Abstract B15: Tumor genotype dependency of checkpoint blockade therapy in EGFR-driven glioblastoma

Abstract

Abstract Background and Rationale: The tumor microenvironment plays a significant role in tumor progression and how patients respond to therapy. Recent work suggests that tumor genotypes significantly influence the immune landscapes of tumors and guide the degree of immune suppression; however, detailed mechanistic insights remain scant. Glioblastoma multiforme (GBM) is an aggressive primary brain cancer with a poor median survival of approximately 15 months. Approximately 60% of GBMs harbor overexpression of EGFR, of which 50% contain a specific mutant of EGFR (EGFRvIII). Downstream signaling of EGFRvIII differs from EGFR. Utilizing genetically engineered mouse models of EGFR- and EGFRvIII-driven GBM, we determined 1) the two GBMs comprise distinct immune landscapes, 2) EGFR GBMs respond to checkpoint blockade immunotherapy whereas EGFRvIII GBMs do not, and 3) how the unique cellular composition of EGFRvIII GBM immune landscapes actively suppresses the effects of checkpoint blockade. Methods: Conditional transgenic EGFR or EGFRvIII, CDKN2A null, PTEN floxed mice with a floxed luciferase reporter were stereotactically injected intracranially with iCre lentivirus to initiate GBMs. Cohorts of mice were imaged using bioluminescence (BLI) to detect and stage GBMs for treatment studies. Mice were randomly enrolled in treatments with antibodies against murine PD-1, CTLA-4 or combination, or matching IgG controls every 3 days for 3 doses. Treated mice were monitored for overall survival and tumor response by BLI. Mice were also sacrificed at predetermined time points before and after therapy to determine the immune landscape by flow cytometry. Results: We identified that, strikingly, only combination blockade of PD-1/CTLA-4 was effective in prolonging survival in EGFR GBMs but not in EGFRvIII GBMs. Flow cytometric studies demonstrate that response to combination blockade is dependent on the balance of CD8 T-cells to immunosuppressive PMN-MDSC infiltrates. Interestingly, EGFRvIII GBMs have higher PMN-MDSC infiltrates, which are associated with insensitivity to checkpoint blockade. Depletion of PMN-MDSCs in EGFR tumors restores sensitivity to monotherapy checkpoint blockade whereas depletion of PMN-MDSCs in EGFRvIII tumors restores sensitivity to combination treatment. Furthermore, EGFRvIII GBMs have higher mRNA expression of ligands for the neutrophil chemokine CXCR2 receptor compared to EGFR GBMs, suggesting a mechanism for the observed higher infiltration of PMN-MDSCs and resistance to checkpoint blockade in EGFRvIII GBMs. Conclusions: We present that combination blockade of PD-1/CTLA-4 provides survival benefit in EGFR GBMs whereas monotherapies are ineffective. All treatment conditions are ineffective in EGFRvIII GBMs. Flow cytometry analysis and depletion studies demonstrate that PMN-MDSCs play a role by limiting the efficacy of PD-1/CTLA-4 response. Our data point to the use of CXCR2 inhibitors in combination with checkpoint blockade as a potential approach for treatment of GBM. Citation Format: Alan T. Yeo, Alain Charest. Tumor genotype dependency of checkpoint blockade therapy in EGFR-driven glioblastoma [abstract]. In: Proceedings of the AACR Special Conference on Tumor Immunology and Immunotherapy; 2019 Nov 17-20; Boston, MA. Philadelphia (PA): AACR; Cancer Immunol Res 2020;8(3 Suppl):Abstract nr B15.