Background: Serum protein S100B determinations have been recently suggested as markers of traumatic brain injury. However, little is known about the effects of extracranial injuries on S100B levels in trauma patients.

Methods: We studied 224 patients with head trauma (54 of whom also had extracranial injuries), 155 patients with various types of extracranial injuries, and 8 healthy pilots exposed to high Gz forces. The head trauma patients had either no brain injury (n = 35), mild brain injury (n = 165), or moderate to severe brain injury (n = 24). The extracranial injuries were divided into small and large injuries. Serum protein S100B levels were determined from samples taken within 6 hours after the trauma event.

Results: The head trauma patients had a significantly higher median S100B (0.17 microg/L) than the patients with extracranial injuries (0.07 microg/L) (p < 0.001). Serum S100B levels also correlated with the severity of brain injury (p < 0.001), the highest values occurring in the patients with moderate to severe brain injury (1.27 microg/L). However, large extracranial injuries also elevated S100B levels (0.35 microg/L), whereas small extracranial injuries in the absence of head trauma did not significantly affect S100B levels (0.07 microg/L). Above the cutoff level of 0.13 microg/L, there were 61% of the head trauma patients and 26% of those with extracranial injuries (Pearson chi test, p < 0.001). However, only 4% of the patients with purely extracranial injuries had a concentration of S100B above the cutoff level of 0.50 microg/L, whereas the head trauma patients with moderate to severe brain injury exceeded this cutoff in 67% of the cases. Exposure to high Gz forces did not influence serum S100B levels in healthy individuals.

Conclusion: We conclude that serum S100B is a sensitive marker of brain injury, which correlates with the severity of the injury. Large extracranial injuries also elevate S100B levels. However, S100B has a high negative predictive power, and the finding of a normal S100B value shortly after trauma should thus exclude significant brain injury with a high accuracy.