Mapping the oxidative stress metabolome in neurology by gas chromatography-mass spectrometry: a systematic review on signature-driven diagnosis and disease monitoring.

Oxidative stress is a pivotal factor in the pathogenesis of neurological conditions; however, its clinical assessment is constrained by the ephemeral nature of reactive species and the variability of analytical methodologies. Gas chromatography coupled with mass spectrometry (GC-MS) provides high sensitivity, molecular specificity, and well-established libraries, positioning it as a promising translational platform for the mapping of oxidative-stress-related metabolites in neurological disorders. We conducted a PRISMA-guided, PROSPERO-registered systematic review of human studies employing GC-MS to quantify oxidative-stress-linked metabolites in central nervous system disorders (January 2014-January 2025). Two independent reviewers screened records from PubMed and Scopus, extracted study and assay characteristics, and evaluated bias using design-appropriate tools. Twenty-four studies met the inclusion criteria, encompassing neurodegenerative, injury-related, infectious, and psychiatric conditions. Blood was the most frequently utilized matrix (14/24), with neurodegenerative diseases being the most represented (10/24). Across these studies, 70 metabolites were identified as significantly altered compared with controls. Consistent findings were associated with lipid peroxidation (e.g., isoprostanes, neuroprostanes, oxysterols), glutathione cycling and amino acid redox pathways (e.g., cystine, pyroglutamate), energy metabolism (e.g. TCA intermediates, lactate, pyruvate), purine turnover and oxidative DNA damage markers, as well as sugars/polyols implicating the pentose-phosphate and polyol pathways. These results underscore oxidative stress as a convergent mechanism linking neuroinflammation, mitochondrial dysfunction, and membrane damage across central nervous system disorders, and highlight GC-MS-derived metabolite panels as emerging candidates for diagnosis and monitoring. Standardized, multi-matrix protocols, untargeted discovery, targeted validation, and longitudinal cohorts are now required to define robust stress-related metabolomic signatures and advance clinical translation in neurology.