Presentation Abstract

Abstract Number: 157
Presentation Title: Mammography interval and breast cancer mortality in the Women’s Health Initiative
Presentation Time: Sunday, Apr 07, 2013, 1:00 PM - 5:00 PM
Location: Hall A-E, Poster Section 10
Poster Board Number: 01
Author Block: Michael S. Simon1, Sylvia Wassertheil-Smoller2, Cynthia A. Thompson3, Roberta M. Ray4, F. Allan Hubell5, Dorothy Lane6, Lawrence Lessin7, Hema Vankayala8, Lew Kuller9. 1Barbara Ann Karmanos Cancer Inst., Detroit, MI; 2Einstein University, Detroit, NY; 3Arizona Cancer Center, Tucson, AZ; 4Fred Hutchinson Cancer Research Center, Seattle, WA; 5University of California at Irvine, Irvine, CA; 6Stony Brook University, Stony Brook, NY; 7Medstar Washington Hospital Center, Washington DC, DC; 8VA Medical Center, Detroit, MI; 9University of Pittsburgh, Pittsburg, PA
Abstract Body: Background: The 2009 U.S. Preventive Services Task Force (USPSTF) statement on screening for breast cancer suggests the lack of evidence for mammography for women 75 years or older based on the premise that disease would be indolent at advanced age. Analysis of the Women’s Health Initiative (WHI) data refute this hypothesis, in that over 60% of breast cancers diagnosed between ages 70 to 79 were moderately or poorly differentiated. In order to better understand the role of screening mammography in older women, we evaluated the relationship between the time interval from breast cancer diagnosis to the last prior mammogram, and mortality due to breast cancer in the WHI. Methods: The study sample included 8,663 women enrolled in the WHI observational study (OS) or clinical trial (CT) who were diagnosed with breast cancer over an average of 12.2 (2.6 SD) years of follow-up (1,670 with insitu disease). Information on mammography was obtained annually or semi-annually from medical record update forms. Mammography interval was defined as the time between the date of diagnosis, and the last self- reported mammogram completed > 6 months prior to diagnosis. Self- and interviewer-administered questionnaires were used to collect information on other risk factors. Cox proportional hazards regression was used to analyze the association between mammography interval and breast cancer mortality stratified by age at diagnosis (50-74 vs 75+). Statistical tests were two-sided. Results: Prior mammograms were reported by 30% of women within > 6 months to < 1 year of diagnosis, 47% between 1 to < 2 years, 17.2% between 2 to 5 years and 5.8 % at > 5 years, or none at all. Longer interval was associated with advanced stage with 23% of women who had an interval of > 5 years diagnosed with regional or distant disease, compared to 20% of women with an interval of > 6 months to < 1 year, p=0.05. Women with a longer interval were also more likely to have estrogen receptor negative disease (22% vs 16%, p=0.03). In a model adjusted for age, race/ethnicity, OS or CT, health insurance, marital status, Charlson co-morbidity index, and body mass index, longer mammography interval as compared to an interval of > 6 months to < 1 year (referent), was significantly associated with increased breast cancer mortality for women > age 75 at diagnosis. A significant association was not seen for women age 50 to 74. For women > age 75, an interval of > 2 to < 5 years resulted in a hazard ratio (HR) of death from breast cancer of 1.87 (95% CI, 1.10 to 3.19), and an interval of > 5 years or no mammogram, resulted in a HR of 3.17 (95% CI, 1.68-5.96), p-trend 0.0001. For women age 50-74, the respective HR’s and 95% CI were 1.05 (0.75-1.46) and 1.28 (0.8-2.06), p-trend 0.43. Conclusions: Extended time between breast cancer diagnosis and the last mammogram had a significant impact on breast cancer mortality for women diagnosed after age 75, suggesting a role for continued screening to reduce breast cancer mortality in older women.