代谢产物对T3期大肠癌淋巴结转移状态的预测

发布时间：2018-12-26 07:53

【摘要】：目的:寻找大肠癌不同淋巴结转移状态之间的差异性代谢产物,通过对代谢产物的鉴定与临床病理资料的结合,建立预测淋巴结转移的危险评分模型,能够运用该模型对不同淋巴结转移状态的患者进行预测。通过对代谢产物在不同淋巴结转移状态患者之间的表达高低不同,对代谢产物与不同淋巴结转移状态患者的预后关系进行研究,期望能够为大肠癌淋巴结转移潜能的预测提供分子标记物,并能为术前进行判断大肠癌患者的淋巴结分期提供客观依据。方法:1、LC-MS分析:运用质谱色谱仪器对大肠癌不同淋巴结转移状态患者的血液上清液进行检测,得到不同淋巴结分期患者代谢产物的响应值,然后对得到的数据运用统计学软件进行处理。2、数据处理:运用主成分分析、t检验、支持向量机、聚类分析、逻辑回归等统计方法对相关数据进行分析,寻找具有差异性的代谢产物,进一步对寻找到的物质进行验证。结果:在本研究中,得到了以下结果。1、通过主成分分析,发现大肠癌不同淋巴结状态患者中存在差异性的代谢产物,可以将不同淋巴结转移状态的患者区分开。2、经过t检验和聚类分析,寻找出17种物质可以将大肠癌按淋巴结转移状态的患者分为不同的类别。3、通过判别分析建立了良好的预测模型,用10种物质对不同组别判别分析(训练组和验证组)均得到了良好的精度、特异度以及敏感性。4、运用逻辑回归建立的模型和以及用3种代谢产物和临床病理参数绘制的危险评分对不同淋巴结转移状态的大肠癌患者具有良好的预测能力。5、选择的3种代谢产物中有2种对大肠癌不同淋巴结转移状态的患者的预后有影响。表达的响应值不同,患者的预后不同。结论:1、大肠癌不同淋巴结转移状态的患者之间有差异性代谢产物;2、以小分子代谢产物为标记物可以对大肠癌不同淋巴结转移状态进行良好的区分;3、根据临床基本资料以及筛选出的代谢产物所建立的logistic回归模型对大肠癌不同淋巴结转移状态具有良好的预测能力;4、筛选出的产物可以对不同淋巴结状态的大肠癌患者的生存具有提示作用。
[Abstract]:Objective: to search for the different metabolites among different lymph node metastases of colorectal cancer, and to establish a risk score model for predicting lymph node metastasis by combining the identification of metabolites with clinicopathological data. The model can be used to predict different lymph node metastases. The relationship between the expression of metabolites and the prognosis of patients with different lymph node metastases was studied. It is expected to provide molecular markers for predicting lymph node metastasis potential of colorectal cancer and provide objective basis for preoperative diagnosis of lymph node staging in patients with colorectal cancer. Methods: 1LC-MS analysis: the serum supernatants of patients with different lymph node metastasis status of colorectal cancer were detected by mass spectrometry, and the response values of metabolites in patients with different lymph node stages were obtained. Then the obtained data are processed with statistical software. 2. Data processing: using principal component analysis, t-test, support vector machine, cluster analysis, logical regression and other statistical methods to analyze the relevant data. Search for metabolites with differences and further verify the substances found. Results: in this study, the following results were obtained. 1. By principal component analysis, it was found that there were different metabolites in patients with different lymph node status of colorectal cancer. Through t test and cluster analysis, 17 substances were found to be able to classify colorectal cancer patients into different categories according to the status of lymph node metastasis. 3. A good predictive model was established by discriminant analysis. The accuracy, specificity and sensitivity of discriminant analysis of different groups (training group and validation group) were obtained by 10 substances. The model established by logical regression and the risk scores plotted by three metabolites and clinicopathological parameters had good predictive ability for colorectal cancer patients with different lymph node metastasis status. Two of the 3 metabolites selected had an effect on the prognosis of colorectal cancer patients with different lymph node metastases. The response value of expression was different, the prognosis of patients was different. Conclusion: 1, there are different metabolites among patients with different lymph node metastases of colorectal cancer, 2, small molecule metabolites can be used as markers to distinguish different lymph node metastases of colorectal cancer. 3. The logistic regression model based on the basic clinical data and the selected metabolites has a good predictive ability for different lymph node metastasis status of colorectal cancer. 4. The selected products may be helpful to the survival of colorectal cancer patients with different lymph node status.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：R735.34

【相似文献】