新型血清肝脏肿瘤标志物研究及肝癌诊断模型建立

发布时间：2018-02-06 02:25

本文关键词： 肝脏恶性肿瘤原发性肝细胞癌血清学肿瘤标志物诊断模型　出处：《中国人民解放军医学院》2016年博士论文　论文类型：学位论文

【摘要】：目的：肝癌是发病率和致死率都很高的消化道恶性肿瘤,但目前该病的检测标志物尚无法满足临床需求。本课题拟评价血清学标志物在肝癌预警、早期诊断及预后评估中的价值,采用血清学标志物和蛋白指纹图谱建立肝癌早期诊断和鉴别诊断数学模型,并鉴定新发现的肝癌多肽标志物。方法：第一部分：基于血清标志物的肝癌诊断模型建立及标志物诊断价值评价。(1)采用回顾性队列研究方法对109例肝硬化(LC)患者进行24个月的随访,分析血清学标志物水平与肝细胞癌(HCC)发病风险之间的关系。(2)检测142例HCC,93例亚临床肝癌(SCHCC),182例肝硬化(LC)患者血清学标志物水平,分别建立多层感知机(MLP),径向基函数(RBF)和判别分析(DA)早期HCC诊断模型,通过联合多项血清学标志物实现HCC早期诊断。(3)采用生存分析方法对36例HCC患者术后随访24个月,分析血清学标志物水平与HCC术后复发风险的关系。第二部分：MALDI-TOF MS联合MB-WCX磁珠建立肝脏恶性肿瘤诊断模型及标志物的鉴定。(1)收集43例肝脏恶性肿瘤和52例肝脏良性肿瘤患者血清,采用基质辅助激光解吸电离飞行时间质谱(MALDI-TOF MS)结合弱阳离子交换磁殊(MB-WCX)技术检测血清中的低分子量多肽,分别建立监督神经网络(SNN),遗传算法(GA)和快速分类器(QC)鉴别诊断模型,并采用50例临床样本进行盲样验证和评价。(2)采用qTOF Synapt G2-S串联质谱对筛选出的差异肽进行鉴定,确定具有诊断价值的差异肽片段来源。结果：第一部分(1)109例LC患者共有29例发展为HCC(26.6%)。甲胎蛋白(AFP)、甲胎蛋白异质体L3 (AFP-L3)、丙氨酸氨基转移酶(ALT)三项指标在发生HCC组中浓度显著高于未发生HCC组(t值分别为-2.299,-2.384,-2.847,p值均0.05)；AFP、AFP-L3阳性为发展为HCC的潜在危险因素(相对危险度(RR)分别为2.41、2.26)；高尔基体蛋白73(GP73)血清水平在发展为HCC后显著下降(t=2.212；p=0.041)。(2)三类肝癌早期诊断模型对HCC训练和验证样本分类正确率分别如下：MLP模型为66.7%和63.6%,RBF模型为62.1%和61.4%,DA模型为61.6%和60.7%。(3)单因素分析发现,仅有术前GP73浓度对患者术后复发具有预测价值(X2=4.695,P=0.03),COX多因素分析发现AFP-L3、凝血酶原时间(Pt(s))、凝血酶原活动度(Pt(a))是患者术后复发的独立预测因子(风险比(HR)分别为1.003,3.465,1.137,p值均0.05)。第二部分(1)通过ClinPro Tools分析,共发现27个血清多肽峰在良恶性肝肿瘤患者血清中有差异表达,GA,SNN,QC模型的重识别率分别为100%,89.38%,80.84%,交叉验证率分别为81.67%,81.11%,86.11%,盲样验证准确率分别为78%,84%,84%。(2)分子量为2860.34Da,2881.54Da,3155.67Da的差异肽峰被成功鉴定,它们分别是FIBA、FIBB和ITIH4片段。结论：我们通过随访研究证实检测AFP、AFP-L3、GP73、ALT血清水平可对HCC发生进行早期预警,而检测GP73、AFP、AFP-L3、Pt(s)、Pt(a)血清水平可对HCC术后复发进行预测；我们建立了基于血清标志物的肝癌早期诊断模型和蛋白指纹图谱肝癌鉴别模型,并通过串联质谱鉴定发现FIBA、FIBB、ITIH4三个血清肽可能成为新的肝癌标志物。
[Abstract]:Objective: liver cancer incidence and mortality rate are high malignant tumor of digestive tract, but the detection of the disease markers cannot meet clinical needs. This project intends to evaluate the serological markers of liver cancer in early warning, early diagnosis and prognosis in value by serological and establishment of protein fingerprint and early diagnosis of hepatocellular carcinoma the mathematical model of differential diagnosis of hepatocellular carcinoma markers, marker polypeptide and identification of new discoveries. Methods: the first part: the establishment of liver cancer diagnosis model of serum markers and markers to evaluate the diagnostic value. (1) based on a retrospective cohort study of 109 cases of liver cirrhosis (LC) patients with 24 months of follow-up, serological analysis markers of hepatocellular carcinoma (HCC) and the relationship between risk. (2) detected 142 cases of HCC, 93 cases of subclinical hepatocellular carcinoma (SCHCC), 182 cases of liver cirrhosis (LC) patients with serological markers, were established Layer perceptron (MLP), radial basis function (RBF) and discriminant analysis (DA) model for the early diagnosis of HCC, through multiple serological markers for early diagnosis of HCC implementation. (3) using the method of survival analysis of 36 cases of HCC patients were followed up for 24 months, analysis of the relationship between serum markers