脑脊液蛋白质指纹图谱技术在结核性脑膜炎诊断中的作用

Application of Cerebrospinal Fluid Protein Fingerprint Model in Diagnosis of Tuberculous Meningitis

[目的]研究脑脊液蛋白质指纹图谱技术在结核性脑膜炎(tubercular meningitis,TBM)诊断中的作用。[方法]应用表面增强激光解吸电离飞行时间质谱技术(surface-enhanced laser desorption/ionization time of flight mass spectrometry,SELDI-TOF-MS)检测80例脑脊液标本,其中TBM30例、化脓性脑膜炎30例及同期排除中枢感染的上呼吸道感染病例20例,用Biomarker Wizard软件分析找出差异蛋白峰,蛋白峰差异的比较使用Biomarker Patterns 5.0软件进行统计判别,建立决策树诊断模型,并通过盲法验证评价其准确性和有效性。[结果]蛋白质组谱图分析发现有9个差异显著的蛋白峰(P<0.01),4个在TBM组中高表达,m/z分别为15920.00,32540.50,7967.66,8046.85;有4个在TBM组中低表达,其m/z为8762.66,9411.40,6840.92,6649.47。由m/z为8046.85,15920.00及32540.50建立的决策树模型对TBM的诊断准确率为90%。通过盲法验证模型对TBM的诊断准确率为85%。[结论]由3个差异蛋白峰构成的TBM蛋白质指纹图谱诊断模型为TBM的诊断提供了新的借鉴和参考。

[Objective] To find new biomarkers and to establish cerebrospinal fluid protein fingerprint model for early diagnosis of tuberculous meningitis. [Methods] 80 cerebrospinal fluid samples（including 30 cases of tuberculous meningitis, 30 cases of purulent meningitis, and 20 excluded central infection） were tested by surface enhance laser desorption/ionization time of flight-mass spectrometry（SELDI-TOF-MS） .The data of spectra were analyzed by bioinformat-ics tools like Biomarker Patterns 5.0 and discriminant analysis to establish diagnostic model. [Results] 9 protein features were stably detected in 80 cere-brospinal fluid samples, in which 4 protein features had higher expression（m/z 15920.0, 32540.5,7967.66 and 8046.85, respectively） and lower expression（m/z 8762.66, 9411.4, 6840.92 and 6649.47, respectively）. The detective model combined with 3 biomarkers（m/z 8046.85,15920.0 and 32540.5） could differentiate tuberculous meningitis with accuracy of 90%. The double blind validation showed an accuracy of 90%. [Conclusion] The detective model established by 3 protein features may be a novel method for diagnosis of tuberculous meningitis.