Breast density and breast cancer risk

Breast cancer is the most common cancer in women. Mammography is the only imaging technique that has been proven to reduce breast cancer mortality. The sensitivity of mammography for a nearly fatty breast is approximately 80%; for a dense breast, however, sensitivity is reduced to 40%. 

Recent reports on the increase of breast cancer risk and breast density call for a review of the developments in the measurement of breast density and the additional radiological imaging methods for improving the detection of breast cancer.

The mammographic image of the breast reflects the tissue composition of the breast – epithelial, connective and adipose tissue. The amount and distribution of the different structures depend on several factors, such as age, parity status and body weight. These factors account for only 20-30% of the variation in percentage of mammographic density observed in the population.(1) 

Pathological research by Bartow et al showed that a greater percentage of mammographic density was associated with a significantly higher total nuclear area of both epithelial and non-epithelial cells and a greater proportion of the area of collagen.(2) This collagen is responsible for 29% of the variance in percentage of density. 

A tumour starts in the glandular tissue, but the growth factors and the stromal matrix proteins in the breast tissue also have an impact on the mammary carcinogenesis. 

Li et al reported that breast tissue of extensive densities had a greater nuclear area (p=0.007) as well as a larger stained area of total collagen (p=0.003) and a greater stained area on immunohistochemistry for tissue inhibitor of metalloproteinase (TIMP)-3 and insulin-like growth factor (IGF)-I. IGF-I is also influenced by age and parity in the same way as these factors are associated with mammographic density.(3) This illustrates the complex interaction of different factors on the composition of the breast.

Wolfe first described a relation between density and breast cancer risk.(4) More than 10 studies (six case-control studies and four cohort studies) conducted between 1982 and 1995 found an increased risk for breast cancer in the category with the most extensive dense tissue that was 1.8 to six times higher than that with the least extensive dense tissue.(5) 

McCormack and dos Santos Silva reviewed a systematic meta-analysis of data for more than 14,134 cases and 226,871 non-cases from 42 studies and concluded that the percentage of mammographic density (PMD) was consistently associated with risk of breast cancer. 

This association was stronger in studies in the general population than in symptomatic patients and stronger for quantitative measurements than for Wolfe’s categories. The breast cancer risk cannot be explained by the masking of cancers by dense breast tissue.(6)

In order to investigate the presence of additive genetic heritable factors, studies on monozygotic and dizogyotic twins in the US and Canada were set up and concluded that the population variation in the percentage of dense tissue on mammography at a given age has high heritability. Finding the genes responsible for this phenotype would be important in understanding the cause of this disease.

Different methods are used to assess the mammographic density. In 1976, Wolfe described four densities (see Table 1). The goal of this classification by John Wolfe was initially to determine whether certain mammographic patterns were associated with an increased risk for breast cancer. A fifth classification was suggested by Wolfe, but this classification (QDY) was not widely used. However, questions about this relationship with breast cancer risk have been raised.

Another classification of mammographic parenchymal patterns, called the Tabar classification, had been recommended in recent years, but so far it has not been used frequently. 

In 1998, the Breast Imaging Reporting and Data System (BI-RADS) described four categories, depending on the percentage of dense tissue and not linking this pattern to the risk of breast cancer (see Table 2).(7)

The previous categories are quite subjective. In order to link a percentage of breast cancer risk to a percentage of mammographic density, more correct descriptions of breast density are necessary. Therefore, computer-assisted methods of measurements based on interactive thresholding were developed. 

Cumulus is such a program: an observer places thresholds at the edge of the breast and at the edge of the density and the areas are recorded and calculated by the computer.(8) Due to the introduction of digital mammography and – more specifically – tomosynthesis, information of the digital data can be used for the quantification of the density of the breast. 

All these different methods have advantages and disadvantages. The question arises: what is the reliability between the readers? Some reports showed a modest reliability for the BI-RADS, intermediate for quantitative estimation and good for Cumulus.(9) 

However, a recent study on the inter-/intra-rater agreement on mammographic density measurement in Spain by Pérez-Gómez et al concluded that the quadratic-weighted kappa values for inter-/intra-rater agreement were excellent (higher than 0.80).(10) A short review of our own data, where the rating of density between first and second reader in a screening programme is investigated, showed a very good correlation between the readers: 80% agreement on a population of 183,830. 

Although high density is associated with an increased risk of breast cancer, it also lowers the sensitivity of mammography for the detection of breast cancer. 

Digital mammography has several advantages over film-screen mammography (FSM). Despite the lower spatial resolution, several studies showed that digital mammography is as accurate as FSM in the diagnosis of breast cancer. 

The DIMST trial reported similar overall accuracy of FSM and full-field digital mammography (FFDM), but for heterogeneously dense or extremely dense breasts and for patients younger than 50 years, FFDM showed a better accuracy than FSM.(11) Some studies reported a higher detection of in situ carcinoma with FFDM.(12)

For more than 20 years, ultrasound (US) has been an additional investigation technique to mammography in the diagnostic evaluation of the breast. The place of US as an adjunct to mammography in the screening of asymptomatic women is, in terms of mortality reduction, not yet proven. 

