Brain injury biomarkers as outcome predictors in pediatric severe traumatic brain injury.

BACKGROUND: To systematically review the literature on brain injury biomarkers, defined as any injury biomarker detected in cerebrospinal fluid (CSF) or blood injury biomarkers primarily expressed in the brain parenchyma, to determine outcome prediction in pediatric severe traumatic brain injury (sTBI).

METHODS: A search of MEDLINE(®), EMBASE(®), PsycINFO(®), Pubmed(®), and the Cochrane Database, as well as grey literature sources, personal contacts, hand searches, and reference lists. The search terms used were traumatic brain injury, biomarkers, prognosis, and children. No language, publication type, or publication date restrictions were imposed. All articles were critically reviewed by two clinicians independently.

RESULTS: A total of 7,150 articles were identified initially with 16 studies identified for review. Eighteen different biomarkers were examined; 11 in CSF and 7 in blood. Outcomes assessed included either in-hospital mortality or functional state (hospital discharge, 3-months or 6-months; Glasgow Outcome Scale or Pediatric Cerebral Performance Category). Significant correlations were established between sTBI outcomes and various biomarkers in CSF (IL-6, IL-8, IL-1β, S100β, NGF, NSE, DCX, ET-1, HMGB-1, cytochrome C) and blood (GFAP, NF-H, UCH-L1, SBDP-145, leptin). Mixed results were obtained for blood S100β. Outcome did not correlate with several biomarkers in either CSF (BDNF, GDNF, α-Syn) or blood (NSE, MBP). The Class of Evidence was considered II in 1 study and III in the remaining 15 studies.

CONCLUSIONS: Based on the status of current sTBI biomarker research, we recommend that future research should be directed at both novel biomarker discovery and validation of biomarker panels in large, well-designed longitudinal studies.