Occurrence and Prognostic Implications of Lymph Node Micrometastases in Patients with Dukes' B Colorectal Carcinoma

Choi, Yun Young; Choi, Hong Jo; Jung, Ghap Joong; Kim, Sang Soon; Hong, Sook Hee

Abstract

PURPOSE
Whereas lymph node metastases in colorectal carcinoma are important prognostic factor, the prognostic relevance of occult tumor cells in lymph nodes has not been elucidated yet. Aims of this study were to assess the incidence of micrometastases of lymph nodes in patients with Dukes' B colorectal carcinoma and to investigate the correlation between nodal micrometastases and malignant potential to determine whether micrometastases of lymph nodes have prognostic significance, using an anti-cytokeratin immunohistochemical technique.
METHODS
To evaluate the incidence of lymph node micrometastases in patients with Dukes' B colorectal carcinoma, 1160 lymph nodes taken from 65 patients (17.9 per case) were assessed by immunohistochemical technique using a monoclonal anti-human cytokeratin (MNF 116). Clinicopathologic parameters and survival rate were compared between patients with and without micrometastases.
RESULTS
The incidence of nodal involvement by tumor cells in 65 patients with Dukes' B colorectal carcinoma (41 colon, 22 rectum, 2 synchronous cancer) was 30.8% (20 cases) and nodal positivity 3.2% (33/1423 nodes) by the immunostaining. No correlations were observed between the incidence of cytokeratin positivity and various clinicopathologic parameters, including preoperative CEA level, tumor site and size, histologic differentiation, pT stage, vascular invasion and lymphatic invasion and rate of recurrence. There was no difference in five-year survival estimated by Kaplan-Meier lifetable method between the micrometastases negative and positive groups (94.8% and 94.1%, respectively).
CONCLUSIONS
The presence of nodal micrometastases detected by anti-cytokeratin immunohistochemistry is an interesting phenomenon but clinically seems to be of little prognostic value in patients with Dukes' B colorectal carcinoma. Thus, this immunostaining technique does not offer a significant benefit over conventional pathologic staging using hematoxylin-eosin staining.