US adjunct to mammography in dense breasts can detect 0.27% to 0.52% more breast cancers. However, all studies also report a higher recall rate, biopsy rate and false-positive rate, which remain the main limitation of additional US next to the time-consuming aspect of the investigation. 

Recent reports in which US is added to mammography in dense breasts reported a lower number of interval cancers (at one-year follow-up). This suggests a possible benefit from US screening in women with mammographically negative dense breasts.(13) Larger prospective studies on the cancer detection of US in screening are ongoing.

Magnetic resonance imaging (MRI) is another additional investigation technique for the evaluation of dense breasts. MRI has a superior sensitivity to mammography, but the specificity is low and counts for a lot of false positives. Since MR imaging is three-dimensional, it may be more accurate in depicting the amount of fibroglandular tissue. 

When cases of breast cancer were matched with normal controls of similar patient ages and MR imaging dates, breast parenchymal enhancement (BPE) was found to be a highly significant predictor of breast cancer. The odds of breast cancer increased with an increasing BPE level.(14) 

As high-risk women are undergoing an MRI, BPE can serve as an additional tool for risk stratification. On the other hand, this BPE is also influenced by the hormonal status or by hormonal therapies, and the effect on the breast cancer risk needs to be investigated. 

Recently, tomosynthesis has found its way in the radiological imaging of the breast. The primary potential of breast tomosynthesis is the reduction of the overlap of normal breast structures with breast abnormalities. This can lead to better lesion detection and delineation and finally a reduction of the recall rate in a screening population. 

In recent studies, better visibility of masses and areas of architectural distortion has been shown for tomosynthesis.(15) Smaller lesions that are normally not visible in a two-dimensional imaging of the breast now become visible.

Implementation of newer techniques, such as dual energy mammography, has just started.

Breast density has proven to be an important factor for the risk of breast cancer. Correct classification of the density of the breast is important and, with digital mammography, more quantitative measurements will be possible. 

Radiologists need to be aware of this higher risk: as the sensitivity of mammography is low in this group of patients, additional imaging such as ultrasound is recommended. The place of additional ultrasound in the screening of women with dense breasts needs to be considered.

The place of newer techniques such as tomosynthesis, more specifically in screening, in dense breasts is now under investigation to substantiate an observed lower recall rate and higher detection of small lesions. Owing to the low specificity of MRI, it cannot be used routinely in the evaluation of dense breasts.

1. Boyd NF et al. Mammographic density: a heritable risk factor for breast cancer. Methods Mol Biol 2009;472:343–60.
2. Bartow SA et al. Correlations between radiographic patterns and morphology of the female breast. Rad Patterns Morph 1997;13:263–75.
3. Li T et al. The association of measured breast tissue characteristics with mammographic density and other risk factors for breast cancer. Cancer Epidemiol Biomarkers Prev 2005;14(2):343–9.
4. Wolfe JN. Breast patterns as an index of risk for developing breast cancer. Am J Roentgenol 1976;126:1130–9.
5. Boyd NF et al. Heritability of mammographic density, a risk factor for breast cancer. N Engl J Med 2002;347(12):886–94.
6. McCormack VA, dos Santos Silva I. Breast density and parenchymal patterns as markers of breast cancer risk: A meta-analysis. Cancer Epidemiol Biomarkers Prev 2006;15:1159–69.
7. BI-RADS (1998) Illustrated breast imaging reporting and data system. Updated draft 8/11/2011. Am College of Radiology, Resten/VA.
8. Byng FW et al. The quantitative analysis of mammographic densities. Phys Med Biol 1994;39:1629–38.
9. Boyd NF et al. Mammographic density. http://breast-cancer-research.com/supplements/11/S3/S4
10. Pérez-Gómez B et al. Women’s features and inter-/intra-rater agreement on mammographic density assessment in full-field digital mammograms (DDM-SPAIN) Breast Cancer Res Treat DOI 10.1007/s 10549-011-1833-3
11. Pisano ED et al. Diagnostic performance of digital versus film mammography for breast-cancer screening. N Engl J Med 2005;353(17):1773–83.
12. Del Turco MR et al. Full-field digital versus screen-film mammography: comparative accuracy in concurrent screening cohorts. Am J Roentgenol 2007;189(4):860–6.
13. Houssami N,  Ciatto S. The evolving role of new imaging methods in breast screening. Prev Med 2011 DOI: 10.1016/j.ypmed.2011.05.003
14. King et al. Background parenchymal enhancement at breast MR imaging and breast cancer risk. Radiology 2011;260(1):50–60.
15. Gur D et al. Digital breast tomosynthesis: observer performance study. Am J Roentgenol 2009;103:586–91.