Is genomic evaluation feasible in endoscopic studies of Barrett's esophagus? A pilot study

Background: The global gene expression in Barrett's esophagus (BE) in comparison to adjacent or histologically similar tissues has not been extensively studied. Objective: To test the feasibility of conducting gene arrays in endoscopically obtained mucosal specimens. Design: Cross-sectional feasibility study. Setting: The Houston Department of Veterans Affairs Medical Center. Patients: We collected endoscopic biopsies from BE, normal esophagus, antrum, duodenum, and sigmoid colon from 5 patients with BE. RNA was extracted and subjected to cDNA microarrays and gene expression was compared between BE and control tissues. Reverse transcription-PCR was conducted to confirm some of the findings. Main Outcome Measurements: Gene expression profiles in BE tissues and 4 control sites: squamous esophagus, antrum, duodenum, sigmoid colon. Results: On average, 2 biopsies by disposable jumbo biopsy forceps provided approximately 5 μg required for microarrays. From the original number of 22,283 gene probes, 13,805 genes had a quality score of P < .05 and were subjected to further comparison. BE gene expression clustered most closely with that of antrum and least closely with squamous esophagus. Of the 587 genes that had significantly different expression between BE and duodenum, 246 were upregulated and 341 were downregulated in BE. The expression of genes involved in apoptosis, negative regulation of apoptosis, and inflammatory response was significantly lower in BE compared to squamous esophagus. None of the gene groups were significantly overexpressed in BE compared to squamous esophagus or antrum. The reverse transcription-PCR confirmed the results of microarrays. Limitations: Small sample size. Conclusions: Microarray-based studies are feasible in endoscopically obtained tissues. Differences in gene expression could identify potential markers and shed light on the pathogenesis of BE.