Toward a molecular classification of human gliomas

描述（由申请人提供）：常见恶性胶质瘤的病理分类问题使预测患者预后和治疗反应复杂化。这项研究的长期目标是提供对神经胶质瘤发生特征的遗传事件的全面了解，并将遗传分析引入常规分类。在此资助的前两个资助期间，我们已经阐明了神经胶质瘤形成的遗传事件，将基因型与临床病理参数相关联，并将遗传分析引入临床实践。这项工作现在提出了三个假设，可以使用经过验证的转化研究策略进行测试。1)为了验证分子谱可以将组织学上典型和非典型恶性胶质瘤实际分为预后改善亚组和预后预测亚组的假设，我们提出定义能够稳健区分化疗敏感和预后较好的恶性胶质瘤与化疗耐药和预后较差的恶性胶质瘤的标志物。2)为了验证细胞侵袭增加与预后无关的组织学外观相关的假设，并且这种分子表型可以很容易地检测到，我们建议在恶性胶质瘤患者中定义与临床病理终点相关的“侵袭基因型/表型”。3)最后，为了验证持续的选择压力决定肿瘤基因型的假设，以及分子诊断方法应该考虑到这种异质性，我们建议表征胶质母细胞瘤周围坏死性假性突起的瘤内选择，以定义与间变性星形细胞瘤患者临床病理终点相关的分子特征。因此，每个目标都测试了一个相关的假设，即特定的分子分析可以增强当前胶质瘤分类的方法，每个目标也遵循类似的实验设计。进一步阐明人类胶质瘤的遗传基础将继续有助于胶质瘤分类系统，该系统将比目前的组织病理学方案更准确地反映肿瘤行为和对治疗的反应。

DESCRIPTION (provided by applicant): Problems in the pathological classification of the common malignant gliomas complicate predicting patient prognosis and response to therapy. The long-term goal of this research is to provide a comprehensive understanding of the genetic events that characterize glioma tumorigenesis and to introduce genetic analyses into routine classification. During the two prior funding periods of this grant, we have clarified genetic events that underlie glioma formation, have correlated genotype with clinicopathological parameters, and have introduced genetic analyses into clinical practice. This work now raises three hypotheses that can be tested using proven translational research strategies. 1) To test the hypothesis that molecular profiles can divide histologically classic and non-classic malignant gliomas into improved prognostic and predictive subgroups in a practical manner, we propose to define markers capable of robust distinction of chemosensitive and better prognosis malignant gliomas from chemoresistant and poor prognosis malignant gliomas. 2) To test the hypothesis that increased cellular invasion correlates with prognosis independent of histological appearance and that such a molecular phenotype can be readily detected, we propose to define an "invasion genotype/phenotype" that correlates with clinicopathological endpoints in patients with malignant gliomas. 3) Finally, to test the hypothesis that ongoing selection pressures operate to determine tumor genotype, and that molecular diagnostic approaches should take such heterogeneity into account, we propose to characterize intratumoral selection in glioblastoma perinecrotic pseudopalisades to define a molecular signature that correlates with clinicopathological endpoints in patients with anaplastic astrocytomas. Each aim thus tests a related hypothesis that a particular molecular analysis can augment current approaches to glioma classification, and each aim also follows a similar experimental design. The further clarification of the genetic basis of human gliomas will continue to contribute to a glioma classification system that will more accurately reflect tumor behavior and response to therapy than current histopathological schemes.