Acute Traumatic Endotheliopathy in Isolated Severe Brain Injury and Its Impact on Clinical Outcome

摘要

Study design: Prospective observational cohort. Objective: To investigate the difference in plasma levels of syndecan-1 (due to glycocalyx degradation) and soluble thrombomodulin (due to endothelial damage) in isolated severe traumatic brain injury (TBI) patients with/without early coagulopathy. A secondary objective was to compare the effects of the degree of TBI endotheliopathy on hospital mortality among patients with TBI-associated coagulopathy (TBI-AC). Methods: Data was prospectively collected on isolated severe TBI (sTBI) patients with Glasgow Coma Scale (GCS) ≤8 less than 12 h after injury admitted to a level I trauma centre. Isolated sTBI patients with samples withdrawn prior to blood transfusion were stratified by conventional coagulation tests as coagulopathic (prothrombin time (PT) ≥ 16.7 s, international normalized ratio (INR) ≥ 1.27, and activated partial thromboplastin time (aPTT) ≥ 28.8 s) and non-coagulopathic. Twenty healthy controls were also included. Plasma levels of thrombomodulin and syndecan-1 were estimated by ELISA. With receiver operating characteristic curve (ROC) analysis, we defined endotheliopathy as a syndecan-1 cut-off level that maximized the sum of sensitivity and specificity for predicting TBI-AC. Results: Inclusion criteria were met in 120 cases, with subjects aged 35.5 ± 12.6 years (88.3% males). TBI-AC was identified in 50 (41.6%) patients, independent of age, gender, and GCS, but there was an association with acidosis (60%; p = 0.01). Following isolated sTBI, we found insignificant changes in soluble thrombomodulin (sTM) levels between patients with isolated TBI and controls, and sTM levels were lower in coagulopathic compared to non-coagulopathic patients. Elevations in plasma syndecan-1 (ng/mL) levels were seen compared to control (31.1(21.5–30.6) vs. 24.8(18.5–30.6); p = 0.08). Syndecan-1(ng/mL) levels were significantly elevated in coagulopathic compared to non-coagulopathic patients (33.7(21.6–109.5) vs. 29.9(19.239.5); p = 0.03). Using ROC analysis (area under the curve = 0.61; 95% Confidence Interval (CI) 0.50 to 0.72), we established a plasma syndecan-1 level cutoff of ≥30.5 ng/mL (sensitivity % = 55.3, specificity % = 52.3), with a significant association with TBI-associated coagulopathy. Conclusion: Subsequent to brain injury, elevated syndecan-1 shedding and endotheliopathy may be associated with early coagulation abnormalities. A syndecan-1 level ≥30.5 ng/mL identified patients with TBI-AC, and may be of importance in guiding management and clinical decision-making.