Claudine Isaacs

Georgetown University

PLCO (Learn more about this study)

2009-0121

Jun 18, 2009

RISK FACTORS FOR INVASIVE BREAST CANCER BY HORMONE RECEPTOR STATUS AND FOR DUCTAL CARCINOMA IN SITU

Breast cancer is a heterogeneous disease with a number of different well recognized subtypes. These subtypes can be subdivided by hormonal and HER2 status and have different natural histories and treatments, and possibly different risk factors. Additionally, chemoprevention with selective estrogen receptor modulators such as tamoxifen and raloxifene has been shown to decrease the incidence of hormone receptor positive breast cancer by 50% among women at increased risk, but has no effect on hormone receptor negative breast cancer. Currently used breast cancer risk prediction models, such as the Gail and Tyrer-Cuzick models, do not distinguish between the risks of developing hormone receptor positive versus hormone receptor negative breast cancer. Additionally, little is known about risk factors for ductal carcinoma in situ (DCIS) as compared with invasive breast cancer.The PLCO Cancer Screening Trial is an ideal source of data to examine the differential impact of standard breast cancer risk factors on hormone receptor defined invasive breast cancer and ductal carcinoma in situ. The collection of risk factor data at baseline, followed by prospective ascertainment of cancer eliminates the concern of recall bias which would be a significant issue in other studies. Data on characteristics of the breast tumors diagnosed during follow-up among PLCO participants are currently being collected as part of the breast cancer supplemental data collection. No additional data will need to be collected for this analysis. Specific Aims: The goals of this study are to determine in a prospective cohort if established and suggested risk factors for invasive breast cancer differ by hormone receptor status and between invasive breast cancer and pure ductal carcinoma in situ (DCIS). Method: This study will include all accrued cases of breast cancer with receptor status information and the non-cases in the cohort. The associations with the breast cancer risk factors and method of detection of cancer, for invasive breast cancer and non-invasive breast cancer, will be compared using unconditional polychotomous logistic regression with associated 95% confidence intervals. To formally test heterogeneity, we will fit adjusted case-case models. Implications: The data will be important for tailoring development of risk modification and prevention strategies for different forms of breast cancer. It will facilitate the selection of the most appropriate candidates for existing chemopreventive agents, and would also identify appropriate subjects for investigation of novel agents to prevent hormone receptor negative breast cancer.