Department of Neurosurgery, University of Florida, Gainesville, FL, USA.
University of Florida Brain Tumor Immunotherapy Program, University of Florida, Gainesville, FL, USA.
J Neurooncol. 2018 Apr;137(2):249-257. doi: 10.1007/s11060-017-2732-1. Epub 2018 Jan 4.
Glioblastoma (GBM) generates a varied immune response and understanding the immune microenvironment may lead to novel immunotherapy treatments modalities. The goal of this study was to evaluate the expression of immunologic markers of potential clinical significance in primary versus recurrent GBM and assess the relationship between these markers and molecular characteristics of GBM. Human GBM samples were evaluated and analyzed with immunohistochemistry for multiple immunobiologic markers (CD3, CD8, FoxP3, CD68, CD163, PD1, PDL1, CTLA4, CD70). Immunoreactivity was analyzed using Aperio software. Degree of strong positive immunoreactivity within the tumor was compared to patient and tumor characteristics including age, gender, MGMT promoter methylation status, and ATRX, p53, and IDH1 mutation status. Additionally, the TCGA database was used to perform similar analysis of these factors in GBM using RNA-seq by expectation-maximization. Using odds ratios, IDH1 mutated GBM had statistically significant decreased expression of CD163 and CD70 and a trend for decreased PD1, CTLA4, and Foxp3. ATRX-mutated GBMs exhibited statistically significant increased CD3 immunoreactivity, while those with p53 mutations were found to have significantly increased CTLA4 immunoreactivity. The odds of having strong CD8 and CD68 reactivity was significantly less in MGMT methylated tumors. No significant difference was identified in any immune marker between the primary and recurrent GBM, nor was a significant change in immunoreactivity identified among age intervals. TCGA analysis corroborated findings related to the differential immune profile of IDH1 mutant, p53 mutant, and MGMT unmethylated tumors. Immunobiologic markers have greater association with the molecular characteristics of the tumor than with primary/recurrent status or age.
胶质母细胞瘤（GBM）会产生多样化的免疫反应，了解免疫微环境可能会导致新的免疫治疗方法。本研究的目的是评估原发性和复发性 GBM 中潜在临床意义的免疫标志物的表达，并评估这些标志物与 GBM 分子特征之间的关系。使用免疫组织化学对多种免疫生物标志物（CD3、CD8、FoxP3、CD68、CD163、PD1、PDL1、CTLA4、CD70）进行了人类 GBM 样本的评估和分析。使用 Aperio 软件对免疫反应性进行分析。肿瘤内强阳性免疫反应的程度与患者和肿瘤特征（包括年龄、性别、MGMT 启动子甲基化状态以及 ATRX、p53 和 IDH1 突变状态）进行了比较。此外，还使用 TCGA 数据库通过期望最大化的 RNA-seq 对这些因素进行了类似的分析。使用优势比，IDH1 突变型 GBM 具有统计学意义上降低的 CD163 和 CD70 表达，并且 PD1、CTLA4 和 Foxp3 的表达呈下降趋势。ATR X 突变型 GBM 表现出统计学上显著增加的 CD3 免疫反应性，而具有 p53 突变的 GBM 则发现 CTLA4 免疫反应性显著增加。在 MGMT 甲基化肿瘤中，具有强 CD8 和 CD68 反应的可能性显著降低。在原发性和复发性 GBM 之间未发现任何免疫标志物存在显著差异，也未发现年龄间隔内的免疫反应性发生显著变化。TCGA 分析证实了 IDH1 突变、p53 突变和 MGMT 未甲基化肿瘤免疫谱差异的发现。免疫生物标志物与肿瘤的分子特征相关性更大，而与原发性/复发性状态或年龄相关性较小。
