1. Brain Cooling in Traumatic Brain Injury and Stroke
Bridget Harris, PhD, RGN
Clinical Research Specialist, NHS Lothian
Research Fellow, University of Edinburgh

2. Content
Therapeutic temperature reduction after acute cerebral insults - evidence and practice - normothermia, hypothermia
Terminology and scope
Systemic versus brain cooling – why brain cooling?
Methods of non-invasive brain cooling – pros and cons
Temperature measurement
Effect of therapeutic brain cooling on temperature – clinical studies
Future directions

3. Terminology and scope
Selective brain cooling vs therapeutic brain cooling – terminology
Therapeutic brain cooling methods
Invasive
neuroprotection during surgery e.g. antegrade cerebral perfusion for aortic arch surgery
Non-invasive
Nasal/pharyngeal cooling
External head cooling
Therapeutic brain cooling – acute cerebral insults - global and focal – normothermia, hypothermia
Selective brain cooling and therapeutic brain cooling – clinicians often conflate the terminology but they aren’t the same thing – it’s possible to therapeutically force intracranial temperature below body temperature but that isn’t natural selective brain cooling. But some of the therapeutic methods may utilise heat loss physiology better than others.
Non-invasive brain cooling (nasal/pharyngeal cooling and external head cooling) – for acute cerebral insults either focal e.g. TBI, stroke or global e.g. cardiac arrest in which cooling goal may be normothermia or hypothermia – not invasive methods employed during surgery e.g. antegrade cerebral perfusion and direct irrigation which are employed pre-insult
NB there are other conditions in which head cooling may be used therapeutically e.g. scalp cooling to prevent hair loss with chemotherapy (Breed et al, 2011), multiple sclerosis [Reynolds et al., 2011], epilepsy [Bajic et al, 2008;Motamedi et al 2013], headache [Ucler et al., 2006] and insomnia [Nofzinger and Buysse, 2011] – are some of the areas in which it is being researched

4. Evidence for therapeutic temperature reduction in acute global cerebral insults
Comatose survivors of cardiac arrest (VF), neonatal hypoxic ischaemic injury
therapeutic hypothermia reduces mortality, improves functional outcome
(Arrich et al. Cochrane Database Syst Rev 2012;9:CD004128;
Jacobs et al. Cochrane Database Syst Rev 2013;1:CD00331)
In these conditions therapeutic hypothermia is recommended as part of standard care
In acute global cerebral insults -
The only 2 conditions in which there is sufficient evidence for therapeutic hypothermia are comatose survivors of cardiac arrest – in particular after ventricular fibrillation - and babies with birth asphyxia – neonatal hypoxic ischaemic encephalopathy - these are systemic ischaemic insults in these conditions therapeutic hypothermia is currently part of standard care

5. Evidence for therapeutic temperature reduction in acute focal cerebral insults
Traumatic brain injury (TBI), stroke
experimental evidence
improved outcome with normothermia and hypothermia
multifactorial neuroprotective effects (early)
prevention and reduction of secondary insults
(Dietrich & Bramlett Prog Brain Res 2007;162:201-17; van der Worp et al. Brain 2007;130: )
human evidence
increased temperature is common and associated with worse outcome - death and disability (e.g. Greer et al Stroke 39: )
insufficient evidence that therapeutic temperature modulation - normothermia or hypothermia - improves outcome
(Saxena et al. 2008; Sydenham et al. 2009; den Hertog et al Cochrane Database Syst Rev; Lakhan & Pamplona Stroke Res Treat 2012:295906; Georgiou & Manara Br J Anaesth 2013;110:357-67)
normothermia is standard practice +/- hypothermia for refractory raised intracranial pressure
(Johnston et al. Resuscitation 2006;70:254-62; Thompson et al. J Neurosci Nurs 2007;39:151-62)
shivering
In acute focal cerebral insults i.e. stroke and traumatic brain injury
ANIMAL - considerable experimental evidence of the neuroprotective effects of hypothermia if given early enough - these effects include reduction of metabolic rate, modulation of cerebral blood flow and the inflammatory response, reduction of excitotoxic damage and cerebral oedema, it’s considered one of the interventions with most promise for translation to humans
HUMAN - TBI and stroke, same certainty regarding benefits of temperature reduction doesn’t hold – evidence is still inconclusive - one of the difficulties is delivering temperature interventions early enough - in animal studies cooling is usually delivered before, during or immediately after the insult
But, raised temperature after brain injury is common and there is evidence associating it with worse outcome, though that may simply reflect worse injury or development of infection which are independently associated with worse outcome
Despite the lack of definitive evidence the usual clinical aim is maintenance of normothermia, with hypothermia being used to treat raised intracranial pressure if other interventions fail – it does reduce pressure but whether improved functional outcome follows is uncertain - Eurotherm3235 trial (Edinburgh) is investigating that
Difficult to reduce temperature to normothermia or hypothermia if the pt is defending against this – shivering can be a significant problem – and sedation and measures to prevent and treat shivering are essential. Brain cooling doesn’t necessarily reduce shivering problems.
A number of trials in progress e.g. POLAR-RCT in Australia looking at early hypothermia in TBI, EuroHyp – hypothermia in stroke.

