Application Of Saliva And Plasma Protein Fingerprint Model In Diagnosis Of Oral And Maxillofacial Malignant Tumors

Objective: This study aims to explore the presence of informative protein biomarkers in thehuman saliva and serum proteome to discover potential biomarkers in oral and maxillofacialmalignant tumors. We use surface-enhanced laser desorption/ionization-time of flight-massspectrometry SELDI-TOF-MS) to establish the serum and saliva proteins diagnosis model toidentify serum and saliva protein biomarkers specific for oral malignant tumors.Methods：A total of53patients came from the Oral and Maxillofacial Department of Tongjihospital of Tongji Medical College, Huazhong University of Science&Technology (HUST)including21patients with malignant tumor of oral and maxillofacial region,21patients withsalivary gland tumors,11patients with epulis and sebaceous cyst. The serum and saliva werecollected before operation, after high speed centrifugalized, the samples were frozen andpresented in-80℃refrigirater until thawed specifically for SELDI-TOF-MS analysis.Proteomic fingerprinting of serum were identified and analyzed by Biomarker WizardSoftware (BWS) and Biomarker Patterns Software(BPS), SPSS16.0was used for statisticalanalysis. After the diagnosis model has been established with BPS software, a decision treemodel which can differentiate oral and maxillofacial malignant tumor from the benign tumorcontrol was selected.Results: We have detected76protein peaks, including3peaks which have significantdifference between oral malignant tumors and the benign tumor control by analyzing serumsamples. The sensitivity and specificity of the detective model which differentiate malignanttumors from the benign control were up to96.9%and86.7%respectively. Furthermore, we have detected148protein peaks including6protein peaks which have significant differencebetween oral malignant tumors and the benign tumor control by analyzing saliva samples.The sensitivity and specificity of the detective model which differentiate malignant tumorsfrom the benign control were81.0%and90.3%respectively.Conclusion：Plasma and saliva protein fingerprint models are novel effective screeningdiagnosis method for oral and maxillofacial malignant tumor. Our discriminative modelcould distinguish malignant tumors between the benign tumor controls with encouragesensitivity and specificity. Potentially, it could be used for early diagnosis of oral andmaxillofacial malignant tumors in the future.