, depth, and rate of change during TTM will continue to be of unclear value in comparing outcomes.
Conflict of interest statement
Dr. Sawyer has no conflicts of interest to disclose. Dr. Kurz wishes to disclose that he has received speaking honoraria from ZOLL Medical Corp in the last year. In addition he receives ongoing research support from the following industry partners: Rapid Pathogen Screening Inc., Boehringer-Ingelheim, and Abbott.
References