Using Microarrays to Estimate Breast Cancer Risk

Institution: Stanford University
Investigator(s): Bradley Ekstrand, M.D., Ph.D. -
Award Cycle: 2002 (Cycle VIII) Grant #: 8FB-0108 Award: $34,233
Award Type: Postdoctoral Fellowship
Research Priorities
Etiology and Prevention>Prevention and Risk Reduction: ending the danger of breast cancer



Initial Award Abstract (2002)
Over the past decade, researchers have developed a better understanding of the characteristics (risk factors) that predispose certain women to breast cancer. Knowledge of these risk factors has led to methods, such as the Gail model, that can inform some women of their odds of developing breast cancer. The Gail model, while perhaps the most commonly used, has its limitations. Perhaps most importantly, less than half of the women who develop breast cancer would have been identified as high risk by the Gail method.

Here, I propose to develop a new method of determining a woman's risk for breast cancer. My approach is based on the well-known link between exposure to x-rays and the subsequent development of breast cancer. For example, women who are exposed to high doses of x-rays as part of the treatment for Hodgkin's disease (a cancer of the immune system), have a markedly increased risk for breast cancer. In addition, all of the known breast cancer genes, including BRCA1 and BRCA2, normally function in the system that repairs damage from these x-rays.

There are certainly other genes, or more likely combinations of genes, involved in both the repair of DNA damage from x-rays and the development of breast cancer. These genes can be identified using microarrays, a novel technology capable of analyzing the function of thousands of genes in a single experiment. I hypothesize that the development of breast cancer is intimately associated with an abnormal genetic response to exposure to x-rays, as detected on microarrays. This abnormal response can then be used to identify women at high risk of developing breast cancer.

I will develop this method by performing a case-control study. I will identify and enroll a group of women with breast cancer (cases) as well as a similar group of women without breast cancer (controls). Each participant will donate a tube of their blood, and the blood cells will be grown in the laboratory. The cells from each patient will be exposed to x-rays in the lab, and the genetic response to the x-rays will be measured using a microarray. The responses from the women without breast cancer will be compared to those with breast cancer. A computer-based statistical method developed by my mentor will be used to identify the genes whose responses to x-rays predict the patients with breast cancer.

This study represents a novel approach to the prediction and prevention of breast cancer. My strategy would allow breast cancer risk assessments to be tailored to each woman. It is based not on general risk factors applicable to many women, but instead on genetic factors that are specific to each individual. This method utilizes the powerful technology of microarrays, which was developed largely by colleagues of ours here at Stanford. When compared to other breast cancer risk assessment tools, our method should better identify high-risk women in the general population. The ability to identify women at elevated risk enables patients and their physicians to make highly educated choices among options for risk-reduction.


Final Report (2005)
The aim of this project was to develop a new method of determining a woman’s risk for breast cancer. I hoped to capitalize on the well-established link between X-ray exposure and the subsequent development of breast cancer. I hypothesized that the development of breast cancer is related to an abnormal genetic response to X-ray exposure. Such an abnormal genetic response can be detected using a microarray, a novel technology capable of analyzing the function of thousands of genes simultaneously in a single experiment. In Specific Aim 1, I planned to use microarrays to develop a method to predict breast cancer in women who had been exposed to X-rays as part of the treatment of Hodgkin’s disease. In Specific Aim 2, I planned to see if a similar method could also predict breast cancer in otherwise healthy women without a history of X-ray exposure.

As part of Specific Aim 1, I completed my proposed case-control study. I obtained blood samples from women who long ago had received radiation to the chest as part of the treatment of Hodgkin’s disease. Some of these women had subsequently developed breast cancer (“cases”); others had not (“controls”). I collected blood samples from 19 “cases” and 32 “control” women who were matched to the cases. The cells in these blood samples were grown in the lab, and exposed to radiation. The cells were then analyzed on microarrays, and data on the response of thousands of genes was collected using specialized software programs. Using statistical software, I then analyzed whether or not there was a fundamental difference in the way the “case” cells handled radiation compared to the “control” cells. Unfortunately, as of this writing, I have not been able to find a set of genes whose responses could reliably differentiate between the “cases” and the “controls”.

Clearly, this methodology requires extensive refinement. We are in the process of examining other features of these cells to help broaden the search for informative genes. It is possible that some of the difficulty is with the statistical software, and not the primary data. We therefore plan to try other statistical methods to address this. While I certainly hope this troubleshooting is successful, it is also possible that my original hypothesis is flawed. The development of breast cancer after exposure to X-rays may not be related to an inherent abnormal genetic response to radiation. Because we have not been able to satisfy Specific Aim 1, work has not yet begun on Specific Aim 2.

Other members of my mentor’s laboratory will be continuing this research, as I will be moving on to another position. I thank the California Breast Cancer Research Program for their generous support of this work.