Breast cancer is a heterogeneous disease with many histologic subtypes. Even if mammography is the best proven method for finding breast cancer in its early stages, it is only natural that a single method will never be able to detect all of the subtypes. In addition, the composition of the breast tissue surrounding the lesion(s) will significantly influence our ability to demonstrate the presence of a benign or malignant lesion. It has long been recognized that a cancer is more likely to be missed if it is located in a dense breast.
Dr. Hixson
is a Women's Imaging Specialist in Chattanooga, TN.
Mammography is the most effective screening modality for
detecting early breast cancer. A landmark study by Tabár et al
1
demonstrated a 42% reduction in breast cancer deaths in women who
were screened versus those who were not screened. Dr. Tabár states,
our goal is to detect invasive tumor masses when they are as small
as possible, preferably <10 mm in diameter. Women with
mammographically detected 1- to 10- mm invasive breast carcinoma
have a 93% 16-year survival.
2
Breast cancer is a heterogeneous disease with many histologic
subtypes. Even if mammography is the best proven method for finding
breast cancer in its early stages, it is only natural that a single
method will never be able to detect all of the subtypes. In
addition, the composition of the breast tissue surrounding the
lesion(s) will significantly influence our ability to demonstrate
the presence of a benign or malignant lesion. It has long been
recognized that a cancer is more likely to be missed if it is
located in a dense breast.
Kolb et al
3
have performed an important study of 27,825 mammographic
examinations and 13,547 bilateral whole-breast screening ultrasound
examinations in conjunction with the mammograms. Ultrasound was
able to detect cancers missed the same day by mammography, and
these were most frequently in dense breasts. Their results showed a
mammographic sensitivity of 78% for all breast density categories.
The sensitivity for mammography was 98% for BIRADS density category
1, breast composed almost entirely of fat. However, the sensitivity
was only 48% for BIRADS density category 4, extremely dense breast.
This is a useful study that documents the progressive decrease in
mammographic sensitivity as the breast density increases. However,
well-trained radiologists using state-of-the-art film screen
mammography should achieve higher sensitivities than reported in
this study.
Birdwell et al
4
studied 115 cancers missed by screening mammography. The majority
of missed cancers (57%) were located in the middle of the breast
where compression is suboptimal with a conventional horizontally
oriented compression paddle. A total of 28% of the missed cancers
were located in the posterior third of the breast, and 11% were
missed in the anterior third of the breast, which receives
negligible compression with a conventional paddle.
The American College of Radiology (ACR) is striving to improve
mammographic image quality so suboptimal images will not cause
cancer to be missed. The ACR reports that common causes for failure
of the clinical imaging evaluation process are unsharpness, poor
contrast, and failure to spread apart the fibroglandular tissues.
5
Inadequate compression is usually the underlying cause for all of
these deficiencies. Better compression prevents motion unsharpness
by reducing breast thickness to allow shorter exposure times and by
immobilizing the breast. Patient motion is the most common cause of
readily detectable image unsharpness.
5
The previous limitations of film-screen mammography led to a
search for alternative screening modalities. Whole-breast digital
mammographic units have been developed and are now available.
However, no studies to date have demonstrated any increased
sensitivity of whole-breast digital mammography. Furthermore, the
performance of film-screen mammography has significantly improved
due to the recent introduction of several new products.
Improved compression paddles
A standard screening mammogram is obtained by compressing a
cone-shaped structure between two flat, parallel surfaces. It is
not possible to achieve uniform compression from the chest wall to
the nipple with a conventional compression paddle. Images obtained
by expert mammographers commonly exhibit subtle motion blur in the
mid-breast and anterior breast. Conventional screening paddles are
oriented horizontally and produce optimum compression of only the
thick posterior portion of the breast. The middle third of the
breast is incompletely compressed, and the anterior breast receives
negligible compression.
Compression paddles have recently been developed that produce
increased compression of the anterior half of the breast. These
paddles typically rotate as compression is applied, causing the
nipple end of the paddle to go downward and the chest wall end to
go upward. As the compression force increases, there is increasing
compression of the breast anteriorly with a corresponding decrease
in compression applied posteriorly. However, this can be
detrimental since most cancers are located in the posterior half of
the breast, which is also the most common location for a cancer to
be missed.
