Background Glioblastoma (GBM) remains one of the deadliest cancers, with a five-year survival rate of less than 10%. Standard of care treatment has remained unchanged for decades and involves surgical resection, radiotherapy (RT), and the chemotherapeutic, temozolomide. Preclinical testing of immunotherapy in mouse models has generated promising results; however, clinical patients showed limited responses with no survival advantage highlighting the poor translatability of the current mouse models.

Methods We developed a new platform for generating immunocompetent, autochthonous mouse models of GBM. For initial treatment studies, we developed a model with the most common mutation, epidermal growth factor receptor (EGFR), to drive tumor growth, along with loss of function of Pten and Cdkn2a, two of the top mutated tumor suppressor genes in GBM. To examine the effects of RT and anti-PD1 immunotherapy, we gave a single dose of 8-Gy radiation, followed by bi-weekly anti-PD-1 treatment.

Results Similar to patients, anti-PD1 did not increase survival compared to control mice, whereas RT alone or in combination with anti-PD-1 did significantly increase survival compared to both control and anti-PD-1 alone (p < 0.01 and p < 0.001, respectively). Using multiplex flow cytometry, we confirmed immune presence of mostly brain-resident myeloid cells (microglia) in the tumor, with a small population of peripherally derived immune cells, recapitulating what is typically found in patients. Within the nearby skull bone marrow immune niche, a location not easily accessible in patients, we identified striking differences compared to the tumor, including increased Ly6G+ neutrophils and Ly6C-hi monocytes, along with both B and T cells. Although there were no differences in cell numbers between treatment groups at endpoint, expression of PD-1 on Ly6C-low and high monocytes, for instance, increased in aged tumor samples of both RT alone and combined treatment groups. This may indicate important mechanisms of immune suppression in the skull bone marrow niche that would not be evaluable in patients. Additionally, there was a trend (p = 0.056) for increased CD19+ B cells in the skull bone marrow with combined treatment. Interestingly, a pilot study looking at five days post-RT identified a significant increase (p < 0.05) in Ly6G+ cells within the tumor, with a significant decrease in the skull, both compared to control. This observation may identify early immune involvement following treatment.

Conclusions In conclusion, using this unique modeling system, we can better explore the mechanisms of therapeutic resistance and identify more relevant targets to improve outcomes in this challenging disease.

Ethics Approval All procedures were performed at Cedars Sinai Medical Center in accordance with the Cedars Sinai Institutional Animal Care and Use Committee (IACUC), protocol #7333.