ABSTRACT
This study was designed to evaluate the correlation between serum levels of protein S100B and neuron-specific enolase [NSE] and the severity and outcome of traumatic brain injury [TBI] so as to be used as prognostic markers for cases admitted to Intensive care unit [ICU] after TBI. The study comprised 40 patients with head injury of varied severity and 10 volunteers [control group]. Inclusion criteria were head injury and presentation to the emergency department within 6 hours of injury. Initial injury severity was assessed using the Glasgow coma score [GCS] and all patients had cranial CT scans and lesions were evaluated with respect to lesion topography and territories of vascular supply. A venous blood sample was collected at admission for estimation of serum protein S100B and NSE levels. All patients received measures to decrease intracranial pressure [ICP] and phyntoin for posttraumatic seizures and underwent the appropriate neurosurgical procedure according to type of post-traumatic lesion. Follow-up was conducted monthly and the final outcome at six months was assessed using the Expanded Disability Status Scale [EDSS]. The mean initial GCS score was 11.1 +/- 3; 14 patients [35%] had mild, 17'patients [42.5%] had moderate and 9 patients [22.5%] had severe trauma. Normal CT was reported in 14 patients [35%]; however, CT scanning detected extradural hemorrhage in 5 patients [12.5%] fissure skull fracture in 5 patients [12.5%], fissure basal skull fracture in 3 patients [7.5%] and depressed skull fracture in 2 patients [5%]; while 3 patients [7.5%] had intracerebral hemorrhage and the other 8 patients had subdural hemorrhage, subdural hemorrhage with contusion, subdural hematoma and subarachnoid hemorrhage. Throughout follow-up, 18 patients [45%] had favorable outcome [EDSS<5]; while 22 patients had unfavorable outcome with EDSS >/= 5. Serum levels of S100B and NSE were significantly [P[1]<0.05] increased in patients compared to control levels, moreover, mean serum level of S100B was significantly [P[2]<0.05] and of NSE was non-significantly [P[2]>0.05] higher in patients with unfavorable outcome [EDSS >/=] compared to those with favorable outcome [EDSS<5] with a positive significant correlation between serum levels of S100B and NSE, [r = 0.485, p = 0.002]. Moreover, serum levels of both parameters showed a negative significant correlation with the initial GCS while showed a positive significant correlation with EDSS. However, there was a negative correlation between both parameters and the final outcome as favorable or unfavorable; such correlation was significant with S100B and non-significant with NSE. Using Logestic regression analysis serum S100B was the most significant predictor of the final outcome, [beta =-0.371, p = 0.018]. Receiver operator characteristics [ROC] curve analysis for serum levels of both S100B and NSE for prediction of favorable outcome found serum levels of S100B more specific with an area under curve [AUC] =0.379 than serum levels of NSE that found to be more sensitive with an AUC = 0.294. It could be concluded that estimation of serum S100B and NSE immediately after traumatic brain injury could define patients who will develop unfavorable outcome and posttraumatic disability with high sensitivity with NSE and specificity with S100 and must be used for the initial evaluation of TBI irrespective of the extent of severity of inflicted trauma