1. Carlson, Robert H.

Article Content

Women with mammographically dense breast tissue are known to be at increased risk for breast cancer, regardless of hormone use. Assessing the degree of breast density with automated mammography may be useful as an indicator of breast cancer risk, said Jennifer A. Harvey, MD, Professor of Radiology and Head of Breast Imaging at the University of Virginia Cancer Center, reporting a study at the San Antonio Breast Cancer Symposium.

Figure. No caption a... - Click to enlarge in new windowFigure. No caption available.

"We've known for a long time that breast density is a risk factor for breast cancer, but until we developed these automated methods of measuring that are consistent and not categorized, that hasn't really been helpful in a risk model. But by adding this automated form, that's what has done it."

Figure. No caption a... - Click to enlarge in new windowFigure. No caption available.

The automated mammogram gives a percentage of the breast that is occupied by breast tissue.


She and her colleagues evaluated mammograms for approximately 3,400 women, who were or were not diagnosed with breast cancer. Breast density was calculated using the automated software program, and additional risk factor information was self-reported by study participants or taken from existing records.


The study showed that including breast density into other models, such as the Gail and Tyrer-Cusick models, and adding in a woman's other risk factors, provides a more accurate reading of their individual risk.


Harvey said the Gail model discrimination index (assessed using area under the receiver operating curve) is 0.6, and the Tyrer-Cusick, which she said is the best model available, performs at only 0.74 discrimination (1.0 would be perfect discrimination of dense tissue.)


"We've pushed this by adding a volumetric, automated measure of breast density to a woman's other risk factors, to get up to 0.86 (C-statistic 0.86)," Harvey said. "I'm hoping that this information can help women make better health decisions.