胰腺癌血浆诊断标志物谱的筛选及验证

发布时间：2018-09-07 10:04

【摘要】：胰腺癌(pancreatic cancer, PC)是一种恶性程度很高的肿瘤,其发病率位列肿瘤发病率第十。由于早期发病缺乏明显的症状,没有较好的早期诊断工具难以早期发现,进展期患者没有有效的治疗方式使得胰腺癌的预后很差,五年生存率低于10%。当前胰腺癌常用的诊断标志物CA19-9对早期患者的诊断不够稳定,灵敏度和特异性均不够理想。因此,为了提高胰腺癌患者的生存期急需发现新的诊断标志物。本课题的主要目的就是为胰腺癌寻找有价值的血浆诊断工具,分析鉴定其诊断灵敏度和特异性为临床应用奠定前期基础。本课题采用407例血浆标本(包括胰腺癌、慢性胰腺炎、正常对照、结直肠者、胃癌和肝癌),检测了七个候选标志物和CA19-9的表达情况。七个候选标志物(miR-20A, miR-21, miR-25, miR-155, miR-196A, miR-210和巨噬细胞抑制因子-1(Macrophage Inhibitory Cytokine-1, MIC-1))均有报道在胰腺癌中的高表达,是可能的诊断标志物。本实验设计包含一个训练组、一个验证组和一个独立的双盲实验。训练组的目的是从候选标志物中筛选特异的诊断标志物,并利用筛选出的标志物建立诊断标志物谱的联合方程。研究结果发现：与正常对照相比,所有候选标志物均在胰腺癌患者的血浆中高表达(P0.001),但miR-155, miR-196A, miR-210在胰腺炎患者中亦高表达,与胰腺癌患者的表达水平无显著性差异。只有miR-20A, miR-21, miR-25, MIC-1和CA19-9在胰腺癌中高表达且可以区别胰腺癌和其他消化道肿瘤(包括结直肠癌、胃癌和肝癌),但miR-155, miR-196A, miR-210在其他消化道肿瘤也高表达,对胰腺癌无组织特异性。因此,miR-20A, miR-21, miR-25, MIC-1和CA19-9是在胰腺癌特异性上调表达的标志物,利用多因素logistic回归分析我们建立2个包含多个标志物的诊断方程用于诊断胰腺癌,他们是：Index1=1.971×ln(MIC-1)+1.795×ln(miR-21)+1.020×ln(CA19-9)-42.305;Index2=2.138×ln(MIC-1)+1.772×ln(miR-25)+1.125×ln(CA19-9)-45.006.两个方程用于诊断胰腺癌的阈值均为0。训练组中,比较胰腺癌和所有非胰腺癌,Index1的曲线下面积(AUC)是0.968(95%CI,0.947-0.989), Index2的AUC是0.967(95%CI,0.945-0.989),单独CA19-9的AUC是0.895(95%CI,0.838-0.952)。Index1、Index2和CA19-9的灵敏度分别是0.895、0.895、0.816,特异度分别是0.909、0.915、0.933,准确度分别是0.904、0.908、0.892。可见多个标志物的联合应用能提高单一标志物的诊断灵敏度和准确度。验证组用于鉴定所选择建立的联合方程的诊断价值。验证组中,Index1的AUC是0.933(95%CI：0.874-0.992),而Index2的AUC是0.921 (95%CI:0.856-0.986),CA19-9的AUC是0.793(95%CI,0.687-0.899)。Index1、Index2和CA19-9的灵敏度分别是0.892、0.838、0.649,特异度分别是0.913、0.925、0.913。从训练组和验证组的结果可见,两个联合方程的诊断灵敏度和准确度比单一标志物CA19-9好。双盲实验用于分析所筛选出的联合方程的临床应用可行性。在双盲实验中,两个联合方程与CA19-9的诊断灵敏度和特异度相同。相关性分析发现多数标志物的表达水平与患者的临床指标如性别、年龄、高血压史、糖尿病史等均无相关性。Index1和Index2的表达水平与性别和糖尿病史有一定相关性,但相关系数很小。随访了100例胰腺癌患者的生存状态并分析候选标志物表达与胰腺癌预后之间的相关性。结果发现,6个miRNA、MIC-1、CA19-9和两个联合方程的表达水平与胰腺癌患者的预后均无相关性。本课题的创新点为：采用多种标志物构建联合方程诊断胰腺癌；分析了候选标志物和联合方程的组织特异性；设计上除了训练组、验证组外还有双盲实验组,系统地筛选验证获得诊断联合方程。综上所述,本课题鉴定出两个诊断灵敏度和特异性较好的胰腺癌特异诊断联合方程。为实现其临床应用,还需要深入研究其作为诊断标志物的价值和潜能。
[Abstract]:Pancreatic cancer (PC) is a highly malignant tumor with the tenth highest incidence. Because of the lack of obvious symptoms in the early stage, it is difficult to detect early without good early diagnostic tools. The current diagnostic marker of pancreatic cancer, CA19-9, is not stable, sensitive and specific enough for early diagnosis. Therefore, in order to improve the survival time of pancreatic cancer patients, it is urgent to find new diagnostic markers. In this study, 407 plasma specimens (including pancreatic cancer, chronic pancreatitis, normal control, colorectal cancer, gastric cancer and liver cancer) were used to detect the expression of seven candidate markers and CA19-9. Seven candidate markers (Mi-20A, Mi-21, Mi-25, Mi-155, Mi-196A, Mi-210 and Giant) Macrophage Inhibitory Cytokine-1 (MIC-1) has been reported to be highly expressed in pancreatic cancer and is a potential diagnostic marker. This study included a training group, a validation group and an independent double-blind trial. The aim of the training group was to