RESUMO
BACKGROUND: Mesencephalic astrocyte-derived neurotrophic factor (MANF) is released under endoplasmic reticulum stress, thereby exerting neuroprotective effects. We determined whether serum MANF may be a prognostic biomarker of human severe traumatic brain injury (sTBI). METHODS: Serum MANF concentrations of 137 sTBI patients and 137 controls were quantified in this prospective cohort study. Patients with extended Glasgow outcome scale (GOSE) scores of 1-4 at post-traumatic 6 months were considered to have poor prognosis. Relationships between serum MANF concentrations and severity plus prognosis were investigated using multivariate analyses. Area under receiver operating characteristic curve (AUC) was calculated for reflecting prognostic efficiency. RESULTS: As compared to controls, there was a significant increase of serum MANF concentrations after sTBI (median, 18.5 ng/ml versus 3.0 ng/ml; P < 0.001), which was independently correlated with Glasgow coma scale (GCS) scores [ß, -3.000; 95% confidence interval (CI), -4.525--1.476; VIF, 2.216; P = 0.001], Rotterdam computed tomography (CT) scores (ß, 4.020; 95% CI, 1.446-6.593; VIF, 2.234; P = 0.002) and GOSE scores (ß, -0.056; 95% CI, -0.089--0.023; VIF, 1.743; P = 0.011). Serum MANF concentrations substantially distinguished risk of poor prognosis with AUC of 0.795 (95% CI, 0.718-0.859) and its concentrations > 23.9 ng/ml was predictive of poor prognosis with 67.7% sensitivity and 81.9% specificity. Serum MANF concentrations combined with GCS scores and Rotterdam CT scores displayed markedly higher prognostic predictive ability than each of them (all P < 0.05). Using restricted cubic spline, there was a linear correlation between serum MANF concentrations and poor prognosis (P = 0.256). Serum MANF concentrations > 23.9 ng/ml was independently associated with poor prognosis (odds ratio, 2.911; 95% CI, 1.057-8.020; P = 0.039). A nomogram was built, where serum MANF concentrations > 23.9 ng/ml, GCS scores and Rotterdam CT scores were integrated. Hosmer and Lemeshow test, calibration curve and decision curve analysis demonstrated such a prediction model was comparatively stable and was of relatively high clinical benefit. CONCLUSIONS: Substantially increased serum MANF concentrations after sTBI are highly correlated with traumatic severity and are independently predictive of long-term poor prognosis, suggesting that serum MANF may represent a useful prognostic biochemical marker of human sTBI.