BACKGROUND AND AIM: Although several molecular biomarkers for esophageal adenocarcinoma (EAC) have been shown to be useful disease indicators, none has been established as a reliable indicator for risk of EAC or have progressed to routine use. The aim was to identify biomarkers of high risk for EAC in patients with Barrett's esophagus (BE). METHODS: Following endoscopic observation by magnified endoscopy with narrow band imaging (ME-NBI), brushing was followed by obtaining biopsy samples from columnar-lined esophagus (CLE) and from EAC lesions of EAC patients, and from age- and sex-matched non-EAC controls with BE. Total RNA was extracted for microarray analysis using Affymetrix GeneChip Human Genome U133 plus 2.0 Array. Real-time-PCR analysis of identified candidate genes was used to confirm the results. RESULTS: Overall, 9 EAC patients and 50 patients with BE were studied. Seventy-nine candidate genes were identified by microarray analysis based on a proportional hazards model (P < 0.005). Six genes exhibited significantly differential expressions in both BE and cancer lesions of the EAC group compared to BE of the controls. In the brushing samples, median CD55 relative expression levels in cancer lesions were highest and decreased in BE of EAC group and BE of the controls, in that order (P < 0.001). CONCLUSION: Over expression of CD55 in brushing samples taken from BE may be associated with the risk of EAC.
BACKGROUND AND AIM: Although several molecular biomarkers for esophageal adenocarcinoma (EAC) have been shown to be useful disease indicators, none has been established as a reliable indicator for risk of EAC or have progressed to routine use. The aim was to identify biomarkers of high risk for EAC in patients with Barrett's esophagus (BE). METHODS: Following endoscopic observation by magnified endoscopy with narrow band imaging (ME-NBI), brushing was followed by obtaining biopsy samples from columnar-lined esophagus (CLE) and from EAC lesions of EACpatients, and from age- and sex-matched non-EAC controls with BE. Total RNA was extracted for microarray analysis using Affymetrix GeneChip Human Genome U133 plus 2.0 Array. Real-time-PCR analysis of identified candidate genes was used to confirm the results. RESULTS: Overall, 9 EACpatients and 50 patients with BE were studied. Seventy-nine candidate genes were identified by microarray analysis based on a proportional hazards model (P < 0.005). Six genes exhibited significantly differential expressions in both BE and cancer lesions of the EAC group compared to BE of the controls. In the brushing samples, median CD55 relative expression levels in cancer lesions were highest and decreased in BE of EAC group and BE of the controls, in that order (P < 0.001). CONCLUSION: Over expression of CD55 in brushing samples taken from BE may be associated with the risk of EAC.