TMIC-58. ALTERATIONS IN GLIOBLASTOMA TUMOR MICROENVIRONMENT TRANSCRIPTOME AFTER TREATMENT WITH PEMBROLIZUMAB

Abstract

Abstract BACKGROUND Immune checkpoint inhibitors (CPI) have failed to show survival benefit in glioblastoma. Understanding the alterations in the tumor microenvironment after treatment with CPI is crucial for development of immunotherapy in glioblastoma. We previously reported our window-of-opportunity trial of pembrolizumab in GBM (NCT02337686). In the current study, we obtained 8 paired glioblastoma tumors pre- and post-treatment with pembrolizumab and conducted RNA sequencing analysis to understand the alterations in the tumor microenvironment transcriptome in response to pembrolizumab. Method: We conducted bulk RNA-sequencing profiling and differential gene expression analysis to compare the expression levels of patient-derived glioblastoma pre- and post-treatment with pembrolizumab, while controlling for paired information. We performed gene enrichment analysis on the differentially expressed genes. We employed the single sample mww-GST method to assess the enrichment of various immune cell types that were previously identified, across the two groups. RESULTS Our analysis revealed 294 down-regulated and 259 up-regulated genes in the post treatment samples. The gene enrichment analysis demonstrated that the up-regulated genes were enriched in pathways related to the immune system. The single sample mwwGST analysis indicated a consistent overall up-regulation of expression in immune cell infiltration. Notably, T cell-related signatures, such as CD4, CD8, and NKT exhibited significantly higher infiltration in the post-treatment group. Interestingly, gene ontology analyses showed upregulation of phagocytosis and antigen presentation gene modules in the post-treatment samples. Additionally, we found neuronal components and genes related to hyperexcitability and synaptic connectivity with an overrepresentation of AMPA and DLG4 pathways in post-treatment tissues. CONCLUSION Our data show that in addition to the anticipated infiltration of myeloid and lymphoid cell populations, checkpoint blockade can also unexpectedly influence glioma-neuron interactions, which may represent additional immunotherapy targets in GBM. Further proteomic validation of infiltration of different cell types and their interactome in pre- and post-CPI treated tissue is warranted.