【作者】 郭静会；

【导师】 张顺财； 王文静；

【作者基本信息】 复旦大学 ， 内科学， 2010， 博士

【摘要】 一.研究背景和目的胰腺癌是一种恶性程度极高的消化系统恶性肿瘤,发病率在国内外均呈上升趋势,发病隐匿,进展快,预后差。胰腺癌早期诊断困难,确诊时大多已属晚期,手术切除率为10%-20%,其中根治者仅为5%-7.5%,可进行外科手术切除术的患者的5年生存率也仅达15%-40%,最近20年来,胰腺癌患者5年生存率仅从3%增长到4%。胰腺癌的预后与是否早期诊断和选择合理治疗方案密切相关,术前评估是正确选定综合治疗方案及预测根治性手术后生存率和正确制定术后随访计划的关键因素,手术后的局部复发和远处转移是胰腺癌根治术后治疗失败的主要原因,因此深入对胰腺癌分期、局部复发、远处转移的研究,对改善预后具有重要的现实意义。CA19-9是最常用的肿瘤标志物,但对于可以切除的早期胰腺癌诊断价值不大,对于胰腺癌筛查,分期和检测复发敏感性特异性低,因此寻找新的胰腺癌血清标志物非常必要。研究表明,胰腺癌是在环境和遗传因素共同作用下,通过多基因、多步骤、多阶段的复杂的生物学演变而形成的疾病,在这一分子事件过程中包括了抑癌基因功能的丧失、癌基因的活化等。因此,用单个或数个因子检测对胰腺癌的早期诊断、预后评估及随访过程中复发转移的检测不可避免地存在敏感性和特异性的矛盾。蛋白质是基因表达的产物,而肿瘤性疾病从蛋白质的角度看,是一种蛋白质缺陷病,其发生过程中有多种蛋白质会发生异常变化。这种变化不仅包括蛋白表达量的增加或减少,还包括蛋白翻译后的不同修饰,从而导致肿瘤组织或体液中相应蛋白质表达谱的改变。为了深入研究恶性肿瘤的发生发展过程,客观上需要在蛋白质组的水平进行进一步探索。蛋白质组学正是以蛋白质组为研究对象,能够识别鉴定细胞、组织或机体的全部蛋白质,提供一组蛋白质的功能及模式信息,反映细胞内部的遗传特性和外界因素影响的结果。近年来发展起来的表面增强激光解析/离子化飞行时间质谱(surface enhanced laser desorption/ ionization time-of-flight mass spectrometry, SELDI-TOF-MS)是一种新的蛋白质组学研究方法,由于SELDI-TOF-MS技术的高灵敏度及高通量特性,能反映被检测样本中蛋白质的全貌,进而能同时找到多个蛋白质标志物,客观上为肿瘤的早期诊断、预后评估和复发转移的早期检测提供了可能。在本项研究中,我们应用SELDI-TOF-MS技术对胰腺癌、胰腺良性疾病患者和健康正常人的血清进行质谱分析,应用生物信息学统计方法寻找差异蛋白峰和明确差异蛋白峰临床应用价值,应用免疫芯片技术和酶联免疫吸附试验(enzyme linked immunosorbent assay, ELISA)鉴定并验证了血清差异蛋白峰的表达,应用免疫组化方法在胰腺癌组织中进一步验证。二.研究方法采用SELDI-TOF-MS技术应用强阴离子交换芯片(strong anionic exchange chromatography, SAX2)分析58例胰腺癌患者、18例胰腺良性疾病患者和51例健康正常人的血清蛋白质指纹图谱,从中筛选出与胰腺癌相关的血清差异蛋白峰,通过建立决策树诊断模型和Logistic回归模型并盲法验证该模型的可靠性,和CA19-9比较对于胰腺癌诊断的敏感性和特异性,评估差异蛋白峰在胰腺癌术前诊断、分期诊断和术后检测中的价值。应用免疫芯片技术鉴定差异蛋白峰M28068,ELISA方法测定血清差异蛋白C14orf166浓度,免疫组化方法验证C14orf166在胰腺癌组织中的表达。三.研究结果1.58例胰腺癌患者和51例健康人的血清蛋白质指纹图谱比较分析,定义m/z值在2000-30000范围内共检测到61个蛋白峰,其中26个蛋白峰在两组中的差异有明显统计学意义(p<0.001)。交叉证实率最高的10个决策树模型预测胰腺癌正确率为0.929,敏感性为0.833,特异性为1.000；7个差异蛋白峰建立的Logistic回归模型预测胰腺癌的正确率为0.857,AUC为0.976(P<0.001),敏感性为0.914-0.776,特异性为0.922-1.000。2.58例胰腺癌患者和18例胰腺良性疾病患者的血清蛋白质指纹图谱比较分析,定义m/z值在2000-30000范围内共检测到61个蛋白峰,其中16个蛋白峰在两组中的差异有统计学意义(P<0.05)。交叉证实率最高的10个决策树模型预测胰腺癌正确率为0.929,敏感性为0.833,特异性为1.000；3个差异蛋白峰建立的Logistic回归模型预测胰腺癌的正确率为0.857,AUC为0.933(p<0.001),敏感性为0.958-0.708,特异性为0.769-1.000；联合CA19-9预测正确率为0.929,AUC为0.976(P<0.001),敏感性为0.958-0.750,特异性为0.923-1.000；CA19-9诊断胰腺癌的敏感性为0.813,特异性为0.770。3.58例胰腺癌患者和69例非胰腺癌患者的血清蛋白质指纹图谱比较分析,发现5个差异蛋白峰在胰腺癌和另外两组中的差异有统计学意义(p<0.05),5个差异蛋白峰在胰腺癌和正常对照组中诊断胰腺癌AUC为0.763,在胰腺癌和胰腺良性疾病组中诊断胰腺癌AUC为0.865,联合CA19-9诊断胰腺癌AUC为0.971,敏感性为0.923-1.000,特异性为0.927-0.917,预测胰腺癌正确率为1.000。4.6个差异蛋白峰和胰腺癌的分期相关(p<0.01),对ⅠⅡ期、ⅡⅢ期和ⅢⅣ期分别建立的Logistic回归模型对胰腺癌的分类能力分别为86.1%、91.9%和77.3%,其AUC值分别为0.897、0.978和0.792(p<0.05),敏感性和特异性分别为0.839-0.871/1.000-0.400、0.903-1.000/0.968-0.833和0.750-0.833/0.938-0.500。5.1个胰腺癌高表达的差异蛋白峰M4016在术后1周、2周、4周和半年随访中峰值强度有一度下降趋势(p<0.05)。6.免疫芯片验证差异蛋白峰M28068为C14orf166,5个胰腺癌患者峰值强度为3.33±1.76,5个正常人对照峰值强度为0.60±0.43。7. ELISA法测定58例胰腺癌患者、18例胰腺良性疾病患者和51例健康正常人的血清中C14orf166浓度分别为24.21±10.42、9.11±4.57和7.78±3.69μg/ml(P<0.001),诊断胰腺癌和正常对照组与胰腺癌和胰腺良性疾病组的AUC分别为0.938和0.917(p<0.001),诊断界值为14.56μg/m1时,诊断胰腺癌的敏感性和特异性分别为0.828/0.922和0.828/0.889。8.免疫组化法测定35例胰腺癌组织和正常胰腺组织C14orf166显示,C14orf166定位在细胞浆,半定量分析胰腺癌组织和正常组织表达量,两者有明显统计学差异(5.05±3.00/0.83±0.95,p<0.001)。四.研究结论1.