Translational evidence for two distinct patterns of neuroaxonal injury in sepsis: a longitudinal, prospective translational study

BACKGROUND: Brain homeostasis deteriorates in sepsis, giving rise to a mostly reversible sepsis-associated encephalopathy (SAE). Some survivors experience chronic cognitive dysfunction thought to be caused by permanent brain injury. In this study, we investigated neuroaxonal pathology in sepsis. METHODS: We conducted a longitudinal, prospective translational study involving (1) experimental sepsis in an animal model; (2) postmortem studies of brain from patients with sepsis; and (3) a prospective, longitudinal human sepsis cohort study at university laboratory and intensive care units (ICUs). Thirteen ICU patients with septic shock, five ICU patients who died as a result of sepsis, fourteen fluid-resuscitated Wistar rats with fecal peritonitis, eleven sham-operated rats, and three human and four rat control subjects were included. Immunohistologic and protein biomarker analysis were performed on rat brain tissue at baseline and 24, 48, and 72 h after sepsis induction and in sham-treated rats. Immunohistochemistry was performed on human brain tissue from sepsis nonsurvivors and in control patients without sepsis. The clinical diagnostics of SAE comprised longitudinal clinical data collection and magnetic resonance imaging (MRI) and electroencephalographic assessments. Statistical analyses were performed using SAS software (version 9.4; SAS Institute, Inc., Cary, NC, USA). Because of non-Gaussian distribution, the nonparametric Wilcoxon test general linear models and the Spearman correlation coefficient were used. RESULTS: In postmortem rat and human brain samples, neurofilament phosphoform, β-amyloid precursor protein, β-tubulin, and H&E stains distinguished scattered ischemic lesions from diffuse neuroaxonal injury in septic animals, which were absent in controls. These two patterns of neuroaxonal damage were consistently found in septic but not control human postmortem brains. In experimental sepsis, the time from sepsis onset correlated with tissue neurofilament levels (R = 0.53, p = 0.045) but not glial fibrillary acidic protein. Of 13 patients with sepsis who had clinical features of SAE, MRI detected diffuse axonal injury in 9 and ischemia in 3 patients. CONCLUSIONS: Ischemic and diffuse neuroaxonal injury to the brain in experimental sepsis, human postmortem brains, and in vivo MRI suggest these two distinct lesion types to be relevant. Future studies should be focused on body fluid biomarkers to detect and monitor brain injury in sepsis. The relationship of neurofilament levels with time from sepsis onset may be of prognostic value. TRIAL REGISTRATION: ClinicalTrials.gov, NCT02442986. Registered on May 13, 2015. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (doi:10.1186/s13054-017-1850-7) contains supplementary material, which is available to authorized users.