The New S.O.F.T. Paddle™
The S.O.F.T. Paddle™ developed by American Mammographics
(Chattanooga, TN) is different. It has a gently curved compression
surface that slopes downward toward the nipple. It does not rotate.
As the paddle is lowered, there is almost simultaneous contact with
the breast from the chest wall to the nipple. Table 1 shows the
benefits of S.O.F.T. Paddle™. Optimum compression of all areas of
the breast is achieved. S.O.F.T. Paddle™ produces a significant
improvement in image quality, particularly in dense breasts
(Figures 1 and 2). Dr. László Tabár is now using the S.O.F.T.
Paddle™ for screening mammograms interpreted at Falun Central
Hospital in Falun, Sweden. Dr. Tabár says of the S.O.F.T. Paddle™,
the image details that are so essential in making the first and
most important decision (callback or normal) are visualized so well
and with such a sharpness as I have never experienced it
previously. (L. Tabar, personal communication).
Other helpful tools
The use of darker films and brighter view boxes is now
recommended, and this is essential for proper interpretation of
dense breasts. The ACR QC Manual recommends an optical density (OD)
of at least 1.40 for the background density for the phantom image.
5
A background OD of 1.80 is much better and is optimal for dense
breasts. The OD should be set high enough to prevent the OD from
falling below 1.0 in the densest area of any breast. The ACR
requires a minimum viewbox luminance of 3000 nits.
5
However, a luminance of at least 5000 nits is necessary to evaluate
the darker peripheral areas of a film. The cones in the retina are
responsible for perceiving fine details, and they function best at
high light levels.
Mammographic films and screens have significantly improved in
the last several years. For example, the Kodak MIN-R 2000 film and
MIN-R 2000 screens produce very high-quality images. The MIN-R 2190
screen is 25% faster. It can be used routinely or only selectively
for dense breasts to improve contrast and reduce motion blur.
Contrast is further increased with Kodak X-OMAT EX II developer and
companion fixer. The use of the 2190 screen and the S.O.F.T.
Paddle™ together can produce considerable improvement in image
quality, especially with dense breasts (Figure 2).
There have been many significant improvements in mammographic
equipment. For example, the Diamond Unit (Instrumentarium Imaging,
Milwaukee, WI) incorporates numerous technological advances. This
unit was selected as the best new product at RSNA 2000. It contains
the Diamond X-ray tube with a special microfocusing cup that
improves spatial resolution. It also utilizes a proprietary grid
(ROC Grid) with a 6:1 ratio and a low radiation absorption
interspace material, which provides the same X-ray transmission as
a 5:1 grid. Figure 3 illustrates the improved image quality when
the Instrumentarium Diamond X-ray tube, the ROC grid, and the
S.O.F.T. Paddle™ are used together.
Discussion
Early detection by mammographic screening is a proven method of
decreasing mortality from breast cancer. The present generation of
physicians is the first to offer a significantly better outcome for
cancer patients, by significantly decreasing the rate of advanced
cancers through early detection and treatment in an early stage.
Yet, paradoxically, there is a growing shortage of radiologists who
are expert in mammography, and many radiologists would prefer not
to have to read mammograms. The reasons are probably reflected in
the results of a recent survey of radiology residents from 211
accredited radiology residencies in the U.S. and Canada by Bassett
et al.
6
Their results showed that 63% of the residents would not want to
spend one-fourth or more of their time interpreting mammograms. The
most common reasons were: not interesting enough, fear of lawsuits,
and too much stress. One surprising result was that 87% of the
residents rated interpretation of mammograms as being more
stressful than other types of imaging. This is partly
understandable, taking the complexity of breast diseases into
account, the need for learning histopathologic-mammographic
correlation, the need for interdisciplinary cooperation, the steady
demand for making perfect decisions, and because of the need for
continuous quality and product control. There are a few solutions
to this problem. One is continuous, high-quality training and
further education. In addition, improvement in image quality and
standardization of the method would help to make mammography more
attractive for the future generation of professionals.
Conclusion
Simple, cost-effective tools and inventions have recently become
available that can significantly improve image quality,
particularly in dense breasts. The use of these products should
result in fewer missed cancers and fewer callbacks for supplemental
imaging. Better quality images also should reduce the time for
interpretations.