血清小分子代谢产物与胃癌淋巴结转移状态的相关性研究

发布时间：2018-08-08 13:09

【摘要】：目的:研究胃癌患者血清中小分子代谢产物与淋巴结转移状态的相关性;初步研究小分子代谢产物预测胃癌淋巴结转移状态的相关机制;尝试利用小分子代谢产物与临床病理资料建立预测胃癌淋巴结转移的风险评分模型。方法:1、LC-MS分析:应用液相色谱-质谱仪分析血清样本代谢产物谱;2、应用酶联免疫法测定各组血清样本中VEGF-C、VEGF-D含量;3、应用免疫组织化学染色技术检测各组癌组织样本中VEGF-C、VEGF-D的表达、癌旁组织淋巴管密度(LVD);4、数据分析:运用主成分分析、t检验、支持向量机判别分析等统计学方法进行数据分析,寻找T1-2N1-3组与T3N0组胃癌患者血清中的差异性代谢产物;并以差异性代谢产物为分组依据对独立样本进行分组;而后运用t-检验、卡方检验比较独立样本分组后的组间VEGF-C、D水平和癌旁组织淋巴管密度;单因素、多因素逻辑回归选出与淋巴结转移相关的独立危险因素;进一步结合临床病理资料和代谢产物建立预测胃癌淋巴结转移的风险评分模型。结果:1.第一部分:血清样本的代谢组研究中,从Peakview软件得出的血清总离子流图中,可明显观察出组间代谢产物的差异。PCA得分图结果提示组T1-2N1-3组与T3N0胃癌之间能很好区分开,t检验可筛选出T1-2N1-3组与T3N0组之间差异的代谢产物共31种,进一步运用支持向量机方法进行判别分析,选出权重值为100%的代谢产物3种。聚类分析的热图结果表明这3种差异性代谢产物能将t1-2n1-3组与t3n0区分开。2.第二部分:以3种差异性代谢产物为依据对另一35例独立样本进行聚类分析,再次验证其对t1-2n1-3组与t3n0组血清的区分结果。结果表明独立样本仍能区分为a组和b组。a、b两组间血清样本的vegf-c、vegf-d浓度,差异有统计学意义(p值分别为0.032、0.047);癌组织中vegf-c、vegf-d表达,a组显著高于b组(p值分别为0.046、0.025);癌旁组织淋巴管密度测定a组显著高于b组(p=0.007);生存分析亦存在显著差异。3.第三部分:对年龄、性别、分化程度、肿瘤大小、3种差异性代谢物等资料进行逻辑回归,结果表明:年龄、phenylpyruvicacid、threoninyl-isoleucine为胃癌淋巴结转移的独立危险因素,结合性别和2种代谢产物建立预测胃癌淋巴结转移的风险评分模型,该模型能够将淋巴结转移阴性与阳性样本区分开,随着评分的增加,淋巴结转移阳性病例占组内病例的百分比也增高,此模型作为诊断性实验的roc曲线图显示相比于单个独立危险因素,其能够更好的预测淋巴结转移。结论:1.不同淋巴就转移状态的胃癌患者血清中,代谢产物存在差异;2.通过液相色谱-质谱技术筛选出的差异性代谢物对不同淋巴结转移状态的胃癌患者有较强的区分能力;代谢产物与淋巴结转移之间有一定的相关性;3.基于差异性代谢产物对另一35例独立样本行聚类分析后,仍能有效分组,组间血清VEGF-C、D浓度水平、癌组织VEGF-C、D表达、癌旁组织淋巴管密度均存在差异,差异有统计学意义;4.代谢产物是可能是通过促进癌组织高表达VEGF-C、D,促进癌旁组织新生淋巴管形成,进而促进淋巴结转移;也可能是VEGF-C、D的高表达,造成了血清中代谢产物的差异;5.逻辑回归结果确定了年龄、Phenylpyruvic acid、Threoninyl-Isoleucine为胃癌淋巴结转移的3个独立危险因素,以该3个危险因素建立的预测胃癌淋巴结转移的风险评分模型可能成为临床预测淋巴结转移的有效手段。
[Abstract]:Objective: To study the correlation between small molecular metabolites and lymph node metastases in patients with gastric cancer, and to study the mechanism of small molecular metabolites to predict lymph node metastasis in gastric cancer, and to establish a risk scoring model for predicting the metastasis of gastric cancer by using small molecule metabolites and clinicopathological data. Method: 1, LC-MS Analysis: using liquid chromatography mass spectrometry to analyze the metabolites of serum samples; 2, VEGF-C, VEGF-D content in serum samples were measured by ELISA; 3, the expression of VEGF-C, VEGF-D, and lymphatic density (LVD) in the para cancer tissue samples were detected by immunohistochemistry. 4, data analysis: using principal component analysis T test, support vector machine discriminant analysis and other statistical methods for data analysis to find the differential metabolites in the serum of patients with gastric cancer in group T1-2N1-3 and T3N0, and group the independent samples on the basis of the differential metabolites, and then using the t- test to compare the VEGF-C and D levels of the group after the independent sample grouping. And the lymph node density in the paracancerous tissue; single factor, multiple factor logic regression to select independent risk factors associated with lymph node metastasis; further combined with clinicopathological data and metabolites to establish a risk scoring model for predicting lymph node metastasis of gastric cancer. Results: 1. Part 1: in the metabolic group study of serum samples, from Peakview software In the serum total ion flow chart, the difference of.PCA score between groups was clearly observed. The results of.PCA score showed that the group T1-2N1-3 group was well separated from the T3N0 gastric cancer. The t test could screen out a total of 31 different metabolites between the T1-2N1-3 group and the T3N0 group, and the discriminant analysis was carried out by the support vector machine method. The weight value was selected to be 100. 