BACKGROUND: Reduction of mortality associated with bacterial meningitis and postsurgical cerebral ventriculitis is dependent on early diagnosis and institution of appropriate therapy. Metabonomics rapidly defines metabolic profiles of biological fluids through the use of high-throughput analytical techniques combined with statistical pattern recognition tools. METHODS: Proton nuclear magnetic resonance (1H NMR)-based metabonomics was applied to (1) lumbar cerebrospinal fluid samples collected prospectively from a cohort of patients with bacterial, fungal, or viral meningitis and from control subjects without neurological disease and (2) ventricular cerebrospinal fluid samples from patients with ventriculitis associated with an external ventricular drain and from control subjects. 1H NMR spectra were analyzed by the unsupervised statistical method of principal components analysis. RESULTS: Metabonomic analysis clearly distinguished patients with bacterial or fungal meningitis (11 patients) from patients with viral meningitis (12) and control subjects (27) and clearly distinguished patients with postsurgical ventriculitis (5) from postsurgical control subjects (10). Metabolites of microbial and host origin that were responsible for class separation were determined. Metabonomic data also correlated with the onset and course of infection in a patient with 2 episodes of bacterial ventriculitis and with response to therapy in another patient with cryptococcal meningitis. CONCLUSIONS: Metabonomic analysis is rapid, requires minimal sample processing, and is not targeted to specific microbial pathogens, making the platform potentially suitable for use in the diagnostic laboratory. This pilot study indicates that metabonomic analysis of cerebrospinal fluid is feasible and a potentially more powerful diagnostic tool than conventional rapid laboratory indicators for distinguishing bacterial from viral meningitis and for monitoring therapy. This should have important implications for early management, reduced empirical use of antibiotics, and treatment duration.
BACKGROUND: Reduction of mortality associated with bacterial meningitis and postsurgical cerebral ventriculitis is dependent on early diagnosis and institution of appropriate therapy. Metabonomics rapidly defines metabolic profiles of biological fluids through the use of high-throughput analytical techniques combined with statistical pattern recognition tools. METHODS: Proton nuclear magnetic resonance (1H NMR)-based metabonomics was applied to (1) lumbar cerebrospinal fluid samples collected prospectively from a cohort of patients with bacterial, fungal, or viral meningitis and from control subjects without neurological disease and (2) ventricular cerebrospinal fluid samples from patients with ventriculitis associated with an external ventricular drain and from control subjects. 1H NMR spectra were analyzed by the unsupervised statistical method of principal components analysis. RESULTS: Metabonomic analysis clearly distinguished patients with bacterial or fungal meningitis (11 patients) from patients with viral meningitis (12) and control subjects (27) and clearly distinguished patients with postsurgical ventriculitis (5) from postsurgical control subjects (10). Metabolites of microbial and host origin that were responsible for class separation were determined. Metabonomic data also correlated with the onset and course of infection in a patient with 2 episodes of bacterial ventriculitis and with response to therapy in another patient with cryptococcal meningitis. CONCLUSIONS: Metabonomic analysis is rapid, requires minimal sample processing, and is not targeted to specific microbial pathogens, making the platform potentially suitable for use in the diagnostic laboratory. This pilot study indicates that metabonomic analysis of cerebrospinal fluid is feasible and a potentially more powerful diagnostic tool than conventional rapid laboratory indicators for distinguishing bacterial from viral meningitis and for monitoring therapy. This should have important implications for early management, reduced empirical use of antibiotics, and treatment duration.
Authors: Ville-Petteri Mäkinen; Pasi Soininen; Carol Forsblom; Maija Parkkonen; Petri Ingman; Kimmo Kaski; Per-Henrik Groop; Mika Ala-Korpela Journal: MAGMA Date: 2006-12-15 Impact factor: 2.310
Authors: Aifric O'Sullivan; Rodney E Willoughby; Darya Mishchuk; Brisa Alcarraz; Cesar Cabezas-Sanchez; Rene Edgar Condori; Dan David; Rafael Encarnacion; Naaz Fatteh; Josefina Fernandez; Richard Franka; Sara Hedderwick; Conall McCaughey; Joanne Ondrush; Andres Paez-Martinez; Charles Rupprecht; Andres Velasco-Villa; Carolyn M Slupsky Journal: J Proteome Res Date: 2012-11-30 Impact factor: 4.466
Authors: Uwe Himmelreich; Richard Malik; Till Kühn; Heide-Marie Daniel; Ray L Somorjai; Brion Dolenko; Tania C Sorrell Journal: PLoS One Date: 2009-04-24 Impact factor: 3.240