these descriptors with no corresponding quantitative percentages for breast composition [ 5].
The often-quoted statement that women with dense breasts
have a four- to six-fold greater risk of developing breast cancer
refers only to those with extremely dense breasts as opposed to
those with completely fatty breasts [ 5]. However, because 10 percent of women have extremely dense breasts and only 10 percent
have fatty breasts, the statement is incorrect when applied to the
remaining 80 percent of the population. The relative risk for the 40
percent of women with heterogeneously dense breasts compared
with the 40 percent with scattered densities is less than 1. 5 times as
high [ 5]. Compared with women with breasts of average density—
that is, those with approximately 50-percent dense tissue—the relative risk for women with heterogeneously dense breasts is less
than 1. 2 times as high, and the relative risk for those with extremely dense breasts is less than 2. 1 times as high. In light of these data,
breast density is not a major risk factor except for women with extremely dense breasts, who constitute 10 percent of the population.
Compared with its relatively modest influence on risk, breast
density has a much greater effect on detection sensitivity of mammography, particularly for early breast cancer. For example, the
ACR Imaging Network (ACRIN) 6666 Investigators trial of
screening breast ultrasound showed that mammographic sensitivity among women with dense breasts is only 50 percent [ 6].
This restricted sensitivity of mammography explains why even in
the Swedish seven-county screening study (with all breast densities combined) breast cancer mortality was reduced by only 45
percent [ 7]. This limitation of mammography has led to the investigation of other imaging modalities, such as ultrasound and MRI,
for supplementary screening of women with dense breasts.
Multiple studies conducted in the 1990s showed substantially
improved detection when ultrasound was added to mammography as a supplementary screening modality for women with
dense breasts. In light of these encouraging studies, the ACRIN
6666 Investigators trial was initiated in 2005 and provided three
annual screenings using both mammography and handheld ultrasound for women with dense breasts who were at high risk [ 6, 8].
The most recent ACRIN results showed that the addition of ultrasound to mammography increased the cancer detection rate by
54 percent [ 6]. However, false-positive biopsies, owing to findings
of ultrasound alone, increased disproportionately. Cancer was
found in 34.6 percent of all biopsies performed on the basis of
mammography versus 14.0 percent of biopsies performed on the
basis of either mammography or ultrasound. In the context of desirable clinical outcomes from screening, a positive predictive
value (PPV) in the range of 20 to 40 percent is considered acceptable; a PPV of below 14 percent is considered too low.
If ultrasound screening were to be performed on all women with
dense breasts, rather than on only women with dense breasts who
Legislation pertaining to dense breasts has
created a dilemma by requiring radiologists
to recommend supplementary screening for
which there is no reimbursement.
The American Cancer Society recommends
supplementary screening with MRI for
women with a 20 percent or higher lifetime
risk of developing breast cancer.
are at high-risk as in the ACRIN 6666 Investigators trial, PPV as
determined by biopsy results would probably be even lower. This
has been confirmed by two separate studies in Connecticut involving screening ultrasound of women in low-, medium-, and high-risk
groups with dense breasts [ 9, 10]. Cancer detection rates were increased by 3.2/1,000, but the PPV for biopsies performed on the
basis of ultrasound alone was only 6. 5 percent.
Some may consider the true-positive biopsy rate to be too low
to justify supplementary screening ultrasound. However, ultra-sound-guided biopsies are minimally invasive, are performed rapidly (in < 30 minutes), and result in negligible post-biopsy scarring.
Many women consider detection of one cancer for every 10 biopsies to be acceptable. Some third-party payers, however, may see
the consequent costs per cancer detected to be too high.
Nevertheless, further experience with screening ultrasound may
allow adjustment of biopsy criteria so that less-suspicious lesions
could be followed clinically, rather than biopsied.
There are many reasons automated whole-breast ultrasound
screening (ABUS) would be preferable to handheld transducers
if screening ultrasound were to be widely performed. ABUS examinations have the following advantages:
1. Faster to perform ( 10 minutes versus 20 minutes for
hand-held ultrasound of both breasts)
2. Highly standardized and only minimally operator dependent,
ensuring a technically high-quality study for every patient
3. Able to be performed by any mammography technologist,
general ultrasound technologist, or even a technical assistant (unlike handheld breast ultrasound, which requires a
specialized breast ultrasound technologist or a radiologist
with expertise in breast imaging)
4. Can be read in batches by the radiologist because images
are obtained of each breast in its entirety (in contrast to
screening with a handheld transducer, in which only images of potential lesions are recorded).
To date, there are no published prospective studies to compare
detection rates and false-positive biopsy rates for ABUS versus
handheld transducers. It is also not clear how many lesions detected with ABUS need to be subsequently reimaged with a handheld
transducer. Nevertheless, ABUS screening is highly promising;
several units are now commercially available, and one has already
received U.S. Food and Drug Administration (FDA) approval.
More than 10 studies in which women at extremely high risk
having a full range of breast densities were screened with both
mammography and MRI have shown that the addition of MRI
doubled the cancer detection rate beyond that from mammography alone [ 11]. This increase of 100 percent was far greater than