Biocellular aspects of high mammographic density as a risk factor for breast cancer

© 2016 Dr. Grace Li Ling ChewHigh mammographic density (MD) is one of the strongest risk factors for breast cancer after high-risk mutations, with a 4-6 fold increased risk comparing the highest to lowest MD quartiles. MD is of great clinical relevance, given that the attributable risk of breast cancer (BC) due to high MD in the population may be as high as 30%, and the already widespread use of mammography for breast assessment. However, the biological basis for high MD and its associated cancer risk is poorly understood. A validated xenograft model where the dynamic effects of drug interventions and gene perturbations on human MD tissue can be investigated in the preclinical setting will be valuable. Prior to commencement of the PhD thesis, there were no animal models of human MD that could maintain the MD differential of tumour-free breast tissues. In the first published study, we developed a xenograft model of human MD, where matched high and low MD human tissues were precisely-sampled under stereotactic-guidance from fresh mastectomy tissues and maintained in separate vascularized biochambers in SCID mice. This study demonstrated that the high and low MD biochamber tissues retained their differential radiographic density and histologic features of the original human tissue. The high compared to low MD biochamber tissues were composed of increased stromal and decreased adipose percentage areas, reflecting the MD phenotype of the original human breast samples. The MD xenograft model was then extended to examine the changes in radiographic density and histology that occurred during murine pregnancy, lactation and postpartum involution states, and after exposure to exogenous endocrine treatment. These studies demonstrated the dynamic nature of the MD xenograft model, with decreased stromal and increased adipose tissue percentage areas observed in high MD biochamber tissues during murine lactation and postpartum involution, and also in Tamoxifen-treated compared to placebo-treated mice. High and low MD biochamber tissues had increased radiographic density with postpartum involution, and increased duration of implant. The radiographic density decreased in high MD biochamber tissues of Tamoxifen-treated compared to oestrogen-treated mice. As increased Cox-2 levels have been observed in breast tumours and stromal regions of high MD tissues, we investigated the expression of Cox-2 in the epithelial and stromal cells of matched high and low MD breast tissues. We demonstrated increased staining in both epithelial and stromal cells of high MD breast tissues. We then showed that the differential Cox-2 expression in high and low MD human breast tissues was maintained in murine biochambers, and was sensitive to hormonal supplementation. Collectively, this thesis indicates that the MD model will be valuable for investigating the mechanisms of how modifying factors, such as lifestyle behaviours and endocrine treatments like Tamoxifen, can reduce MD in cancer-free human breast tissues, and would be useful in research aiming to develop preventative therapies to reduce MD-related risk, such as Cox-2 inhibition