Magnetic resonance imaging of breast lesions - A pathologic correlation

Article (PDF Available)inBreast Cancer Research and Treatment 103(1):1-10 · June 2007with4 Reads
DOI: 10.1007/s10549-006-9352-3 · Source: PubMed
Abstract
Magnetic resonance imaging of the breast is useful in assessing breast lesions. An understanding of the pathologic characteristics of the tumors may help to understand these magnetic resonance imaging observations. Large lesional size (>10 mm), ill-defined margin, and irregular outlines are associated with malignancy. These correlate with the pathological features of breast tumor, characterized by rapid growth rate, large size, and infiltrative growth pattern, invasion into stroma resulting in desmoplasia, and hence irregular outline and margin. The detection and estimation of tumor extent of invasive lobular carcinoma is problematic, even with magnetic resonance imaging, which is considered the most sensitivity. This inaccuracy likely derives from the characteristic linear, single cells infiltration growth pattern of the tumor, which is also often underestimated by clinical examination. Estimation of tumor extent after neoadjuvant chemotherapy is also essential but problematic by imaging, as the shrunken tumor becomes fibrotic, with stromal hyalinization, diminished microvasculature and tumor break up causing size underestimation. Non-enhancement of breast tumors occurs in about 8% of cases correlates with diffuse growth pattern, particularly of infiltrative lobular carcinoma. The observation of disproportionately high non-enhancing ductal carcinoma in situ remains an enigma. Finally, early rim enhancement correlates with small cancer nests, low ratio of peripheral to central fibrosis and high ratio of peripheral to central microvessel density. These may be related to increased vascular endothelial growth factor mediated increased microvessel density as well as increased permeability, which manifest as increased rapid contrast uptake and dissipation.
Abstract Magnetic resonance imaging of the breast is
useful in assessing breast lesions. An understanding of
the pathologic characteristics of the tumors may help
to understand these magnetic resonance imaging
observations.Large lesional size (>10 mm), ill-defined
margin, and irregular outlines are associated with
malignancy. These correlate with the pathological
features of breast tumor, characterized by rapid growth
rate, large size, and infiltrative growth pattern, invasion
into stroma resulting in desmoplasia, and hence irreg-
ular outline and margin. The detection and estimation
of tumor extent of invasive lobular carcinoma is
problematic, even with magnetic resonance imaging,
which is considered the most sensitivity. This inaccu-
racy likely derives from the characteristic linear, single
cells infiltration growth pattern of the tumor, which is
also often underestimated by clinical examination.
Estimation of tumor extent after neoadjuvant chemo-
therapy is also essential but problematic by imaging, as
the shrunken tumor becomes fibrotic, with stromal
hyalinization, diminished microvasculature and tumor
break up causing size underestimation. Non-enhance-
ment of breast tumors occurs in about 8% of cases
correlates with diffuse growth pattern, particularly of
infiltrative lobular carcinoma. The observation of dis-
proportionately high non-enhancing ductal carcinoma
in situ remains an enigma. Finally, early rim enhance-
ment correlates with small cancer nests, low ratio of
peripheral to central fibrosis and high ratio of periph-
eral to central microvessel density. These may be
related to increased vascular endothelial growth factor
mediated increased microvessel density as well as in-
creased permeability, which manifest as increased
rapid contrast uptake and dissipation.
Keywords Breast Æ Cancer Æ Magnetic resonance
imaging
Introduction
Magnetic resonance imaging (MRI) of the breast,
including dynamic enhanced contrast MRI, is an
accurate diagnostic modality in the assessment of
breast lesions since its introduction approximately
20 years ago, especially in the diagnosis of malignant
G. M. K. Tse (&)
Department of Anatomical and Cellular Pathology, Prince
of Wales Hospital, The Chinese University of Hong Kong,
Ngan Shing Street, Shatin, NT, Hong Kong
e-mail: garytse@cuhk.edu.hk
K.-T. Wong Æ A. L. M. Pang
Diagnostic Radiology and Organ Imaging, Prince of Wales
Hospital, The Chinese University of Hong Kong,
Shatin, Hong Kong
D. K. Yeung
Diagnostic Radiology and Organ Imaging, Prince of Wales
Hospital, The Chinese University of Hong Kong, Shatin,
Hong Kong
B. Chaiwun
Department of Pathology, Chiang Mai University,
Chiang Mai, Thailand
A. P. Y. Tang
Department of Radiology, North District Hospital,
Fanling, Hong Kong
H. S. Cheung
Department of Radiology, International Islamic University,
Kuantan, Malaysia
Breast Cancer Res Treat (2007) 103:1–10
DOI 10.1007/s10549-006-9352-3
123
REVIEW
Magnetic resonance imaging of breast lesions—a pathologic
correlation
Gary M. K. Tse Æ Benjaporn Chaiwun Æ
Ka-Tak Wong Æ David K. Yeung Æ Amy L. M. Pang Æ
Alice P. Y. Tang Æ Humairah S. Cheung
Received: 27 June 2006 / Accepted: 19 July 2006 / Published online: 11 October 2006
Springer Science+Business Media B.V. 2006
lesions [15]. This is particularly so when the
assessment is coupled with mammography and ultra-
sonography [6, 7]. Both lesion morphology and
enhancement kinetics are useful parameters in identi-
fying malignant breast lesions. Apart from differenti-
ating between malignant and benign lesions, there are
other specific indications for the use of contrast MRI.