6. Systemic cooling versus brain cooling – why brain cooling?
Systemic methods – drugs (e.g. acetaminophen), cooling blankets/pads, intravenous cooling catheters – side effects
Brain cooling – nasal/pharyngeal cooling and external head cooling – rationale
Brain cooling has fewer side-effects than systemic hypothermia e.g. infection – some studies use body warming (Feigin et al. J Clin Neurosci 2002;9:502-7; Gluckman et al. Lancet;365:663-70; Harris et al. J Neurosurg 2009;110: )
Brain rather than body temperature is important in cerebral protection (Busto et al. J Cereb Blood Flow Metab 1987;7:729-38; Busto et al. Crit Care Med 1989;20:1113-4)
Preferential cooling of cortices (external cooling) ?of benefit (Wityk Crit Care Med 1994;8: )
Little evidence in humans (Harris et al. HTA 2012;16;1-175)
Systemic cooling – side effects e.g. shivering – which can lead to failure of cooling – increased sedation requirements, increased infection e.g. pneumonia is common with hypothermia
Assumption has been that brain cooling has fewer side-effects than systemic hypothermia, although in fact there is no good evidence in humans to show whether it actually does
Some studies use body warming with head cooling the theory being this will reduce cooling side-effects
In babies it’s easy to induce hypothermia with head cooling (they don’t shiver) and body warming is needed to prevent over cooling
In adults body warming may be counter productive. There is experimental evidence that considerable intracerebral gradients occur with head cooling plus body warming (Laptook et al 2001 Pediatrics 108:1103–1110) – not known if this is detrimental.
Experimental evidence shows that it’s brain rather than body temperature which is important for neuroprotection in cerebral trauma and ischaemia – therefore logically cooling should be targeted at the brain.
External brain cooling could preferentially cool the more metabolically active cortices - gray matter – gray matter may be less resistant to hypoperfusion than white matter.
When cooling is clinically warranted systemic cooling is the norm – our systematic review of head cooling found no good quality evidence that it improves functional outcome - and it remains a research intervention at present

7. Methods of non-invasive brain cooling – pros and cons
Nasal/pharyngeal cooling – induce heat loss from the upper airways by
convection +/- evaporation e.g. nasal gas flow, nasal lavage
conduction e.g. nasal or pharyngeal balloons
External head cooling – induce heat loss through the skull by
convection +/- evaporation e.g. fanning,
conduction e.g. circulating liquid cooling helmet (active), ice packs/frozen gel helmet (passive)
Pros and cons
Two non-invasive avenues of cooling – nasal/pharyngeal cooling induces heat loss from the upper airways – either by convection or conduction - and external head cooling induces heat loss through the skull, also by convection or conduction +/- evaporation
Pros and cons
Nasal and pharyngeal methods –
relatively invasive compared to external methods – contraindicated with base of skull fracture (which may be hard to rule out conclusively in TBI). Rhinochill the only commercially available device, requires a protected airway, intended for short-term use only (about 1 hr) and expensive (uses nebulised perfluorcarbon) – nasal and pharyngeal balloons are another method – not in commercial production, these are inflated to conform them to the nasal passages or pharynx which with longer term use may cause damage from pressure
External head cooling –
convective methods using water or air are uncommon but arguably utilise physiological avenues of heat loss from the head better than conductive methods – scalp veins are not compressed, evaporative heat loss is possible – whether this makes them more effective in reducing temperature is not known yet – methods using air relatively noisy
conductive methods – helmets circulating cold fluid – temperature of the circulating fluid can be set and maintained, heat is removed by the circulating fluid – but to be really effective require pressurising to conform them to the head which is contraindicated with skull fractures and may increase intracranial pressure
Ice or gel helmets – thaw out and require replacing frequently – no control over the cooling temperature