of recurrence risk level and HCC after the operation. The second part: the identification of MALDI-TOF MS combined with MB-WCX beads to establish liver cancer model and diagnosis markers. (1) serum collected from 43 patients with liver cancer and 52 cases of benign liver tumors, using matrix assisted laser desorption ionization time of flight mass spectrometry (MALDI-TOF MS) with weak cation exchange (MB-WCX magneticscreening low molecular weight polypeptide) detected in serum, respectively establish the supervised neural network (SNN), genetic algorithm (GA) and fast classifier (QC) differential diagnosis model, and the 50 cases of clinical samples of blind sample validation and evaluation (2) by qTOF. Mass spectrometry to identify differentially expressed peptide tandem Synapt G2-S, to determine the diagnostic value between the peptide fragment sources. Results: the first part (1) of 109 cases of LC patients with a total of 29 patients developed HCC (26.6%). Alpha fetoprotein (AFP), alpha fetoprotein (AFP-L3), alanine L3 acid aminotransferase (ALT) three indicators in HCC group were significantly higher than non HCC group (t = -2.299, -2.384, -2.847, P 0.05); AFP, AFP-L3 positive for potential risk factors for the development of HCC (relative risk (RR) respectively; 2.41,2.26). Protein 73 (GP73) serum levels decreased significantly in the development of HCC (t=2.212; p=0.041). (2) three early diagnosis model for HCC training and testing sample classification accuracy are as follows: the MLP model was 66.7% and 63.6%, 61.4% and 62.1% for the RBF model, DA model and 60.7%. (61.6% 3) single factor analysis showed that only A preoperative GP73 concentration have predictive value for the patients after operation (X2=4.695, P=0.03), COX multi factor analysis showed that the AFP-L3, prothrombin time (Pt (s), prothrombin activity (Pt) (a)) is an independent predictor of recurrence after surgery (hazard ratio (HR) respectively for 1.003,3.465,1.137. The p value was 0.05). The second part (1) by ClinPro Tools analysis, we found 27 differentially expressed serum peptide peaks, in the serum of patients with benign and malignant hepatic tumors in GA, SNN, heavy QC model recognition rates were 100%, 89.38%, 80.84%, the cross validation rate were 81.67%, 81.11%, 86.11%. Blind sample validation accuracy were 78%, 84%, 84%. (2) with molecular weight of 2860.34Da, 2881.54Da, 3155.67Da between peptide peaks were successfully identified, they are FIBA, FIBB and ITIH4 fragment. Conclusion: our study confirmed by follow-up detection of AFP, AFP-L3, GP73, ALT levels of HCC in early stage early warning, The detection of GP73, AFP, AFP-L3, Pt (s), Pt (a) serum levels can predict the postoperative recurrence of HCC; we establish the early model and protein fingerprint identification model of liver cancer diagnosis based on serum markers of hepatocellular carcinoma, and through tandem mass spectrometry showed that FIBA, FIBB, ITIH4 three serum peptides may as a new tumor marker.

【学位授予单位】：中国人民解放军医学院
【学位级别】：博士
【学位授予年份】：2016
【分类号】：R735.7;R730.43

【相似文献】