In cases where the patients present with axillary lymph
node metastases, but when the possible occult primary
is not detectable by mammography or ultrasound, MRI
can be useful in the detection of such occult primary
[8]. In addition, for some specific breast tumors, like
invasive lobular carcinoma, MRI is generally consid-
ered more accurate in the assessment of tumor extent
[9, 10]. MRI is generally accepted to be more sensitive
to detect small second primaries in the breast [1113]
as well as to differentiate tumor recurrence from scar
tissue in those patients undergone breast conservation
therapy [14, 15].
There are many issues about MRI observations in
relation to breast tumors, which carry important clini-
cal significance, including MR lesional morphological
characteristics, measurement of tumor sizes, and
enhancement of the tumors. The emphasis of this re-
view will be on the pathological correlations, which
provide the basis for accurate interpretation of MRI in
diagnosis and management of breast carcinoma, even
though many aspects are only partially understood.
Lesional morphological characteristics
Basing on morphology alone, shape and margin type
are the key factors to assess malignancy. Features with
highest positive predictive value for malignancy
include spiculated margin, rim enhancement, and
irregular shape for masses, and segmental or clumped
ductal enhancement for non-mass lesions [16]. Some
authors also assessed the lesional size [17]. In general,
large lesional size (with a diameter >10 mm)
(compared to < 10 mm) (Fig. 1), ill-defined margin
(compared to rounded margin) and irregular, linear,
branching or stellate lesional outlines compared to
regular outlines (round, oval or polygonal) (Fig. 2)
are associated with higher predictive values for
malignancy [17].
These morphological features are readily correlated
with the pathological characteristics of breast tumors.
While basing on lesional size alone, one may not be
able to predict whether a lesion is malignant or benign,
one may extrapolate and postulate that the larger
lesions are likely associated with a more rapid growth
rate, so that within the time frame of a regular interval
clinical examination, the lesion has a greater likelihood
to have attained a larger size. Increased proliferation is
well established in breast cancer, and there are
approaches that are available to pathologists to access
the proliferation rate, including mitotic count which
is embedded in the widely adopted Bloom and
Richardson grading of breast cancer [18], as well as
others like assessing S-phase fraction by flow cytometry
[19] or immunohistochemical staining against antibody
MIB1, directed at the antigen of Ki67 [20]. In general,
Fig. 1 Axial subtracted fat-suppressed T1-weighted MR image
following gadolinium enhancement showing a large phyllodes
tumor with heterogeneously enhancement with ill-defined and
lobulated margin in the right breast (arrow)
Fig. 2 Axial subtracted fat-suppressed T1-weighted MR image
following gadolinium enhancement showing a fibroadenoma
with well-circumscribed margin, round contour and homoge-
neous contrast enhancement in retroaerolar area of right breast
(arrow)
2 Breast Cancer Res Treat (2007) 103:1–10
123
the higher the proliferation rate, the higher grade is the
tumor, and a worse outcome. There are however
exceptions, as there are benign lesions showing rapid
growth rate or large size at presentation, as well as
malignant lesions that are small or slow growing.
Belonging to the former category are benign lesions
that occur during pregnancy, when the internal milieu
has high sex hormonal content. The so-called lacta-
tional adenoma is totally benign, has a rapid growth
rate, and regresses after pregnancy. The other benign
lesions that may occur with rapid growth rate in the
non-pregnant state are juvenile fibroadenoma and
phyllodes tumors, both of which tend to grow to large
size, and may be associated with slightly increased
mitotic activity [21]. On the other hand, low-grade
malignancies, particularly those occurring in the
elderly population, may be very slow growing, and
hence may not present as a large mass. The typical
example that comes to mind is carcinomas with neu-
roendocrine differentiation [22, 23] (Fig. 3) or those
with solid papillary carcinoma component [24]. Hence,
attention has also to be paid to the age of the patients;
in younger patients, even large tumors may be benign,
and by the same corollary, small tumors in elderly
patients should be viewed with suspicion.
Ill-defined margin and irregular shape usually indi-
cate invasive malignancy, as the tissue desmoplasia
resulting from stromal invasion tends to contract be-
cause of the presence of myofibroblasts, resulting in a
radiating shape, hence the name cancer (can-
cer = crab) (Fig. 4). The exception to this rule, again
are derived from specific subtypes of cancers of the
breast. Examples of such specific cancers including
invasive cancers like medullary and mucinous
carcinomas (Fig. 5), which are usually rounded in
shape and possess a smooth margin. As margin irreg-
ularity is most likely due to desmoplasia secondary to
stromal invasion, malignancies at the pre-invasive stage
(carcinoma in situ) tend also to grow in a rounded
manner with regular and smooth outline, surrounded
by an intact albeit attenuated myoepithelial cell layer
and a continuous basement membrane (Fig. 3). This
includes the usual ductal carcinoma in situ as well as
encysted papillary carcinoma (Fig. 6). Conversely,
some benign lesions have a growth pattern that mimics
invasive malignant lesions, with dense stromal fibrosis
within the lesion, characterized by a group of fibro-
sclerosing lesions, with the commonest examples being
Fig. 3 Photograph of a whole mount section showing nodular
proliferation of carcinoma in situ with distension of the ducts by
the carcinoma cells (H&E, 1·)
Fig. 4 Axial subtracted fat-suppressed T1-weighted MR image
following gadolinium enhancement showing an infiltrative ductal
carcinoma with ill-defined and spiculated margin in lateral
portion of right breast (arrow)
Fig. 5 Axial subtracted fat-suppressed T1-weighted MR image
following gadolinium enhancement showing well-defined margin
in a medullary carcinoma of left breast (arrow)
Breast Cancer Res Treat (2007) 103:1–10 3
123
sclerosing adenosis, radial scars or complex sclerosing
lesions. This group of lesions shows typically radiating
growth pattern, with a dense central, radially distrib-
uted fibrosis resulting in architectural distortion, and a
hard consistency on clinical examination, virtually
indistinguishable from cancer (Figs. 7 and 8).
