To cool, or not to cool: that is the severe TBI question

‘Effect of Early Sustained Prophylactic Hypothermia on Neurologic Outcomes Among Patients With Severe Traumatic Brain Injury: The POLAR Randomized Clinical Trial.’ Cooper DJ et al. JAMA. 2018 Dec 4;320(21):2211-2220. doi: 10.1001/jama.2018.17075.


  • Traumatic brain injury (TBI) is a leading cause of disability and mortality (Mas et al, 2017). Prevention of secondary neurological injury is paramount in TBI management (Sahuquillo and Vilalta, 2007).
  • Hypothermia has been postulated as a promising therapy for prophylactic neurological protection. 
  • Several studies have been carried out looking at whether hypothermia may improve neurological outcomes in TBI with no clear consensus.
    • RCT (Clifton et al, 2001) demonstrated no benefit. 
    • Meta-analysis (American Association of Neurological Surgeons, 2007) demonstrated some benefit using prophylactic hypothermia. However, a significant limitation was the slow induction of hypothermia (up to 8hours post primary TBI) potentially masking any benefit.
    • EUROtherm3235 (Andrews et al, 2015) showed  worsening mortality and neurological disability with therapeutic hypothermia
    • Meta-analysis 2018 (Olah et al) demonstrated a decrease in mortality when using prophylactic hypothermia.


  • To definitively answer the question as to whether or not prophylactic hypothermia is of clinical benefit in severe TBI.

Trial Design/Setting

  • Randomised Control Trial. Worldwide (6 countries: New Zealand, Qatar, Saudi Arabia, France, Australia, Switzerland) multicentred. Study between 2010 and 2018.
  • Use of block randomisation, equal (1:1) assignment to either prophylactic hypothermia intervention or controlled normothermia. Stratification and randomisation occurred pre-hospital or at the emergency department and by geographical regions.
  • Partially blinded, sealed envelopes used to allocate, clinical staff not blinded but independent outcome assessors who were following up via telephone questionnaires were blinded. Follow up was carried out up to 6months post randomisation.
  • 511 patients were randomised (study required n of 364 for an 80% two-sided power calculation to detect a 15% absolute risk increase (from 50% to 65%) of a favourable 6months neurological outcome post TBI)
  • Intervention
    • Cooling: IV infusion bolus of 0.9% Normal saline at 40C up to 2000mls. Maintenance with surface temperature control. Aim for core temperature of 330C±0.50C (or 350C if clinical bleeding concerns) maintained for 3 to 7days. Slow individualised rewarming based on ICP measurements.
    • Normothermia control: Target temperature 37.00C ± 0.50C using cooling vests/paracetamol.
    • Clinical management as per local practice and clinician judgement.


  • Inclusion criteria: TBI, age range 18 to 60years, Glasgow Coma Sore <9, intubated or “imminently” intubated. Any patients not enrolled pre-hospital were included up to 3hours post TBI.
  • Exclusion criteria: Indication of severe bleeding judged by clinical parameters/assessment (sustained tachycardia >120min-1and or systolic hypotension <90mmHg); GCS 3 and unreactive pupils, suspected pregnancy, destination hospital not one of the multicentre study sites.  
  • 511 patients randomised, 500 analysed; 266 hypothermia (256 included in 6month mortality analysis), 245 normothermia (239 in 6month follow up). 
  • Time from TBI to initial hypothermia target temperature of 350C was 2.5hours (median) with final core temperature of 330C reached in 10.1hours in 71.5% of patients.
  • Clear use of validated and established TBI scores (Marshall, IMPACT-TBI, Glasgow Outcome Scale-Extended Scores) and baseline characteristics near equivalent across both interventions (approximately 80% male, mean age 35, GCS mean 6).


  • Primary outcome. 
    • No significant favourable outcome (a GOS-E score of 5 to 8) between hypothermia and normothermia: favourable outcome 48.8% in hypothermia vs 49.1% in normothermia; absolute difference (95% CI) 2.7 (-4.3 to 9.7), p value 0.45.  Similarly there was no significant difference when adjusting for the IMPACT-TBI model prediction.
  • Secondary outcomes. 
    • No significant difference in mortality at 6months: 21.1% in hypothermia vs 18.4% in normothermia; absolute difference (95% CI) -0.4 (-9.4 to 8.7), p value 0.94 with a similar time to death. No significant difference in GOS-E at 6months remained when using sliding dichotomy and proportional odds models.
  •  Additional outcomes: 
    • No significant change between groups in days of mechanical ventilation, ICU length of stay and unfavourable GOS-E score <5.
    • No statistically significant difference in adverse events such as intracranial bleeding.
    • No significant effect of hypothermia on intracranial pressure.
    • Per-protocol analysis and as-treated analysis demonstrated significant increase in pneumonia in the hypothermia cohort.



  • Largest RCT to date and accounted for previous study limitations.
  • Achieved early induction of hypothermia 350C (median 2.5hours). 


  • Median time to achieve 330C was much greater (10hours) with 32% of the hypothermia cohort not achieving the target temperature of 330C (19% excluded as not suitable and 13% unable to achieve).  Nevertheless this likely reflects the clinical limitations in safely cooling in trauma and, although a limitation of the study, it exposes the clinical reality of hypothermia as a therapeutic intervention.
  • Bias may have been introduced as clinicians and patient families were not blinded. Clinicians were also able to veto and exclude the patient from the trial in keeping with ethical conduct.


  • Hypothermia does not improve neurological outcomes post severe
  • TBI.

Visual Abstract taken from JAMA Network: 

Summary by Dr Nadia Abid

Presented October 2nd 2019


Andrews PJ, Sinclair HL, Rodriguez A, et al; Eurotherm3235 Trial Collaborators. Hypothermia for intracranial hypertension after traumatic brain injury. N Engl J Med. 2015;373(25):2403-2412. doi:10.1056/NEJMoa1507581

Brain Trauma Foundation; American Association of Neurological Surgeons. Guidelines for the management of severe traumatic brain injury, Third Edition.J Neurotrauma. 2007;24(suppl 1):S1-106

Clifton GL, Miller ER, Choi SC, et al. Lack of effect of induction of hypothermia after acute brain injury. N Engl J Med. 2001;344(8):556-563. doi:10 .1056/NEJM200102223440803

Maas AIR, Menon DK, Adelson PD, et al; InTBIR Participants and Investigators. Traumatic brain injury: integrated approaches to improve prevention, clinical care, and research. Lancet Neurol.2017;16(12):987-1048. doi:10.1016/S1474-4422(17) 30371-X

Olah E, Poto L, Hegyi P, et al. Therapeutic whole-body hypothermia reduces death in severe traumatic brain injury if the cooling index is sufficiently high: meta-analyses of the effect of single cooling parameters and their integrated measure.J Neurotrauma. 2018;35(20):2407-2417. doi:10.1089/neu.2018.5649Sahuquillo J, Vilalta A. Cooling the injured brain: how does moderate hypothermia influence the pathophysiology of traumatic brain injury? Curr Pharm Des. 2007;13(22):2310-2322. doi:10.2174/138161207781368756

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s