ID #941 Tumoroids as models of pediatric high-grade glioma: comparative sequencing, in vitro drug testing, and co-culture modeling

Abstract Background Pediatric high-grade gliomas (PHGGs) are aggressive, incurable brain tumors, with 40% occurring in the cerebral hemispheres. Radiation-induced gliomas (RIGs) are a subtype of PHGG occurring as a late complication in 1-4% of children who undergo cranial radiation for a prior cancer. We and others have previously developed patient-derived cell culture and xenograft models of PHGG and RIG, though more personalized models are needed to address individual treatment response. Here, we show feasibility of creating tumoroid models from patient biopsies and perform a comparative analysis vs. matched tumor tissue and a 2D cell line. We also describe co-culture allograft models in which we engraft PHGG cells into forebrain organoids. Methods Disaggregated cells from a RIG biopsy at Children’s Hospital Colorado were used to establish 3D tumoroids and a 2D cell line. We performed single-cell RNA-seq (scRNA-Seq) on the biopsy tissue and resulting tumoroids. Drug testing was performed in the 2D cell line and tumoroids. To show feasibility of engrafting PHGG cells into organoids, we co-cultured iPSC-derived forebrain organoids with a PHGG GFP-tagged cell line, HSJD-GBM01, tracked them visually over time with IncuCyte, and used histology and immunofluorescence to analyze infiltration. Results scRNA-seq revealed the patient-derived tumoroids recapitulated cellular heterogeneity of the biopsy sample. Drug testing with the proteasome inhibitor ixazomib left a resistant population in the tumoroids not observed in the cell line, suggesting the tumoroids more realistically model clinical treatment outcomes. Co-culture experiments showed PHGG cells congregate around forebrain organoids over time. Subsequent tumoroid fixation and immunofluorescent staining revealed evidence of PHGG infiltration. Conclusions The biopsy-derived PHGG/RIG tumoroid model accurately represents the original tumor’s cellular makeup and is a clinically relevant model for testing treatments. Our co-culture model is a first step in developing a realistic allograft host that can pave the way for personalized treatment testing for PHGG and RIG patients.