Size measurement of breast tumors
It is generally accepted that MRI has a high sensitivity
in detecting ductal carcinoma of the breast [25, 26].
Breast MRI provides a more accurate assessment of
tumor size than mammography and ultrasound. In a
series of 61 breast cancer [9], mammography and
ultrasound underestimated tumor size by 14% and
18%, respectively, while MRI showed no significant
difference in size compared with pathologic examina-
tion. Other authors found MRI measurements had the
best correlation coefficient, compared to lower corre-
lation for ultrasound and mammography [27].
The detection and estimation of tumor extent of
invasive lobular carcinoma (ILC) is well known to be
problematic, with a low sensitivity in mammography
[28] and in ultrasound, especially for lesions smaller
than 1 cm [29, 30]. Recently many authors have dem-
onstrated that MRI is the most sensitive method in the
detection of ILC, with a sensitivity of 100% [31, 32]. In
many cases, it is essential that radiological estimation
of tumor extent to be done, particularly in cases for
clinical staging before neoadjuvant chemotherapy [33].
In the same study [32], the accuracy of size estimation
was reported to be about 75%, with the remaining 25%
showing either overestimation or underestimation.
This inaccuracy likely derives from the peculiar growth
pattern of the ILC. Histologically ILC is characterized
by a typical linear, single cells infiltration into the
breast stroma and frequently into fat, visible only
under microscopy (Figs. 9 and 10). The extent of tumor
involvement is usually more than that obtained by
clinical examination or palpation of the excised speci-
men. Even with MRI, ILC may show variable patterns,
including solitary mass with irregular margins, multiple
small enhancing foci that may be non-contaiguous or
with interconnecting enhancing strands, or with
enhancing septae only without dominant tumor focus
[34]. These patterns correlate well with the growth
pattern of ILC, which may occasionally form a solid
Fig. 6 Photomicrograph showing an encysted papillary carci-
noma, with proliferation of low-grade carcinoma cells within a
large duct, with the cells adopting a papillary configuration
(H&E, 20x)
Fig. 7 Photomicrograph showing a sclerosing adenosis, with
hyalinization of the stroma resulting in compression of the ductal
element, resulting in an irregular outline (H&E, 40x)
Fig. 8 Higher magnification of Fig. 7 (H&E, 200x)
4 Breast Cancer Res Treat (2007) 103:1–10
123
mass, or small masses interlinked by strands of infil-
trating tumor cells [35].
Another issue on measuring tumor extent is in pa-
tients undergoing neoadjuvant chemotherapy, aiming
at reducing the primary tumor size to allow surgical
excision to be done [36, 37], notably breast conserva-
tion resection as compared to mastectomy. As patients
may show variable levels of responses, ranging from
complete disappearance of the tumor, to partial
response to no response, estimation of the tumor size
become critical during therapy. This in vivo assessment
of tumor response allows the oncologists to tailor the
therapy regime for the patient, omitting the ineffective
agents. Although pathologic assessment remains the
gold standard, this is only applicable to excised speci-
men; during therapy, assessment of tumor response has
to be based on clinical examination or imaging, but it
appears that mammography and sonography are sub-
optimal modalities for this purpose [38, 39]. Contrast-
enhanced MRI has been evaluated as a more reliable
imaging technique for assessment of tumor extent after
neoadjuvant therapy [4043]. There is still, however,
some degree of inaccuracy, up to 30% in some series
[44], and with underestimation occurring more com-
monly than overestimation. Some authors postulated
that this might be related to fibrosis in the tumor bed
post chemotherapy [44]. In one study, using a contin-
uum of histologic changes including the degree of
fibrosis and sclerosis, differential loss of invasive over
in situ components, cytopathic effects of the tumor and
inflammatory cells infiltrate, it was demonstrated that
MRI estimation of tumor sizes in the neoadjuvant
setting was most accurate when there was complete or
absent response. For those tumors with partial re-
sponses, MRI estimation with tumor sizes was less
accurate [45]. This can be well correlated with the
known pathological changes of breast tumor post
chemotherapy. In cases with response, there will be
tumor necrosis, apoptosis, followed by inflammatory
cells infiltrate including acute inflammatory cells, and
histiocytes. In some cases there is granulation tissue
formation associated with capillary proliferation and
stromal fibrosis and hyalinization. The tumor cells may
show degeneration with changes in the tintoral prop-
erty upon histological staining, with nuclear and cyto-
plasmic vacuolations [46]. It is very likely that an
underestimation of the tumor may occur when the
shrunken tumor becomes fibrotic, with stromal hyali-
nization and diminished microvasculature. The loss of
tumor cells will also contribute to the loss of accom-
panying microvasculature, again, resulting in dimin-
ished contrast enhancement and hence
underestimation of tumor size (Fig. 11). Necrosis of
the tumor resulting in fragmentation of the overall
tumor may also result in under estimation of tumor size
on imaging. On the other hand, particularly in the
earlier stage, tumor cell death may result in tissue
organization and granulation tissue formation. As the
latter is associated with neovascularization, and also
the newly formed capillaries tend to be leaky (in-
creased permeability), these may contribute towards
increased enhancement (Fig. 12). Another fact that
may contribute is non-specific enhancement of the
adjacent breast tissue [47], particularly when there is
benign proliferative disease in the background.
