No abstract

Therapeutic options to treat primary glioblastoma (GBM) tumors are scarce. GBM tumors with epidermal growth factor receptor (EGFR) mutations, in particular a constitutively active EGFRvIII mutant, have extremely poor clinical outcomes. GBM tumors with concurrent EGFR amplification and active phosphatase and tensin homolog (PTEN) are sensitive to the tyrosine kinase inhibitor erlotinib, but the effect is not durable. A persistent challenge to improved treatment is the poorly understood role of cellular, metabolic, and biophysical signals from the GBM tumor microenvironment on therapeutic efficacy and acquired resistance. The intractable nature of studying GBM cell in vivo motivates tissue engineering approaches to replicate aspects of the complex GBM tumor microenvironment. Here, we profile the effect of erlotinib on two patient-derived GBM specimens: EGFR+ GBM12 and EGFRvIII GBM6. We use a threedimensional gelatin hydrogel to present brain-mimetic hyaluronic acid (HA) and evaluate the coordinated influence of extracellular matrix signals and EGFR mutation status on GBM cell migration, survival and proliferation, as well as signaling pathway activation in response to cyclic erlotinib exposure. Comparable to results observed in vivo for xenograft tumors, erlotinib exposure is not cytotoxic for GBM6 EGFRvIII specimens. We also identify a role of extracellular HA (via CD44) in altering the effect of erlotinib in GBM EGFR+ cells by modifying STAT3 phosphorylation status. Taken together, we report an in vitro tissue engineered platform to monitor signaling associated with poor response to targeted inhibitors in GBM.

The extracellular matrix (ECM) is increasingly recognized as having a key role in cancer development. We have developed a biomaterial tumor mimic that recapitulates systematically the main characteristics and heterogeneities of glioblastoma ECM. With the use of these 3D in vitro platforms we seek to understand the key features that make this cancer so challenging to treat. We are using this device in three different ways: (1) to further understand how particularities of brain tumor ECM affect cancer cells and assist in the diagnostic of different glioblastoma (GBM) tumor types, (2) provide new models for drug screening to afford more personalized therapies and (3) study the mechanisms of resistance to current therapies in order to identify more effective combinatorial treatments and new therapeutic targets. Hyaluronic acid (HA), is the main component of the brain ECM and GBM is associated with aberrant HA secretion and overexpression of receptors associated with HA, such as CD44 and EGFR. We cultured GBM patient-derived xenograft (PDX) cells within gelatin based hydrogels that contained variable concentrations of HA. Cell proliferation and gene expression analyses demonstrate that biomimetic hydrogels support xenograft culture and cells upregulate matrix remodeling genes and others related to tumor growth in response to matrix biophysical properties. We also used these platforms to evaluate the mechanisms of cell resistance to tyrosine kinase inhibitors. For this purpose, we first focused on the epidermal growth factor receptor (EGFR), which has been identified as a molecular target and associated with worse clinical outcomes. We characterized the 3D in vitro behavior of 3 PDX that represent these EGFR variants: GBM10 (EGFR, wild type), GBM12 (EGFR+) and GBM6 (EGFRvIII). We studied the relationship between the HA contained in the surrounding matrix and response of GBM cells to a tyrosine kinase inhibitor (TKI), erlotinib. Results indicate that while EGFR+ cells are sensitive to TKI in HA hydrogels, HA seems to collaborate with EGFRvIII signaling to stir cell activity. Immunoblots demonstrate that this enhanced cell activity is related to a significant increase in PDGFR concentration. Blockade of the CD44 receptor, in combination with erlotinib treatment, does not affect phosphorylation rates of PI3K or ERK, however, it leads to a significant decrease of STAT3 phosphorylation. This is translated into a lower tumor cell metabolic activity in HA matrices. Finally, we show that, patient-derived xenograft cells resistant to erlotinib in vivo (GBM10), become sensitive when CD44 is blocked in HA-containing matrices, as demonstrated by a decrease of phosphorylated PI3K. In summary, we highlight the importance of extracellular HA in EGFR inhibition efficiency. We demonstrate that this biomaterial tumor model can be used as a valuable tool in the mechanistic studies of tumor development and prediction of tyrosine kinase inhibitors efficacy. Citation Format: Sara Pedron, Gabrielle L. Wolter, Emily Chen, Jann N. Sarkaria, Brendan A. Harley. Glioblastoma tumor model to analyze the mechanisms of resistance to tyrosine kinase inhibitors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 5776. doi:10.1158/1538-7445.AM2017-5776