8. Temperature Temperature measurement sites
Intracranial
Tympanic
Magnetic resonance spectroscopy
Core body – PA, oesophagus, bladder, rectum – best proxy
Effectiveness of brain cooling in reducing temperature (Harris et al. HTA 2012;16(45)1-175)
Sites of temperature measurement
Intracranial temp (either in parenchyma – brain tissue - or a ventricle) is gold standard for brain temperature measurement but is generally used only in severe TBI - cardiac arrest and stroke patients don’t usually have invasive brain monitoring
Tympanic temperature even when measured correctly on the ear drum (rather than with an infrared thermometer) greatly overestimates intracranial temperature reduction with non-invasive head cooling [Abou-Chebl et al., 2011;Poli et al., 2013]
MRS – is not routinely used for temperature measurement in pts – has been used in healthy volunteers and is undergoing development to improve reliability, better for measuring temp change than absolute temperature
Core body temperature is probably the best proxy for intracranial temperature for assessing temperature reduction with brain cooling – if it’s reduced this means the effect of head cooling is more than simply superficial – whether brain cooling can reduce deeper brain temperature at all, let alone sufficiently, has been a concern among clinicians – some discount it as a therapy for that reason
Effectiveness of temperature reduction:
An aim of our systematic review (Harris et al HTA 2012;16(45)1-175) was to assess the effectiveness of brain cooling in reducing temperature – either intracranial or core body – disappointingly little information – temperature reduction was poorly reported – often not at all – the target temperature usually reported but not baseline temperature and whether target temperature was achieved
Jugular bulb temp occasionally measured – but the jugular bulb is not within the cranium and receives some extracranial blood (Kety and Schmidt. J Clin Invest 1948;27:476-83) and pts after TBI and with intracranial pressure problems are nursed head up which means the vertebral veins together with the spinal epidural veins may be receiving the majority of cerebral venous outflow (Valdueza et al. Lancet 2000;355:200-1).

9. Summary of average temperature reduction
with therapeutic cranial cooling
(studies reporting temperature reduction achieved)
This is a summary of studies in which temperature change was reported –
These studies were in cardiac arrest, severe stroke and TBI - mainly feasibility studies of cranial cooling methods and devices, only two (Harris et al 2007 and 2009) were randomised trials. Core body temp was measured in the oesophagus, bladder or rectum
The Rhinochill intranasal device and the liquid circulating devices could reduce temperature by around 1 °C or more, within 1 hour. This is a reasonable effect if normothermia is the aim, it may not be enough if hypothermia is require
References
 Andreas J, Losert H, Bayegan K, Haugk M, Arrich J, Krizanac D, Laggner AN, Barbut D, Sterz F (2008) Nasal cooling with a new cooling device in patients after cardiac arrest and successful resuscitation. Resuscitation 77: S29
Busch HJ, Eichwede F, Fodisch M, Taccone FS, Wobker G, Schwab T, Hopf HB, Tonner P, Hachimi-Idrissi S, Martens P, Fritz H, Bode C, Vincent JL, Inderbitzen B, Barbut D, Sterz F, Janata A (2010) Safety and feasibility of nasopharyngeal evaporative cooling in the emergency department setting in survivors of cardiac arrest. Resuscitation 81:
Gaida BJ, Yaldizli OO, Mnk S, Muroi C, Mudra R, Fröhlich J (2008) Treatment of resistant fever with local cerebral cooling: P 029. European Journal of Anaesthesiology 25: 11
Harris BA, Andrews PJD, Murray GD (2007) Enhanced upper respiratory tract airflow and head fanning reduce brain temperature in brain-injured, mechanically ventilated patients: a randomized, crossover, factorial trial. Br J Anaesth 98: 93-99
Harris OA, Muh CR, Surles MC, Pan Y, Rozycki G, Macleod J, Easley K (2009) Discrete cerebral hypothermia in the management of traumatic brain injury: a randomized controlled trial. J Neurosurg 110:
Miller E (2009) Determination of the rate and degree of selective brain cooling in adults with the TraumaTec NeuroWrap (abstract P94). Journal of neurotrauma 26: A25
Poli S, Purrucker J, Priglinger M, Diedler J, Sykora M, Popp E, Steiner T, Veltkamp R, Bosel J, Rupp A, Hacke W, Hametner C (2013) Induction of cooling with a passive head and neck cooling device: effects on brain temperature after stroke. Stroke ( ) 44:
Springborg JB, Springborg KK, Romner B. (2013) First clinical experience with intranasal cooling for hyperthermia in brain-injured patients. Neurocrit Care 18(3):400-5
Sung G, Torbey M, Abou-Chebl A (2009) Rhinochill: A Novel Brain Hypothermia Delivery Device. Neurology 72: A75
Wang H, Olivero W, Lanzino G, Elkins W, Rose J, Honings D, Rodde M, Burnham J, Wang D (2004) Rapid and selective cerebral hypothermia achieved using a cooling helmet. J Neurosurg 100:
* includes mean and median data, all other temperatures are mean reductions

