Attenuation of Immune Suppression to Complement Glioma Immunotherapy

Glioblastoma Multiforme (GBM) is a malignant primary brain tumour with an extremely poor prognosis. Following surgical resection, radiotherapy, and concomitant and adjuvant chemotherapy, median survival is only 12-15 months. New therapeutic approaches are therefore desperately needed. Accumulating evidence suggests that activated T cells are capable of selectively targeting and eliminating tumour cells, even in the brain, making vaccine-mediated immunotherapy a promising candidate for the treatment of brain cancers. However, cancer vaccination has generally been disappointing in the clinic, and is unlikely to bestow long-term survival unless suppressive mechanisms are overcome. Checkpoint blockade is a recent treatment modality that enhances naturally occurring T cell responses to cancer by relieving suppression mediated by immune checkpoints – molecular signals that prevent T cell function. While significant clinical responses are often seen, it is clear that most patients fail to respond to checkpoint blockade alone. Therefore, there is considerable interest in combining the different immunotherapeutic strategies, with vaccines providing an immunogenic stimulus to induce anti-tumour T cells, and checkpoint blockade to ensure T cell function is retained. An orthotopic murine model of glioma was utilised to examine this form of combined treatment. Immune responses induced with a unique whole-cell vaccine that utilises the adjuvant properties of invariant natural killer T cells (iNKT cells) were able to resist tumour challenge, but failed to eradicate established tumours. When the vaccine was combined with blocking antibodies to the immune checkpoint molecule cytotoxic T lymphocyte antigen-4 (α-CTLA-4) regression of established intracranial tumours was observed, whereas α-CTLA-4 was ineffective as a monotherapy. In contrast, combining the vaccine with antibodies to programmed death-1 (α-PD-1) or lymphocyte activation gene-3 (LAG-3) failed to provide any survival advantage. This was despite α-PD-1 being effective against the same tumour implanted subcutaneously, suggesting efficacy in the orthotopic setting was limited by poor access of α-PD-1 to effector T cells within the brain. The effective combination of vaccine and α-CTLA-4 was associated with enhanced proliferation and accumulation of T cells in the lymphoid tissues without any obvious changes in the adjuvant function of iNKT cells or altered numbers of regulatory T cells, suggesting recently primed T cells were the targets of checkpoint inhibition. While tumours regressing under this combined treatment were highly infiltrated with a variety of leukocytes, tumour eradication was strictly dependent on CD4⁺ T cells. Further interrogation of the cell-types responsible for anti-tumour activity revealed that CD11b⁺ cells were required for therapy, although it remains to be established whether these cells were involved in T cell priming or served as anti-tumour effectors in their own right, possibly under the influence of activated CD4⁺ T cells. In addition, therapy was hampered, although not entirely eliminated, in hosts deficient in interferon-γ. Therapy was also reduced significantly, but not entirely, in hosts deficient in perforin. In vitro studies showed that restimulated splenocytes from animals that had received the combined therapy were able to kill glioma cells in a perforin and MHC-II dependent manner, suggesting that cytotoxic CD4⁺ T cells were important effector cells. Overall, these results demonstrate that immunotherapeutic vaccination can be combined effectively with checkpoint blockade to induce effective immune responses against glioma. The immune response induced in combination with CTLA-4 blockade differs from many other cancer models, with a strict dependence on CD4⁺ T cells that can serve either as cytotoxic effector cells, or potentially as modulators of other accessory cells. Furthermore, the tumour location presents new challenges, with access of inhibitors to the brain, particularly important if immune checkpoints on intratumoural effector cells are to be targeted. In this context, strategies to improve access of checkpoint inhibitors like α-PD-1 and α-LAG3 to the brain warrant further investigation.