gms | German Medical Science

Objective: Glioblastoma (GBM) is one of the most challenging tumor entities to overcome. Recent immunotherapeutic approaches showed promising effects in vitro and in preclinical animal models. However, tumor heterogeneity, antigen escape mechanisms and complex interactions with the tumor microenvironment (TME) hinder their clinical breakthrough. Robust models for ex vivo trials of new therapeutic agents reflecting those interactions are desperately needed. Here, we present a modified method to establish patient derived organoids and investigate whether organoids reflect the surface texture of their parental tumors.

Methods: Tissue of patients, who underwent surgery at our institution, was obtained, cleared from debris and divided into six different portions under the microscope. From each of the six subsections acute tissue was embedded in paraffin (t1). The remaining tissue was carefully minced to 500 µm sized pieces and cultured on an orbital shaker for 2-4 weeks until round organoids formed (t2). In order to freeze the organoids, they were minced to 100 µm sized pieces. After thawing, organoids were cultured for 4-8 weeks (t3). Immunohistological analysis of GFAP, Ki67 and three surface antigens (GD2, Podoplanine, ROR2) was performed at t1, t2 and t3 in order to investigate organoid stability and the ability of reflecting interpatient heterogeneity.

Results: Intraoperative material from nine different patients (GBM n=6, oligodendroglioma n=2, astrocytoma, IDH mutant WHO grade 3 n=1) was obtained. Within 2-4 weeks of culture, stable, round organoids formed independently of the WHO grading (t2), which were still viable after a freeze-thaw cycle (t3). The organoids regained their prior size 4-8 weeks after thawing (t3) and showed no significant difference in expression of GFAP, Ki67 and two surface antigens (GD2, Podoplanine) compared to the acute tissue (t1) or the first culture period (t2) (Friedman´s test: GFAP: p=0.662, Ki67: p=0.08, GD2: p=0.196, Podoplanine: p=0.104). In contrast, ROR2 expression was significantly altered for t1-t3 (p=0.002) but the effect size r was low (r=0.27).

Conclusion: The results suggest that organoids are a very promising ex vivo model for drug development. Patient derived glioma organoids can be established, frozen and re-cultured without losing their ability to reflect the surface texture of the parental tumor. Therefore, organoids might represent a promising ex vivo model for individual immunotherapeutic drug testing.