P12.03.B TRACKING GLIOMA PROGRESSION: BLOOD-BASED METABOLIC MARKERS FROM PRE-DIAGNOSIS TO SURGERY

Abstract BACKGROUND Gliomas are highly invasive brain tumors with poor prognosis, particularly glioblastoma, the most aggressive subtype. Late detection significantly worsens prognosis, underscoring the need for early diagnostic markers. MATERIAL AND METHODS This study investigates metabolic alterations in blood samples collected at surgery and up to 20 years before diagnosis to identify potential markers for early glioma detection. We analyzed blood samples from 52 glioma patients using untargeted UHPLC-MS/MS. Metabolite profiling targeted 773 annotated metabolites and 164 unknown compounds, capturing both established and novel metabolic alterations. Multivariate data analysis, including OPLS Effect Projection and Discriminant Analysis, was applied to compare pre-diagnostic and surgery samples. Subtype-specific (glioblastoma vs. lower grade glioma) and sex-specific analyses were conducted. RESULTS Our study reveals significant metabolic shifts over time, with 518 detected metabolites showing substantial changes. Amino acid metabolism, particularly N-lactoyl amino acids, was notably elevated at surgery. Glioblastoma patients exhibited distinct metabolic profiles, including reduced long-chain acylcarnitine levels, compared to non-glioblastoma cases. While metabolic changes were largely consistent between sexes, some sex-specific differences emerged. Notably, a panel of eight metabolites demonstrated strong potential for early glioma detection, with levels rising well before diagnosis. CONCLUSION These preliminary findings highlight the promise of blood-based metabolic markers for earlier glioma detection. Further validation in larger cohorts could pave the way for a non-invasive tool, improving early diagnosis and patient survival.