Understanding the role of mammographic density in a population based breast cancer screening program: A step towards stratified screening for breast cancer in Norway?

Abstract

Mammographic density represents the amount of the epithelium and fibrous tissue in the breast and refers to the radiographic density of the breast visible on mammography. The epithelium and fibrous tissues are radiodense, and appear as white or light gray areas on a mammogram, whereas the fatty tissue is radiolucent, and appears as black or dark gray areas. Mammographic density has been shown to be an independent risk factor for breast cancer. Women with high density (>75% dense tissue) have a four- to six-fold increased risk of breast cancer compared to women with entirely fatty breasts (<25% dense tissue). High mammographic density also significantly decreases the sensitivity of mammography. Furthermore, dense breasts can negatively influence early performance measures of a screening program, resulting in a higher recall rate, missed tumors due to masking and an increased risk of interval cancer. In this thesis, I examined the role of mammographic density in breast cancer screening in Norway. The goal was to contribute with knowledge that would be helpful in determining whether mammographic density could be used for future stratified screening. The articles in this thesis are based on information about women screened in the Norwegian Breast Cancer Screening Program. We investigated whether mammographic density affects early performance measures of breast cancer screening, which compression parameters are associated with density estimates, and how different mammographic density classifications correspond to each other. Mammographic density was assessed both subjectively, by breast radiologists working in the screening program, and objectively, using a fully automated method. We found that positive predictive values for recall examinations and invasive procedures decreased with increasing mammographic density among women screened in the program. We also determined that high mammographic density was associated with large (>15 mm) tumor size and positive lymph node status in women with screen-detected invasive breast cancer. Further, we identified correlations between compression force, pressure, compressed breast thickness, breast volume and fibroglandular volume, and volumetric breast density. The strongest associations were observed between compression pressure, breast volume and fibroglandular volume, and between compressed breast thickness and volumetric breast density. We found that subjective mammographic density classifications used by the screening program in Norway corresponded well to estimates of the fully automated density assessment method. The results of our studies indicated that mammographic density could be a useful tool in stratification of breast cancer screening. However, the obtained evidence is currently not sufficient to support stratified screening for breast cancer based on mammographic density in Norway.