Clincal Aspects of Biological Brain Damage Markers

University dissertation from Clinical Sciences, Lund University

Abstract: Biomarkers for organ damage and/or dysfuntion are used in almost all areas of medicine. The brain has eluded this technological development for some time. Recently, the S100B protein has been shown to be a promising marker of brain damage. However, before S100S can reach clinical reality, several problems must be solved. The specificity of S100B requires special investigative attention. Although the sensitivity has seldom been criticised in the literature and assumed to be very high for S100B, this also craves further analysis. Furthermore, other clinical applications of biomarkers such as S100B need to be examined.

Aim/brief methods

This thesis therefore looks at these aspects in 6 clinical studies comprising 294 separate patients. The main aims were to; examine the specificity of serum S100B in clinically relevant situations; to examine the sensitivity of serum S100B in head injury; to investigate serum S100B levels in infectious disease; to investigate the source of serum S100B in patients with multitrauma and brain injuries; to examine the clinical utility potential for serum S100B as a daily measurement parameter in neurointensive care; and finally, to examine the ability of a biomarker panel (including the three most promising brain biomarkers; S100B, GFAP and NSE, with a biomarker of coagulation system activation; PCI-APC) to differentiate ischemic from haemorrhagic acute stroke. Biomarker analysis for S100B and NSE was performed with commercial assays. Analysis of GFAP and APC-PCI was performed by non-commercial assays.


The results show that elevated levels of S100B were seen, in absence of brain injury, in isolated orthopaedic fractures, extracerebral infectious disease and in extracranial surgery after total brain herniation. These levels were generally consistent with levels seen after minor head injury. Epidural haematomas showed near-normal levels of S100B, even soon after the head trauma has occurred. Serum S100B is generally elevated in CNS infections, for instance in bacterial meningitis. S100B was especially raised in encephalitis. Daily S100B levels did not predict secondary neurological complications or outcome in the neurointensive care unit. Serum GFAP and APC-PCI measurements, prior to neuroimaging, had a 97% negative predictive power for haemorrhagic stroke and a 100% value if only samples within 12 hours were considered.


The clinical specificity of S100B is low. Elevated levels of S100B, in patients with clinical evidence of extracerebral tissue damage, should not be interpreted as brain damage. The sensitivity of serum S100B in head injury is high. However, epidural haematomas showed near-normal levels of S100B. Therefore, the magnitude of S100B levels in head trauma patients should not be directly related to the risk and/or magnitude of intracranial pathology. S100B is elevated in cerebral infections, especially in encephalitis where levels were found to be high. Most of the circulating S100B in serum after multitrauma injuries originates from the brain. Daily serum S100B levels are not clinically useful in the neurointensive care setting. Serum GFAP and APC-PCI measurements, prior to neuroimaging, can accurately rule out haemorrhagic stroke in a mixed population of stroke patients.