Proteomics Diagonosis Models From Tissue, Serum, And Saliva Of Patient With Oral Squamous Cell Carcinoma

Early detection of oral squamous cell carcinoma (OSCC) is critical to avoiddeformity and malfunction of swallowing, speech, head and neck activity, which mayresult from the radical treatment for a delayed later diagnosis of OSCC, a diseasemoreover usually is marching to recurrence and metastasis and whose 5-year surviverate is only 50%, while, the unique local resection for early stage of OSCC willsurvive 90% of these patients.Rather than the genome, it is proteome who control and participate directly in thephysiologic and pathologic life procedures. In proteomics, the technology ofmicro-array and bioinformatics carved out a new way to effectively seek tumorbiomarkers. Surface enhanced laser desorption/ionization-time of flight-massspectrometry (SELDI-TOF-MS) is such a new path to provide high throughput proteinprofiling. Combination of ProteinChip array with time-of-flight mass spectrometryoffers the advantages of speed, simplicity, sensitivity and suitability for a comparativestudy. And because of the multifactorial nature of cancer, it is inevitable that acombination of several markers will be necessary. Tons of data could be obtained inone experiment, so, furthermore, sophisticated bioinformatics methods such asdiscriminant analysis, support vector machines (SVM), ANN (artificial neuralnetworks) etc. were needed to collect and analyze as well as pattern recognize thesecomplex data. In this study, the differential proteomics patterns in serum, saliva andtissue between OSCC patients or tissues and normal ones or precancerous lesionrespectively were detected by SELDI-TOF-MS technology, CM10 ProteinChip, andZhejiang University-Cancer Institute-ProteinChip Data Analysis System(ZUCI-PDAS). The stereoscopic protein patterns of OSCC patients as well as detailedinvestigations were obtained and studied to build the OSCC diagnosis models. Thecapability of each peak in distinguishing different groups of data was estimated by the p value of Rank test. The top ten peaks with the smallest p value were selected forfurther analysis. Each of the 1023 combinations of the 10 peaks was analyzed by theleave-one-out crossvalidation SVM. Combinations with the highest accuracy indistinguishing different groups of data were selected as potential biomarkers. TheSVM model with the highest Youden's index was selected as the model for detectingdiseases regarding OSCC. All these bioinformatics studies were integrated in theZUCI-PDAS available at www.zlzx.netSection 1: Differential proteomics patterns in serum regarding OSCCBy applying above technologies to sera from 28 cases of local OSCC, 8 cases ofregionally metastasisferred OSCC, 6 cases of Oral leukoplaque (OLK), and 32 cases ofhealthy volueenteers, 4 diagnostic models were founded, which were presented in thefollowing first and fourth tables with the sensitivity, specificity and accuracy.Section 2: Discrepant proteomics patterns in saliva regarding OSCCAlso by applying above technologies to saliva from 17 cases of local OSCC, 7cases of regionally metastasisferred OSCC, 8 cases of Oral leukoplaque (OLK), and 15cases of healthy volueenteers, 3 diagnostic models were founded, being presented inthe following second and fourth tables with the sensitivity, specificity and accuracy.Section 3: Different proteomics patterns in lesion mucosal tissues and adjacentnormal mucosal tissuesIn addition to applying above technologies, Laser Capture Microdissection wasapplied to local tissues from 21 cases of OSCC and 7 cases of Oral leukoplaque (OLK),3 diagnostic models were founded, being presented in the following third and fourthtables with the sensitivity, specificity and accuracy. Sera proteomic diagnosis models of OSCCSaliva proteomic diagnosis models of OSCCTissue proteomic diagnosis models of OSCCNote: All the Italics represent high expressions in group 0, regular scripts represent highexpression in groupl, blacken writings represent the consistency in the same row.The sensitivities, specificities and accuracies of serum, saliva, and tissue proteomicdiagnosis models in OSCC patients. Summary: Taken together, the SELDI-TOF-MS and LCM technique combined withbioinformatics approaches can not only facilitate the discovery of better biomarkersfor OSCC but also provide a useful tool for molecular diagnosis in the future. The factthat the OSCG diagnosis models of serum, saliva, and tissue of molecular and cellularlevel from these data were found to be of significant sense is highly encouraging.Progressive study of such prognostic biomarkers which were based on tumorphenotype and biologic behavior would allow clinicians not only to diagnose adisease involving OSCC more early but also to select the most efficacious treatmentmodalities.