蛋白质芯片SELDI-TOF-MS技术重复性高、稳定性好,是比较理想的蛋白质组学研究技术平台；样品的质量控制和操作过程的标准化将直接影响实验结果的可靠性。2.应用SELDI-TOF-MS技术可以从血清中筛选出胰腺癌相关的标志蛋白。建立的胰腺癌诊断预测模型可能对早期诊断胰腺癌具有重要的意义和对胰腺癌的预后评估有一定价值,同时可以评估疗效,早期发现肿瘤复发。3.免疫芯片技术可以作为SELDI-TOF-MS筛选出的差异蛋白峰的验证方法。4.鉴定出的差异蛋白C14orf166有望在胰腺癌的特异性诊断中发挥作用,为今后胰腺癌的发病机制研究奠定了基础。目的：探讨趋化因子受体CCR7在人胰腺癌组织中的表达与淋巴结转移之间的关系以及与趋化因子受体CXCR4和血管内皮生长因子VEGF-C之间的关系。方法：应用荧光实时定量PCR (realtime polymerase chain reaction)方法,检测CCR7、CXCR4和VEGF-C在24例胰腺癌组织及正常组织样本中的基因水平；应用免疫组化方法,检测CCR7、CXCR4和VEGF-C在65例胰腺癌组织及癌旁组织样本中的表达,IPP图像分析软件计算平均光密度和累积光密度比较蛋白表达差异。结果：(1)实时定量PCR分析显示CCR7和VEGF-C在伴有淋巴结转移的胰腺癌组较淋巴结阴性组基因相对表达量明显增加(P<0.05), CXCR4在两组中基因相对表达量无明显差异(P>0.05), CCR7、CXCR4和VEGF-C在癌组织较正常组织基因相对表达量明显增加(p<0.05)。(2)免疫组化分析显示CCR7平均光密度值在伴有淋巴结转移的胰腺癌组较淋巴结阴性组明显高(P<0.001),在胰腺癌组织中CXCR4和VEGF-C平均光密度值与是否淋巴结转移无关(p>0.05)；CCR7和VEGF-C累积光密度在伴有淋巴结转移的胰腺癌组较淋巴结阴性组明显高(p<0.05), CXCR4累积光密度与是否淋巴结转移无关(P>0.05),CCR7和CXCR4平均光密度和累积光密度在胰腺癌组织中较癌旁组织低(P＜0.05), VEGF-C在两组中无差异(P>0.05)。结论：CCR7和VEGF-C可以作为评估胰腺癌淋巴结转移的一个观测指标,CXCR4与胰腺癌是否淋巴结转移无明显相关性。更多还原

【Abstract】 一. Background and objectivePC is one of the most malignant tumors in digestive system, due to its rising incidence, difficulty in early diagnosis, rapid progress and poor prognosis.Currently, there are no ideal methods for the early detection of PC. Only 10-20% of patients (5-7.5% can be under radical correction) are resectable at the time of diagnosis and 5-year survival rate after operation is still 15-40%.The early diagnosis and right treatment are closely related to death. The preoperative assessment is one key factor of the choice of treatment, the prediction of survival and the plan of postoperative follow-up and local recurrence or distant metastasis confirms the failure of treatment. It’s important for the improving of prognosis to achieve the early and right diagnosis and to monitor the recurrence and metastasis. CA19-9 is the most commonly used tumor marker but value little for the early diagnosis, prognosis evaluation and detection of recurrence. So it’s necessary to seek for new biomarkers.By now, PC is proved not a single disease, but an accumulation of sophisticated biological evolutions include polygene, muti-procedure and multistage under the role of environment and heredity. Also in the procedure exists the function lost of cancer suppressor gene and activation of oncogene. Therefore, the attempt of early diagnosis, prognosis evaluation and monitoring for local recurrence and/or metastasis by the detection of single or several factors has unavoidablely the conflict between the sensitivity and specificity. Protein is the production of gene expression and tumor is believed to be a disease of protein’s pitfall. Numerous of proteins changed during the included different modifications after protein translation as well as the expression level, resulted in the change of protein expression profiles of tumor tissue or body fluid. Therefore there is much need in the proteome level for further research of the development of malignant tumor. It detects the functional units of expressed genes using proteomic methods to