P01.02.B CLONAL ARCHITECTURE, 3D TUMOR GROWTH, TRANSCRIPTIONAL STATES AND IN-VITRO DRUG RESPONSE IN GLIOBLASTOMA

Abstract BACKGROUND Glioblastoma (GBM) is one of the deadliest types of cancer with median overall survival of little more than a year. It is the prototypic example of a molecularly diverse tumor associated with multiple subclones and high transcriptional plasticity. We here set out to map the clonal architecture and link it with growth patterns, transcriptional states and drug response in-vitro. MATERIAL AND METHODS Seventy-eight multiregional samples from 24 GBM patients (F/M ratio = 0.60, median age= 65.5ys) were profiled using deep whole-exome sequencing with a nominal coverage of >300x. Clonal complexity was assessed by the Euclidean distance based on somatic mutations between regions. Molecular distance was integrated with 3D neuronavigation-based physical distances and patient survival. Fresh tumor slices of 8 patients were cultured in minimal essential media and treated with 4.2Gy plus 200µM temozolomide for 48h. Spatial Transcriptomics was performed on matched tumor slices with /without treatment. RESULTS Integrative characterization revealed two evolutionary trajectories termed expansive and stochastic diversification based on the correlation between physical and molecular distance (RE=0.6, RS=-0.2). High molecular distance was associated with reduced PFS (R=-0.5538, p=0.026). Tumors with expansive evolution demonstrated volume-based growth that directly correlated with increased molecular diversification, while stochastic models showed irregular and discordant patterns. Patients with stochastic evolution were characterized by unfavorable prognosis (p= 0.35) and tumors appeared more frequently in the frontal lobe. In-vitro drug response varied considerably between patients based on cell proliferation (minus 0.5-13.8%), apoptosis (30.1-44.4%), and viability (minus 7.8-34.1%). Under treatment an increase of mesenchymal states with peculiar redistribution of transcriptional patterns and increased clonal complexity was observed. CONCLUSION In sum, we report on an ongoing study in glioblastoma that seeks to link the clonal architecture with tumor growth, transcriptional patterns and in-vitro drug response. Our study will provide new insights into the clinical relevance of the clonal composition of this deadly cancer.