Project 4 (Hartmann)

点击翻译按钮获取中文摘要

Most of the 250,000 U.S. women diagnosed annually with in situ or invasive breast cancer (BC) were not recognized as being at increased risk for the disease. Both the NCI and Institute of Medicine have identified accurate individualized risk assessment as central to improving early detection and prevention. The most widely used risk prediction model for BC, the Gail model, generally performs well in predicting BC risk across groups of women, but is limited when predicting individualized risk. In general, the cancers for which we can predict susceptibility best are those where the at-risk tissue can be examined, e.g., cervix, colon, endometrium. In breast, benign tissue is available for risk assessment from women who have had a benign biopsy. Women with so-called benign breast disease (BBD), numbering 1-2 million/year In the US alone, are a common and clinically important group, with a known increased risk of BC. About 25% of women with BC report having had a prior benign biopsy. Our purpose with this project is to develop a risk prediction model for women with BBD. We hypothesize that features in benign breast tissue, in particular from subgroups known to be at increased risk for BC, can help to identify those women who are at highest likelihood of progression of BC. Two recognized higher-risk groups within BBD are women with atypia, and women in whom normal, age-related regression (or involution) of breast lobules has not occurred. Utilizing a cohort of over 11,000 women with BBD at the Mayo Clinic, we will identify novel BC-predictive biomarkers obtained from transcriptional profiling of women with atypia who either did or did not progress to BC and from dissection of the mechanisms underlying lobular involution, a physiological process which we recently found to be associated with decreased cancer risk. Using a newly-developed technique, we will also quantitate the specific extent of lobular involution that has occurred In these women to generate a continuous risk feature. Then in a nested case: control series within our BBD cohort, we will build a risk prediction model that incorporates the top predicting elements from histology, clinical-epidemiologic features, mammographic density, quantitation of involution and molecular biomarkers. Finally, building upon a strong collaboration fostered by Breast SPORE developmental funds, we will assess our BC risk model using the Nashville BBD cohort supported by the Vanderbilt Breast SPORE. Altogether these studies will test the ability to use tissue based features to enhance risk prediction of BC and potentially provide insights about important early events in breast carcinogenesis.