A new approach to the quantitative analysis of aromatic metabolites in cerebrospinal fluid samples of neurosurgical patients based on microextraction by packed sorbent coupled with derivatization and GC–MS was developed. Analytical characteristics such as recoveries (40–90%), limit of detection (0.1–0.3 μm) and limit of quantitation (0.4–0.7 μm) values, accuracy (<±20%), precision (<20%) and linear correlations (R2 ≥ 0.99) over a 0.4–10 μm range of concentrations demonstrated that microextraction by packed sorbent provides results for the quantitative analysis of target compounds comparable with those for liquid–liquid extraction. Similar results were achieved using 40 μl of sample for microextraction by packed sorbent instead of 200 μl for liquid–liquid extraction. Benzoic, 3‐phenylpropionic, 3‐phenyllactic, 4‐hydroxybenzoic, 2‐(4‐hydroxyphenyl)acetic, homovanillic and 3‐(4‐hydroxyphenyl)lactic acids were found in cerebrospinal fluid samples (n = 138) of neurosurgical patients in lower concentrations than in serum samples (n = 110) of critically ill patients. Analysis of the cerebrospinal fluid and serum samples taken at the same time from neurosurgical patients (n = 5) revealed similar results for patients without infection and multidirectional results for patients with central nervous system infection. Our preliminary results demonstrate the necessity of further evaluating the aromatic compound profile in cerebrospinal fluid for its subsequent verification for potential diagnostic markers.
Indole-containing acids—tryptophan metabolites—found in serum and cerebrospinal fluid (CSF) samples of patients with diseases of the central nervous system (CNS) were determined with the use of microextraction by packed sorbent (MEPS) followed by silylation and gas chromatography–mass spectrometry (GC–MS) analysis. MEPS with the following silylation led to the reproducible formation of derivatives with an unsubstituted hydrogen ion in the indole ring, the chromatographic peaks of which are symmetric and can be used for GC–MS analysis without additional derivatization. The recoveries of analytes at the limit of quantitation (LOQ) levels were 40–80% for pooled CSF and 40–60% for serum. The limit of detection (LOD) and LOQ values were 0.2–0.4 and 0.4–0.5 µM, respectively, for both CSF and serum. The precision (the reproducibility, RSD) value of less than 20% and the accuracy (the relative error, RE) value of less than ±20% at the LOQ concentrations meet the Food and Drug Administration (FDA) recommendations. Linear correlations for all analytes were determined over a potentially clinically significant range of concentrations (0.4–10 µM for serum, R2 ≥ 0.9942, and 0.4–7 µM for CSF, R2 ≥ 0.9949). Moreover, MEPS significantly reduced the matrix effect of serum compared to liquid–liquid extraction (LLE), which was revealed in the example of reducing the amount of cholesterol and its relative compounds.