Abstract Introduction: Drug testing of Glioblastoma multiforme (GBM) has revealed that many promising preclinical therapies fail when translated to the clinic. Serious limitations in preclinical models (often monolayer, serum-treated immortalized cell models) likely contribute to this translational failure. Although patient-derived xenografts (PDX), in which patient tumors are serially passaged in immunocompromised mice, are potentially superior model systems of human disease, high throughput drug studies in GBM PDX tumors are not feasible due to time, cost and throughput. To overcome these challenges, we have developed a GBM PDX MicroTumor system in which GBM PDX cells are grown in a novel 3D extracellular matrix material (HuBiogel™, Vivo Biosciences Inc., Birmingham, AL). We hypothesized that this 3D MicroTumor system would better emulate in vivo PDX tumor growth while providing a high throughput assay system for drug combination studies that can be performed in a more rapid timeframe. Methods: Six GBM PDX tumor “xenolines” were selected from the UAB Brain Tumor Animal Model Core to represent the 4 known molecular subtypes of GBM: Classical (JX10, X1016, X1046); Proneural (XD456); Neural (JX10) and Mesenchymal (JX22P). Tumor cells were prepared as single cell suspensions from subcutaneously passaged tumors and embedded in HuBiogel™ beads to form MicroTumors using Neuro-basal media devoid of serum. Single and combinatorial drug cytotoxicity studies were performed using MTT mitochondrial viability assays and Calcein AM imaging. Small molecule inhibitors selected for drug testing were: 1) selumetinib-targeting MEK1/2; 2) crizotinib-targeting c-MET and ALK; 3) cediranib-targeting VEGFR, FLT-1, FLT-4, c-KIT, and PDGFR; and 4) WP1066-targeting JAK2/STAT3. Combinatorial interactions were determined by Chou-Talalay combination index (CI) calculation using Calcusyn software. Global kinase signaling (kinomic profiling) was assessed using paired total cell lysates of GBM-PDX from orthotopic tumors grown in athymic nu/nu mice and as 3D MicroTumors. Testing was performed on the PamStation12 high content peptide microarray platform within the UAB Kinome Core. Results: Following optimization of model system conditions, 4 drugs were tested at multiple concentrations to identify effective dose ranges. While crizotinib, cediranib and WP1066 demonstrated significant cytotoxicity in at least 1 of the GBM-PDX MicroTumors tested, selumetinib was ineffective in producing cytotoxicity as a monotherapy. However, in combinatorial drug testing, selumetinib was effective in promoting synergistic activity (CI<1.0) when combined with each of the other drugs in several of the GBM-PDX tested. Conversely, WP1066 demonstrated some antagonistic interactions with the other drugs in certain tumors. Overall, cediranib and crizotinib demonstrated the most consistent level of efficacy both as monotherapies and combinatorial therapies across the GBM-PDX MicroTumors. Kinomic profiles were compared between GBM-PDX MicroTumors and matched GBM-PDX tumor xenolines grown as orthotopic intracranial tumors to confirm kinase signaling fidelity between the model systems. Paired t-test of kinomic phosphopeptide probes revealed only 20 of 246 peptides significantly differed (FDR<0.05) between the model systems suggesting a high level of signaling preservation within the MicroTumors. Conclusions: The GBM-PDX 3D MicroTumor provides a relatively high throughput drug screening model system that displays similar kinase signaling pathway activation as compared to their in vivo intracranial counterpart. Specific synergistic drug combinations were identified through MicroTumor testing. Ongoing drug combination testing of intracranially implanted GBM-PDX will determine how robustly the MicroTumor system predicts in vivo efficacy. Citation Format: Christopher D. Willey, Ashley N. Gilbert, Rachael Shevin, Catherine P. Langford, Christine W. Duarte, Raj Singh, Joshua C. Anderson, G. Yancey Gillespie. Combinatorial drug testing using 3D microtumors derived from GBM PDX reveals cytotoxic synergisms in pharmacokinomics-informed pathway interactions. [abstract]. In: Proceedings of the AACR Special Conference: Advances in Brain Cancer Research; May 27-30, 2015; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2015;75(23 Suppl):Abstract nr A16.