10. Rhinochill Intranasal Cooling Device
Nasal/pharyngeal cooling:
Rhinochill device – which uses nebulised perfluorcarbon to induce convective and evaporative heat loss from the upper airways – just under the base of the brain
Rhinochill Intranasal Cooling Device
Benechill, Inc. USA

11. Mean temperature reductions during the 1-hour RhinoChill induction
Fig 1 Abou-Chebl et al. Stroke 2011;42:2164-9
Results with Rhinochill in a study of 15 patients with TBI or stroke – for reduction of fever and intracranial pressure
Intracranial temperature measured in parenchyma in 11 of the pts
Core temperature – rectal in 11 pts – otherwise 2 x PA, 1 oesophagus, 1 bladder
Tympanic temperature was measured in 10 pts – how was not stated – this data demonstrates that tympanic temperature overestimates intracranial temperature reduction
Mean temperature reductions during 1-hour cooling with RhinoChill
ICT = intracranial temperature (Abou-Chebl et al. Stroke 2011;42:2164-9)
Copyright © American Heart Association

12. QuickCool nasal balloons
Nasal/pharyngeal cooling:
QuickCool - uses balloons passed up each nostril with cold saline circulated through them
This has been tested in volunteers (Covaciu et al 2011) and patients (Springborg et al 2013)
QuickCool nasal balloons
QuickCool AB, Lund

13. Springborg et al. Neurocrit Care 2013; 18(3):400-5
Results with Quickcool set at 1oC in hyperthermic pts with brain injuries (TBI and haemorrhagic stroke) - median intracranial temperature reduction measured directly in parenchyma (3 cm depth) was approximately 1oC over 72 hrs. The aim of normothermia was not achieved.
Fig. 2 Median temperature in the cerebrum in the first 72 h of cooling. As indicated by the linear regression line a temperature level of 37°C was not reached within 72 h of cooling
(y = x , R2 = 0.111, p < ) (n=6)
Springborg et al. Neurocrit Care 2013; 18(3):400-5

14. 12 pts – traumatic brain injury or subarachnoid haemorrhage
Moving on to a study in 12 pts with traumatic brain injury or SAH – a factorial randomised trial testing the effect of nasal airflow with NO (for local vasodilation) compared to head fanning i.e. an upper airways and external head cooling study
The interventions were delivered for 30 mins – no intervention, airflow, fanning and both together -
they were delivered in random order, not the order shown here to illustrate temperature differences
You can see, broadly speaking, that both together induced a larger reduction in intracranial temperature than either on their own but the temperature reductions were not large
Harris et al. Br J Anaesth 2007;98:93-9
12 pts – traumatic brain injury or subarachnoid haemorrhage
Intracranial temperature reduction compared to baseline with:
1. no intervention
2. nasal airflow - twice minute volume (≤24 L) + 20ppm NO
3. bilateral head fanning (ambient air approximately 8 m s-1)
4. airflow plus fanning
(Harris et al. Br J Anaesth 2007;98:93-9)

15. MedCool Device (Wandaller et al. Am J Emerg Med 2009;27:460-5)
External head cooling – water:
Example of a device using water – was trialled in cardiac arrest patients. Used with systemic cooling so no temperature data of its effect alone. Not in commercial production
Wandaller C, Holzer M, Sterz F, Wandaller A, Arrich J, Uray T, Laggner AN, Herkner H. Head and neck cooling after cardiac arrest results in lower jugular bulb than esophageal temperature. American Journal of Emergency Medicine 2009 May;27(4):460-5.
MedCool Device
(Wandaller et al. Am J Emerg Med 2009;27:460-5)

