Ki-67-determined growth fraction versus standard staging and grading parameters in colorectal carcinoma. A multivariate analysis.

BACKGROUND. The antibody Ki-67 binds to nuclei in all cell cycle phases except GO and can be used to measure growth fraction. Because proliferative activity has been linked to prognosis in neoplasia, the authors analyzed 100 cases of colorectal carcinoma, each with 3 or more years of follow-up, using Ki-67 immunostaining. METHODS. The Ki-67-positive nuclear area and total nuclear area of carcinoma cells in 20 microscopic fields were measured by computed morphometry. A Ki-67 score (percent positive nuclear area x 100) was calculated. The following characteristics also were recorded for each case: patient age and sex, tumor site and size, modified Dukes' stage, spread beyond bowel wall, lymph node status, tumor grade, histologic type, extramural venous spread, tumor growth pattern, fibrosis, lymphocytic infiltration, and mitotic rate. RESULTS. Ki-67 scores ranged from 1 to 90 (mean, 34.6). Ki-67 scores were higher in Stage A disease (versus Stage B, C, and D disease) but were not associated with survival. Survival curves differed by stage, lymph node metastases, infiltrative growth pattern, lymphocytic infiltration, fibrosis, extramural venous spread, and tumor grade in a univariate analysis. The infiltrative growth pattern (P = 0.04) and lymphocytic infiltration (P = 0.003) were features associated independently with survival after adjusting for modified Dukes' stage. Furthermore, the lack of a significant lymphocytic infiltrate was associated with a death rate 3.4 times greater than that occurring in patients with Stage B disease with a significant infiltrate. CONCLUSIONS. The authors conclude that proliferative activity in colorectal carcinoma as measured by Ki-67 immunostaining was not associated with prognosis.