Predictive Value of Nervous Cell Injury Biomarkers in Moderate-to-Severe Traumatic Brain Injury: A Network...

Area of Science:

• Neuroscience and Neurology
• Biomarker Discovery and Validation
• Clinical Outcome Prediction

Background:

• Traumatic brain injury (TBI) outcome prediction is crucial for guiding clinical management.
• Several biomarkers, including S100B, GFAP, NSE, and UCH-L1, show potential for prognostication.
• The relative effectiveness of these biomarkers for predicting mortality and functional outcomes remains unclear.

Purpose of the Study:

• To conduct a network meta-analysis (NMA) comparing the predictive value of S100B, GFAP, NSE, and UCH-L1.
• To assess the relative effectiveness of these biomarkers for predicting mortality and functional outcomes (GOS/GOSE) in moderate/severe TBI.
• To identify the most effective biomarker for mortality and functional outcome prediction.

Main Methods:

• Systematic literature search of PubMed, EMBASE, and Cochrane Library until June 2024.
• Inclusion of 32 studies (n=2,401) reporting biomarker levels and outcome measures in adult moderate/severe TBI patients.
• Bayesian network meta-analysis to compare indirect effects, assessing sensitivity, specificity, and pooled effect sizes (SMD, MD).

Main Results:

• All four biomarkers (S100B, GFAP, NSE, UCH-L1) demonstrated statistically significant predictive value for mortality and functional outcomes.
• Neuron-specific enolase (NSE) showed the highest sensitivity for mortality prediction (88%), while UCH-L1 had the highest specificity (89%).
• S100B exhibited the highest sensitivity for unfavorable functional outcomes (74%), and GFAP had the highest specificity (84%). NSE ranked highest for mortality prediction, and UCH-L1 for functional outcomes.

Conclusions:

• S100B, GFAP, NSE, and UCH-L1 are valuable biomarkers for predicting mortality and functional outcomes in moderate/severe TBI.
• NSE is the most effective biomarker for predicting mortality, whereas UCH-L1 is superior for predicting unfavorable functional outcomes.
• Standardization of biomarker measurement protocols and integration into clinical predictive models are necessary for wider application.