analyze celluar proteins and provide a protein fingerprint. The proteomic reflects both the intrinsic genetic programme of the cell and the impact of the immediate environment and is therefore valuable in biomarker discovery. Surface enhanced laser desorption/ionization time-of-flight mass spectrometry (SELDI-TOF-MS) is a new method of proteomics developed in recent years. Not only because of its high sensitivity and high throughput, but also its ability of reflecting the full protein profiles of tested sample and eventually finding the protein biomarkers, it is a promising method for early diagnosis, prognosis evaluation and monitoring of local recurrence and metastasis of tumorous diseases.Serum samples were applied to strong anionic exchange chromatography (SAX2) protein chips for protein profiling by SELDI-TOF-MS to distinguish PC, Pancreatic benign disease and normal healthy control. Statistical Methods in Bioinformatics were used to analyze the multiple protein peaks. The valuable protein peak was identified using SELDI Immunoassay and enzyme linked immunosorbent assay (ELISA) and its tissues expression was further confirmed.MethodsSerum samples obtained from subjects with PC (n=58), Pancreatic benign diseases (n=18) and normal healthy control (n=51) were applied to SAX2 chips for protein profiling by SELDI-TOF-MS to screen multiple serum biomarkers. The decision tree or logistic regression classification models were constructed to diagnose PC, evaluate the prognosis and curative effect, compared with CA19-9. The valuable protein peak was identified using SELDI Immunoassay and further confirmed by ELISA and immunohistochemical analysis.Results1. Sixty-one qualified protein peaks between 2 000 and 30 000 m/z ratios were detected and significant differences were detected in the levels of 26 serum protein biomarkers between PC patients (n=58) and the healthy controls (n=51) (P<0.001). The top 10 decision trees with the highest correct validation rate were chosen to establish the classification tree model, which had the high positive predictive value of 0.929, a sensitivity of 0.833, and a specificity of 1.000. It was significantly better to use a combination assay of seven protein biomarkers (AUC,0.976; P<0.001) in diagnostic power, resulting in a specificity of between 0.922 and 1.000 and a corresponding sensitivity of between 0.914 and 0.776. Application of this logistic model classification using combinations of the seven protein biomarkers gave diagnostic accuracies of up to 0.857 in the independent testing set.2. Significant differences were detected in the levels of 16 serum protein biomarkers between PC patients and Pancreatic benign disease patients (P<0.05). The top 10 decision trees with the highest correct validation rate between 0.90 and 0.95 were chosen to establish the classification tree model, which had the high positive predictive value of 0.929, with a sensitivity of 0.833 and a specificity of 1.000. It was significantly better to use a