Therapeutic hypothermia: Not so hot for cerebral protection after cardiac arrest or traumatic brain injury
The role of therapeutic hypothermia in the management of traumatic brain injury and cardiac arrest has been called into question. Mehnaz Khan and Simon Stacey review the evidence of recent trials to determine what can be perceived as best practice.
Therapeutic hypothermia (TH) has played a major role in the management of out-of-hospital cardiac arrest (OOHCA) and often in-hospital cardiac arrest following seminal publications in the New England Journal of Medicine which led to its incorporation into the ESC guidelines in 2002 [1,2]. Similarly, TH has been a part of the recommendations issued by the Brain Trauma Foundation (BTF) for the management of traumatic brain injury (TBI) since 2007 ; however, since then, a systematic review published in the British Journal of Anaesthesia in 2013 , did not support these recommendations, and now the role of TH in the management of cardiac arrests has also come into question. In this review we look at the evidence of previous and more recent trials in order to determine what can currently be perceived as best practice.
It is clear how, theoretically, TH may benefit patients following cardiac arrest, despite its significant negative physiological effects. The use of induced hypothermia can be traced as far back as 1937 when patients were cooled in an attempt to alleviate symptoms of metastatic cancer. It has routinely been used for a number of years in certain cardiothoracic surgical procedures, and recently some neurosurgical procedures. Physiologically, hypothermia reduces the metabolic rate such that a one degree Celcius drop in core temperature is thought to decrease cerebral metabolic rate by 6–7%. As a result of this, the cerebral blood flow, and therefore (potentially) intracranial pressure (ICP) are also reduced. Due to improved regional cerebral perfusion, it is also thought that ischaemic areas are better perfused, and there may be some anticonvulsant effects of this therapy. At a molecular level, it is thought that there is suppression of a number of chemical actions associated with reperfusion injury, such as excitatory amino acid production, free radical release and calcium shifts causing mitochondrial damage .
In TBI, animal models have pointed to multiple pathways involved in neuronal injury which may be influenced by the use of TH positively .
Evidence leading to previous guidelines
In 2002 the advanced life support task force (ALS) of the internal liaison committee on resuscitation (ILCOR) recommended that unconscious adults with a ROSC following OOHCA should be cooled to 32–34°C for 12–24 hours post VF arrest. In 2010, the same group went further, suggesting that primary therapeutic hypothermia (PTH) may be useful in initial non-shockable rhythms also . These guidelines largely followed two major trials published in the New England Journal of Medicine in 2002. The first was a multicentre European trial with 273 patients with OOHCA due to ventricular fibrillation (VF) or ventricular tachycardia (VT) . They found improved survival in 59% of patients at six months, compared with 45% after standard care, as well as more favourable neurologic outcome (55% vs 39%). The second trial was smaller, with only 77 patients from four hospitals in Melbourne, Australia . In this trial, they looked at survival to hospital discharge with good neurologic outcome, and found that 49% in the hypothermia group had a good outcome, compared with 26% of the non-hypothermia group. It is worth noting that the particular recommended range of 32–34°C used in both trials has been extrapolated from experiments in animals.
Similarly, following the results of animal models of TBI as well as a number of clinical trials, the BTF issued a level III recommendation for the use of PTH in TBI management in 2007 . Between 2000 and 2007, three trials followed up patients for over 12 months and found the benefit of PTH on neurological outcome. Ten very different trials over the period of 1993–2007 found reduction in ICP associated with PTH (adults and children). Of these, two found improvement in mortality and six in neurologic morbidity.
What’s changed? Current evidence
A recent trial published in the New England Journal of Medicine in December 2013, was conducted looking at 939 patients who had an out-of-hospital cardiac arrest (between November 2010 and January 2013), and received targeted temperature management of either 33°C or 36°C. Results showed no difference in mortality rates in either group, or even any difference in neurologic function at the 180 day follow-up (evaluated by the cerebral performance category scale and modified Rankin scale) . Not only was there no statistical significance, there wasn’t even a non-statistical trend that might suggest inadequate power of the study to detect a difference. This particular randomised controlled trial (RCT) recruited patients from 36 ICUs in Europe and Australia, and the targeted temperature management period was 36 hours from randomisation, with gradual re-warming from 28 hours. Following the intervention period, temperature was still maintained at below 37.5°C until 72 hours post-arrest. This factor may have been the key difference between the previous and current trials. Neurologic evaluation by a physician blinded to treatment was performed at 72 hours. This trial was a large, high-quality RCT, adequately powered to detect a 20% reduction in hazard ratio for death, as well as a 20% relative risk reduction of mortality in the hypothermic compared with the normothermic group, yet it failed to do so. Sub-group analysis of only the VF/VT group also showed no difference. There was, however, no evidence of harm in the hypothermic group either.
