基于外周血KIF1B和STAT4mRNA表达以及基因表达式编程对肝癌患者辅助诊断的研究

发布时间：2018-04-27 09:24

本文选题：肝细胞癌 + 驱动蛋白家族成员1B　；参考：《青岛大学》2015年硕士论文

【摘要】：【目的】探讨肝癌患者、肝硬化患者、慢性肝炎患者、正常人群外周血驱动蛋白家族成员1B(kinesin family member 1B,KIF1B)m RNA和信号转导与转录激活因子4(signal transducers and activators of transcription,STAT4)m RNA表达情况及临床意义,评价其在肝癌(hepatocellular carcinoma,HCC)诊断中的应用价值。同时,应用基因表达式编程(gene expression programming,GEP)方法,利用常见的肝癌血清肿瘤标记物AFP、癌胚抗原(CEA)和糖类抗原199(CA199),以及肝功等生化指标数据信息来建立肝癌预测模型,对肝癌患者进行辅助性诊断。【方法】1.收集在2013年9月到2014年7月青岛大学医学院附属医院就诊的HCC患者(126例)、肝硬化患者(80例)、慢性肝炎患者(46例)、健康体检者(100例)共352例外周血样本。应用实时荧光定量PCR法检测各组外周血KIF1Bm RNA与STAT4m RNA的表达情况。应用SPSS 17.0软件分析KIF1B和STAT4m RNA在各组的表达水平,并绘制受试者工作特征曲线(receiver operating characteristic curve,ROC曲线),分析KIF1B与STAT4m RNA单独检测及联合AFP检测对肝癌的诊断价值。2.收集2011年1月到2014年7月于青岛大学医学院附属医院就诊的肝癌病人(500例)和健康体检者(500例)的肿瘤标记物与生化指标。采用APS 3.0软件对其进行GEP算法,并建立肝癌早期诊断模型。并对其可靠性还引入了人工神经网络(artificial neural network,ANN)的分析方法。【结果】1.肝癌患者血清CA199高水平组与低水平组相比,其外周血KIF1B m RNA出现较低表达(P0.05),与肝癌患者的性别、年龄、肿瘤的TNM分期、HBs Ag是否阳性、淋巴结转移、门脉癌栓、血清AFP和CEA水平无显著差异(P0.05)。而STAT4 m RNA的表达在肝癌患者高年龄组中出现高表达(P0.05),与肝癌患者的性别、肿瘤的TNM分期、HBs Ag是否阳性、淋巴结转移、门脉癌栓、血清AFP、CEA和CA199水平无显著差异(P0.05)。2.组间两两比较,其结果显示肝癌组外周血KIF1Bm RNA水平显著低于肝硬化组、慢性肝炎组以及正常对照组(F值分别为22.27、17.73、20.40,P0.05)。而外周血KIF1Bm RNA水平在肝硬化组、慢性肝炎组以及与正常对照组三组之间相互比较,其结果没有统计学意义(P0.05)。肝癌组外周血STAT4m RNA表达水平与正常对照组比较,其结果无统计学差异(F值为2.48,P0.05。肝癌组外周血STAT4m RNA表达水平与肝硬化组以及慢性肝炎组相比较,其结果无统计学差异(P0.05);肝硬化组、慢性肝炎组以及正常对照组三者外周血STAT4m RNA水平相互比较,其结果亦无统计学意义差异(P0.05)。3.在诊断肝癌方面,KIF1B、STAT4、AFP单独检测时的AUC分别为0.706、0.536、0.838,因STAT4曲线下的面积为0.536,接近0.5,不能用于临床肝癌诊断价值的评价。而KIF1B与AFP两者联合检测,其ROC曲线下面积为0.875,与AFP单独检测曲线下的面积相比无统计学差异(P0.05)。4.使用APS 3.0软件建立的GEP模型1包括AFP、CEA、CA199、ALB、ALT、Cl-六项指标。其训练集的准确性为95.75%,测试集的准确性为92.5%。GEP模型2包括AFP、CA199、ALB、ALT四项指标,其训练集的准确性为94%,测试集的准确性为92%。在ANN方法中,模型1同样包括AFP、CEA、CA199、ALB、ALT、Cl-六项指标。其训练集与测试集在多层感应器中的准确性分别为91.6%与89.1%,在径向基函数中的准确性分别是91%和83.3%。模型2同样包括AFP、CA199、ALB、ALT四项指标。其训练集与测试集在多层感应器中的准确性分别为90.5%与85.4%,在径向基函数中的准确性分别是90%和77.8%。【结论】1.KIF1Bm RNA在肝癌患者外周血中呈现低表达,且在血清CA199高水平组中出现低表达;2.STAT4m RNA在四组研究对象外周血中表达无统计学差异,但在肝癌患者高年龄组STAT4m RNA表达水平较高。3.外周血中KIF1Bm RNA与STAT4m RNA对肝癌的诊断效能均低于AFP,说明外周血KIF1Bm RNA与STAT4m RNA不能作为新型肿瘤标记物用于肝癌的辅助诊断。4.GEP方法联合六项指标建立的预测模型对肝癌的早期诊断有较高的准确性。5.GEP模型对肝癌的辅助诊断要优于ANN方法,更适合临床数据的分析。可以用于肝癌的早期辅助诊断。
[Abstract]:[Objective] to investigate the expression and clinical significance of 1B (kinesin family member 1B, KIF1B) m RNA and signal transduction and transcription activator 4 (signal transducers and activators) in patients with liver cancer, cirrhosis, chronic hepatitis, and normal population. Patocellular carcinoma, HCC) diagnostic value. At the same time, using gene expression programming (GEP) method, using the common liver cancer serum tumor markers AFP, carcinoembryonic antigen (CEA) and carbohydrate antigen 199 (CA199), and liver function data information to establish the prediction model of liver cancer, to the liver cancer patients 1. HCC patients (126 cases), cirrhosis patients (80 cases), chronic hepatitis (46 cases), healthy persons (100 cases) and 352 peripheral blood samples were collected from September 2013 to July 2014 at the Affiliated Hospital of Qiingdao University medical college. The peripheral blood KIF1Bm RNA and STAT4m R were detected by real time fluorescence quantitative PCR method. The expression of NA. The expression level of KIF1B and STAT4m RNA in each group was analyzed with SPSS 17 software, and the working characteristic curves of the subjects (receiver operating characteristic curve, ROC curve) were plotted. The value of the diagnosis of liver cancer was collected from January 2011 to July 2014 in Qingdao. The tumor markers and biochemical indexes of liver cancer patients (500 cases) and healthy persons (500 cases) were treated in the Affiliated Hospital of University Medical College. The GEP algorithm was carried out by APS 3 software, and the early diagnosis model of liver cancer was established. And the reliability of the artificial neural