| [ Summary ] |
Aim ; The aim of this study was to clarify whether serrated adenoma (SA) is neoplastic or not, compared with hyperplastic polyp and villous adenoma. Subject and method ; All lesions were completely resected endoscopically and the histological diagnosis confirmed between July 1992 and December 1999, at the National Cancer Center Hospital East. The subjects consisted of 101 lesions, 201 hyperplastic polyps, and 109 villous adenomas. SA lesions were endoscopically classified into two sub-groups: 56 serrated hyperplastic lesions (SH) and 16 serrated villous lesions (SV), and 29 SH+SV lesions. Immunohistochemical studies were performed using antiproliferative cell nuclear antigen (PCNA) monoclonal antibody, anti-cycline D1 polyclonal antibody, anti-COX-2 polyclonal antibody, and anti-p53 antibody. K-ras point mutations were analyzed by using the PCR-SSCP method. Results : PCNA positive-cells, indicating cell division, were found in the lower part of hyperplastic and SH lesions but in the upper part of SV and villous lesions. More villous adenomas (41.2%) and SV (93.5%) showed positive nuclear staining in cycline D1 than hyperplastic polyps (14.3%) and SH (13.0%). COX-2 protein expression was higher in SV>SH than in hyperplastic polyps. Except for the low incidence rate of 14.3% in hyperplastic polyps, the rates of K-ras point mutations were similar for SH (44.4%), SV (41.7%), and villous adenomas (52.9%). The rate of p53 over expression was low in SV (19.4%), SH (56 %), and hyperplastic lesions (0%) compared with villous adenomas (38.2%). In addition, histological examination revealed that tubular glands similar to SH coexisted in 69% (31/45) of SV lesions. Conclusion : serrated villous lesions are similar to those of villous adenomas and should be considered neoplastic. Serrated hyperplastic polyps had features in common with both hypei-plastic and serrated adenomas suggesting that they represent an intermediate step in the development of some hyperplastic polyps into serrated villous lesions. |