Combining proteomics, serum biomarkers and bioinformatics to discriminate between esophageal squamous cell carcinoma and pre-cancerous lesion.

OBJECTIVE

Biomarker assay is a noninvasive method for the early detection of esophageal squamous cell carcinoma (ESCC). Searching for new biomarkers with high specificity and sensitivity is very important for the early detection of ESCC. Serum surface-enhanced laser desorption/ionization-time of flight mass spectrometry (SELDI-TOF-MS) is a high throughput technology for identifying cancer biomarkers using drops of sera.

METHODS

In this study, 185 serum samples were taken from ESCC patients in a high incidence area and screened by SELDI. A support vector machine (SVM) algorithm was adopted to analyze the samples.

RESULTS

The SVM patterns successfully distinguished ESCC from pre-cancerous lesions (PCLs). Also, types of PCL, including dysplasia (DYS) and basal cell hyperplasia (BCH), and healthy controls (HC) were distinguished with an accuracy of 95.2% (DYS), 96.6% (BCH), and 93.8% (HC), respectively. A marker of 25.1 kDa was identified in the ESCC patterns whose peak intensity was observed to increase significantly during the development of esophageal carcinogenesis, and to decrease obviously after surgery.

CONCLUSIONS

We selected five ESCC biomarkers to form a diagnostic pattern which can discriminate among the different stages of esophageal carcinogenesis. This pattern can significantly improve the detection of ESCC.