应用血清蛋白质指纹谱鉴别结直肠癌与其他恶性肿瘤

Differential diagnosis of colorectal cancer from other malignant tumors by serum protein fingerprinting

目的建立结直肠癌与其他恶性肿瘤的血清蛋白质指纹谱鉴别诊断模型。方法收集血清标本235份,其中结直肠癌58例,乳腺癌、胃癌、食管癌、肝癌、肺癌及肾癌各15例,结直肠良性病30例及正常人57例组成建模组,应用表面增强激光解吸/电离飞行时间质谱检测其蛋白质指纹谱。用Biomarker Patterns软件分析结直肠癌与其他恶性肿瘤患者血清中的差异蛋白后,建立结直肠癌鉴别诊断最优分类树模型。再从同期住院或健康体检人群中随机抽取以上诸病种及正常人血清标本各10例组成测试组,盲法验证该模型对结直肠癌的鉴别诊断效能。结果成功建立了由16种蛋白组成的结直肠癌鉴别诊断最优分类树模型。测试模式下对结直肠癌的鉴别诊断准确率83.8%,灵敏度和总特异性分别为86.2%和83.1%。盲法验证显示,预测总准确率为74.4%。结论应用血清蛋白质指纹谱技术建立的结直肠癌与其他恶性肿瘤的鉴别诊断模型具有较高的敏感性与特异性,值得进一步研究。

[ Objective] To establish the differential diagnostic model for colorectal cancer from other malignant tumors by serum protein fingerprinting. [Methods] A total of 235 serum samples from patients of eoloreetal cancer （n =58）, other malignant tumors （breast, gastric, esophagus, liver, lung and kidney, each of them 15 eases）, coloreetal benign diseases （n =30） and healthy people （n =57） were collected and formed the training group. Their serum protein fingerprinting were read by surface-enhanced laser desorption/ionization-time of flight-mass spectrometry. Using Biomarker Patterns software to analyze the distinct proteins between colorectal cancer and other malignant tumors and create a best decision classification tree model for differential diagnosing eolorectal cancer. Then the model was blindly validated by the protein fingerprinting of the test group which was composed of all of the diseases mentioned above and healthy people （each of them 10 cases）, who were randomly selected at the same period. [ Results ] The best decision classification tree model for differential diagnosis of colorectal cancer was successfully established, which including 16 kinds of proteins. In the testing model, the accuracy of diagnostic was 83.8%, the sensitivity and specificity were 86.2% and 83.1%, respectively. Blinded validation suggested that the total accuracy for prediction was 74.4%. [Conclusion] The serum protein fingerprinting differential diagnostic model for eolorectal cancer yields rather a high sensitivity and specificity, which deserves furfiler research.