Admission S100B fails as neuro-marker but is a good predictor for intrahospital mortality in major trauma patients

Background: S100 B is an extensively studied neuro-trauma marker, but its specificity and subsequently interpretation in major trauma patients might be limited, since extracerebral injuries are known to increase serum levels. Thus, we evaluated the potential role of S100B in the assessment of severe traumatic brain injury (TBI) in multiple injured patients upon emergency room (ER) admission and the first days of intensive care unit (ICU) stay. Methods: Retrospective study employing trauma registry data derived from a level 1 trauma center. Four cohorts of patients were grouped: isolated TBI (iTBI), polytrauma patients with TBI (PT + TBI), polytrauma patients without TBI (PT-TBI) and patients without polytrauma or TBI (control). S100B-serum levels were assessed immediately after admission in the emergency room and during the subsequent ICU stay. Values were correlated with injury severity score (ISS), Glasgow Coma Score (GCS) and in-hospital mortality. Results: 780 predominantly male patients (76 %) with a median age of 48 (30-63) and a median ISS of 24 (17-30) were enrolled in the study. Admission S100B correlated with ISS and TBI severity defined by the GCS (both p < 0.0001) but not with head abbreviated injury score (AIS) (p = 0.38). Compared with survivors, non-survivors had significantly higher median S100B levels in the ER (6.14 mu g/L vs. 2.06 mu g/L; p < 0.0001) and at ICU-day 1 (0.69 mu g/L vs. 0.17 mu g/L; p < 0.0001). S100B in the ER predicted mortality with an area under curve (AUC) of 0.77 (95 % CI 0,70-0,83, p < 0.0001), vs. 0.86 at ICU-day 1 (95 % CI 0,80-0,91, p < 0.0001). Conclusion: In conclusion, S100B is a valid biomarker for prediction of mortality in major trauma patients with a higher accuracy when assessed at the first day of ICU stay vs. immediately after ER admission. Since S100B did not correlate with pathologic TBI findings in multiple injured patients, it failed as predictive neuro-marker because extracerebral injuries demonstrated a higher influence on admission levels than neurotrauma. Although S100B levels are indicative for injury severity they should be interpreted with caution in polytrauma patients.