Metabolomics of cerebrospinal fluid reveals prognostic biomarkers in pediatric status epilepticus

Abstract

Aims: Status epilepticus (SE) is the most common neurological emergency in pediatric patients. This study aimed to screen for prognostic biomarkers of SE in the cerebrospinal fluid (CSF) using metabolomics. Methods: Ultra-performance liquid chromatography quadrupole time-of-flight tandem mass spectrometry (UPLC-QTOF-MS) was conducted to identify prognostic biomarkers in CSF metabolomics by comparing the poor outcome group (N = 13) with the good outcome group (N = 15) of children with SE. Differentially expressed metabolites were identified using Mann–Whitney U test corrected by Benjamini-Hochberg and partial least squares discriminant analysis (PLS-DA). Results: The PLS-DA model identified and validated significant metabolic differences between the poor and good outcome groups of children with SE (PLS-DA with R2Y = 0.992 and Q2 = 0.798). A total of 49 prognosis-related metabolites were identified. Of these metabolites, 20 including glutamyl-glutamine, 3-iodothyronamine, and L-fucose had an area under the curve (AUC) ≥ 80% in prognostic prediction of SE. The logistic regression model combining glutamyl-glutamine and 3-iodothyronamine produced an AUC value of 0.976, with a sensitivity of 0.863 and specificity of 0.956. Pathway analysis revealed that dysregulation of the citrate cycle (TCA) and arginine biosynthesis may contribute to poor SE prognosis. Conclusions: This study highlighted the prognosis-related metabolomic disturbances in the CSF of children with SE and identified potential prognostic biomarkers. A prognostic prediction model combining glutamyl-glutamine and 3-iodothyronamine with high predictive value was established.