screen specific diagnostic markers from candidate markers and utilize them. The results showed that all the candidate markers were highly expressed in plasma of pancreatic cancer patients (P 0.001), but the expressions of microRNAs-155, microRNAs-196A and microRNAs-210 were also high in pancreatitis patients. There was no significant difference in the expression level between pancreatic cancer patients and pancreatic cancer patients. R-20A, Mi-21, Mi-25, MIC-1 and CA19-9 are highly expressed in pancreatic cancer and differentiate pancreatic cancer from other gastrointestinal cancers (including colorectal cancer, gastric cancer and liver cancer), but Mi-155, Mi-196A, and Mi-210 are also highly expressed in other gastrointestinal cancers and are not tissue-specific for pancreatic cancer. Using multivariate logistic regression analysis, we established two diagnostic equations containing multiple markers for the diagnosis of pancreatic cancer: Index1 = 1.971 *ln (MIC-1) + 1.795 *ln (Mi-21) + 1.020 *ln (CA19-9) - 42.305; Index2 = 2.138 *ln (MIC-1) + 1.772 *ln (Mi-25) + 1.125 *ln (CA19-9) - 45.006. In the training group, the AUC of Index1 was 0.968 (95% CI, 0.947-0.989), the AUC of Index2 was 0.967 (95% CI, 0.945-0.989), and the AUC of CA19-9 was 0.895 (95% CI, 0.838-0.952). Index1, Index2, and CA19-9 were 0.895 (0.895, 0.895, 0.895, 0.8, 0.989) respectively. 16, specificity 0.909, 0.915, 0.933, accuracy 0.904, 0.908, 0.892, respectively. It can be seen that the combination of multiple markers can improve the diagnostic sensitivity and accuracy of a single marker. The ACU of CA19-9 was 0.793 (95% CI, 0.687-0.899). The sensitivity of Index1, Index2 and CA19-9 were 0.892, 0.838, 0.649, and the specificity was 0.913, 0.925, 0.913, respectively. The results of the training group and the verification group showed that the diagnostic sensitivity and accuracy of the two combined equations were better than that of the single marker CA19-9. In the double-blind experiment, the sensitivity and specificity of the two equations were the same as that of CA19-9. Correlation analysis showed that the expression levels of most of the markers were not correlated with clinical parameters such as sex, age, history of hypertension and diabetes mellitus. The survival status of 100 patients with pancreatic cancer was followed up and the correlation between the expression of candidate markers and the prognosis of pancreatic cancer was analyzed. The innovations of this study are as follows: using multiple markers to construct the joint equation for the diagnosis of pancreatic cancer; analyzing the tissue specificity of candidate markers and joint equation; designing a double-blind experimental group besides training group and validation group, and systematically screening and validating to obtain the joint equation of diagnosis. In order to realize its clinical application, it is necessary to study its value and potential as a diagnostic marker.
【学位授予单位】：北京协和医学院
【学位级别】：博士
【学位授予年份】：2015
【分类号】：R735.9

【相似文献】