P05.13.A DECIPHERING GLIOBLASTOMA METABOLIC SIGNATURE: MOLECULAR PROFILING AND NADH-FLIM IMAGING

Abstract INTRODUCTION Recent breakthroughs in single-cell analysis reveal distinct metabolic cellular states of glioblastoma (GBM) and allow the detection of GBMs which are dependent on oxidative phosphorylation (OXPHOS). The aims of the study was to explore the prevalence of molecular/metabolic subtypes of GBM and to evaluate the efficacy of Fluorescence Lifetime Imaging (FLIM) microscopy of NADH in discerning the metabolic subtype of FFPE GBM tumor tissue. METHODS We analyzed 3’ mRNA NGS FFPE tumor tissue of IDHwt glioblastomas. We selected OXPHOS and glycolytic/plurimetabolic (GPM) cases for the NADH-FLIM microscopy study on FFPE slides. Specific pharmacological vulnerability were investigated in vitro in patients derived GBM-Explants using NADH-FLIM to mesure metabolic shift induced by tumor sensitivity to drugs (Morelli et al. 2022; PMID: 36132155). RESULTS RNA Sequencing of 95 newly diagnosed IDHwt GBM patients found 37 OXPHOS (39%), 16 GPM (15%), 4 NEU (4%), 27 not classifiable NC (28%) and 12 PPR cases (14%). Sixteen out of the 37 OXPHOS patients exhibited top-scoring mitochondrial activity and were defined as OXPHOS_high. Longitudinal analysis of 11 pairs of GBM untreated and recurrent tumors found that 3 OXPHOS_high patients maintained the OXPHOS status at recurrence. We assessed FGFR3-TACC3 (F3T3) status in 62 cases and we found that all 4 FGFR3-TACC3+ overexpressed OXPHOS genes. Four OXPHOS patients and 3 GPM cases were selected for the NADH-FLIM study: NADH bound/free curves significantly differed between metabolic subtypes [p<0.0001]. PCA analysis of NADH FLIM features showed different clustering according to the OXPHOS and GPM subtype. In vitro pharmacological assay showed that prazosin (an FDA-approved drug which inhibits glioblastoma growth through a PKCδ-dependent mechanism) treatment induced specifically a higher percentage of NADH-bound increase in PPR+ GBM explants compared to PPR-(OXPHOS+) explants. Of note PPR GBM’s proliferation is based on DNA-PK activity. Conversely, metformin (anti-mitochondrial agent), either alone and in combination with TMZ, induced specifically an increase in the percentage of NADH-bound primarily in OXPHOS+ GBM explants, with a less pronounced effect in PPR+ GBM explants. CONCLUSION F3T3 fusions are tightly associated with the OXPHOS signature and the OXPHOS status can be maintained at recurrence. OXPHOS and GPM glioblastoma clusters significantly differed in NADH bound/free distribution curves. Preliminary in vitro analysis suggest specifical pharmacological vulnerabilities. Supplementary cases are currently under study.