combination assay p of three protein biomarkers (AUC,0.933; P<0.001) in diagnostic power, resulting in a specificity of between 0.769 and 1.000 and a corresponding sensitivity of between 0.958 and 0.708. Application of this logistic model classification using combinations of the three protein biomarkers gave diagnostic accuracies of up to 0.857. However, there was a strong trend for a superior discrimination of PC from Pancreatic benign disease combining SELDI profiling and CA19-9 (P<0.001). The combination p1 of CA19-9 and the discriminating peaks had an AUC of 0.976 (P<0.001), resulting in a specificity of between 0.923 and 1.000 and a corresponding sensitivity of between 0.958 and 0.750, and diagnostic accuracies of up to 0.929. So the classification tree model was superior to CA19-9 in the discrimination of PC from Pancreatic benign disease. The latter had a sensitivity of 0.813 and a specificity of 0.770.3. The five most discriminating protein biomarkers were detected through the comparison of the PC group and the noncancer group (i.e. healthy controls and the Pancreatic benign disease group), which yielded an AUC of 0.763 between the group of PC and healthy controls and 0.865 between another group of PC and Pancreatic benign disease. There was a strong trend for a superior discrimination of PC from Pancreatic benign disease when combining SELDI profiling and CA19-9 (P<0.001). The combination p of CA19-9 and the five discriminating peaks had an AUC of 0.971 resulting in a sensitivity of between 0.923 and 1.000 and a specificity of between 0.927 and 0.917.4. A panel of the six most discriminating protein biomarkers could classify patients with PC in different stages (P<0.01). There was a significant improvement when predicting different PC stages by combining the SELDI six protein peaks. The AUC were 0.897 (between stage I and stageⅡ),0.978 (between stageⅡand stageⅢ), and 0.792 (between stageⅢand stageⅣ) (P<0.05) in the diagnosis of different PC stages resulting in a sensitivity of between 0.839-0.871/0.903-1.000/0.750-0.833 and a specificity of 1.000-0.400/0.968-0.833/0.938-0.500.5. There was a down-regulated trend (P<0.05) in the most discriminating protein biomarker(M4 016)through analyzing the pre-operative cancer group and the postoperative cancer group (1,2,4 weeks, and 6 months after operation).6. We performed a SELDI-based immunoassay with a specific anti-C14orf166 monoclonal antibody on 10 pancreatic serum samples:five for which the 28 068 Da peak was present and five for which a 28 068 Da peak was nearly absent on the SAX2 chip. We found that a specific peak of mean mass at 28 068 Da with an intensity of 3.33±1.76 was present in all five samples that displayed a peak on the SAX2 chip, and the peak had an intensity of 0.60±0.43 in the other five samples.7. In 127 individual serum samples including 58 PC,18 Pancreatic benign diseases, and 51 healthy controls, C14orf166 levels were detected. C14orf16 serum concentrations were significantly higher in patients