Fig. 9 Photomicrograph showing an infiltrating lobular carci-
noma showing cords and strands of malignant cells showing
bland morphology, permeating within a rather fibrotic stroma.
No tumor mass formation is noted (H&E, 200·)
Fig. 10 Higher magnification of Fig. 9, demonstrating the bland
nuclear morghology of the tumor cells, devoid of mitotic activity
(H&E, 400·)
Breast Cancer Res Treat (2007) 103:1–10 5
123
Contrast enhancement on MRI
Non-enhancement of malignant breast lesions
Internal enhancement of breast mass is one of the
important morphologic criteria for detection of
malignant lesion. Masses displaying rim enhancement
is a particular finding for malignancy (Fig. 13). Other
suspicious findings include heterogeneous ductal and
branching enhancement (Fig. 14). Enhancing septa-
tions and central enhancement occur less frequently
in breast carcinoma. The high sensitivity of contrast
MRI in the detection of breast cancer had initially led
some to assume that non-enhancing lesions are be-
nign, such as fibroadenoma that have a high hyaline
content, and may not warrant biopsies [48]. Other
authors have demonstrated that non-enhancement oc-
cur in about 8% of invasive malignant lesions [4952].
While in a minority the non-enhancement was due
to technical reasons including faulty contrast infusion,
motion artifact and non-visualization of tumor outside
the examination field, the majority of non-enhance-
ment was caused by tumor factors. These include the
nature and type of tumor, the growth pattern, and
even strong enhancement of the adjacent tissue rela-
tive to tumor enhancement. Diffuse growth and small
size (diameter of 0.5 cm or less) have been cited as the
underlying reasons for non-enhancement [52]. In an-
other study [53] of 104 tumors, the 9 tumors that did
not enhance include 4 cases with significant lobular
pattern, either mixed with ductal carcinoma or as ILC
alone, and another 3 cases with significant carcinoma
in situ components [53]. It would thus appear that the
growth pattern of tumors with a diffuse pattern may
result in false negativity for contrast enhancement, as
it has been previously demonstrated that lobular
carcinoma was associated with lower angiogenesis
activity than other more common ductal carcinoma
[54]. In addition, other studies have demonstrated that
in lobular carcinoma, the angiogenetic activity as
measured by microvessel density did not correlate
with other prognostic parameters, suggesting that an-
giogenetic activity may play a minor role in lobular
carcinoma, in contradistinction to ductal carcinoma
[55, 56].
Another entity that is likely to produce false nega-
tive results by contrast-enhanced MRI is ductal carci-
noma in situ (DCIS), or cases in association with
extensive in situ component (more than 25% carci-
noma in situ within the tumor). In one study, the sen-
sitivity of MRI detection of ductal in situ spread
adjacent to an invasive ductal carcinoma was about
80% [57], while in another, the detection rate of
extensive in situ components was 71% [58]. These
findings indicate a lower MRI detection rate for DCIS
compared to invasive ductal carcinoma. Pathologically
it has been demonstrated that microvessel density dif-
fers with different grades of DCIS, with a higher mi-
crovessel density in the higher-grade lesions with
comedo necrosis [59, 60], and this is related to VEGF
expression [61]. It would thus be reasonable to postu-
late that for low-grade DCIS, it is likely that absent
contrast enhancement on MRI may be due to the lower
angiogenetic activity.
Fig. 11 Photomicrograph of a carcinoma post chemotherapy,
with loss of the tumor cell mass and the remaining tumor cells
are present singly within the stroma (H&E, 200·)
Fig. 12 Photomicrograph of a carcinoma post chemotherapy,
proliferation of granulation tissue and a lymphocytic infiltrate
permeating the stroma (H&E, 200·)
6 Breast Cancer Res Treat (2007) 103:1–10
123
Dynamic enhancement pattern
Dynamic contrast-enhanced MRI represents a feature
on breast MRI, which supplement morphological
findings for more accurate differentiation between
benign and malignant breast conditions. It has been
reported that early enhancement is suggestive of
malignancy [25, 6264]. In general three types of time-
intensity curves are described for morphologically
suspicious focal breast lesions [65]. These are as follow:
Type I (steady)—curve corresponds to a straight or
slightly curved enhancement pattern with the
enhancement progressively increasing over time;
Type II (plateau)—curve levels off after initial sharp
level of enhancement;
Type III (washout)—curve has a drop in signal inten-
sity after the initial upstroke indicating washout of
contrast.
In general a Type I curve is regarded as benign,
Type II is suggestive of malignancy and Type III is
indicative of a malignant lesion.
