PO.CL07.01 · 临床研究
胶质母细胞瘤的新型精准治疗
New precision treatments for glioblastoma
作者与单位 Authors & Affiliations
摘要 Abstract
中文摘要
本研究的主要目标是开发针对胶质母细胞瘤（GBM）的精准治疗。GBM是一种致命的脑肿瘤，治疗选择很少，主要原因在于其高度的瘤内和瘤间异质性，表现为单个肿瘤内或不同患者间存在多样的基因改变和细胞成分。此前针对分子特征的努力集中于三种分子亚型——经典型（CL）、间充质型（ME）和前神经型（PN）——但尚未针对每种亚型开发出有效的治疗方法。为克服这一挑战，我们最近采用机器学习算法识别出31个必需生存基因，构成新的GBM进展基因特征（GBM-PGS），将患者分为不良预后高风险或低风险（HR对比LR）。由于这些基因对癌细胞生存/生长具有功能上的重要性，GBM-PGS比EGFR、MGMT和IDH1等现有生物标志物更准确地预测GBM患者的进展。鉴于GBM-PGS与肿瘤细胞生存和疾病进展的功能相关性，我们假设将GBM-PGS与分子分型相结合将把GBM患者分层为不同的治疗组，从而在这一致命疾病中实现对FDA批准药物的重新利用以用于精准医学。我们利用GBM-PGS和分子分型将TCGA GBM患者和DepMap胶质瘤细胞系分层为六个亚组：HR-CL、HR-ME、HR-PN、LR-CL、LR-ME和LR-CL。使用Kaplan-Meier分析评估TCGA亚组的生存，并使用基因本体论分析HR和LR组的潜在药物靶点。此外，从DepMap PRISM重利用药物筛选数据集中识别出每个亚组的候选精准治疗药物，并在一系列GBM细胞系中进一步验证。LR-PN显示出比其他亚组显著更好的预后。针对HR型GBM的药物在功能上不同于针对LR型GBM的药物。六个治疗亚组对包括FDA批准药物在内的化学化合物表现出多样的反应。每个亚组的候选药物均在GBM细胞系中得到验证。我们的结果验证了GBM-PGS和分子分型在患者分层以及开发亚组特异性精准治疗中的关键作用。总体而言，本研究有望改变GBM的治疗格局。
查看英文原文 English abstract
The main goal of this study is to develop precision treatments for glioblastoma (GBM). GBM is a deadly brain tumor with few treatment options, mainly due to high intra- and inter-tumoral heterogeneity manifested by diverse genetic alterations and cellular components within an individual tumor or among different patients. Previous efforts targeting molecular features have focused on three molecular subtypes - Classical (CL), Mesenchymal (ME), and Pro-Neural (PN) - but no effective treatments have been developed for each subtype. To overcome this challenge, we recently employed a machine learning algorithm to identify 31 essential survival genes, forming a new GBM progression gene signature (GBM-PGS) that divided patients into high- or low-risk (HR vs LR) of poor prognosis. Because these genes are functionally vital for cancer cell survival/growth, GBM-PGS predicts the progression of GBM patients more accurately than existing biomarkers such as EGFR, MGMT, and IDH1. Given the functional relevance of GBM-PGS to tumor cell survival and disease progression, we hypothesize that combining GBM-PGS with molecular subtyping will stratify GBM patients into distinct treatment groups, enabling the repurposing of FDA-approved drugs for precision medicine in this fatal disease. We stratified TCGA GBM patients and DepMap glioma cell lines into six subgroups: HR-CL, HR-ME, HR-PN, LR-CL, LR-ME, and LR-CL using GBM-PGS and molecular subtyping. The survival of TCGA subgroups was assessed using Kaplan-Meier analysis, and potential drug targets for the HR and LR groups were analyzed using Gene Ontology. Moreover, candidate precision treatments for each subgroup were identified from the DepMap PRISM Repurposing Drug Screen dataset and further verified in a range of GBM cell lines. The LR-PN displayed a significantly better prognosis than the other subgroups. Drugs targeted for HR GBMs were functionally different from those for LR GBMs. The six treatment subgroups exhibited diverse responses to chemical compounds, including FDA-approved drugs. Candidate drugs for each subgroup were validated in GBM cell lines. Our results have verified the crucial roles of GBM-PGS and molecular subtyping in patient stratification and in the development of subgroup-specific precision treatments. Overall, this study has the potential to transform the GBM therapeutic landscape.
利益披露 Disclosure
L. O. Anifowose, None..
Z. Sheng, None.