Expansion of the Ki-67 proliferative compartment correlates with degree of dysplasia in Barrett's esophagus.

Background: Barrett's esophagus is a histologically defined premalignant lesion of the esophagus in which normal squamous epithelium is replaced by intestinalized columnar epithelium. In a multistep progression from Barrett's esophagus to fully developed carcinoma, accelerated proliferation may indicate or precede genomic instability and, therefore, may be an important factor in the pathogenesis and/or prediction of malignant transformation. Ki-67 is a nuclear antigen expressed in proliferating cells, (G1, S, G2, and M phases) but not in resting cells (G0 phase). This study was undertaken to determine if Ki-67 expression correlates with the degree of dysplasia and if Ki-67 expression can help to differentiate those patients with or without dysplasia.

Methods: The Ki-67 proliferation fraction in 87 paraffin embedded esophageal biopsies from 43 patients with the Ki-67 antibody (MIB-1) was analyzed using immunohistochemistry. Using a computerized proliferation index program (QNA v2.54, Becton Dickinson Cellular Imaging Systems, Inc., Elmhurst, IL), a Ki-67 score was derived for the luminal surface, upper esophageal crypt, lower crypt, and underlying glandular zone of the columnar-lined esophagus.

Results: Significant differences in Ki-67 scores were noted in each zone among different histologic categories: normal gastric ([NG] n = 17); Barrett's without dysplasia ([ND] n = 17); low grade dysplasia ([LG] n = 21); high grade dysplasia ([HG] n = 14); and adenocarcinoma ([CA] n = 5). The pattern of Ki-67 expression was associated strongly with each histologic category. The percentage of Ki-67 positive nuclei in each mucosal zone statistically separated high grade from low grade dysplasia (P < 0.001). In high grade dysplastic tissues, the Ki-67 positive nuclei were found predominantly on the surface epithelium and upper crypt zones, whereas in low grade dysplasia, the majority of Ki-67 positive nuclei were found in the lower crypt zone. The number of Ki-67 positive nuclei in each mucosal component also was significantly different in Barrett's esophagus without dysplasia when compared with Barrett's esophagus with low grade dysplastic tissues. (P < 0.001) Staining patterns of indefinite for dysplasia by H & E staining separated into several distinct patterns (five LG, seven ND, one NG) whereas six biopsies with low grade dysplasia had a Ki-67 expression pattern more consistent with that of high grade dysplasia.

Conclusion: The Ki-67 staining pattern correlated with histologic findings in Barrett's esophagus and may represent an additional parameter for differentiating patients with or without dysplasia.