16. Forced convective device - soft, fabric helmet
External head cooling – convection – air:
This is a convective cooling helmet (14oC air) – it was used in cardiac arrest as an adjunct to systemic therapeutic hypothermia – time to target temperature was shorter with systemic cooling combined with head cooling than with systemic cooling alone. Intracranial temperature wasn’t measured. This device is also not commercially available
Forced convective device - soft, fabric helmet
(Wass et al. J Cardiothorac Vasc Anesth 2013;27: )

17. CoolSystem Discrete Cerebral Hypothermia Device
External head cooling – active conduction:
E.g. of a cooling helmet which actively circulates cold liquid through channels in the helmet and requires pressurising (15 mmHg). Used in a RCT in TBI
Harris OA, Muh CR, Surles MC, Pan Y, Rozycki G, Macleod J, Easley K. Discrete cerebral hypothermia in the management of traumatic brain injury: a randomized controlled trial.[Erratum appears in J Neurosurg Jun;110(6):1322]. Journal of Neurosurgery 2009 June;110(6):
CoolSystem Discrete Cerebral Hypothermia Device
(Harris et al. J Neurosurg 2009;110: )

18. 12 pts – traumatic brain injury – 12 head cooled, 13 controls
Harris et al. J Neurosurg 2009;110: results:
RCT – traumatic brain injury – 12 head cooled pts, 13 controls
Brain temperature – parenchyma or a ventricle
Intracranial temperature reduced below bladder temperature but bladder temperature was being artificially maintained above 36oC with body warming to prevent systemic hypothermia
12 pts – traumatic brain injury – 12 head cooled, 13 controls
Mean intracranial temperatures in cooled (cap) vs not cooled (no cap)
After 24 hours, cooled group intracranial temperature 1.2°C lower than controls
(Harris et al. J Neurosurg ; )

19. Sovika head and neck cooling device (Sovika GmbH)
External head cooling – passive conduction:
Sovika head and neck cooling helmet – containing frozen gel, helmet requires replacing as it absorbs heat from head and warms up
Has been used in severe stroke
Sovika head and neck cooling device (Sovika GmbH)
(Poli et al. Stroke 2013;44:708-13)

20. Brain, bladder, and tympanic temperatures – stroke patients
Results with Sovika in severe stroke
11 stroke patients but the device was used on more than one occasion in a number of them – hence 51 applications
Intracranial temperature was measured in parenchyma (3 cm below dura)
Body temp in the bladder
Tympanic – continuous with probe in the ear
Top left graph shows temp reductions in the 11 pts on the initial application – brain and bladder started off similar, brain temperature reduced commensurately more with cooling, again tympanic temperature greatly overestimates the effect of cooling on intracranial temperature
Brain, bladder, and tympanic temperatures. A, Mean curves of the 11 initial applications (n=11 patients). B, Mean curves including all 51 applications (n=11 patients). C, Individual and mean curves of the 11 initial applications (n=11 patients). D, Individual and mean curves of all 51 applications (n=11 patients).
(Poli et al. Stroke 2013;44:708-13)
Copyright © American Heart Association

21. Future directions
Non-invasive methods of measuring intracranial temperature – continuous measurement
Device development
Higher quality studies – temp reporting, outcome
complications of cranial cooling vs systemic cooling
cranial cooling to reduce intracranial pressure
Standardardised terminology for therapeutic cranial cooling and methods
Improved non-invasive methods of measuring intracranial temperature – suitable for continuous measurement
Zero heat flux thermometer – available in Japan and under development by Arizant - may be a way forward
Device development – more effective non-invasive devices
Higher quality studies of brain cooling - better temperature reporting, functional outcome assessment
Unanswered questions include whether cranial cooling:
- Has fewer side effects and complications than systemic cooling?
Can reduce intracranial pressure?
Improves functional outcome?
Develop a standard terminology for therapeutic cranial cooling and methods – will facilitate clarity and help enormously with systematic reviewing

22. Thank you Systematic review of head cooling in adults
after traumatic brain injury and stroke
Harris, Andrews, Murray, Forbes, Moseley
Health Technology Assessment 2012;16(45):1-175
can be downloaded without charge from:
Project funded by the UK National Institute for Health Research Health Technology
Assessment Programme, project number