AJR 2004; 183:1805-1809
© American Roentgen Ray Society
Contrast-Enhanced High-Resolution MRI of Invasive Breast Cancer: Correlation with Histopathologic Subtypes
Kakuya Kitagawa1,
Hajime Sakuma1,
Nanaka Ishida1,
Tadanori Hirano2,
Akinori Ishihara3 and
Kan Takeda1
1 Department of Radiology, Mie University, School of Medicine, 2-174 Edobashi,
Tsu, Mie 514-8507, Japan.
2 Department of Radiology, Matsusaka Central Hospital, 102 Kobo Kawaimachi,
Matsusaka, Mie 515-8566, Japan.
3 Department of Pathology, Matsusaka Central Hospital, Mie 515-8566,
Japan.
Received February 3, 2004;
accepted after revision May 11, 2004.
Address correspondence to K. Kitagawa
(kakuya{at}clin.medic.mie-u.ac.jp).
Abstract
OBJECTIVE. We sought to determine whether contrast-enhanced MRI
could aid in the identification of the histopathologic subtypes of invasive
ductal carcinoma.
MATERIALS AND METHODS. We evaluated the contrast-enhanced MR images
obtained in 62 women with invasive ductal carcinoma of no special type. The
presence or absence of three distinct MRI findingslinear enhancement, a
serrated border, and delayed rim enhancementwas evaluated.
Classification and regression tree analyses were performed to construct the
most efficient algorithm for predicting histopathologic subtype on the basis
of dynamic MRI features.
RESULTS. Histopathologic subtypes of the invasive ductal carcinomas
were scirrhous carcinoma in 22 patients, solid tubular carcinoma in 14, and
papillotubular carcinoma in 26. A lesion with a serrated border was observed
in 28 (45.2%) of the 62 patients. Delayed rim enhancement was seen in 23
(37.1%) and linear enhancement in 20 (32.3%). Scirrhous carcinomas were
closely associated with a serrated border (20/22 or 90.9%, p <
0.0001). Delayed rim enhancement was frequently observed in solid tubular
carcinomas (12/14 or 85.7%, p < 0.0001) but was not typically seen
in scirrhous carcinomas (1/22 or 4.5%, p < 0.0001). Linear
enhancement showed relatively high prevalence in papillotubular carcinomas
(13/26 or 50%) and low prevalence in solid tubular carcinomas (1/22 or 7%,
p < 0.02). Histopathologic subtypes of invasive breast carcinoma
of no special type could be correctly identified in 47 (75.8%) of 62 lesions
using the diagnostic algorithm generated by the classification and regression
tree analyses.
CONCLUSION. MRI features showed a close relationship with
histopathologic subtypes of invasive ductal carcinoma of no special type.
Contrast-enhanced MRI can be a noninvasive diagnostic tool for histopathologic
subtypes of invasive breast cancer.
Introduction
Contrast-enhanced MRI has become an important method for evaluating breast
cancer because of its excellent accuracy
[14].
MRI can be useful not only for detecting breast cancer but also for
determining its biologic properties because morphologic features and patterns
of contrast enhancement on contrast-enhanced MRI are representative of the
histologic features of tumor such as desmoplastic reactions, angiogenesis, and
the presence of intraductal components
[57].
Invasive ductal carcinoma of no special type, which accounts for more than
75% of all breast cancer, can be classified by histologic patterns into three
subgroups: papillotubular carcinoma, solid tubular carcinoma, and scirrhous
carcinoma
[811]
(Table 1). These histologic
subtypes are known to be related to patient prognosis: good for papillotubular
carcinoma, intermediate for solid tubular carcinoma, and poor for scirrhous
carcinoma. Furthermore, the usefulness of this classification has been
indicated with regard to the status of estrogen, progesterone, and epidermal
growth factor receptors; DNA ploidy; p53 protein; vascular proliferation; and
bcl-2 gene expression
[1214].
However, correlation between MRI features and histopathologic subtypes of
invasive breast cancer has not been fully elucidated.
The purpose of our study was to determine whether contrast-enhanced MRI
could provide characterization of histopathologic subtypes of invasive ductal
carcinoma.
Materials and Methods
Patients
This study was approved by our institutional review board. In our hospital,
all the patients who are candidates for surgical resection of breast cancer
undergo contrast-enhanced breast MRI to determine the extent of the tumor and
whether it is multifocal. Informed consent for contrast-enhanced breast MRI
was obtained from all patients. Sixty-two women between 31 and 83 years old
(mean, 58 years) were selected for this retrospective study from 114 patients
who underwent preoperative MRI. Criteria for inclusion in the study were that
the patients had contrast-enhanced MRI performed before undergoing either a
simple mastectomy or quadrantectomy, that histopathologic proof of the
diagnosis as invasive ductal carcinoma of no special type had been obtained,
that focal-enhancing lesions were smaller than 40 mm in maximum diameter at
histopathologic examination, and that no neoadjuvant chemotherapy had been
administered. A breast pathologist determined histopathologic subtypes of the
invasive ductal carcinomas according to General Rules for Clinical and
Pathological Recording of Breast Cancer formulated by the Japanese Breast
Cancer Society [8]. If two or
more histologic patterns were identified in a tumor, the pathologist made the
diagnosis on the basis of the rule of predominance. If the pathologist
encountered difficulty in judging the predominant type of pattern, the least
differentiated type was chosen as a histologic type.
MRI Technique
MRI of the breast was performed with a 1.5-T system (Magnetom Vision,
Siemens) with a dedicated bilateral breast surface coil. Contrast-enhanced
T1-weighted images were acquired using a 3D fast low-angle shot pulse sequence
with a TR/TE of 30/4 and a flip angle of 35°. Fat suppression was achieved
with a chemical shift saturation method. MR images covering the whole
unilateral breast were obtained with imaging parameters that included an
effective section thickness of 3.0 mm, a field of view of 220 x 220
x 96 mm, and acquisition matrices of 192 x 256 x 32. The
breast was imaged before and at 90 sec and again at 10 min after a bolus
injection of 0.1 mmol/kg of gadopentetate dimeglumine (Magnevist, Schering) on
coronal imaging planes [15].
The contrast medium was injected via a 20-gauge IV cannula into an antecubital
vein and was followed by a 20-mL saline flush. With the same parameters as
those used for the coronal sequences, additional contrast-enhanced MR images
were obtained on transverse imaging planes approximately 5 min after injection
of the contrast medium.
Image Interpretation
Two radiologists interpreting in consensus characterized the lesions as
either displaying or lacking linear enhancement, a serrated border, and
delayed rim enhancement. Both interpreters were unaware of pathologic results
and other imaging results. The enhancement pattern was assessed by visually
comparing the signal intensity on the images acquired at 90 sec and at 10 min
using the same window and level settings. Linear enhancement was defined as
linear or bandlike enhancement on early or delayed images that did not taper
and arose and extended from the mass for longer than 5 mm (Fig.
1A,
1B). A serrated border was
defined as an enhancement observed on early or delayed images that tapered as
it separated from the mass (Fig.
2A,
2B). Delayed rim enhancement
was defined as an enhancement that did not appear on the early image but
appeared on the delayed image rimming the lesion (Fig.
3A,
3B).

