Objectives: Biomarkers have been suggested as potential prognostic predictors following a moderate or severe traumatic brain injury but their prognostic accuracy is still uncertain. The objective of this systematic review is to assess the ability of the glial fibrillary acidic protein to predict prognosis in patients with moderate or severe traumatic brain injury.
Data sources: MEDLINE, Embase, CENTRAL, and BIOSIS electronic databases and conference abstracts, bibliographies of selected studies, and narrative reviews were searched.
Study selection: Pairs of reviewers identified eligible studies. Cohort studies including greater than or equal to four patients with moderate or severe traumatic brain injury and reporting glial fibrillary acidic protein levels according to the outcomes of interest, namely Glasgow Outcome Scale or Extended Glasgow Outcome Scale, and mortality, were eligible.
Data extraction: Pairs of reviewers independently extracted data from the selected studies using a standardized case report form. Mean levels were log-transformed, and their differences were pooled with random effect models. Results are presented as geometric mean ratios. Methodologic quality, risk of bias, and applicability concerns of the included studies were assessed.
Data synthesis: Seven-thousand seven-hundred sixty-five citations were retrieved of which 15 studies were included in the systematic review (n = 1,070), and nine were included in the meta-analysis (n = 701). We found significant associations between glial fibrillary acidic protein serum levels and Glasgow Outcome Scale score less than or equal to 3 or Extended Glasgow Outcome Scale score less than or equal to 4 (six studies: geometric mean ratio 4.98 [95% CI, 2.19-11.13]; I = 94%) and between mortality (seven studies: geometric mean ratio 8.13 [95% CI, 3.89-17.00]; I = 99%).
Conclusions: Serum glial fibrillary acidic protein levels were significantly higher in patients with an unfavorable prognosis. Glial fibrillary acidic protein has a potential for clinical bedside use in helping for prognostic assessment. Further research should focus on multimodal approaches including tissue biomarkers for prognostic evaluation in critically ill patients with traumatic brain injury.