A. M. Crawford1, S. Yang1, C. L. Ramirez1, P. Hu1, Y. Li1, H. Li1, T. M. Scalea1, D. M. Stein1 1University Of Maryland,Shock Trauma And Anesthesiology Research (STAR)-Organized Research Center,Baltimore, MD, USA
Introduction: Identification of prognostic adverse outcomes after traumatic brain injury (TBI) could preclude immediate long distance military air evacuation and determine if a patient is “Fit-to-fly” to a neurosurgical-capable facility. In the study, various biomarkers were tested to predict adverse intracranial pressure (ICP) changes in severe traumatic brain injury (TBI) prior to occurrence.
Methods: Adult direct admitted trauma patients with severe TBI were prospectively enrolled. Continuously measured VS and biomarker levels were obtained on admission and every 6 hours for 72 hours. Systemic vital signs (SVS), such as blood pressure and heart rate, and intracerebral monitoring (ICM), such as ICP and cerebral perfusion pressure (CPP), were recorded. Fifteen individual biomarkers including cytokines and its associated principal components (BMPC1-15) [Fig 1] were used in a boosting decision tree model for the prediction of elevation of ICP and hypoperfusion in the following next 6 hours. Area Under the Receiver Operating Characteristic Curve (AUROC) was used to evaluate the outcome prediction. Variable importance was ranked based on contribution to the outcome prediction
Results: 50 patients were enrolled in the study. The mean age was 40±18.9 years and 78.7% were male. Median admission motor Glasgow Coma Score was 3, median Marshall Classification score was 3, and in-hospital mortality rate was 22.9%. A Total of 491 biomarker measurements were available. The models demonstrated a prediction of ICP > 20mmHg for > 30min with an AUROC =0.75 (95% CI: 0.73-0.78) and an AUROC of 0.76 (95% CI: 0.74-0.78) in predicting a CPP < 50mmHg for > 15 min in the next 6 hours. Respectively, BMPC9 contributed by IL8 and S100β ranked as an important variable [Fig.1]. Individual biomarker IL-8 alone had the highest contribution for prediction of ICP elevation.
Conclusion: Cumulative biomarker patterns with elevated ICP can confer a pattern of importance with ICP as the most direct predictor of neurological worsening. Biomarker patterns provide indicative data in TBI and could provide early prediction of intracranial insult, thus expediting care of critically injured and improve patient outcomes.