Histological correlation has not been extensively
studied. In one interesting study, early rim enhance-
ment was correlated with small cancer nests, low ratio
of peripheral to central fibrosis and high ratio of
peripheral to central microvessel density, whereas de-
layed rim enhancement was associated with a high
degree of peritumoral fibrosis and inflammatory
changes [66]. Early rim enhancement is associated with
high peripheral microvessel density. Intra-tumoral
variation of microvessel density has been well de-
scribed in the literature [67]. Nevertheless, the areas
with high microvessel density are expected to be those
at the invasive edge of the tumor, where most of the
tumor activity is. Also, in many tumors, there are
central areas of fibrosis with or without hyalinization
(Figs. 1518). Both of these biological and histological
characteristics will understandably give rise to a higher
microvessel density in the tumor periphery compared
to the central portion. This is a possible explanation for
the early rim enhancement. It should also be noted that
in tumors with large area of central necrosis, there
might also be rim enhancement particularly in the early
phase. The pathologically demonstrated intratumoral
variability of microvessel density may also result in
further variability in the enhancement pattern in the
Fig. 13 Axial subtracted fat-suppressed T1-weighted MR image
following gadolinium enhancement showing infiltrative carci-
noma with rim enhancement (arrow)
Fig. 14 Axial subtracted fat-suppressed T1-weighted MR image
following gadolinium enhancement showing infiltrative ductal
carcinoma and ductal carcinoma in situ with heterogeneous
ductal and branching enhancement (arrowheads)
Fig. 15 Photograph of a whole mount section showing a
carcinoma with prominent central fibrosis, and with the carci-
noma cells present mostly in the periphery of the tumor with an
infiltrating edge (H&E, 1·)
Breast Cancer Res Treat (2007) 103:1–10 7
123
time intensity curves. An increased level of vascular
endothelial growth factor (VEGF) has been reported
to correlate with the pattern of early rim enhancement.
Vascular endothelial growth factor (VEGF) is a potent
angiogenic peptide that stimulates angiogenesis, acting
as a highly specific and potent mitogen for endothelial
cells, inducing formation of new blood vessels as well
as enhancing vascular permeability. Prominent mem-
bers of VEGF include VEGF A, B, C and D. VEGF A
is probably the most extensively studied member in
relation to breast cancer, and has been demonstrated to
be significantly related to the disease activity [68, 69].
Both the formation of blood vessels as well as the in-
crease in permeability will result in early rim
enhancement.
Delayed rim enhancement has been defined as
enhancement occurring after 300 s. Carcinomas that
demonstrated this pattern showed more fibrosis and
inflammatory changes. This attests to the fact that
stromal reaction in terms of inflammation or fibrosis
possesses a lower level of microvessel density. There
appears to be no correlation with VEGF staining
[66], indicating that the action of VEGF will result in
more immediate increase vascularity, and probably
increased permeability of the neovessels. A delayed
enhancement suggests minimal increase in microvas-
culature, as the flow of contrast is only marginally
increased, correlating with the understanding that
there is no increase in microvessel density in benign
breast lesions.
To summarize, MRI appears to be a superior
modality in the imaging of breast diseases, corre-
lating well with the lesion morphology. It is partic-
ularly useful in the specific settings of invasive
lobular carcinoma and in assessing tumor responses
in patients undergoing neoadjuvant therapy. Con-
trast enhancement and the time intensity dynamic
patterns appear to be functional derivatives of
microvessel density and VEGF driven increased per-
meability, both of which are known to be altered (in-
creased) in malignancy. Recognizance and
understanding of the underlying changes in pathology
may allow one to avoid pitfalls in the interpretation
of MRI findings.
Fig. 18 Photomicrograph showing the low microvessel density in
the fibrotic peripheral area of the tumor (CD31, 400·, same
magnification as in Fig. 17)
Fig. 16 Photomicrograph showing the same tumor with the
highly cellular peripheral tumor cells on the left and the more
fibrotic and hypocellular central area present on the right (H&E,
20·)
Fig. 17 Photomicrograph showing the high microvessel density
in the cellular peripheral area of the tumor (CD31, 400·)
8 Breast Cancer Res Treat (2007) 103:1–10
123
References
1. Heywang SH, Hahn D, Schmidt H et al (1986) MRI imaging
of the breast using gadolinium-DTPA. J Comput Assist
Tomogr 10:199–204
2. Bone B, Pentek Z, Perbeck L et al (1997) Diagnostic accu-
racy of mammography and contrast-enhanced MR imaging
in 238 histologically verified breast lesions. Acta Radiol
38:489–496
3. Stomper PC, Winston JS, Herman S et al (1997) Angiogen-
esis and dynamic MR imaging gadolinium enhancement of
malignant and benign breast lesions. Breast Cancer Res
Treat 45:39–46
4. Boetes C, Barentsz JO, Mus RD et al (1994) MR charac-
terization of suspicious breast lesions with a gadolinium-
enhanced TurboFLASH subtraction technique. Radiology
193:777–781
5. Heiberg EV, Perman WH, Herrmann VM et al (1996) Dy-
namic sequential 3D gadolinium-enhanced MRI of the whole
breast. Magn Reson Imaging 14:337–348
6. Muller-Schimpfle M, Stoll P, Stern W et al (1997) Do
mammography, sonography, and MR mammography have a
diagnostic benefit compared with mammography and
sonography? AJR Am J Roentgenol 168:1323–1329
7. Heywang-Kobrunner SH, Viehweg P, Heinig A et al (1997)
Contrast-enhanced MRI of the breast: accuracy, value, con-
troversies, solutions. Eur J Radiol 24:94–108
8. Morris EA, Schwartz LH, Dershaw DD et al (1997) MR
imaging of the breast in patients with occult primary breast
carcinoma. Radiology 205:437–440
9. Boetes C, Mus RD, Holland R et al (1995) Breast tumors:
comparative accuracy of MR imaging relative to mammog-
raphy and US for demonstrating extent. Radiology 197:
743–747
10. Kepple J, Layeeque R, Klimberg VS et al (2005) Correlation
of magnetic resonance imaging and pathologic size of infil-
trating lobular carcinoma of the breast. Am J Surg 190:623–627
11. Rieber A, Merkle E, Bohm W et al (1997) MRI of histo-
logically confirmed mammary carcinoma: clinical relevance
of diagnostic procedures for detection of multifocal or con-
tralateral secondary carcinoma. J Comput Assist Tomogr
21:773–779
12. Orel SG, Schnall MD, Powell CM et al (1995) Staging of
suspected breast cancer: effect of MR imaging and MR-
guided biopsy. Radiology 196:115–122
13. Mumtaz H, Hall-Craggs MA, Davidson T et al (1997) Stag-
ing of symptomatic primary breast cancer with MR imaging.
