大肠癌肝转移诊断模型的建立并分析其预测（隐性）肝转移的价值

Colorectal cancer is one of the most common malignancies in the world. Despite treatment improved, postoperative survival rate is still low. The main reason of the treatment failed is liver metastasis. In the early stage of liver metastases, when colorectal cancer cells colonize to the liver, biological actions and proteins has already changed. However, it can't be detected from the imaging methods. We will define this phenomenon as colorectal recessive liver metastasis in our project. Many of different protein will participate in the process of liver metastasis, and the difference of specific protein expression will likely be predicted. Our aim is to find the specific protein by using the technology of 2D Difference Gel electrophoresis among the none liver metastasis, the recessive liver metastasis and the dominant liver metastasis. Furthermore, the diagnosis model of liver metastases will be set up by using the artificial neural network methods. At first, the diagnosis model in mice will be established by using above method and the phenomenon of colorectal recessive liver metastasis will be comfirmed by using thin layer slice pathological examination. The second, the diagnosis model of liver metastases in human will be set up by using the methods of 2D Difference Gel electrophoresis and artificial neural network. The data will follow-up to verify the accuracy. Through this topic, our aim is to set up the diagnosis method of liver metastasis in colorectal cancer and explore the the characteristics of the recessive liver metastases.

大肠癌是最常见的恶性肿瘤之一，尽管治疗措施不断改进，术后生存率仍较低，治疗失败的主要原因为肝转移。在发生肝转移的初期阶段，现有影像学手段难以检出，本课题将该现象定义为"大肠癌隐性肝转移"。大肠癌肝转移的过程是多种蛋白协同参与完成的，而特异蛋白的差异表达将可能是其预测的风向标。本项目拟采用蛋白组学方法分离大肠癌无肝转移、隐性肝转移和显性肝转移患者血清蛋白质，筛选并鉴定肝转移相关差异蛋白质，利用数据挖掘方法构建大肠癌肝转移的诊断模型。首先使用该方法构建小鼠大肠癌肝转移诊断模型，通过全肝脏薄层切片病理检查的方法验证"大肠癌隐性肝转移"假说，为临床实验提供理论和实验依据；接着使用上述蛋白组学和数据挖掘的方法构建人大肠癌肝转移诊断模型，横断面盲法验证其准确性，并前瞻性研究其预测隐性肝转移的价值。本课题旨在建立诊断大肠癌肝转移的新方法，探索大肠癌隐性肝转移的机制。

大肠癌是最常见的恶性肿瘤之一，尽管治疗措施不断改进，术后生存率仍较低，治疗失败的主要原因为肝转移。在发生肝转移初期阶段，现有影像学手段难以检出，本课题将该现象定义为"大肠癌隐性肝转移"。大肠癌肝转移的过程是多种蛋白协同参与完成的，而差异蛋白将可能是预测的风向标。研究内容包括：1）改良方法构建小鼠大肠癌肝转移诊断模型；2）通过全肝脏薄层切片病理检查，验证“大肠癌隐性肝转移” 假说；3）采用蛋白组学方法分离大肠癌无肝转移、隐性肝转移和显性肝转移间的差异蛋白质；4）筛选并鉴定肝转移相关差异蛋白质；5）利用数据挖掘方法构建大肠癌肝转移诊断模型；6）使用上述蛋白组学和数据挖掘的方法，构建人大肠癌肝转移诊断模型；7）横断面盲法验证其准确性等。研究发现：1）通过Transwell等方法可获得高迁移力的CT26大肠癌细胞；2）改良盲肠造疝原位接种瘤块小鼠大肠癌肝转移模型对比其他同类模型，能更真实模拟大肠癌肝转移生物行为，具有更好的稳定性；3）通过全肝薄层切片病理检查，发现显性肝转移率是27.7%，隐性肝转移率58.3%；通过2D DIGE分离3组小鼠肝脏蛋白质，利用DecyderTM2D 6.5软件分析蛋白差异位点CD组与NC组存在107个，CR组与NC组存在67个，CR组与CD组存在23个；通过全肝脏薄层切片病理检查和2D DIGE等方法验证了“大肠癌隐性肝转移”的假说；4）2D DIGE分离后选取其中55个差异位点进行蛋白质谱分析，挑选Mascot Score值较高的抗体（Annexin V、Prohibitin、spectrin α II、Syne-1、FBP1、PNP），分别进行Western Blot验证并发现Syne-1、FBP1可能与隐性肝转移相关，并使用FBP1、AnnexinV和PNP构建ANN模型；5）使用2D等方法筛选出有/无肝转移的大肠癌患者血清差异蛋白，根据cut-off值以Transferrin和Haptoglobin联合构建的ANN模型的预测准确率最高及约登指数最大，筛选该ANN模型作为大肠癌肝转移诊断模型，具有一定临床诊断价值。通过上述方法，初步论证“大肠癌隐性肝转移”假说，为进一步探索大肠癌肝转移机制提供基础。利用差异蛋白质构建人工神经网络大肠癌肝转移诊断模型具有较高准确度，可为临床早期诊断大肠癌肝转移提供一个有效手段。