3 kinds of metabolites were found. The results of cluster analysis showed that the 3 different metabolites could separate the t1-2n1-3 group and the T3N0 region from the second part of.2.: 3 different metabolites were used to cluster analysis of the other 35 independent samples, and the results of the difference between the t1-2n1-3 group and the T3N0 group were verified again. The results showed that the independent sample was independent. The VEGF-C and VEGF-D concentration of serum samples between group A and group B, B two were statistically significant (P value was 0.032,0.047), VEGF-C, VEGF-D expression, a group were significantly higher than that of B group, and lymphatic density in paracancerous tissue was significantly higher than that in group two; survival analysis was also significantly worse. The third part of.3.: age, sex, degree of differentiation, tumor size, and 3 different metabolites. The results show that age, phenylpyruvicacid, threoninyl-isoleucine are independent risk factors for lymph node metastasis of gastric cancer, and the risk scoring model for predicting lymph node metastasis of gastric cancer combined with sex and 2 kinds of metabolites is established. The model can separate the negative lymph node metastasis from the positive sample area. As the score increases, the percentage of lymph node metastasis positive cases also increases. As a diagnostic test, the model shows that compared with a single independent risk factor, the model can predict lymph node metastasis better than a single independent risk factor. Conclusion: 1. different types of lymph node metastasis can be predicted. Conclusion: 1. different lymphatic vessels can be used to predict lymph node metastasis. The metabolic products were different in the serum of patients with gastric cancer with metastatic state; 2. the differential metabolites screened by liquid chromatography mass spectrometry had a strong distinguishing ability for gastric cancer patients with different lymph node metastases; there was a certain correlation between the metabolites and lymph node metastasis; and 3. based on the difference of metabolic products to the other 35. After cluster analysis, the serum VEGF-C, D concentration level, VEGF-C, D expression and lymphatic density in the para cancerous tissue were different, and the difference was statistically significant. 4. the metabolites could promote the high expression of VEGF-C, D, promote the formation of lymphatic tube in the paracancerous tissues, and then promote the drenching. The high expression of VEGF-C and D resulted in the difference in the metabolites in the serum; 5. the logical regression results confirmed the age, Phenylpyruvic acid, Threoninyl-Isoleucine as 3 independent risk factors for lymph node metastasis of gastric cancer, and the risk score model for predicting lymph node metastasis of gastric cancer by the 3 risk factors It is an effective means to predict lymph node metastasis in clinical practice.
【学位授予单位】：吉林大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R735.2

【相似文献】