ABSTRACT
Objective To evaluate the potential differences in serum proteomic profiles between patients with esophageal squamous cell carcinoma(ESCC)and precancerous lesions in order to establish proteomic pattern model for diagnosis of ESCC and precancerous lesions in high risk area,and to investigate its value in screening ESCC.Methods The serum and endoscopic biopsy samples were obtained from 38 normal controls,63 patients with atypical hyperplasia(class Ⅰ 26 cases,class Ⅱ 26 cases,class Ⅲ 11 cases)and 36 patients with advanced esophageal carcinoma.The serum proteomic patterns were examined using surface enhanced laser desorption/ionization time of flight mass spectrometry(SELDI-TOF-MS)and CM10 protein chip.The data was analyzed and disease diagnostic models were established using support vector machine(SVM).The diagnostic model was evaluated and validated by leave one cross validation.Results ①The diagnostic model could differentiate advanced esophageal carcinoma from normal controls with a specificity of 89.47%and a sensitivity of 83.33%.②The results delivered 92.31%,80.77% and 90.91%specificity,and 80.56%,83.33%and 94.44%sensitivity for discrimination of atypical hyperplasia Ⅰ,Ⅱand Ⅲ,respectively,using diagnostic models.③Four(4291,5644,5664,8775)m/z peaks observed repeatedly using diagnostic models.Conclusions The SELDI-TOF-MS and SVM provide a new approach for discrimination of ESCC and precancerous lesions in high risk area.Four(4291,5644,5664,8775)m/z peaks may considered as potential biomarkers which related to the ESCC and esophageal precancerous lesions.