Abstract 3962: High-dimensional flow cytometric immune profiling of malignant gliomas

Abstract

Abstract INTRODUCTION: Immune therapy is emerging as a treatment modality for patients with solid tumors including intracranial metastases. Similar immune strategies are being tested in patients with malignant gliomas; diseases with uniformly poor prognosis. Whereas, recent work has demonstrated a correlation between the immune landscape of solid tumors and response to immune therapy, the immunologic landscape of CNS tumors has not been thoroughly defined. To capture the phenotypic diversity of immune cells in CNS tumors, we developed an 18-color flow cytometry panel and applied it to mononuclear cells isolated from acutely resected brain tumor, matched patient blood, and healthy donor blood specimens. METHODS: Linear discrimination analysis (LDA) and T-distributed stochastic neighbor embedding (tSNE) dimensionality reduction techniques were used to analyze flow cytometry data acquired from brain tumor, patient-matched peripheral blood, and healthy donor peripheral blood specimens. RESULTS: LDA of 13 parameters collected by flow cytometry revealed that CD8 tumor-infiltrating lymphocytes (TILs) differed notably from peripheral blood. CTLA-4 (TILs) and CD3 (peripheral blood) provided the greatest degree of separation. Using the tSNE algorithm to analyze the same 13 parameters on CD8 T cells, we rapidly reduced 78 potential parameter combinations by creating a clustering pattern representing lymphocytes with similar staining patterns. Lymphocyte clusters created from TILs were easily distinguished from clusters enriched in peripheral blood. Analysis of TIL clusters identified markers consistent with activated/memory populations as well as elevated levels of CTLA-4, LAG-3, and PD-1 relative to peripheral blood; due to a higher percentage of cells expressing these markers rather than increased expression on individual cells. Identifying individual specimen points as well as clusters of specimens (e.g. patients with the same tumor classification), enabled comparisons across individual patient specimens (intra-patient blood vs tumor), across tumor type or grade, and across disease state (patient vs healthy donor). CONCLUSIONS: Dimensionality reduction techniques provided a rapid means to identify markers that were differentially expressed in TILs relative to peripheral blood. Importantly, these analyses were unrestricted by canonical gating strategies. CD8 TILs express multiple markers of activation and maturation, suggesting reactivity to tumor antigens and potential for immune therapy. Application of these analyses to a broader patient population will reveal the immunologic landscape of primary CNS tumors and provide insight into the immunologic consequences of current therapies. Further, this may lead to identification of specific immune signatures that may predict therapeutic response to immune therapy. Note: This abstract was not presented at the meeting. Citation Format: Amber J. Giles, Caitlin M. Reid, Deric M. Park, Mario Roederer, Mark R. Gilbert. High-dimensional flow cytometric immune profiling of malignant gliomas [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2017; 2017 Apr 1-5; Washington, DC. Philadelphia (PA): AACR; Cancer Res 2017;77(13 Suppl):Abstract nr 3962. doi:10.1158/1538-7445.AM2017-3962