Methods We performed an electronic search of the National Library of Medicine’s MEDLINE and Biomedical Library of University of Pennsylvania database in March 2008 using a search heading of traumatic head injury and protein biomarkers. The search was focused especially on protein degradation products (spectrin breakdown product, c-tau, amyloid-β1–42) in the last 10 years, but recent data on “classical” markers (S-100B, neuron-specific enolase, etc.) were also examined.
Conclusions We conclude that only S-100B in severe traumatic brain injury has consistently demonstrated the ability to predict injury and outcome in adults. The number of studies with protein degradation products is insufficient especially in the pediatric care. Cohort studies with well-defined end points and further neuroproteomic search for biomarkers in mild injury should be triggered. After critically reviewing the study designs, we found that large homogenous patient populations, consistent injury, and outcome measures prospectively determined cutoff values, and a combined use of different predictors should be considered in future studies.