Cell cycle phase abnormalities do not account for disordered proliferation in Barrett's carcinogenesis

Abstract

Barrett's esophagus (BE) epithelium is the precursor lesion for esophageal adenocarcinoma. Cell cycle proteins have been advocated as biomarkers to predict the malignant potential in BE. However, whether disruption of the cell cycle plays a causal role in Barren's carcinogenesis is not clear. Specimens from the Barrett's dysplasia-carcinoma sequence were immunostained for cell cycle phase markers (cyclin D1 for G1; cyclin A for S, G2, and M; cytoplasmic cyclin B1 for G2; and phosphorylated histone 3 for M phase) and expressed as a proportion of proliferating cells. Flow cytometric analysis of the cell cycle phase of prospective biopsies was also performed. The proliferation status of nondysplastic BE was similar to gastric antrum and D2, but the proliferative compartment extended to the luminal surface. In dysplastic samples, the number of proliferating cells correlated with the degree of dysplasia (P < .001). The overall levels of cyclins A and B1 correlated with the degree of dysplasia (P < .001). However, the cell cycle phase distribution measured with both immunostaining and flow cytometry was conserved during all stages of BE, dysplasia, and cancer. Hence, the increased proliferation seen in Barrett's carcinogenesis is due to abnormal cell cycle entry or exit, rather than a primary abnormality within the cell cycle.