Exploiting Molecular Subtype Cell Plasticity As Novel Strategy for Targeting Glioma Stem Cells Through Alternating Therapy

Abstract

Glioblastoma (GBM) is the most commonly diagnosed glioma and has poor median survival (14 months) despite aggressive trimodal therapy, which includes resection, radiation therapy (RT) and Temozolomide (TMZ). Recent genome-wide expression profiling has separated GBM into several distinct subtypes including proneural (PN) and mesenchymal (MES). PN and MES brain tumor initiating cells (BTICs) have mutually exclusive gene signatures and opposed sensitivities towards specific therapies. Interestingly, radiation triggers a shift from PN toward MES subtype, which is accompanied by increased sensitivity towards TMZ. Similarly, we have reported that exposure of MES BTICs to TMZ results in a loss of MES features and gain in PN characteristics, including upregulation of Nestin and down-regulation of TGM2. In addition, MES BTICs treated with TMZ exhibited greater response to subsequent chemoradiation compared to upfront concurrent RT and TMZ. We hypothesize that therapy-induced switches in BTICs subtypes drives alternating vulnerabilities to components of standard care (TMZ/RT), which therefore should be applied in an alternating mode rather than concomitantly. Furthermore, we propose that such alternating therapeutic regimen should be designed based on tumor subtype, such that PN tumors should receive neoadjuvant RT while MES should be treated with neoadjuvant TMZ to match the expected predominant subtypes of BTICs We utilised an orthotopic mouse model of GBM using a panel of patient-derived BTICs of PN and MES identity and the XRAD-225cx irradiation platform, which allows precise imaging and RT delivery to recapitulate clinically relevant protocols in a pre-clinical setting. Using this model, we have assessed a panel of BTIC xenografts for response to neo-adjuvant TMZ or RT protocols to identify predictors of response. Furthermore, we investigated the mechanism driving improved response from neo-adjuvant TMZ protocols. We demonstrate that PN and MES have preferential sensitivity to RT or TMZ, respectively. Exposure of MES BTICs to TMZ caused loss of MES features (TGM2) and a gain of PN features (Nestin, CD15), which was accompanied by an increase in radiation sensitivity. We demonstrate that neo-adjuvant TMZ is beneficial towards MES xenografts and is driven by an initial MES to PN shift, which improves subsequent response to chemoradiation. We also explore the clinical effectiveness of protocols alternating between therapies and the impact of RT versus TMZ upfront for MES and PN xenografts. Neo-adjuvant TMZ is preferentially efficacious in MES tumors where initial exposure to TMZ drives evolution towards a PN-like status, which has a greater response to subsequent chemoradiation. Development of a personalized protocol based upon initial assessment of molecular subtype is warranted.