Computational Approach To Identifying Neuroinflammation In Glioblastoma Multiforme (Gbm)

Abstract INTRODUCTION Glioblastoma multiforme (GBM) is an aggressive brain cancer with dismal prognosis, despite aggressive surgery, radiation and chemotherapy. Therapies directed to the immune system are an exciting prospect in oncology and require an understanding of the interaction between tumors and immune cells. Our objective was to use computational tools to estimate immune infiltration in GBM and determine whether specific immune cell types are associated with clinical outcomes. METHODS RNA sequencing and targeted DNA sequencing (to 981 oncology genes) was performed from 37 surgically-resected GBM tumors. Tumor mutations were identified, and gene-level transcript counts were used to estimate tumor-associated cell types using bioinformatics tools. Clinical variables, including survival from surgery and diagnosis, were collected and tested for associations with molecular data. RESULTS We detected leukocyte fractions (i.e., immune infiltration) ranging from 2% to 50% in GBMs, with an average of 10.2%. Specifically, we found a statistically significant association between high Th2 cell estimates and reduced overall survival (from both surgery and diagnosis). Nine patients with high Th2 tumors had a median OS from surgery of 187 days, compared with a median of 454 days for 28 patients with low Th2 tumors (log-rank Matel-Cox test; p = 0.0023, HR = 3.1). We also found an association between NF1 mutant tumors, which were enriched for a KRAS signaling signature, and high immune infiltration (p < 0.05). CONCLUSION Our computationally-driven approach predicted significant immune infiltration in GBM and a potential association between poor prognosis and Th2 cells. The specific class of CD4+ helper T-cells is generally associated with poor anti-tumor immunity and a Th2-bias has been reported in gliomas. Our data adds to the collective understanding of the molecular landscape of GBM, as well as the complex immune environment, which will have important implications in tumor treatment and prognosis.