Authors: Cooper DJ, Nichol AD, Bailey M, Bernard S, Cameron, PA, Fili-Floury S, Forbes A, Gantner D, Higgins AM, Huet O, Kasza J, Muray L, Newby L, Presneill JJ, Rashford S, Rosenfeld JV, Stephenson M, Vallance S, Varma D, Webb SAR, Trapani T, McArthur C. “Effect of early sustained prophylactic hypothermia on neurological outcomes among patients with severe traumatic brain injury: the POLAR randomized clinical trial.” JAMA
2018 Oct 24. doi:10.1001/jama.2018.17075
. Published online ahead of print.
Reviewed by: Sara Stern-Nezer, MD, MPH, Assistant Clinical Professor, University of California, IrvineRead the article.**You will need to log in to US National Library of Medicine National Institutes of Health to read this article.
Hypothermia mitigates multiple pathways implicated in secondary damage after neurological injury, but data on its potential role for treatment of TBI remains unclear, especially as many prior studies had delayed onset of hypothermia. POLAR-RCT enrolled patients with severe TBI within three hours of presentation who were 18-60 years of age with GCS<9. They excluded patients with hemorrhagic shock, suspected pregnancy, GCS=3 and nonreactive pupils. Patients were randomized to prophylactic hypothermia (33°C, or 35°C if bleeding concerns persisted) or controlled normothermia (37±0.5°C). Hypothermia was induced using 4°C fluid and surface-cooling. Normothermia was maintained using surface-cooling wraps if necessary. Temperature was maintained for 72 hours after randomization then rewarming was done at 0.25°C/hr. If ICP was >20mm Hg before or during rewarming, patients were re-cooled and assessed periodically for rewarming for up to seven days post randomization; once rewarmed, patients were kept at 37°C for seven days to avoid hyperthermia. Primary outcome was “favorable outcome” at 6 months defined as GOS-E 5 to 8; secondary outcomes included GOS-E as ordinal variable, mortality at discharge and 6 months, adverse events within 10 days of randomization. Outcomes were compared using unadjusted chi-squared tests. Study was powered to show a 15 percent difference between groups. Hierarchic multi-variable log-binomial regression was used to adjust for the IMPACT-TBI score. Patient survival was analyzed using Cox proportional hazard regression censored at six months.
Of 511 patients enrolled, 466 were included in the final analysis. Most patients were excluded because of missing data; 11 withdrew consent and five were lost to follow-up. There were no significant differences between the hypothermia and normothermia groups in terms of age, gender, GCS, pupillary response, hypotension or hypoxia, CT Marshall score, IMPACT-TBI score or cause of injury. Median time from injury to randomization was 1.9 hours (IQR 1.0-2.7) in hypothermia group and 2.0 hours (IQR 1.1-2.8) in the normothermia group, and time to target temperature of 33°C was a median of 10.1 hours (IQR 6.8-15.9). Thirty-three percent of the hypothermia group received <48 hours of hypothermia, and 27 percent never reached the target temperature due to complications or physician decisions. At 6 months post-injury, there was no significant difference in favorable outcomes between the groups (p=0.94), even after adjusting for injury severity using IMPACT-TBI (p=0.75). This held true when GOS-E was analyzed as an ordinal variable. There was also no difference in mortality at discharge nor at six months nor any statistically significant difference in infections or bleeding risk between the two groups. Bradycardia was more common in the hypothermia group (18.8 percent vs. 4.2 percent, p<0.0001) but there was no difference in other cardiovascular events including hypotension or other cardiac arrhythmias. To account for the patients in the hypothermia group that were rewarmed prematurely because of significant bleeding, physician discretion, or realization that TBI was not as severe as initially indicated, study authors underwent secondary as-treated and per-protocol analyses which excluded patients for whom hypothermia was withdrawn and who did not undergo cooling for the 72-hour protocol. There was no difference in outcomes except that pneumonia was significantly more likely in both the per protocol (p=0.02) and the as-treated (p=0.003) analyses.
This large, multi-site, international randomized controlled trial of early prophylactic hypothermia for severe TBI did not show any benefit for mortality nor functional outcome at six months after injury. Interestingly, there were no significant differences in complications between groups in the intention-to-treat analysis except for bradycardia, which was not symptomatic. This was a well-designed, large study with two groups that were reasonably similar and added to previous literature by initiating temperature control or hypothermia within hours of initial injury. The main limitation of this study is that many patients were not cooled for the full 72 hours, leading to considerable variability in the duration patients remained at their target temperature, with significant overlap at times between treatment and control patient temperatures. To account for these variations, study authors performed a secondary per protocol analysis that included only patients who underwent the full treatment protocol. This also failed to show any benefit of early prophylactic hypothermia on either mortality or outcome, however it included only 175 patients and as such was not powered to show the proposed 15 percent treatment effect. Overall, this study does not support the use of early prophylactic hypothermia for severe TBI and the per protocol analysis suggests it may worsen risk of pneumonia. With that exception, this study does support the relative safety of hypothermia to 33°C from a bleeding and cardiovascular standpoint.
#LiteratureWatch #NEWSReview #SaraStern-Nezer