• William F Anderson, Kenneth C Chu, and Susan S Devesa.
• DHHS/NIH/NCI/Division of Cancer Prevention, EPN, Suite 2141, 6130 Executive Boulevard, Bethesda, MD 20892, USA. wanders@mail.nih.gov
• Breast Cancer Res. Treat. 2004 Nov 1; 88 (2): 149-59.

BackgroundIncidence patterns are well-established for invasive breast carcinoma (InvBC) overall and for InvBC defined by estrogen receptor (ER) expression, but are not as well-defined for breast carcinoma in situ (CIS).MethodsWe, therefore, examined and compared the incidence patterns for CIS and InvBC in the SEER program to define these patterns and to generate etiologic hypotheses. Data were stratified by age < 50 and > or =50 years to approximate menopause.ResultsDuring the years 1973-2000, annual age-adjusted incidence rates rose 660% for CIS and 36% for InvBC, with the most rapid increases occurring in women age > or =50 years. Age-specific incidence rate curves for CIS increased until age 50 years, and then flattened, irrespective of ER expression. On the other hand, rates for InvBC overall and for InvBC defined by ER-positive expression increased continuously with aging, whereas rates for InvBC defined by ER-negative expression flattened after 50 years. Age frequency distribution for CIS and for ER-negative InvBC demonstrated bimodal populations, with a predominant early onset peak incidence at age 50 years. Age frequency distribution for ER-positive InvBC showed bimodal populations with a predominant late-onset mode at age 71 years.ConclusionOver the last three decades, age-adjusted incidence trends differed for CIS and InvBC in the United States, possibly due to screening mammography and/or etiologic diversity. Indeed, age-specific incidence patterns suggested that carcinogenic events operating early in reproductive life had greater impact upon CIS and InvBC defined by ER-negative expression than upon InvBC overall and InvBC defined by ER-positive expression.

30 keywords...

hide…