Phenotypic Therapy and Immune Escape in Cancer

Project 3Targeting clonally expanding drug-resistant tumor cells in recurrent glioblastoma

Based on recently developed methods for predictive modeling of intra-tumor subclone dynamics, we have identified in clinical samples a potentially relapse-driving subpopulation of brain tumor cells in glioblastoma. Through molecular and pharmacological investigation using tailored ex vivo and PDX models of disease as well as paired vital cell samples from naïve and post-therapeutic settings, we could demonstrate that these cells expand clonally under the influence of standard temozolomide (TMZ) chemotherapy. Further in situ validation in paired sample cohorts confirmed that specifically these cells develop resistance to TMZ-exposure via selective co-expression and hyperactivation of the AKT pathway at disease relapse. This is important, because TMZ-based treatment schemes represent the ‘gold’ standard in current glioblastoma care and because every patient on these schedules develops a relapse (even those that initially respond to treatment) usually within 6-9 months of primary therapy, i.e. during TMZ exposure. In the proposed work, we will scrutinize the clinically relevant, yet previously unconsidered subclonal evolution of target cell populations. Underlying candidate mechanisms of mutational and/or phenotypic drug resistance will be thoroughly explored in paired naïve vs. post-therapeutic human samples using targeted genome-, transcriptome- and methylome analysis, and by investigating epi-metabolic circuitries. Identified driver mechanisms, e.g., affecting the AKT-mTOR pathways, deletion of PTEN, or autophagy will be probed in selected gain- or loss-of function assays, while alternative (epigenetic) mechanisms and assays for therapy-induced E-M transitions and epigenetic modulators like the H3K4 demethylase JARID1B/KDM5B will be conducted in appropriate 2D/3D culture models. Lastly, cell state-specific drug regimens identified by our approach will be applied in orthotopic PDX approaches towards straightforward translational rationales for the design of early clinical trials and for the establishment of new biomarkers/companion diagnostics alongside.



Prof. Dr. med. Björn Scheffler
DKFZ Division of Translational Neurooncology at the West German Cancer Center (WTZ)
German Cancer Consortium (DKTK)
University Hospital Essen

Lab homepage