血清蛋白质指纹图谱诊断模型在肾癌中的应用  

Application of serum protein fingerprint model in diagnosis of renal cell carcinoma

作　　者：陈新洲[1] 王玉杰[1] 张琼[2] 王庆荣[2] 余捷凯[3] 

机构地区：[1]新疆医科大学第一附属医院泌尿外科,新疆乌鲁木齐830011 [2]新疆医科大学第一附属医院临床医学检验中心,新疆乌鲁木齐830011 [3]浙江医科大学肿瘤研究所,杭州310009

出　　处：《新疆医科大学学报》2010年第2期141-144,148,共5页Journal of Xinjiang Medical University

摘　　要：目的检测肾癌患者血清蛋白质,筛选特异的蛋白质标记物,构建用于肾癌早期诊断的血清蛋白质指纹图谱模型。方法应用蛋白质芯片CM10及表面增强激光解吸/离子化飞行时间质谱(SELDI-TOF-MS)技术测定168例血清标本(其中肾癌53例,肾良性占位性病变47例,健康志愿者68例)的蛋白质质谱,用随机抽取的118例标本(肾癌38例,肾良性占位病变30例,健康志愿者50例)作为训练组,应用支持向量机进行训练和交叉验证,建立肾癌诊断模型;其余50例标本进行盲法验证。结果利用质荷比分别为5350、4100、3446、5027和6115的5个蛋白峰建立区分肾癌和正常人的诊断模型,其敏感性为94.74%,特异性为92%。盲法验证显示敏感性为93.33%,特异性为88.89%。区分肾癌和肾良性占位病变的诊断模型,敏感性为92.11%,特异性为90%。区分透明细胞癌和其他病理类型的肾癌的诊断模型对透明细胞癌的判别率为92.84%,对其他病理类型的肾癌的判别率为81.82%。结论表面增强激光解吸电离/飞行时间质谱技术结合支持向量机建立肾癌血清蛋白质指纹图谱模型对诊断肾癌具有较高的敏感性与特异性。Objective To detect new biomarkers and to establish a serum protein fingerprint model for early detection and diagnosis of renal cell carcinoma.Methods The serum samples of 53 renal cell carcinoma patients,47 benign renal masses patients,and 68 healthy volunteers were randomly divided into 2 sets：training set（n=118,including 38 renal cell carcinoma patients,30 benign renal masses patients,and 50 healthy volunteers） and test set（n=50）.The fingerprint expressions of protein chips were obtained by using surface enhanced laser desorption/ionization time-of-flight mass spectrometry（SELDI-TOF-MS） and CM10 protein chip.The data of spectra were analyzed by support vector machine（SVM） to establish a diagnostic model.Results Five peaks with the molecular weight of 5350,4100,3446,5027 and 6115 were detected and the detective model combined with 5 biomarkers could differentiate the serum of renal cell carcinoma from that of healthy volunteers with a specificity of 92% and a sensitivity of 94.74%.The diagnostic model combined with 3 biomarkers could differentiate renal cell carcinoma from benign renal masses with a specificity of 90% and a sensitivity of 92.11%.The positive predictive value to differentiate clear cell renal cell carcinoma from the renal cell carcinoma of other types was 92.86%,and the positive predictive value of renal cell carcinoma of other pathological types was 81.82%.Conclusion The predictive models built by the differences of serum protein fingerprint could be a novel,effective,highly specific and sensitive diagnostic tool in renal cell carcinoma.

关 键 词：肾细胞癌 表面加强激光解吸电离-飞行时间-质谱技术 蛋白质组 支持向量机 诊断模型 

分 类 号：R445[医药卫生—影像医学与核医学] R737.11[医药卫生—诊断学]