Abstract
INTRODUCTION: There is a lack of studies examining the most promising blood biomarkers for traumatic brain injury (TBI) in relation to gross pathology types. RESEARCH QUESTION: To examine whether the admission levels of blood biomarkers can discriminate patients with different combinations of traumatic intracranial findings from patients with negative computed tomography (CT) scans. MATERIAL AND METHODS: One hundred thirty patients with all severities of TBI were studied. Seventy-five had CT-positive and 55 CT-negative findings. CT-positive patients were divided into three clusters (CL) using the Helsinki CT score: focal lesions (CL1), mixed lesions (CL2) and mixed lesions + intraventricular haemorrhage (CL3). CT scans were obtained upon admission and blood samples taken within 24 h from admission. S100 calcium-binding protein B (S100B), glial fibrillary acidic protein (GFAP), heart fatty-acid binding protein (H-FABP), neurofilament light (NF-L), interleukin-10 (IL-10), total-tau (t-tau), and β-amyloids 1-40 (Aβ40) and 1-42 (Aβ42) were analysed from plasma samples. CT-negative cluster was used as control. RESULTS: GFAP, Aβ40 and Aβ42 levels differed between the clusters, but not significantly. NF-L and t-tau discriminated CL1 from CT-negative cluster with AUCs of 0.737 and 0.771, respectively. NF-L, t-tau and GFAP discriminated CL2 from CT-negative cluster with AUCs of 0.839, 0.781 and 0.840, respectively. All biomarkers analysed were able to discriminate CL3 and CT-negative cluster. DISCUSSION AND CONCLUSION: All studied biomarkers distinguished the most severely injured cluster, CL3, from CT-negative cluster. The results may reflect the severity of TBI but also show that biomarkers have a variable ability to identify patients with combinations of intracranial traumatic lesions in the examined time window.