应用SELDI-TOF-MS技术建立小肾癌的血清蛋白质谱筛选模型

Proteomic analysis of serum identifies protein changes present in small renal cell carcinoma using surface enhanced laser desorption/ionization mass spectrometry

目的探讨SELDI-TOF-MS技术在小肾癌的筛查诊断中的应用。方法收集了111例血清标本,分别为小肾癌患者30例,肾脏良性肿瘤34例,健康人47例。应用SELDI-TOF-MS技术获得所有患者的IMAC-Cu2+蛋白芯片的表达图谱,随机采用19例小肾癌患者和26例健康人的血清蛋白质谱建立决策树模型,并用其余11例小肾癌患者和21例健康人进行双盲验证;再随机采用21例小肾癌患者和16例肾脏良性肿瘤患者的血清蛋白质谱建立决策树诊断模型,并用余下9例小肾癌和18例肾脏良性肿瘤患者的血清标本进行双盲验证。结果小肾癌与健康人的决策树诊断模型的敏感性和特异性均为100%,双盲验证后的敏感性和特异性分别为81.8%(9/11)和100%(21/21)。小肾癌与肾脏良性肿瘤的决策树诊断模型的敏感性为95.2%(20/21),特异性为100%(16/16),双盲法验证后的敏感性为77.8%(7/9),特异性为88.9%(16/18)。结论本次试验中建立的两个决策树有望通过进一步的验证用于肾癌的早期诊断和鉴别诊断中。在质荷比分别为15282.4、4215.96这2个蛋白峰中可能会发现肾癌的特异性肿瘤标记物。

Objective To identify the proteomic changes in small （≤3 cm in diameter） renal cell carcinoma （RCC） using surface enhanced laser desorption/ionization mass spectrometry technology. Methods One hundred and eleven serum samples were collected, including 30 small RCC cases, 34 renal benign mass cases and 47 healthy persons. The samples were analyzed using IMAC-Cu^2＋ ProteinChip system by SELDI-TOF-MS technology. Two decision trees were generated by Biomark Pattern 5.0.2 software to distinguish small RCC cases from healthy persons and from benign renal masses cases, respectively. Double blind confirmation was applied to each decision tree. Results Both the sensitivity and specificity of the decision tree, which distinguishes small RCC cases from healthy persons, were 100%. The sensitivity and specificity of double blind confirmation procedure were 81.8% （9/11） and 100%（21/21）, respectively. The sensitivity and specificity of the second decision tree, which distinguish small RCC cases from benign renal masses cases, were 95.2 % （20/21） and 100% （16/16）, respectively. After the double blind confirmation, the sensitivity and specificity were 77.8%（7/9） and 88.9% （16/18）, respectively. Conclusions Two decision trees have been established in this experiment, which can hopefully be used for early diagnosis and deferential diagnosis of small RCC. Potential biomarkers of RCC can possibly be found from 2 proteins with molecular masses of 15 282.4 and 4 215.96.