AJR Am J Roentgenol 169:417–424
14. Fischer U, Kopka L, Grabbe E (1999) Breast carcinoma:
effect of preoperative contrast-enhanced MR imaging on the
therapeutic approach. Radiology 213:881–888
15. Muuller RD, Barkhausen J, Sauerwein W et al (1998)
Assessment of local recurrence after breast-conserving
therapy with MRI. J Comput Assist Tomogr 22:408–412
16. Liberman L, Morris EA, Lee MJ et al (2002) Breast lesions
detected on MR imaging: features and positive predictive
value. AJR Am J Roentgenol 179:171–178
17. Siegmann KC, Muller-Schimpfle M, Schick F et al (2002)
MR imaging-detected breast lesions: histopathologic corre-
lation of lesion characteristics and signal intensity data. AJR
Am J Roentgenol 178:1403–1409
18. Elston C, Ellis IO (2000) Assessment of histological grade.
In: Elston CW, Ellis IO (eds), Symmets W (Emeritus Editor)
The breast, systemic pathology, Col 13. Churchill Livingston,
Edinburgh, pp 365–384
19. Walker RA, Camplejohn RS (1986) DNA flow cytometry of
human breast carcinomas and its relationship to transferrin
and epidermal growth factor receptors. J Pathol 150:37–42
20. Bouzubar N, Walker KJ, Griffiths K et al (1989) Ki67 im-
munostaining in primary breast cancer: pathological and
clinical associations. Br J Cancer 59:943–947
21. Fechner RE (1987) Fibroadenoma and related lesions. In:
Page DL, Anderson TJ (eds) Diagnostic histopathology of
the breast. Churchill Livingstone, Edinburgh, pp 72–85
22. Tsang WY, Chan JK et al (1996) Endocrine ductal carcinoma
in situ (E-DCIS) of the breast: a form of low-grade DCIS
with distinctive clinicopathologic and biologic characteristics.
