Biomedical Engineering Reference
In-Depth Information
meta-analysis performed in 2006 reviewed data from 42 studies and supported the association between
PMD and breast cancer risk [4].
The risk associated with high PMD is significant and robust among many age and ethnic groups and is
independent of which breast was used for estimation of PMD [3,5]. Most studies place the increased risk of
a PMD more than 75% at four to six times the risk of a PMD <10% [2,6]. This correlation is density-depen-
dent with increasing PMD related to increased risk. Additionally, the risk associated with PMD persists
for 8−10 years from the time of the mammogram [4]. While PMD is associated with age and body mass
index (BMI), it represents an independent risk factor even after adjustment for these and other risk factors.
Early in the study of PMD and breast cancer risk, much of the correlation was thought to be due to
masking; many researchers believed that cancers were difficult to detect in patients with high PMD
due to the similar appearance of cancers and high-density tissue. However, studies that have shown
an increased risk of breast cancer related to high PMD up to 10 years after the initial screening, which
challenges the masking hypothesis and suggests that high PMD has a distinct and real correlation with
breast cancer risk [3]. It remains true, however, that women with high PMD have two disadvantages—
the higher risk associated with PMD as well as the difficulty of imaging cancers in highly dense breast
tissue, which can delay detection of breast cancers [4].
Clinically, the association between PMD and breast cancer has not been fully utilized. The most
widely used method of predicting breast cancer risk in the clinic is currently the Gail model. This model
uses six risk factors—age, age at menarche, age at first live birth, number of first-degree relatives with
breast cancer, number of biopsies, and presence of atypia on biopsy [3]. However, PMD is more strongly
correlated with breast cancer risk than any of the factors in the Gail model. Current efforts are under-
way to incorporate additional risk factors, including PMD, into clinical risk prediction models.
In addition to predicting breast cancer risk, PMD may also be a useful marker of therapy efficacy.
PMD can change in response to hormone therapy (increase), menopause (decrease), and tamoxifen ther-
apy (decrease). While a causal connection has not been shown between tamoxifen treatment, reduced
PMD and reduced breast cancer risk, it may be possible to use PMD as a marker of therapy efficacy, thus
assessing treatments early and easily [1]. This might be especially important for high-risk patients who
take tamoxifen prophylactically and do not have a cancer lesion to monitor. PMD could also be used to
determine how often a patient should receive mammograms, how likely the mammograms might be to
miss early disease, or if other tests should be used to estimate breast cancer risk.
Before PMD is used routinely in a clinical setting, improvements in and standardization of imaging
must be established. Radiologists often use subjective tools that lack automation or quantification, making
comparisons between mammographic images difficult. New methods are being developed including semi-
automated or user-assisted procedures. In particular, the development of a standardized classification sys-
tem, the Breast Imaging Reporting and Data System (BI-RADS) of the American College of Radiology
(ACR) allows comparison of clinical findings across users and treatment sites. The BI-RADS assessment
includes numerical description of mammographic density from ACR 1 (fatty) to ACR 4 (extremely dense).
(For a recent review, see [7]) Additionally, other imaging modalities could be used to better understand
and quantify PMD such as ultrasound, magnetic resonance imaging, or full-field digital mammography
(FFDM) [3,5]. These techniques will provide more complete and standardized measurements that will
allow for more accurate predictions of breast cancer risk and more useful suggestions for treatment options.
Because of the link between PMD and breast cancer risk, it is important to study the mechanisms
underlying the correlation. Mechanistic understanding will both improve the ability to use PMD as a
risk predictor, as well as improve the choice of appropriate therapies that take into account the effect of
the connective tissue environment on breast cancer progression.
Mammographic density is most strongly linked to an increase in stromal collagen [8], but a direct
causal link between density and breast cancer formation has not been established in humans. A func-
tional role for increased collagen in breast cancer progression is suggested by a transgenic mouse model
of higher collagen density, where increased stromal collagen increases tumor formation and produces
more invasive cancers [9]. The effect is at least in part a direct effect of the increased extracellular matrix