| Objective: Due to small lesions and early limitation of the lymphoma, the lesions detection rate is low, At the same time, the occult lesions or lesions that hidden deep in the body and patients can not tolerate or unwilling to Surgical biopsies , cause the enormous difficulty for clinical diagnosis.We used the SELDI technology to detect serum protein of lymphoma patients and normal controls. protein abundance and quality of the Netherlands, Belgium, (M/S) were analysis by the Biomarker Wizard 3.1 software,which expecte to selecte specific serum biomarkers used in early diagnosis and Condition monitoring of clinical lymphoma, and further serum protein peaks with significant differences of the lymphoma patients , normal controls and non-lymphoma group (lymphadenitis and other tumor) were preferred and established a database, which combinate artificial neural network (ANN) model to establish the diagnostic model. It was breakthrough for application of traditional biopsy techniques diagnosed lymphoma .Method: The serum were detected by surface-enhanced laser desorption ionization time of flight mass spectrometry (SELDI-TOF-MS) in samples of 55lymphoma patients, 108 other tumors patients, 24 lymphadenitis patients and 53 normal controls . CM10 protein chips were used to obtain the expression of serum protein fingerprinting, and test data analyzed by Biomarker Wizard 3.1 software were statistical analysised (mean values, P-values and standard deviation) , P﹤0.05 was considered that the difference has statistics significance.the optimal 9 kind of proteins were Selected from Serum protein fingerprinting with a significant difference in the quality of the Netherlands, Belgium of the lymphoma group, the normal controls and non-lymphoma group (lymphadenitis and other tumor), which apply the BP artificial neural networks to establish lymphoma neural network diagnostic model.Results: 1. we found 76 differentially expressed protein peaks (P <0.05) by comparison of serum protein fingerprint analysis in samples of 55 lymphoma patients, 108 other tumors patients, 24 lymphadenitis patients and 53 normal controls,And 69 proteins have the significance difference(P<0.001).13 protein peaks are highly expressed in lymphoma, 6 protein peaks are lowly expressed in lymphoma. There are 11 protein peaks which are highly expressed and 7 protein peaks lowly expressed in lymphadenitis. 12 protein peaks are highly expressed and 4 protein peaks are lowly expressed in other tumor; There are 5 protein peaks in lymphoma and other tumors and 7 protein peaks in lymphadenitis lymphoma and other tumors which showed high expression ; 5 protein peak presents the low expression in lymphadenitis lymphoma and other tumors; A protein peak showed only low expression in lymphoma. We screened 9 high-quality protein, 6 of them are highly expressed in serum of lymphoma Patients ,the M/S of high expression protein were 4218,4966,9278,7752,8906,10220 respectively.; 3 of them are lowly expressed in serum of lymphoma Patients ,the M/S of low expression protein were 9216,9237,7714. 2 respectively.2. the diagnostic sensitivity , diagnostic efficiency and diagnostic efficiency of BP artificial neural network model are respectively 87.7% , 98.1% and 97.4% in diagnosis of lymphoma , positive predictive value is81.8% , negative predictive value is 98.6%.Conclusion: 1. Serum protein expression has obvious difference in Lymphoma patients ,normal control group and the non-lymphoma patients 2..Screening of differentially expressed biomarkers establish predictive models for the laboratory diagnosis of lymphoma . 3. SELDI-TOF-MS technology may be an effective tool for diagnosis and Screening serum biomarkers of lymphoma... |
PDF Full Text Request