Am J Surg Pathol 20:921–943
23. Tse GM, Ma TK, Chu WC et al (2004) Neuroendocrine
differentiation in pure type mammary mucinous carcinoma is
associated with favorable histologic and immunohistochem-
ical parameters. Mod Pathol 17:568–572
24. Nassar H, Qureshi H, Volkanadsay N et al (2006) Clinico-
pathologic analysis of solid papillary carcinoma of the breast
and associated invasive carcinomas. Am J Surg Pathol
30:501–507
25. Orel SG, Schnall MD, LiVolsi VA et al (1994) Suspicious
breast lesions: MR imaging with radiologic–pathologic cor-
relation. Radiology 190:485–493
26. Weinreb JC, Newstead G (1995) MR imaging of the breast.
Radiology 196:593–610
27. Krecke KN, Gisvold JJ (1993) Invasive lobular carcinoma of
the breast: mammographic findings and extent of disease at
diagnosis in 184 patients. AJR Am J Roentgenol 161:
957–960
28. Davis PL, Staiger MJ, Harris KB et al (1996) Breast cancer
measurements with magnetic resonance imaging, ultraso-
nography, and mammography. Breast Cancer Res Treat
37:1–9
29. Paramagul CP, Helvie MA, Adler DD (1995) Invasive lob-
ular carcinoma: sonographic appearance and role of sonog-
raphy in improving diagnostic sensitivity. Radiology 195:
231–234
30. Butler RS, Venta LA, Wiley EL et al (1999) Sonographic
evaluation of infiltrating lobular carcinoma. AJR Am
J Roentgenol 172:325–330
31. Kepple J, Layeeque R, Klimberg VS et al (2005) Correlation
of magnetic resonance imaging and pathologic size of infil-
trating lobular carcinoma of the breast. Am J Surg 190:623–627
32. Boetes C, Veltman J, van Die L et al (2004) The role of MRI
in invasive lobular carcinoma. Breast Cancer Res Treat
86:31–37
33. AJCC (2002) Cancer staging handbook, 6th edn. Springer,
New York, pp 255–282
34. Qayyum A, Birdwell RL, Daniel BL et al (2002) MR
imaging features of infiltrating lobular carcinoma of the
breast: histopathologic correlation. AJR Am J Roentgenol
178:1227–1232
35. Page DL, Anderson TJ, Sakamoto G (1987) Infiltrating
carcinoma: major histological types. In: Page DL, Anderson
TJ (eds) Diagnostic histopathology of the breast. Churchill
Livingstone, Edinburgh, pp 193–235
36. Bonadonna G, Veronesi U, Brambilla C et al (1990) Primary
chemotherapy to avoid mastectomy in tumors with diameters
of three centimeters or more. J Natl Cancer Inst 82:
1539–1545
37. Calais G, Berger C, Descamps P et al (1994) Conservative
treatment feasibility with induction chemotherapy, surgery,
and radiotherapy for patients with breast carcinoma larger
than 3 cm. Cancer 74:1283–1288
Breast Cancer Res Treat (2007) 103:1–10 9
123
38. Cocconi G, Di Blasio B, Alberti G et al (1984) Problems in
evaluating response of primary breast cancer to systemic
therapy. Breast Cancer Res Treat 4:309–313
39. Segel MC, Paulus DD, Hortobagyi GN (1988) Advanced
primary breast cancer: assessment at mammography of re-
sponse to induction chemotherapy. Radiology 169:49–54
40. Partridge SC, Gibbs JE, Lu Y et al (2002) Accuracy of MR
imaging for revealing residual breast cancer in patients who
have undergone neoadjuvant chemotherapy. AJR Am
J Roentgenol 179:1193–1199
41. Esserman L, Kaplan E, Partridge S et al (2001) MRI phe-
notype is associated with response to doxorubicin and
cyclophosphamide neoadjuvant chemotherapy in stage III
breast cancer. Ann Surg Oncol 8:549–559
42. Delille JP, Slanetz PJ, Yeh ED et al (2003) Invasive ductal
breast carcinoma response to neoadjuvant chemotherapy:
noninvasive monitoring with functional MR imaging pilot
study. Radiology 228:63–69 (Epub 2003 May 29)
43. Yeh E, Slanetz P, Kopans DB, Rafferty E et al (2005) Pro-
spective comparison of mammography, sonography, and
MRI in patients undergoing neoadjuvant chemotherapy for
palpable breast cancer. AJR Am J Roentgenol 184:868–877
44. Dao TH, Rahmouni A, Campana F (1993) Tumor recurrence
versus fibrosis in the irradiated breast: differentiation with
dynamic gadolinium-enhanced MR imaging. Radiology
187:751–755
45. Wasser K, Sinn HP, Fink C et al (2003) Accuracy of tumor
size measurement in breast cancer using MRI is influenced
by histological regression induced by neoadjuvant chemo-
therapy. Eur Radiol 13:1213–1223 (Epub 2002 Nov 30)
46. Elston CW, Ellis IO, Pinder SE (2000) Effects of treatment.
In: Elston CW, Ellis IO (eds), Symmets W (Emeritus Editor)
The breast, systemic pathology, Col 13. Churchill Livingston,
Edinburgh, pp 463–476
47. Knopp MV, Brix G, Junkermann HJ et al (1994) MR
mammography with pharmacokinetic mapping for monitor-
ing of breast cancer treatment during neoadjuvant therapy.
Magn Reson Imaging Clin N Am 2:633–658
48. Harms SE, Flamig DP, Hesley KL et al (1993) Fat-sup-
pressed three-dimensional MR imaging of the breast. Ra-
diographics 13:247–267
49. Bone B, Aspelin P, Bronge L et al (1996) Sensitivity and
specificity of MR mammography with histopathological
correlation in 250 breasts. Acta Radiol 37:208–213
50. Gilles R, Guinebretiere JM, Lucidarme O et al (1994)
Nonpalpable breast tumors: diagnosis with contrast-en-
hanced subtraction dynamic MR imaging. Radiology
191:625–631
51. Boetes C, Strijk SP, Holland R et al (1997) False-negative
MR imaging of malignant breast tumors. Eur Radiol 7:1231–
1234
52. Teifke A, Hlawatsch A, Beier T, Werner Vomweg T,
Schadmand S et al (2002) Undetected malignancies of the
breast: dynamic contrast-enhanced MR imaging at 1.0 T.
Radiology 224:881–888
53. Ghai S, Muradali D, Bukhanov K et al (2005) Nonenhancing
breast malignancies on MRI: sonographic and pathologic
correlation. AJR Am J Roentgenol 185:481–487
54. Hansen S, Grabau DA, Sorensen FB et al (2006) The
prognostic value of angiogenesis by Chalkley counting in a
confirmatory study design on 836 breast cancer patients. Clin
Cancer Res 6:139–146
55. Chhieng DC, Tabbara SO, Marley EF et al (2003) Micro-
vessel density and vascular endothelial growth factor
expression in infiltrating lobular mammary carcinoma.
Breast J 9:200–207
56. Morphopoulos G, Pearson M, Ryder WD et al (1996) Tu-
mour angiogenesis as a prognostic marker in infiltrating
lobular carcinoma of the breast. J Pathol 180:44–49
57. Komatsu S, Lee CJ, Hosokawa Y, Ichikawa D et al (2004)
Comparison of intraductal spread on dynamic contrast-en-
hanced MRI with clinicopathologic features in breast cancer.
