基于microRNA原位杂交和数据挖掘技术的三阴性乳腺癌预后预测模型研究

项目名称： 基于microRNA原位杂交和数据挖掘技术的三阴性乳腺癌预后预测模型研究

项目编号： No.81202079

项目类型： 青年科学基金项目

立项/批准年度： 2013

项目学科： 肿瘤学2

项目作者： 曹志刚

作者单位： 复旦大学

项目金额： 23万元

中文摘要： 三阴性乳腺癌是一种特殊的乳腺癌亚型，最新的研究结果显示其并不是一种单一的疾病，预后差异较大。作为一种新的分子标志，miRNA从转录后水平广泛参与基因表达调控，可用于肿瘤的预后预测。基于"基因组合"的个体化预后预测有重要的科学及社会意义，然而，绝大多数的预后模型都是基于基因芯片或PCR构建的，对标本要求高，稳定性和重复性差，耗时长，费用高昂，不利于推广，传统的统计方法在处理临床上出现的非线性数据方面也存在不足。为此，基于我院大样本量的规范治疗、随访良好的患者样本和技术平台，我们拟利用组织芯片和基于锁核酸探针的miRNA原位杂交技术，在训练集和验证集患者中筛选能够独立预测三阴性乳腺癌预后的临床病理指标及miRNA，以此为基础运用基于数据挖掘方法的支持向量机技术，构建特异性、灵敏度、稳定性和重复性好的三阴性乳腺癌预后预测模型，为三阴性乳腺癌的个体化预后预测及治疗奠定基础。

中文关键词： 乳腺癌；miRNA；原位杂交；数据挖掘；预后预测模型

英文摘要： Triple negative breast cancer (TNBC), a unique subtype of breast cancer, is a very heterogeneous group of breast cancer and there is conflicting information over prognosis for TNBC. MicroRNA (miRNA), as a new molecular biomarker and post-transcriptional regulator of target gene expression, is widely used in cancer prognosis. It's already confirmed that "gene combination" can successfully predict outcome of cancer patients. However, most of the predictive models for prognosis are based in gene chips or PCR technique and have some defect such as low stability and reproducibility, time-consuming and expensiveness. Besides, traditional statistical methods can not handle nonlinear data in the clinic. And basing on some advantages such as high level of technical platform and lots of samples with complete follow-up data in our center, we plan to use a tissue microarray-based miRNA detection in situ hybridization with LNA probe to screen and optimize clinic-pathological parameters, miRNAs, and the combinations related to the prognosis of triple negative breast cancer. Then construct the predictive model for prognosis of TNBC with good repeatability, low cost, easy to spread, suitable for Chinese negative breast cancer by data mining and support vector machine.

英文关键词： breast cancer；microRNA；in situ hybridization；data mining；prognosis predictive model