View larger version (73K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1A. 62-year-old woman with papillotubular carcinoma. Coronal
contrast-enhanced T1-weighted fast low-angle shot (FLASH) image (TR/TE, 30/4)
obtained 90 sec after injection of contrast medium shows irregular mass
(arrow) in lateral part of right breast.
|
|

View larger version (83K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 1B. 62-year-old woman with papillotubular carcinoma. On
transverse contrast-enhanced T1-weighted FLASH image (30/4) obtained 5 min
after contrast injection, linear enhancements (arrowheads) extending
from mass are clearly depicted.
|
|

View larger version (85K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2A. 52-year-old woman with scirrhous carcinoma. Coronal
contrast-enhanced T1-weighted fast low-angle shot (FLASH) image (TR/TE, 30/4)
obtained 90 sec after injection of contrast medium shows polygonal mass
(arrow) with serrated border in lateral part of left breast.
|
|

View larger version (95K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 2B. 52-year-old woman with scirrhous carcinoma. Coronal
contrast-enhanced T1-weighted FLASH image (30/4) obtained at same level as
A 10 min after contrast injection reveals more prominent serrated
border (arrowhead) than seen on A.
|
|

View larger version (92K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3A. 46-year-old woman with solid tubular carcinoma. Coronal
contrast-enhanced T1-weighted fast low-angle shot (FLASH) image (TR/TE, 30/4)
obtained 90 sec after injection of contrast medium shows round mass in medial
part (arrow) of left breast.
|
|

