Distinct breast cancer incidence and prognostic patterns in the NCI's SEER program: suggesting a possible link between etiology and outcome

Background. Breast cancer is a heterogeneous and chronic disease with relapses and death occurring 25 years or more after primary diagnosis. Standard tumor characteristics are used to predict initial relapse or death, but their ability to estimate long-term patterns of failure may be limited. Methods. To further evaluate the significance of standard tumor features, we compared incidence and prognostic patterns in the National Cancer Institutes (NCIs) large-scale population-based Surveillance, Epidemiology, and End Results (SEER) program for high-risk versus low-risk breast cancers, i.e., size > 2.0 versus <= 2.0 cm, lymph node positive versus negative, high versus low histologic grade, and hormone receptor negative versus positive expression, respectively. Data were stratified by age 50 years to approximate menopause. Results. High-risk versus low-risk breast cancers demonstrated two very different incidence and prognostic patterns. Age-specific incidence rates among women with high-risk tumors increased until age 50 years then flattened, whereas rates among women with low-risk tumors increased continuously with aging. Hazard rates for breast cancer death spiked sharply two years following primary breast cancer diagnosis among women with high-risk but not with low-risk tumors. Paradoxically, hazard function crossed over 6-8 years following breast cancer diagnosis, with hazard rates lower for high-risk than for low-risk breast cancers. Conclusion. Distinct incidence and prognostic patterns among high-risk and low-risk breast cancers suggest a possible link between breast cancer etiology and outcome. These epidemiologic results appear to complement emerging molecular genetic techniques, showing distinct genotypes for high-risk and low-risk breast cancer phenotypes.