Jpn J Clin Oncol 34:515–518
58. Ikeda O, Nishimura R, Miyayama H et al (2004) Magnetic
resonance evaluation of the presence of an extensive in-
traductal component in breast cancer. Acta Radiol 45:721–
725
59. Cao Y, Paner GP, Kahn LB et al (2004) Noninvasive carci-
noma of the breast: angiogenesis and cell proliferation. Arch
Pathol Lab Med 128:893–896
60. Ruiz A, Almenar S, Cerda M et al (2002) Ductal carcinoma
in situ of the breast: a comparative analysis of histology,
nuclear area, ploidy, and neovascularization provides dif-
ferentiation between low- and high-grade tumors. Breast
J 8:139–144
61. Guidi AJ, Schnitt SJ, Fischer L et al (1997) Vascular per-
meability factor (vascular endothelial growth factor)
expression and angiogenesis in patients with ductal carci-
noma in situ of the breast. Cancer 80:1945–1953
62. Weinreb JC, Newstead (1995) G.MR imaging of the breast.
Radiology 196:593–610
63. Rubens D, Totterman S, Chacko AK et al (1991) Gado-
pentetate dimeglumine-enhanced chemical-shift MR imag-
ing of the breast. AJR Am J Roentgenol 157:267–270
64. Kerslake RW, Carleton PJ, Fox JN et al (1995) Dynamic
gradient-echo and fat-suppressed spin-echo contrast-en-
hanced MRI of the breast. Clin Radiol 50:440–454
65. Kuhl CK, Mielcareck P, Klaschik S et al (1999) Dynamic
breast MR imaging: are signal intensity time course data
useful for differential diagnosis of enhancing lesions? Radi-
ology 211:101–110
66. Matsubayashi R, Matsuo Y, Edakuni G et al (2000) Breast
masses with peripheral rim enhancement on dynamic con-
trast-enhanced MR images: correlation of MR findings with
histologic features and expression of growth factors. Radi-
ology 217:841–848
67. Ahlgren J, Risberg B, Villman K et al (2002) Angiogenesis in
invasive breast carcinoma a prospective study of tumour
heterogeneity. Eur J Cancer 38:64–69
68. Kato T, Kameoka S, Kimura T et al (2001) Angiogenesis
as a predictor of long-term survival for 377 Japanese pa-
tients with breast cancer. Breast Cancer Res Treat 70:65–
74
69. Santinelli A, Baccarini M, Colanzi P et al (2000) Microvessel
quantitation in intraductal and early invasive breast carci-
nomas. Anal Quant Cytol Histol 22:277–284
10 Breast Cancer Res Treat (2007) 103:1–10
123
    • "For nonmass-like enhancement, features that have the highest positive predicitive value are clumped linear, segmental or regional enhancement. Lesions larger than 10 mm have a higher positive predictive value for being malignant than lesions < 10 mm (Tse et al., 2007). "
    Full-text · Chapter · Mar 2012 · American Journal of Roentgenology
  • [Show abstract] [Hide abstract] ABSTRACT: The system described in this paper is an outgrowth of one described in [1] and is a combination of deterministic and probabilistic speech recognition strategies. It is intended for telephone quality speech and possible single chip implementation with memory requirements of 256 bytes for processing (RAM) and 8k bytes for template storage and programs (ROM). Two major modifications for effecting reduction in memory requirements are described. The first is the use of a one-pass analysis algorithm for processing an utterance. The second is an approach for generation of uniform segment prototypes. Modification of the original system for telephone quality speech included obvious changes in filter bank analysis and parameterization of each phonetically classified segment with more emphasis on the voiced portions of utterances. The system was evaluated on a ten digit vocabulary (21 talkers, over 1000 tokens) and showed an accuracy of 94.4% with a 3.1% error rate and 2.5% rejection rate.
    Conference Paper · Apr 1984 · American Journal of Roentgenology
  • [Show abstract] [Hide abstract] ABSTRACT: The purpose of our study was to evaluate the effectiveness of gadolinium-enhanced MR imaging in imaging arterial, venous, and ureteric anatomy in a group of potential laparoscopic renal donors and to compare our findings with those established at surgery. Sixty-four consecutive patients underwent successful laparoscopic donor nephrectomy. Imaging of the kidneys was performed before surgery with MR imaging and breath-hold three-dimensional gadolinium-enhanced MR angiography. All studies were reviewed prospectively by one of two attending radiologists. Results were compared with findings at the time of laparoscopic nephrectomy. Of the 64 patients, MR imaging and MR angiography identified 30 patients with normal arterial, venous, and ureteric anatomy, and concordance was found at surgery in 29 of these patients. Vascular anomalies were depicted on MR imaging in 34 patients, with complete concordance at surgery in 29 patients. The use of MR angiography for revealing arterial anomalies had a sensitivity of 89.4%, specificity of 94.1%, and accuracy of 90.6%. For venous anomalies, there was a sensitivity of 98.3%, specificity of 100%, and accuracy of 98.4%. No important utereric anomalies were identified at surgery or on MR imaging. Renal MR imaging and gadolinium-enhanced MR angiography provide a safe, accurate, and minimally invasive means of comprehensive assessment of the potential living renal donor.
    Article · Jul 2002
Show more