These results are in stark contrast to those of the earlier papers that led to the incorporation of PTH post-cardiac arrest into the guidelines. Further analysis of these papers, however, goes some way to explaining these differences. First, there was a clear lack of blinding in previous trials, potentially leading to treatment bias. This limitation was also present in the current trial but the major difference was that there was blinded outcome assessment.
But perhaps the main difference (as is suggested by the authors of this current trial) is that the current trial didn’t allow hyperthermia to develop during the intervention period in either group and prevented further fever development over the next three days. Previous trials showed fever developed in many patients in the standard treatment group.
Other unmeasured differences between the trials are those of the pre-hospital and critical care management, which has undoubtedly improved over this last decade since the publication of the previous trials. Clearly this is difficult to evaluate, especially given the number of different centres these trials have involved.
Despite the evidence suggesting that the results of the current trial pose a very real threat to the future of PTH in management post-cardiac arrest, there were flaws in its design also. It was a multicentre trial with non-standardised care at each centre, simply with the advice to manage both patient groups in a similar fashion, despite obvious potential differences between centres. This could lead to outlier effects skewing the overall result; however, this is unlikely given the overall large size of the study and an absence of this signal in statistical analyses.
Another important factor to consider is that the original trials measured a greater temperature difference between groups, delivering a potentially higher ‘dose’ of hypothermia than in the current trial (4–4.5°C instead of 3°C). It has been argued that such strongly negative results refuting previous evidence in such a highly powered study are unlikely to be affected as a result of this small difference in temperature .
Finally, looking at the evidence for hypothermia in TBI, the review printed in the British Journal of Anaesthesia in March 2013 –comparing the multiple papers spanning from as far back as 1966 up until 2011 – found that only three trials could be considered to be of high quality following application of the ‘GRADE’ system of assessment of trials. Analysis of high-quality trials alone indicated no benefit of TH in reducing neurological morbidity after TBI. The authors of this review found that based on detailed analyses of the trials that purported to demonstrate a benefit of TH in TBI, it was not possible to make any firm conclusions outside of the obvious ICP-lowering effects of PTH. Not all trials differentiated between the use of hypothermia as a treatment to lower ICP, compared with hypothermia as a treatment in its own right for TBI. Similar to the findings of the cardiac arrest trial, it was evident following this review that in 11 of the 15 trials, there was pyrexia in the control group which may have favourably biased the outcome in the hypothermia groups. Following in-depth analyses of all of these 15 trials, the authors of this review failed to find any benefit of TH on TBI on both mortality and neurological outcome at one year following TBI. They do suggest, however, that this may be due to the paucity of high-quality trials, which may in future yield more positive results .
So is it cool to cool?
Current guidelines have yet to reflect the most recent results of trials looking at TH so it is likely that, at present, many centres will continue to cool, until clear guidelines are available to the contrary. However, the main message from all of the above trials is to avoid hyperthermia. Whatever the position on the hypothermic part of the management, maintaining normothermia at all other times is perhaps the most critical message of the trials to date.
1. The hypothermia after cardiac arrest group (2002) Mild therapeutic hypothermia to improve neurologic outcome after cardiac arrest. N. Engl. J. Med. 346(8): 549–556
2. Bernard, S.A., Gray, T.W., et al. (2002) Treatment of comatose survivors of out of hospital cardiac arrest with induced hypothermia. N. Engl. J. Med. 346(8): 557–563
3. Brain Trauma Foundation. Guidelines for the management of severe traumatic brain injury (3rd edition). Available at www.braintrauma.org/pdf/protected/Guidelines_Management_2007w_bookmarks.pdf (accessed 24 February 2015)
4. Georgiou, A.P. & Manara, A.R. (2013) Role of therapeutic hypothermia in improving outcome after traumatic brain injury: a systematic review. Br. J. Anaesthesia 110: 357–367
5. European Resuscitation Council (2010) English guidelines. Available at https://www.erc.edu/index.php/doclibrary/en/209/1/ (accessed 24 February 2015)
6. Nielsen, N., Wettersley, J., Cronberg, T., et al. (2013) Targeted temperature management at 33°C versus 36°C after cardiac arrest. N. Engl. J. Med. 369: 2197–2206
7. PulmCCM. Hypothermia did not help after out of hospital cardiac arrest, in largest study yet (NEJM). Available at http://pulmccm.org/main/2013/randomized-controlled-trials/hypothermia-help-hospital-cardiac-arrest-nejm/ (accessed 24 February 2015)
Simon Stacey and Mehnaz Khan are consultant anaesthetists at Barts Health NHS Trust, London.