TISSUE MICROARRAY PERSPECTIVES ON CYCLIN D1 IN BREAST CANCER: Progression, prognosis and prediction

Abstract: Breast cancer is a truly heterogenous disease, reflected in the existence of a wide spectrum of phenotypic subsets of tumours with varying intrinsic aggressiveness as well as ability to respond to a given treatment. Recent advances in large-scale genomic and proteomic screening techniques have confirmed this heterogeneity also at a molecular level, but there is still a bottleneck to be overcome before the discoveries in the laboratory can be translated into the much more complex clinical situation in order to develop better targeted therapies. For this purpose, the tissue microarray (TMA-) technique is a promising tool, since it enables rapid in situ screening of potential prognostic and predictive tumour markers in hundreds of minute tissue samples linked to clinicopathological data and disease endpoints. The cell cycle regulatory protein cyclin D1 is overexpressed in about 50% of primary breast cancers and it has been the prevailing notion that its pivotal role in the G1/S transition of the cell cycle, thus promoting proliferation, also underlies its oncogenic potential. However, the attention is now being turned towards several non-cell cycle associated functions of cyclin D1, among which the ability to interact directly with the estrogen receptor may be fundamental. This thesis is the result of an effort to outline the role(s) of cyclin D1 in different morphological entities of breast cancer, using the TMA-technique. In 64 cases of pure ductal carcinoma in situ (DCIS), treated with breast conserving surgery without postoperative radiotherapy, we found that a high cyclin D1 protein content, defined as the fraction of positive tumour cells and not the staining intensity, was associated with a significantly prolonged time to local recurrence. Further, a comprehensive individual tumour progression-TMA was constructed in the?single patient tissue chip?. In this model, a clonal spread of cyclin D1 (CCND1) gene amplified tumour cell populations could be vizualized throughout the different morphological stages of tumour progression, not least to individual lymph nodes, in 30 patients with primary, node-positive breast cancer. We have also evaluated the utility of routine pathology parameters, including established proliferation markers, as predictors of response to endocrine treatment (tamoxifen) in primary premenopausal breast cancer, using tumour specimens from a randomized trial. Our results indicated a beneficial effect of tamoxifen in estrogen receptor (ER) positive tumours with a high and low proliferation rate, in contrast to an ?intermediate? group, and that lobular carcinomas may benefit less from tamoxifen than ductal carcinomas. In the same trial, CCND1 gene amplification status turned out to be a strong predictor of tamoxifen response, with implications of an agonistic tamoxifen effect in amplified tumours, in contrast to an excellent response in non-amplified tumours. A significant, but non-obligate, association was observed between gene amplification and protein expression, therefore indicating the involvement of one or several other co-amplified genes and their protein products in this agonism. Finally, we have outlined in situ associations between several elements of the Ras/Raf/ERK signalling cascade and cyclin D1 in three different primary breast cancer cohorts. An impaired tamoxifen response could further be demonstrated for ERK1/2 phosphorylated tumours in one of the cohorts, a randomized postmenopausal tamoxifen trial, which is entirely in line with previous findings for cyclin D1-high tumours in this trial.