胶质瘤脑脊液蛋白指纹图质谱仪分析及其在临床诊断中的应用

Fingerprint pattern analysis of glioma CSF protein with mass spectrometry and its clinical significance

目的建立区分胶质瘤和非肿瘤脑脊液的蛋白指纹图诊断模型,利用蛋白组学技术寻找新的肿瘤标志物。方法收集胶质瘤和轻中度脑外伤病人(共46份)的脑脊液,利用表面加强解析/电离吸收-时间飞行质谱(surface-enhancedlaserdesorption/ionizationtime-of-flightmassspectrometry,SELDI-TOF-MS)检测蛋白芯片,得出蛋白指纹图,将46份标本随机分为训练组30例(12例胶质瘤,18例脑外伤)和盲法测试组16例(7例胶质瘤,9例脑外伤),运用Ciphergen公司提供的BiomarkerPatternsTM Software(一种决策树软件)分析收集的数据。结果(1)得到胶质瘤与非肿瘤的蛋白指纹图;(2)利用训练集得出基于决策树的脑脊液蛋白模型,测试集盲法检测,胶质瘤诊断的敏感性和特异性分别为100%(8/8)和88.9%(8/9)。结论建立了胶质瘤的脑脊液蛋白指纹图谱,为以后的胶质瘤蛋白质组学研究奠定了一定基础;建立了区分胶质瘤与非肿瘤的脑脊液蛋白表达质谱诊断模型,盲法检验敏感性和特异性达到100%和88.9%,为胶质瘤的临床诊断提供了一条崭新的途径和方法。

Objective To develop and evaluate a diagnostic model of cerebrospinal protein fingerprint pattern for distinguishing gliomas from non-brain-tumors; to screen the new tumor biomarkers by proteomic technology. Methods Forty-six cerebrospinal samples from gliomas and non-brain-tumors(mild brain traumas) were collected .The peptide maps of these patients were detected by surface-enhanced laser desorption/ionization time-of-flight mass spectrometry(SELDI-TOF-MS).The samples were randomly divided into training set(n=30) and test sets(n=16)and the collected data were analyzed with Biomarker PatternsTM Software provided by Ciphergen Inc.(a software of decision tree algorithms). Results The protein fingerprints of glioma and non-brain-tumor were set up by Ciphergen ProteinChip Software 3.0(Ciphergen), the diagnostic model of cerebrospinal protein profiling for separating gliomas from non-brain-tumors was developed using Biomarker Patterns 3.1 Software(a decision tree algorithm), then the model was challenged with the test set randomly, the sensitivity and specificity were 100% and 88.9% respectively. Conclusions We have developed a cerebrospinal protein fingerprint of glioma for the proteomic fundamental study for glioma, moreover we have yielded a diagnostic model of the protein profiling for distinguishing gliomas from non-brain-tumors, then the diagnostic model was cross-validated blindly, and the sensitivity and specificity were 100% and 88.9% respectively. It has provided a new approach in clinical diagnosis of glioma.