Estrogen receptor breast cancer phenotypes in the Surveillance, Epidemiology, and End Results database.
BACKGROUND: Researchers question whether estrogen receptor alpha-negative (ERN) and -positive (ERP) represent different stages of one disease or different breast cancer types. OBJECTIVE: To further examine ERalpha phenotypes, we stratified incident tumor characteristics in the Surveillance, Epidemiology, and End Results (SEER) Database (n = 82,488) by ERN and ERP. METHODS: Study variables included black-white race, age-at-diagnosis, and standard incident tumor characteristics. These characteristics were arbitrarily dichotomized into good versus poor prognostic factor groups, for example, good (tumor size < or = 2.0 cm, negative axillary lymph nodes, and good histologic grade) versus poor (tumor size > 2.0 cm, positive nodes, and poor grade). Age frequency density plots were generated from the corresponding age-at-diagnosis frequency histograms. Average annual age-specific incidence rates (or risks) were adjusted to the 1970 United States standard female population. RESULTS: Age frequency density plots demonstrated bimodal premenopausal and postmenopausal breast cancer populations. ERN was correlated with premenopausal disease, black race, and poor prognostic factor groups, whereas ERP was associated with postmenopausal disease, white race, and favorable tumor characteristics. ERN rates increased premenopausally and then flattened to a nearly constant level after 50 years of age. ERP risk rose for most of a woman's lifetime with the greatest risk occurring between 75 and 79 years. CONCLUSIONS: ERalpha exhibited bimodal age frequency distribution with a dichotomous pattern for age-specific rates, racial, and prognostic factor profiles. Menopause had a greater effect on ERN than ERP. Possible implications for breast carcinogenesis and cancer prevention are discussed in the text.