with PC (24.21±10.42μg/ml) than in patients with Pancreatic benign diseases (9.11±4.57μg/ml) and healthy controls (7.78±3.69μg/ml) (P<0.001). There was no statistically significant difference when comparing levels in patients with Pancreatic benign diseases and healthy controls (P= 0.34). The sensitivity and specificity of a serum C14orf166 level of 14.56μg/ml for predicting PC in healthy controls was 0.828/0.922 and 0.828 /0.889 compared with benign disease, respectively, which yielded an AUC of 0.938/0.917 (P＜0.001).8. We analyzed C14orf166 expression by immunohistochemistry in patients undergoing pancreatectomy including 35 PC. Semi-quantitative scoring showed that C14orf166 expression levels of cancer cells (5.05±3.0) were significantly higher than in the normal pancreas (0.83±0.95, P<0.001). C14orf166 protein was localized in the cytoplasm of tumor cells.Conclusions1. SELDI-TOF-MS was an ideal technological platform for proteomic research because of high reproducibility and stability. Quality control and standardization conditions could be key issue for the reliability of outcome.2. The results suggest that SELDI-TOF-MS serum profiling is helpful to the diagnostic, prognostic or therapeutic effects of PC, which is superior to CA19-9.3. SELDI immunoassay is useful for the identification of the significant protein peak screened by SELDI-TOF-MS.4. The identified protein biomarker C14orf166 is a potential biomarker of PC and lays the foundation for the pathogenesis research. Objective:To investigate the relationship of chemokine receptor CCR7, CXCR4, VEGF-C and the lymph node metastasis of PC. Methods:The transcription levels of CCR7, CXCR4 and VEGF-C in PC(n=24) were measured by Real-time PCR, the expressions of CCR7, CXCR4 and VEGF-C were measured by immuohistochemistry(n=65). The professional software of pathological image manipulation (Image Pro Plus 6.0, IPP 6.0)was used to quantitate the results of the immunohistochimical staining including mean density and integrated option density (IOD). Results:The transcription of CCR7 and VEGF-C in PC with lymph node metastasis increased compared with PC without lymph node metastasis (P<0.05), but that of CXCR4 was irrelevant to lymph node metastasis (P>0.05). The transcription of CCR7, CXCR4 and VEGF-C in PC increased compared with the normal pancreas (P<0.05). The mean dentisy of CCR7 in PC with lymph node metastasis increased compared with PC without lymph node metastasis(P<0.05), but that of CXCR4 and VEGF-C was irrelevant to lymph node metastasis(P>0.05). The IOD of CCR7 and VEGF-C in PC with lymph node metastasis increased compared with PC without lymph node metastasis (P<0.05), but that of CXCR4 was irrelevant to lymph node metastasis (P>0.05). The mean density and IOD of CCR7 and CXCR4 in PC decreased compared with the adjacent pancreas (P<0.05) but there was no diffirence in both group for VEGF-C(P>0.05). Conclusions:CCR7 and VEGF-C not CXCR4 seem to play a pivotal role in lymph node metastasis of PC.更多还原