P06.07.b development of a novel co-culture model of glioblastoma neurospheres with autologous peripheral blood mononuclear cells to investigate the efficacy of oncolytic adenovirus therapy

BACKGROUND The dismal prognosis of glioblastoma (GBM) patients, with a median survival of less than 15 months despite maximal therapy urgently warrants new therapeutic approaches. Cancer immunotherapies such as oncolytic virotherapy have shown encouraging results against GBM, but a large fraction of patients fails to respond to treatment. As the mechanisms that malignant cells employ to avoid elimination by the immune system are diverse, we hypothesized that development of an ex vivo co-culture model that would capture the complex interactions between different immune cell types, would allow us to explore the nature and degree of immune response induced by oncolytic viruses. To this end, we established a co-culture model of primary GBM neurospheres with autologous patient peripheral blood mononuclear cells (PBMCs). MATERIAL AND METHODS Different culture parameters such as co-culture media and cell ratios were explored to select the optimal co-culture setting. Furthermore, we developed a 20-color flow cytometry panel to evaluate the phenotype of the PBMCs; markers such as CD69, HLA-DR, TIGIT on T cells, as well as M1 and M2-related signatures on monocyte-derived macrophages were assessed. Co-culture supernatants were evaluated for the presence of pro-inflammatory cytokines and chemokines. RESULTS We established that viability and effector functions of PBMCs were retained in our co-culture conditions. Co-culture of GBM cells with PBMCs induced a pro-tumoral microenvironment. However, treatment with oncolytic adenovirus led to downregulation of M2-related signatures on monocyte-derived macrophages as well as upregulation of activation markers on CD8+ T cells. We also observed increase in a number of pro-inflammatory cytokines and chemokines such as IFNγ, TNFα, CCL4 as well as granzyme b. The extent of immune activation upon treatment, however, was distinct per patient, indicating that our system may offer a tool to assess patient-specific responses to oncolytic virus treatment. Importantly, we showed that patient exposure to dexamethasone prior to PBMC isolation did not preclude the IFNγ production in our system. CONCLUSION We established an autologous GBM/PBMCs co-culture model which reflects inter-tumoral heterogeneity in response to oncolytic virotherapy. This may allow the establishment of an ex vivo assay to assess and predict individual patient response to oncolytic virotherapy and possibly in the future also to other immunotherapies.