Introduction: Although soy isoflavones intakes have been associated with reduced breast cancer risk in many populations, there is still uncertainty as to the relationships between these compounds and breast cancer clinical characteristics, which may have implications for prognosis and survival.

Methods: Included in this study were women with newly diagnosed, incident breast cancer (n = 683) and women without a history of breast cancer (n = 611) enrolled in Roswell Park Cancer Institute's Data Bank and BioRepository. Epidemiologic and dietary data were collected by self-administered questionnaires, and clinical data from breast cancer cases were abstracted from medical records. Total and specific isoflavones intakes were calculated from the food frequency questionnaire using published food composition data. Polytomous logistic regression was performed to determine the relationship between tertiles of intakes of total and specific isoflavones and tumor characteristics, adjusting for known risk factors and stratifying by menopausal status.

Results: In the sample overall, compared to controls, cases in the highest vs. lowest tertile of total isoflavone intake had approximately 30% decreased odds of having an invasive tumor and an approximately 60% decreased odds of having a grade 1 tumor. Among specific isoflavones, intakes in the highest vs. lowest tertile were associated as follows: for genistein, an approximately 60% decreased odds of having grade 1 tumor; and for glycetein, an approximately 25% decreased odds of being a case, 60% decreased odds of having grade1 tumor, and 30% decreased odds of having HER2 negative tumor. In postmenopausal women only, cases in the highest vs. lowest tertile of glycetein intake had an approximately 30% decreased odds of having luminal A or stage I disease. In premenopausal women only, higher total isoflavone, daidzein, genistein, and glycetein intakes were associated with an approximately 70% decreased odds of having a large (>2cm) tumor. Further, those premenopausal women in the highest vs. lowest tertile of total isoflavone and genistein intakes had an approximately 60% decreased odds of having stage II breast cancer. Higher intakes of total and specific isoflavones were not associated with tumor characteristics indicative of a poor prognosis.

Conclusions: In general, higher total isoflavone intake (and some specific isoflavones intakes) were associated with a reduced risk of tumors with more favorable prognostic characteristics. Intakes of total and specific isoflavones had no association with tumors having less favorable characteristics. In premenopausal women, however, higher intakes of isoflavones appear to be associated with a decreased risk of having large (>2cm) tumors. In this study, there is little evidence that isoflavones are associated with indicators associated with poor prognosis or recurrence, suggesting that these compounds may operate through mechanisms specific to the biologic heterogeneity of the tumor. Further research is warranted to address the impact of tumor heterogeneity in studies of isoflavones and breast cancer.

Citation Information: Cancer Prev Res 2010;3(12 Suppl):A79.