Object: The role of the neuromarkers S100B protein and neuron-specific enolase (NSE) in minor head injury is well established. Moreover, there are sensitive decision rules available in the literature to identify clinically important brain lesions. However, it is not clear if using the biomarkers has an influence on the predictability of the decision rule. The purpose of this study was to determine if a set of preclinical and clinical parameters combined with 2 neuromarker levels could serve as reliable guidance for accurate diagnosis.
Methods: Prospective evaluation of a cohort of head trauma patients with Glasgow Coma Scale scores of 13-15 was performed at an academic, Level I trauma center. Blood samples and cranial CT studies were obtained for all patients within 3 hours after injury. The hypothesis of the study was whether the combination of an increase of S100B and NSE levels in serum and other defined risk factors are associated with a pathological finding on CT. A forward stepwise logistic regression model was used.
Results: The study included 107 head trauma patients with a mean age of 59 ± 23 years. Twenty-five patients (23.4%) had traumatic lesions on CT. Eight patients underwent craniotomy. The analysis provided a model with good overall accuracy for discriminating cases with clinically important brain injury, including the 6 variables of S100B, NSE, nausea, amnesia, vomiting, and loss of consciousness. The area under the curve (AUC) was 0.88 (0.83-0.93). The receiver operating characteristic curve plots detecting clinically important brain injury for the single variables of S100B and NSE showed an AUC of 0.63 and 0.64, respectively. Conclusions The integration of the neuromarker panel as part of a diagnostic rule including the high-risk factors of nausea, vomiting, amnesia, and loss of consciousness is safe and reliable in determining a diagnosis, pending the availability of more brain-specific neuromarkers. CLINICAL TRIAL REGISTRATION NO.: NCT00622778 (ClinicalTrials.gov).