Women who have breasts that appear dense on a mammogram are at increased risk for breast cancer. The reason for the increase in risk is unknown. Denser breasts may be made of tissue that is more likely to become cancerous; however, it is more difficult for a physician to see signs of cancer in a woman with breasts that appear dense on a mammogram. Researchers have developed mathematical equations that can predict a woman’s chance of developing breast cancer. None of these predictions is perfectly accurate or based on information about breast density. Including breast density in tools that predict cancer might improve the accuracy of those predictions.

Why did the researchers do this particular study?
To develop a new equation that uses information about breast density to predict a woman’s chance of developing breast cancer.

Who was studied?
About 1.1 million women who had at least 1 mammogram in which breast density was measured and reported.

How was the study done?
The researchers gathered information about the women’s risk factors for breast cancer, including breast density. They used the information to create a mathematical equation and predicted the women’s chance of developing cancer. To test the accuracy of the predictions, the researchers compared the predictions with what actually happened to the women during the 5 years after their mammogram. They also compared the predictions with those from other equations that do not use breast density.

What did the researchers find?
The predictions from the equation that used breast density closely matched what actually happened to the women. However, the equation did not always accurately predict which individual women would and would not develop cancer. Also, the new equation was not always clearly better than existing equations. Compared with those other equations, the breast-density equation sometimes predicted a higher risk for women who did not develop cancer. It also sometimes predicted a lower risk for women who did develop cancer.

What were the limitations of the study?
Complete information about what happened to the women was available for only a portion of the original 1.1 million women.

What are the implications of the study?
A risk prediction equation that includes measurements of breast density accurately predicts breast cancer risk in some women. It seemed to be less accurate in other women. As a result, it requires further study before individual women can use it to predict their personal chances of developing breast cancer.

The full report is titled “Using Clinical Factors and Mammographic Breast Density to Estimate Breast Cancer Risk: Development and Validation of a New Predictive Model.” It is in the 4 March 2008 issue of Annals of Internal Medicine (volume 148, pages 337-347).
The authors are J.A. Tice, S.R. Cummings, R. Smith-Bindman, L. Ichikawa, W.E. Barlow, and K. Kerlikowsk

American College of Physicians