1073

Ductal carcinoma in situ (DCIS), an alleged precursor of invasive breast cancer, is characterized by neoplastic epithelial cells growing within mammary ducts. One of the most critical stages in the development of breast cancer is the progression from DCIS to invasive carcinoma, when the basement membrane of the duct is breached and tumour cells spill out into the surrounding stroma. Clinical prognostication for DCIS is challenging, as the likelihood of progression to invasion by low-grade DCIS and the potential for low-grade lesions to evolve to high-grade both remain unclear. The small size of lesions, coupled with variations in time to invasion and rates of post-treatment recurrence, have complicated the study of DCIS in humans. The MMTV-neu 202Mul mouse model has been reported to develop DCIS-like lesions in the mammary fat pads. When a tumour is detected by manual palpation, the animal is sacrificed and the mammary fat pads are removed. Tissues are prepared as both snap-frozen and formalin-fixed paraffin-embedded (FFPE) blocks, sectioned, stained with haematoxylin and eosin, and assessed for quality. Immunohistochemical staining for smooth muscle actin (SMA) reveals the myoepithelial cells which are wrapped around the basement membrane of the ducts. Assessing the continuity of the myoepithelial cell line permits classification of tumours as either DCIS, invasive, or focally invasive. FFPE sections of DCIS and fully invasive samples are then microdissected. DNA is extracted and examined using array comparative genomic hybridization. Analysis has revealed differing patterns of genomic amplification and deletion between the two tumour types. Copy numbers of Gas6 and Psmb3 have been found amplified in invasive carcinoma. To assess gene expression levels, RNA is extracted from frozen DCIS and fully invasive tumours and analysed via expression microarray. Array profiling has revealed altered expression of genes including Tsga2, Grb7, Arl1, and Phka2. Preliminary data indicate that these genes have higher expression in DCIS. Gene copy number and gene expression differentials will be further analyzed using quantitative polymerase chain reaction (Q-PCR). Immunohistochemistry is being employed to develop a profile of the tumour types. Targets including Ki67, Laminin, Activated Caspase 3, E-Cadherin, and CD31 have been assessed. These data indicate differences in the expression of certain markers between tumour types. In comparison to DCIS tumours, invasive tumours have shown an increase in expression of Ki67 but a decreased expression of Activated Caspase 3. Additional immunohistochemistry is ongoing, as is collection and analysis of microarray data. Through comparison of data from mouse and human cases, we plan to identify a subset of genes which could be used as prognostic markers and/or therapeutic targets to improve clinical management of human DCIS cases.

[Proc Amer Assoc Cancer Res, Volume 46, 2005]