Fever worsens outcomes in animal models of ischaemic stroke: A systematic review and meta-analysis

BACKGROUND: Subfebrile temperatures and fever in the first days after stroke are associated with a greater risk of a poor outcome. If this relation is causal, prevention of hyperthermia may improve outcome. Causality can be tested in animal models. We therefore assessed the effects of hyperthermia on outcomes in animal models of ischaemic stroke and explored under which conditions prevention of hyperthermia could be most effective. METHODS: We performed a systematic review and meta-analysis of data from animal experiments testing the effect of spontaneous or induced hyperthermia on outcome after focal cerebral ischaemia. Our primary outcome measure was infarct size. Normalised mean differences were combined using the random effects model and stratified meta-analysis was used to explore the impact of study characteristics. RESULTS: We included 19 publications, reporting on 49 comparisons involving 603 animals. Overall, hyperthermia increased infarct size by 43.4% (95% confidence interval, 29.8–56.9%) and worsened neurobehavioral outcomes by 48.5% (17.2–79.8%). The increase in infarct size was larger with higher temperatures. Hyperthermia was most harmful if present for more than 2 h and when started at the time of artery occlusion rather than later. CONCLUSION: Hyperthermia substantially increased infarct size in animal models of ischaemic stroke, suggesting that the relation between fever and poor outcome observed in patients is at least in part causal. These data provide support to trials testing the effect of the prevention of fever with antipyretic drugs in patients with acute stroke.