Background: Traumatic brain injury (TBI) is a major cause of disability and death in young adults worldwide. It results in around 1 million hospital admissions annually in the European Union (EU), causes a majority of the 50,000 deaths from road traffic accidents and leaves a further ≈10,000 people severely disabled.
Objective: The Eurotherm3235 Trial was a pragmatic trial examining the effectiveness of hypothermia (32-35 °C) to reduce raised intracranial pressure (ICP) following severe TBI and reduce morbidity and mortality 6 months after TBI.
Design: An international, multicentre, randomised controlled trial.
Setting: Specialist neurological critical care units.
Participants: We included adult participants following TBI. Eligible patients had ICP monitoring in place with an ICP of > 20 mmHg despite first-line treatments. Participants were randomised to receive standard care with the addition of hypothermia (32-35 °C) or standard care alone. Online randomisation and the use of an electronic case report form (CRF) ensured concealment of random treatment allocation. It was not possible to blind local investigators to allocation as it was obvious which participants were receiving hypothermia. We collected information on how well the participant had recovered 6 months after injury. This information was provided either by the participant themself (if they were able) and/or a person close to them by completing the Glasgow Outcome Scale - Extended (GOSE) questionnaire. Telephone follow-up was carried out by a blinded independent clinician.
Interventions: The primary intervention to reduce ICP in the hypothermia group after randomisation was induction of hypothermia. Core temperature was initially reduced to 35 °C and decreased incrementally to a lower limit of 32 °C if necessary to maintain ICP at < 20 mmHg. Rewarming began after 48 hours if ICP remained controlled. Participants in the standard-care group received usual care at that centre, but without hypothermia.
Main outcome measures: The primary outcome measure was the GOSE [range 1 (dead) to 8 (upper good recovery)] at 6 months after the injury as assessed by an independent collaborator, blind to the intervention. A priori subgroup analysis tested the relationship between minimisation factors including being aged < 45 years, having a post-resuscitation Glasgow Coma Scale (GCS) motor score of < 2 on admission, having a time from injury of < 12 hours and patient outcome.
Results: We enrolled 387 patients from 47 centres in 18 countries. The trial was closed to recruitment following concerns raised by the Data and Safety Monitoring Committee in October 2014. On an intention-to-treat basis, 195 participants were randomised to hypothermia treatment and 192 to standard care. Regarding participant outcome, there was a higher mortality rate and poorer functional recovery at 6 months in the hypothermia group. The adjusted common odds ratio (OR) for the primary statistical analysis of the GOSE was 1.54 [95% confidence interval (CI) 1.03 to 2.31]; when the GOSE was dichotomised the OR was 1.74 (95% CI 1.09 to 2.77). Both results favoured standard care alone. In this pragmatic study, we did not collect data on adverse events. Data on serious adverse events (SAEs) were collected but were subject to reporting bias, with most SAEs being reported in the hypothermia group.
Conclusions: In participants following TBI and with an ICP of > 20 mmHg, titrated therapeutic hypothermia successfully reduced ICP but led to a higher mortality rate and worse functional outcome.
Limitations: Inability to blind treatment allocation as it was obvious which participants were randomised to the hypothermia group; there was biased recording of SAEs in the hypothermia group. We now believe that more adequately powered clinical trials of common therapies used to reduce ICP, such as hypertonic therapy, barbiturates and hyperventilation, are required to assess their potential benefits and risks to patients.
Trial registration: Current Controlled Trials ISRCTN34555414.
Funding: This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 22, No. 45. See the NIHR Journals Library website for further project information. The European Society of Intensive Care Medicine supported the pilot phase of this trial.