Transcriptomic and Proteomic Responses to Targeted Inhibitors in Glioblastoma: In Pursuit of Combination Therapies

Glioblastoma (GBM) is the most common malignant primary brain tumor in adults and is characterized by a dismal median survival of 12-15 months. It is defined histologically by pseudopalisading necrosis and microvascular proliferation, and its diffusively invasive nature precludes complete resection. Standard of care includes maximal safe surgical resection, radiotherapy, and concomitant and adjuvant temozolomide treatment, and has remained unchanged since the addition of temozolomide in the mid-2000s. Large-scale molecular analyses of GBM have revealed extensive intratumoral and interpatient heterogeneity and havecharacterized the frequency of genetic alterations, but these studies have yet to impact therapy. Epidermal growth factor receptor (EGFR) is the most commonly mutated oncogene in GBM and potentiates downstream activation of the PI3K and MAPK signaling pathways. Despite the attractiveness of this therapeutic target, clinical trials targeting EGFR in GBM have repeatedly failed.The aim of this dissertation was to determine how relevant pathway mutations and acquired EGFR tyrosine kinase inhibitor (TKI) resistance influence kinome dynamics and response to targeted kinase inhibition. First, we used immortalized normal human astrocytes with and without mutant RAS and transduced them with wild-type or mutant PIK3CA to address the roles of specific PIK3CA missense mutations on gliomagenesis. We showed that mutations in the helical and kinase domains of PIK3CA activate PI3K signaling and cooperate with mutant RAS in vivo to potentiate malignancy. Cells were sensitive to PI3K and MEK inhibition regardless of PIK3CA status. Next, we used a panel of mouse astrocytes immortalized via Cdkn2a deletion, with varying combinations of EGFRvIII, a common oncogenic variant, and Pten, a tumor suppressor lost in 40% of GBM. This panel was used to identify differences in transcriptome and kinome expression between genotypes and in response to EGFR inhibition and identify kinase targets for combination inhibition with an EGFR TKI. Finally, we used Cdkn2a-null, EGFRvIII-expressing mouse astrocytes with and without acquired EGFR TKI resistance. We used these cell lines to characterize transcriptomic and kinome changes and identify key kinases upregulated in response to EGFR TKI. Taken together, these results should help guide the future development of rational combination therapy and personalized medicine approaches to GBM.