Endocrine determinants of breast density and breast cancer

Abstract

Worldwide, breast cancer is the most common malignancy among females. The total breast area on a mammogram can be dived in a radiologicaly dense area (glandular and stromal tissue) and a non-dense area (mainly fat tissue). Women with a high proportion of dense breast tissue (percent breast density) have increased breast cancer risk. Exposure to high levels of growth factors and hormones with known mitogenic properties could increase breast cancer risk, through an increase in the amount of dense tissue of the breast. In this thesis we aimed to investigate the relationships between growth factors (insulin-like growth factor I (IGF-I)) and hormones, both endogenous (insulin and sex steroids) and exogenous (phytoestrogens), on the one hand, and breast density and breast cancer on the other hand. In Chapter 2, we investigated the effects of premenopausal circulating levels of IGF-I (Chapter 2.1) and genetic variation in the IGF-1 gene (Chapter 2.2) on menopausal changes in breast density, using data of 656 women participating in the Prospect-EPIC cohort (European Prospective Investigation into Cancer and Nutrition). Women with high premenopausal IGF-I levels showed a smaller decrease in percent breast density over menopause, which was, however, largely explained by the smaller increase of the non-dense area, whereas the dense area was hardly affected. Several of a total 16 single nucleotide polymorphisms (SNP), within the IGF-1 gene were significantly associated with higher IGF-I levels. These same SNPs were related, but to a lesser extent and statistically non-significantly, with a smaller decrease of percent density during the menopausal transition phase. In (Chapter 3) the association between circulating levels of sex steroids and breast density in 969 postmenopausal women participating in the Prospect-EPIC cohort was investigated. None of the sex steroids studied was clearly associated with percent breast density or the dense area. High circulating levels of estradiol, DHEAS and androstenedione were however associated with significantly smaller non-dense (fat) area. Within the large EPIC cohort, prediagnostic C-peptide serum levels were analyzed for 1141 women who were diagnosed with breast cancer, and for 2204 matched control subjects. Elevated levels of serum C-peptide -a marker for pancreatic insulin secretion- were not associated with overall breast cancer risk and we concluded that our results do not support the hypothesis that chronic hyperinsulinemia generally increases breast cancer risk (Chapter 4). In Chapter 5 we show that women with high circulating levels of five types of isoflavones had slightly decreased breast cancer risk. The association was strongest and statistically significant for genistein. Results originate from the largest prospective study (383 breast cancer case subjects and 383 control subjects participating in Prospect-EPIC) on circulating levels of phytoestrogens (plant compounds with hormone-like activity) and breast cancer risk. In Chapter 6, we discuss possible explanations for the findings that IGF-I and sex steroids that are known to have mitogenic properties and to increase breast cancer risk, hardly affect the absolute amount of dense breast tissue, which is supposed to be the target tissue for breast cancer, but are related to the non-dense area of the female breast.