• Thomas C Glenn, Daniel Hirt, Gustavo Mendez, David L McArthur, Rachael Sturtevant, Stephanie Wolahan, Farbod Fazlollahi, Matthew Ordon, Arzu Bilgin-Freiert, Ben Ellingson, Paul Vespa, David A Hovda, and Neil A Martin.
• Department of Neurosurgery, David Geffen School of Medicine, UCLA, Los Angeles, CA 90095-7039, USA. tglenn@mednet.ucla.edu
• Acta Neurochir. Suppl. 2013 Jan 1; 118: 115-9.

AbstractProton nuclear magnetic resonance (H-NMR) spectroscopic analysis of cerebral spinal fluid provides a quick, non-invasive modality for evaluating the metabolic activity of brain-injured patients. In a prospective study, we compared the CSF of 44 TBI patients and 13 non-injured control subjects. CSF was screened for ten parameters: β-glucose (Glu), lactate (Lac), propylene glycol (PG), glutamine (Gln), alanine (Ala), α-glucose (A-Glu), pyruvate (PYR), creatine (Cr), creatinine (Crt), and acetate (Ace). Using mixed effects measures, we discovered statistically significant differences between control and trauma concentrations (mM). TBI patients had significantly higher concentrations of PG, while statistical trends existed for lactate, glutamine, and creatine. TBI patients had a significantly decreased concentration of total creatinine. There were no significant differences between TBI patients and non-injured controls regarding β- or α-glucose, alanine, pyruvate or acetate. Correlational analysis between metabolites revealed that the strongest significant correlations in non-injured subjects were between β- and α-glucose (r = 0.74), creatinine and pyruvate (r = 0.74), alanine and creatine (r = 0.62), and glutamine and α-glucose (r = 0.60). For TBI patients, the strongest significant correlations were between lactate and α-glucose (r = 0.54), lactate and alanine (r = 0.53), and α-glucose and alanine (r = 0.48). The GLM and multimodel inference indicated that the combined metabolites of PG, glutamine, α-glucose, and creatinine were the strongest predictors for CMRO2, ICP, and GOSe. By analyzing the CSF of patients with TBI, our goal was to create a metabolomic fingerprint for brain injury.

21 keywords...

hide…