network (artificial neural network, ANN) was also introduced. [results] 1 Compared with the low level group, the CA199 high level group of liver cancer patients had lower expression of KIF1B m RNA in peripheral blood (P0.05). There was no significant difference in sex, age, TNM staging, HBs Ag, lymph node metastasis, portal vein thrombus, serum AFP and CEA levels in the patients with liver cancer (P0.05). High expression (P0.05) was found in patients with liver cancer, the TNM staging of tumor, HBs Ag positive, lymph node metastasis, portal vein tumor thrombus, serum AFP, CEA and CA199 levels were not significantly different (P0.05) in group.2. (P0.05). The results showed that the KIF1Bm RNA level in peripheral blood of liver cancer group was significantly lower than that of liver cirrhosis, chronic hepatitis and normal control. Group (F value was 22.27,17.73,20.40, P0.05). But the level of KIF1Bm RNA in peripheral blood was compared between three groups of liver cirrhosis, chronic hepatitis and normal control group, and the results were not statistically significant (P0.05). There was no statistical difference between the expression level of STAT4m RNA in the peripheral blood of liver cancer group and the normal group (F value of 2.48, P0.05.). The expression level of STAT4m RNA in peripheral blood of liver cancer group was compared with that of liver cirrhosis group and chronic hepatitis group, the results were not statistically different (P0.05). The level of STAT4m RNA in peripheral blood of the three groups of liver cirrhosis, chronic hepatitis and normal control group was compared, and there was no statistically significant difference (P0.05).3. in the diagnosis of liver cancer, KIF1B, STAT4, AFP. The AUC of the single test was 0.706,0.536,0.838, respectively, because the area under the STAT4 curve was 0.536, and was close to 0.5. It could not be used for the evaluation of the diagnostic value of the clinical liver cancer. The joint detection of KIF1B and AFP had an area of 0.875 under the ROC curve, which was not statistically different from the area under the AFP single detection curve (P0.05).4. using APS 3 software. The GEP model 1 includes six indexes, such as AFP, CEA, CA199, ALB, ALT, Cl-. The accuracy of the training set is 95.75%. The accuracy of the test set is the four indexes of AFP, CA199, ALB and ALT. The accuracy of the training set is 94%. The accuracy of the test set is six. The accuracy of the training set and the test set in the multilayer sensor is 91.6% and 89.1% respectively. The accuracy in the radial basis function is 91% and the 83.3%. model 2, respectively, including the four indexes of AFP, CA199, ALB, ALT. The accuracy of the training set and the test set in the multilayer sensor is 90.5% and 85.4% respectively, and the accuracy in the radial basis function is divided. The low expression of 90% and 77.8%. [Conclusion] 1.KIF1Bm RNA in the peripheral blood of the patients with liver cancer and low expression in the high level group of serum CA199; 2.STAT4m RNA in the peripheral blood of the four groups of subjects, there is no statistical difference, but the KIF1Bm RNA in the.3. peripheral blood of the high age group of the liver cancer patients and the KIF1Bm RNA in the.3. peripheral blood The diagnostic efficiency of the liver cancer is lower than that of AFP. It shows that the peripheral blood KIF1Bm RNA and STAT4m RNA can not be used as a new tumor marker for the auxiliary diagnosis of liver cancer, the prediction model established by the.4.GEP method combined with six indexes, the prediction model for the early diagnosis of liver cancer has a higher accuracy. The auxiliary diagnosis of the liver cancer by the.5.GEP model is better than the ANN method. The analysis of bed data can be used for early diagnosis of liver cancer.

【学位授予单位】：青岛大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：R735.7

【相似文献】