Recent Declines in Hormone Therapy Utilization and Breast Cancer Incidence: Clinical and Population-Based Evidence

Abstract

TO THE EDITOR: Long-term use of estrogen- or progestincontaining hormone therapies (HT) is well recognized to be associated with increased risk of breast cancer both in observational studies and the Women’s Health Initiative (WHI) randomized trial. 1 After the early termination of the trial in July 2002 and the subsequent media coverage, reports from clinical series indicated immediate declines of 28% to 46% in prevalences of HT and estrogen-only (ET) use. 2,3 The population-level impact of these declines on breast cancer are uncertain given the 2- to 3-year lag time between diagnosis and data availability in cancer registries. To better understand the possible impact from changes in HT on breast cancer in larger populations, we compared recent secular trends in HT use and breast cancer incidence in available clinical and population-based data resources, including recently released cancer registry data for the year 2004. We calculated the annual prevalence of HT use and the annual incidence of invasive breast cancer for the period from 1994 to 2003 for women ages 50 to 74 years in Kaiser Permanente’s (Oakland, CA) Northern California region (KPNC), a large integrated health-delivery system from which computerized pharmacy, membership, and hospital cancer registry data were obtained. Women filling at least two prescriptions (generally constituting 90 days per prescription) of HT or ET in any calendar year were considered users for that year. First primary invasive breast cancer (International Classification of Disease–Oncology, 3rd edition [ICDO-3] 4 site 50.0-50.9) diagnoses were identified, with histologic subtypes defined as ductal (ICD-O-3 codes 8500), or “with lobular component” (ICD-O-3 codes 8520, 8522, 8524). Yearly denominators included women who were KPNC members for at least 11 months of