Résumé
International audience
Background: Optimisation of brain oxygenation might improve neurological outcome after traumatic brain injury. The OXY-TC trial explored the superiority of a strategy combining intracranial pressure and brain tissue oxygen pressure (PbtO2) monitoring over a strategy of intracranial pressure monitoring only to reduce the proportion of patients with poor neurological outcome at 6 months.Methods: We did an open-label, randomised controlled superiority trial at 25 French tertiary referral centres. Within 16 h of brain injury, patients with severe traumatic brain injury (aged 18-75 years) were randomly assigned via a website to be managed during the first 5 days of admission to the intensive care unit either by intracranial pressure monitoring only or by both intracranial pressure and PbtO2 monitoring. Randomisation was stratified by age and centre. The study was open label due to the visibility of the intervention, but the statisticians and outcome assessors were masked to group allocation. The therapeutic objectives were to maintain intracranial pressure of 20 mm Hg or lower, and to keep PbtO2 (for those in the dual-monitoring group) above 20 mm Hg, at all times. The primary outcome was the proportion of patients with an extended Glasgow Outcome Scale (GOSE) score of 1-4 (death to upper severe disability) at 6 months after injury. The primary analysis was reported in the modified intention-to-treat population, which comprised all randomly assigned patients except those who withdrew consent or had protocol violations. This trial is registered with ClinicalTrials.gov, NCT02754063, and is completed.Findings: Between June 15, 2016, and April 17, 2021, 318 patients were randomly assigned to receive either intracranial pressure monitoring only (n=160) or both intracranial pressure and PbtO2 monitoring (n=158). 27 individuals with protocol violations were not included in the modified intention-to-treat analysis. Thus, the primary outcome was analysed for 144 patients in the intracranial pressure only group and 147 patients in the intracranial pressure and PbtO2 group. Compared with intracranial pressure monitoring only, intracranial pressure and PbtO2 monitoring did not reduce the proportion of patients with GOSE score 1-4 (51% [95% CI 43-60] in the intracranial pressure monitoring only group vs 52% [43-60] in the intracranial pressure and PbtO2 monitoring group; odds ratio 1·0 [95% CI 0·6-1·7]; p=0·95). Two (1%) of 144 participants in the intracranial pressure only group and 12 (8%) of 147 participants in the intracranial pressure and PbtO2 group had catheter dysfunction (p=0.011). Six patients (4%) in the intracranial pressure and PbtO2 group had an intracrebral haematoma related to the catheter, compared with none in the intracranial pressure only group (p=0.030). No significant difference in deaths was found between the two groups at 12 months after injury. At 12 months, 33 deaths had occurred in the intracranial pressure group: 25 (76%) were attributable to the brain trauma, six (18%) were end-of-life decisions, and two (6%) due to sepsis. 34 deaths had occured in the intracranial pressure and PbtO2 group at 12 months: 25 (74%) were attributable to the brain trauma, six (18%) were end-of-life decisions, one (3%) due to pulmonary embolism, one (3%) due to haemorrhagic shock, and one (3%) due to cardiac arrest.Interpretation: After severe non-penetrating traumatic brain injury, intracranial pressure and PbtO2 monitoring did not reduce the proportion of patients with poor neurological outcome at 6 months. Technical failures related to intracerebral catheter and intracerebral haematoma were more frequent in the intracranial pressure and PbtO2 group. Further research is needed to assess whether a targeted approach to multimodal brain monitoring could be useful in subgroups of patients with severe traumatic brain injury-eg, those with high intracranial pressure on admission.