Authors: Andrews PJD et al., N Engl J Med 2015 Oct 7;
A multicenter randomized trial was stopped early because hypothermia worsened outcomes when used as second-line therapy for elevated intracranial pressure.
Induced hypothermia has become popular for treating cerebral edema from traumatic brain injury (TBI), but we do not know the effect on long-term outcomes. Therefore, European investigators performed a multicenter randomized trial of hypothermia for TBI. The authors recruited patients with TBI and intracranial pressure >20 mm Hg for ≥5 minutes despite first-line therapy with mechanical ventilation, sedation, analgesia, head-of-bed elevation, cerebrospinal fluid drainage, and surgical removal of space-occupying lesions. Participants were randomized to second-line therapy with mannitol with or without hypertonic saline (control group) or to induced hypothermia (32–35°C), with hyperosmotic therapy only if needed (hypothermia group). Third-line treatments (e.g., barbiturates, craniectomy) were allowed in both groups if needed. Blinded assessors determined the primary outcome: Glasgow Outcome Scale-Extended (GOS-E) score at 6 months.
After 387 patients at 47 centers in 18 countries were enrolled, the trial was stopped early because of safety concerns. Hypothermia led to higher overall odds of a worse outcome (P=0.04). Good outcomes (GOS-E score 5–8) occurred in 26% of the hypothermia group versus 37% of the control group (P=0.03). Third-line therapy with barbiturates was used in more control than hypothermia patients (41 vs. 20 patients).
This trial leaves many questions unanswered. What about hypothermia as a third-line agent? Do barbiturates improve outcomes after TBI? We need randomized trials to answer these questions. Arguments for why neurocritical care trials are not feasible do not hold up. Patients are too complicated for any single intervention to show an effect? Both DECRA (NEJM JW Neurol Jul 2011 and N Engl J Med 2011; 364:1493) and this trial showed an effect — just not in the direction we expected. Randomizing some patients to not receive an already-used therapy is unethical? This common clinical practice (which requires no informed consent) has been hurting patients. Until further high-quality trials are done, clinicians should avoid hypothermia as a second-line therapy for managing cerebral edema after TBI.