2059 Background: BPM 31510-IV is a drug-lipid conjugate nanodispersion containing oxidized Coenzyme Q10 (CoQ10) in clinical development for glioblastoma multiforme (GBM). In a recently concluded Phase 1 study of BPM 31510-IV (NCT03020602), in addition to safety and tolerability, longitudinal pharmacodynamic samples (20 samples/cycle of 28 days) were collected at various times in patient’s refractory to radiation, temozolomide, and bevacizumab. Methods: Comprehensive multi-omic (proteomic, lipidomic, metabolomic) profiles were generated from buffy coat (proteomics only), plasma, and urine matrices. These data were further analyzed using bAIcis, a Bayesian statistics based artificial intelligence (AI) software, creating causal networks linking clinical information and endpoints to molecular composition of diverse biomatrices of patients prior to, as well as during, treatment with BPM 31510-IV. Twelve subjects comprised the intent to treat population (ITT) which were stratified across days of treatment (DR1; ≤28 days; DLT period; n=6) and (DR2, OS; >28 days; n=6). Bayesian networks and regression analysis were performed on the outputs of the analysis. Molecular analyte panels (combination of proteins, lipids, and metabolites) descriptive of progression free survival (PFS), adverse events (possibly/probably related to BPM 31510-IV), and of overall survival (OS) were generated. Results: Significant alteration (p<0.05) of metabolically associated protein and critical metabolite drivers of intermediary metabolism were identified as causally related to PFS. Significant quantitative changes in levels of several proteins (buffy coat) and metabolites (urine) were identified with probable or possible associations to adverse events in BPM 31510-IV treated subjects. Conclusions: Overall, alterations in proteins and metabolites influencing mitochondrial function and intermediary metabolism that differentiated responders versus non-responders and identified potential markers of adverse events associated with BPM 31510-IV exposure were identified and will be further explored for complementary diagnostic utility.