Tissue microarray analyses of G1/S-regulatory proteins in ductal carcinoma in situ of the breast indicate that low cyclin D1 is associated with local recurrence

Abstract

Ductal carcinoma in situ (DCIS) of the breast constitutes about 10% of all diagnosed breast cancers and, despite surgical removal, it may recur, either as DCIS or invasive breast cancer. Nuclear grade and growth pattern according to Andersen et al as well as surgical margins are factors that have been used to predict local recurrence, but ideally a set of tumour-specific factors should be identified and used as prognostic markers. Many cell cycle regulatory gene products have been shown to be involved in the formation of tumours and are either oncogenes or suppressor genes and involved in key processes in the transformation. We therefore characterised the cell cycle regulators cyclin E, cyclin D1, p27 and p16 in a material of DCIS cases arranged in a tissue microarray. With a manual tissue arrayer, 52% of the initial 177 DCIS samples were successfully targeted allowing immunohistochemical analyses of all four proteins in 92 cases of DCIS. As also observed in invasive breast cancer, there was a trend indicating that DCIS cases with high cyclin D1 were cyclin E low and oestrogen receptor-positive, whereas cyclin E high DCIS cases were cyclin D1 low and oestrogen receptor-negative. For the 64 patients that did not receive postoperative radiotherapy, there were 16 local recurrences (eight DCIS and eight invasive breast cancer) during a mean follow-up time of 63 months. Cyclin E, p27 or p16 were not associated with local recurrence, but interestingly cyclin D1 was significantly and inversely associated with local recurrence, both using univariate and multivariate analyses. In summary, using a tissue array approach we have shown that cyclin D1, besides growth pattern, is a prognostic marker for local recurrence in DCIS.

Figure 1

Examples of immunohistochemical staining of the cell cycle regulatory proteins in different DCIS tumours: (A) cyclin D1 low; (B) cyclin D1 high. (Note that the intensity of the cytoplasmic staining only differs slightly between (A) and (B), in contrast to the nuclear staining.); (C) cyclin E low; (D) cyclin E high; (E) p16 low; (F) p16 high; (G) p27 low; (H) p27 high.

Figure 2

Distribution of the percentage positive cells for the analysed cell cycle regulators in 92 DCIS cases analysed by immunohistochemistry. The cutoffs used in the analyses are indicated in the figure.

Figure 3

Scatter plots showing the associations between cyclin E, cyclin D1 and the OR content in DCIS and in invasive breast cancer.

Figure 4

Kaplan–Meier plots of ipsilateral local recurrence (DCIS and invasive breast cancer) in the material of 64 DCIS divided according to (A) cyclin D1 protein content, (B) growth pattern according to Andersen et al (1988) and (C) a combination of cyclin D1 and growth pattern.