Targeting the glioma associated microenvironment in experimental glioma

Abstract

One of the most aggressive primary tumours of the central nervous system are glioblastomas. Despite of multimodal therapeutic approaches patient’s prognosis of highly selected clinical trial populations is still in the range of 1.5 years. [10-12] Therefore, novel therapeutic strategies are urgently needed. Therapeutic targeting of the glioblastoma-associated microenvironment is a promising approach in this regard.[49, 61] Particularly macrophages represent a highly abundant population of tumour-infiltrating host cells and are key players of the immune suppressive milieu inside human gliomas.[63, 71] In this regard the colony stimulating factor-1/colony stimulating factor-1 receptor (CSF1/CSF1R) axis plays an important role for macrophage differentiation and survival. [67, 88] To assess presence of CSF1R in human glioblastoma, CSF1R staining in human tissue samples from primary and recurrent glioblastoma was performed inside the described study and discussed in the context of TME specific histologic markers. For targeting the TME an anti-CSF1R approach using the antibody clone 2G2 in combination with immune checkpoint inhibition by an anti-PD1 antibody was performed in two preclinical glioma mouse models.[88] Composition of the TME after executed treatments was evaluated by immunohistochemistry. Additionally the study includes characterization of a frequent used PDGFB-driven glioma model using the somatic gen transfer system RCAS/TVA by the means of MRI, FET-PET imaging and IHC. [102] The evaluated target, CSF1R is present in human samples of primary and recurrent glioblastoma. (Table 9)[108] Moreover, analysis show no statistical difference of CSF1R expression linked to received treatment or after first relapse of disease. Use of an established Immunoreactive Score (IRS) confirmed results and pronounced high staining intensity of CSF1R inside evaluated tissue samples.[108] Single case analysis showed that over 50% of cases have stable expression after first relapse with tendency of showing higher CSF1R expression after first diagnosis than after first relapse of disease.(Figure 11) Monotherapy with anti-CSF1R antibody as well as in combination with anti-PD1antibody in the syngeneic, immunocompetent VM/Dk mouse model revealed reduction of CD204 and CD11b positive cells in IHCs, which most likely indicate reduction of TAMs inside the TME.[108]( Figure 17-18) Simultaneously increased numbers of infiltrated CD8+ cells, higher CD8+/CD4+ ratios and increased Caspase 3 cleaved signals were observed in vivo, suggestive for improved immunogenicity of the treated experimental gliomas and recomposition of the TME depending on CSF1R inhibition. (Figure 15-16+ 20-21) Characterization of the pro-neural glioma model using the RCAS/TVA delivery system could reveal stable tumour formation in vivo, showing high reproductivity in regard of onset of neurological symptoms as defined in Table 5. Monitoring tumour growth by MRI present first contrast agent enhancement starting at day 25 to day 30 after surgery, followed by exponential growth dynamics going hand in hand with FET accumulation.(Figure 25) TME consisting of TAMs, particularly CD163+ and CD204+ cells in combination with basal infiltration of T cells was observed in untreated ex vivo experimental glioma samples. (Figure 27-28) However, anti-CSF1R approach could not repeat observed later onset of neurological symptoms (as defined in Table 5) in combination treatment regimens as in the VM/DK model has shown before.(Figure 30) In contrast several animals treated exclusively with anti CSF1R or anti-PD1 showed prolonged time till reaching defined endpoints, building a long tail inside the Kaplan-Meyer curve. Taking all results together and considering the other published results by Przystal and Becker et al., CSF1R is identified as a promising therapeutic target for glioblastoma, probably in combination with PD1 inhibition.[108] However, further studies concerning the question why only some animals per combination treatment show favourable prognosis ought to be addressed, in particularly inside the characterized PDGFB-driven glioma model using the somatic gen transfer system RCAS/TVA. Before clinical translation, it also needs further data to evaluate anti-CSF1R antibody tolerability and toxicity in combination with anti-PD1. Therefore, further clinical studies, especially targeting the timing of therapy and dose finding, are urgently needed for further evaluation.