Abstract P1-07-03: Effect of Exemestane on Mammographic Density in Postmenopausal Women at Risk for Breast Cancer

Abstract Background: A reduction in mammographic density(MD) has been associated with the reduction in breast cancer risk among those taking tamoxifen. Aromatase inhibitors (AIs) show promise for breast cancer prevention in postmenopausal women. We are conducting a phase II trial of exemestane in women at increased risk for breast cancer. The primary endpoint for this study is change in MD. Methods: Subjects were screened and enrolled at two sites: Center for Cancer Research, National Cancer Institute and Lombardi Comprehensive Cancer Center, Georgetown University Hospital. Eligible participants are at increased risk for breast cancer by virtue of: Gail model risk ≥1. 7% over 5 years; high risk pathological lesion (e.g. lobular neoplasia, ductal carcinoma in situ); known BRCA1/2 deleterious mutation or prior stage I/II breast cancer at least 2 years from breast cancer treatment and not treated with AIs. Women were excluded if AP spine T-score was <-2.5 (osteoporosis). Study participants received exemestane 25 mg, calcium carbonate 1200 mg and vitamin D 400 IU daily for two years. Mammograms were conducted at baseline, 1 and 2 years. MD was calculated using the semi-automated Cumulus software. The cranio-caudal view of one breast was selected for analysis. For patients who had had no prior surgery or radiation, the side with greater density at baseline was selected. For patients who had undergone significant breast surgery or radiation, mammograms of the unaffected side were selected. Percent MD was determined by one investigator (C.B.) who read each image twice and was masked to the subject's time on exemestane. The mean change in paired images was tested for a mean change of zero using the Wilcoxon signed rank test. As of June 2010, 42 women have enrolled in the trial and we have analyzed change in MD at one year in the first 22 subjects. Results:For the 22 subjects included in this analysis 15 were eligible due to a high risk pathological lesion, 4 by Gail Model and 3 by prior breast cancers. 18 had film screen mammograms at baseline and 12 months, 4 had film screen mammograms at baseline and digital mammogram at 12 months. Mammographic density declined 7% (p=0.0006, 95% CI −11% to −3%) at one year. The change in MD ranged from −26% to +14% and the standard deviation was 9%. In regard to the four subjects who had film screen mammograms at baseline and digital at follow up, the baseline percent dense areas were slightly higher than average but this did not reach statistical significance (p=0.32 by the Wilcoxon rank sum test). The 12 month MD comparison was similar to the 18 subjects with film images (p=0.90). Conclusions: Exemestane use was associated with a statistically significant decrease in MD at one year. Prior evaluations of AI therapy in high risk women have not shown a significant decline in MD, however time on treatment was shorter and other third generation AIs were used. Although the number of subjects is small, MD did not differ significantly between film screen and digital mammograms. We will further evaluate change in MD in the entire cohort at 1 and 2 years and between film screen and digital mammography. Phase III trials are on-going to evaluate the effect of exemestane on breast cancer prevention. Citation Information: Cancer Res 2010;70(24 Suppl):Abstract nr P1-07-03.