Flow cytometry vs. Ki67 labelling index in breast cancer: a prospective evaluation of 181 cases

Abstract

Background: Excessive proliferation is one of the first steps in oncogenic activation and one of the most important biological features defining the aggressiveness of tumors. Quantifying the proportion of tumor cells in S-phase by means of flow cytometry has shown, in the past, to be useful for defining high-risk subgroups in breast cancer. Several antigens closely associated with proliferation are also detectable by means of immunohistochemistry, offering in theory an easy to perform and cheap alternative to flow cytometry for measuring proliferation. To test this hypothesis, we compared both methods prospectively in a series of breast cancers.

Materials and methods: We studied the proliferation rate of 181 breast cancers (152 ductal infiltrating, 17 lobular infiltrating, 12 other histological varieties), operated upon at our institution, by means of flow cytometry and the Ki67 labelling index, using the MIBI antibody. Ploidy (expressed as DNA content or DNA-index), S-phase fraction and the Ki67 labelling index were the variables of the study. The S-phase fraction was considered separately for diploid and aneuploid tumors, following the 1992 Maine Consensus guidelines and was judged abnormally elevated if higher than the 75th percentile for each group. The Ki67 labelling index was expressed as percent positive tumor cells, positive cells being those showing specific nuclear staining.

Results: DNA-ploidy and the Ki67 labelling index could be evaluated in all tumors. Of the total, 96 (53%) were diploid and 85 (47%) aneuploid. S-phase fraction could be measured in 172 out of the 181 tumors (95%). The 75th percentile cut-offs for diploid and aneuploid tumors were 9.9% and 15.8%, respectively. We found a significant correlation beween rising DNA content and increasing Ki67 index (r = 0.18; p = 0.022), as well as between the percentage of cells in S-phase of the whole tumor population and Ki67 (r = 0.22; p = 0.0055). A Ki67 cut-off of 50% or higher identified most aneuploid tumors, or a small group of diploid ones with a high S-phase fraction (specificity = 96.7%; positive predictive value 92.5%), however at the price of a very low sensitivity (62.6%). This was due to the presence of many aneuploid tumors with a low S-phase fraction.

Conclusion: The Ki67 labelling index and S-phase fraction are significantly correlated. However, flow cytometry provides additional indirect information on tumor aggressiveness associated with DNA-ploidy. Further studies are needed to determine whether Ki67 alone is sufficient as a routinely applicable method.