Applied Research Of SELDI Protein Chip Technology In Meningitis Classification

As a common, frequently-occurring intracranial infectious disease, meningitis is classified mainly into purulent meningitis, tuberculous meningitis and viral meningitis, whose common clinical manifestations are headache, fever, vomiting and even convulsions, coma, etc. Meningitis may leave serious complications and even death if not treated timely and effectively. Early diagnosis of meningitis relies mainly on clinical manifestations, imaging examinations and laboratory tests. Currently, direct detection of pathogenic microorganisms in cerebrospinal fluid remains a "gold standard" for clinical diagnosis. However, traditional pathogen detection methods do not have a clear suggestive significance for the early diagnosis of meningitis types due to low positive rate and long time consumption. Therefore, establishment of an efficient, rapid, sensitive and specific diagnostic method is in urgent need.OBJECTIVE: To establish a classificatory diagnosis model for three types of meningitis(purulent meningitis, tuberculous meningitis and viral meningitis) by surfaceenhanced laser desorption ionization time-of-flight mass spectrometry(SELDI-TOF-MS) combined with protein microarray technology.METHODS: Cerebrospinal fluid(CSF) samples of 62 meningitis patients(16 cases of purulent meningitis, 18 cases of tuberculous meningitis and 28 cases of viral meningitis) were collected, centrifuged at high speed and stored at a-80℃ refrigerator. Analyses were performed within the same freeze-thaw cycle of the samples. The 62 CSF samples were detected by hydrophobic H50 protein chip technology and SELDI-TOF-MS, and differential protein peaks were identified using Biomarker Wizard software. Data were statistically analyzed using SPSS 16.0 software, and a meningitis classificatory diagnosis model was established using Biomarker Patterns 5.0 software.RESULTS: Altogether 129 protein peaks could be collected, of which 40 were different upon software analysis. Meningitis classification decision tree was successfully created using 7 differentially expressed protein peaks, i.e. m/z=2176.205, m/z=2326.05, m/z= 2436.154, m/z=3399.328, m/z=4125.172, m/z=8927.876 and m/z=14675.06, which had an accuracy of 93.548%(58/62). Meningitis classification accuracy of the model was verified by blind test to be 93.33%(28/30).CONCLUSION: The classification decision tree comprising 7 differential protein peaks provides a new reference for the classification of meningitis.