Expression of microRNAs by array analysis provides unique profiles for classifying tissues and tumors. The purpose of our study was to examine microRNA expression in Barrett esophagus and esophageal cancer to identify potential markers for disease progression.MicroRNA was isolated from 35 frozen specimens (10 adenocarcinoma, 10 squamous cell carcinoma, 9 normal epithelium, 5 Barrett esophagus, and 1 high-grade dysplasia). MicroRNA expression was analyzed with Ambion bioarrays (Ambion, Austin, Tex) containing 328 human microRNA probes.Unsupervised hierarchic clustering resulted in four major branches corresponding with four histologic groups. One branch consisted of 7 normal epithelium samples and 1 squamous cell carcinoma sample. The second branch consisted of 7 squamous cell carcinoma samples and 1 normal epithelium sample. The third branch contained 4 Barrett esophagus samples and 1 squamous cell carcinoma sample. The fourth contained all the adenocarcinoma samples and 1 sample each of Barrett esophagus, normal epithelium, squamous cell carcinoma, and high-grade dysplasia. Supervised classification with principal component analysis determined that the normal epithelium samples were more similar to the squamous cell carcinoma tumors, whereas the Barrett esophagus samples were more similar to adenocarcinoma. Pairwise comparisons between sample types revealed microRNAs that may be markers of tumor progression. Both mir_203 and mir_205 were expressed 2- to 10-fold lower in squamous cell carcinoma and adenocarcinomas than in normal epithelium. The mir_21 expression was 3- to 5-fold higher in both tumors than in normal epithelium. Prediction analysis of microarray classified 3 Barrett esophagus samples as Barrett esophagus, 1 as adenocarcinoma, and 1 as normal epithelium.Expression profiles of miRNA distinguish esophageal tumor histology and can discriminate normal tissue from tumor. MicroRNA expression may prove useful for identifying patients with Barrett esophagus at high risk for progression to adenocarcinoma.

MicroRNA expression profiles of esophageal cancer