View larger version (115K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 3B. 46-year-old woman with solid tubular carcinoma. Coronal
contrast-enhanced T1-weighted FLASH image (30/4) obtained at same level as
A 10 min after contrast injection reveals rimlike enhancement
(arrowhead) surrounding mass.
|
|
Statistical Analysis
The mean maximum diameter of tumors was expressed as mean ± SD. Each
imaging parameter was analyzed separately for its association with the
histopathologic subtype of breast cancer using the chisquare test. A
p value of less than 0.05 was considered to be statistically
significant. Classification and regression analyses (Answer-Tree 3.1,
Statistical Package for the Social Sciences) were performed to construct the
most efficient algorithm for predicting histopathologic subtype on the basis
of dynamic MRI features [16].
Construction of the algorithm depended on the improvement scores as calculated
with the Answer-Tree software. The higher the improvement score for an MRI
feature, the better its performance in differentiating histopathologic subtype
of breast cancer. A previously reported proportion of histopathologic subtypes
was used as the prior probability for the analysis
[12].
Results
Histopathologic Subtypes of Breast Cancer
Histopathologic subtypes of the invasive ductal carcinomas were 26
papillotubular carcinomas, 14 solid tubular carcinomas, and 22 scirrhous
carcinomas. The mean maximum diameter of the tumor for scirrhous carcinoma was
20.7 ± 7.4 mm (range, 840 mm), for solid tubular carcinoma was
23.3 ± 9.0 mm (range, 1240 mm), and for papillotubular carcinoma
was 20.8 ± 7.6 mm (range, 1040 mm). No statistically significant
difference in mean maximum diameter was observed among the three
histopathologic subtypes.
Frequency of MRI Features
A serrated border was observed in 28 (45.2%) of the 62 lesions, delayed rim
enhancement in 23 (37.1%), and linear enhancement in 20 (32.3%)
(Table 2). Appearance of a
serrated border showed a high prevalence for scirrhous carcinomas (20/22,
90.9%), whereas the incidence of this finding was low for other subtypes
(8/40, 20.0%, p < 0.0001). Delayed rim enhancement was observed
frequently in solid tubular carcinomas (12/14, 85.7%) and was also found in
papillotubular carcinomas (10/26, 38.5%). However, the incidence of this
finding was quite low in scirrhous carcinomas (1/22, 4.5%; p <
0.0001). Linear enhancement was observed relatively frequently in
papillotubular carcinomas (13/26, 50.0%) but was not commonly seen in solid
tubular carcinomas (1/14, 7.1%; p < 0.02).
Diagnostic Algorithm of Histopathologic Subtype of Breast Cancer
The diagram shown as Figure
4 represents the most efficient algorithm for predicting the
histopathologic subtype of invasive ductal carcinoma of no special type on the
basis of dynamic MRI features generated by the classification and regression
tree analyses. The lesions were divided into five groups (AE) on the
basis of the presence or absence of linear enhancement, a serrated border, and
delayed rim enhancement. Group A (serrated border, present; delayed
enhancement, present) consisted of three solid tubular carcinomas (50.0%), two
papillotubular carcinomas (22.2%), and one scirrhous carcinoma (11.1%).
Nineteen (86.4%) of 22 tumors classified in group B (serrated border, present;
delayed rim enhancement, absent) were scirrhous carcinomas. The 13 lesions in
group C (serrated border, absent; linear enhancement, present) represented 12
papillotubular carcinomas (92.3%) and one solid tubular carcinoma (0.08%). The
12 lesions in group D (serrated border, absent; linear enhancement, absent;
delayed rim enhancement, present) included eight solid tubular carcinomas
(66.7%) and four papillotubular carcinomas (33.3%). In group E (serrated
border, absent; linear enhancement, absent; delayed rim enhancement, absent),
there were five papillotubular carcinomas (55.6%), two solid tubular
carcinomas (22.2%), and two scirrhous carcinomas (22.2%).

View larger version (28K):
[in this window]
[in a new window]
[as a PowerPoint slide]
|
Fig. 4. Diagram shows diagnostic algorithm (generated by Answer-Tree
3.1 software, Statistical Package for the Social Sciences) used for
classifying histopathologic subtype of invasive ductal carcinoma of no special
type on basis of MRI features. IS = improvement score of breast MRI features
in nodes of classification and regression tree.
|
|
Classification and regression analysis of contrast-enhanced MRI findings
yielded correct classification of the histopathologic subtype of breast cancer
in 19 (86.4%) of the 22 scirrhous carcinomas, 11 (78.6%) of the 14 solid
tubular carcinomas, 17 (65.4%) of the 26 papillotubular carcinomas, and 47
(75.8%) of the 62 total lesions.
Discussion
Our study showed that morphologic characteristics of invasive breast cancer
on contrast-enhanced MR images such as linear enhancement, a serrated border,
and delayed rim enhancement are closely related to histopathologic subtypes of
invasive ductal carcinoma of no special type. Using the diagnostic algorithm
presented in Figure 4,
histopathologic subtypes of invasive breast carcinoma of no special type can
be correctly identified in 75.8% of the cases.
Invasive ductal carcinoma of no special type is not a uniform histologic
entity. The extent of neovascularization, fibrosis, and cell density may vary
considerably. Such heterogeneity of breast carcinomas leads to a vast range of
morphologic features and patterns of contrast enhancement on MRI.
Consequently, we evaluated the correlation between the MRI appearance of the
breast carcinoma and its histopathologic subtype.
On contrast-enhanced MR images, scirrhous carcinomas were usually observed
as serrated masses without delayed rim enhancement. Histopathologic studies
indicated that the cells of scirrhous carcinoma exhibit scattered invasion
into the stroma in small clusters or in trabecular structures with
accompanying desmoplasia of varying degrees
[10]. Such histologic
characteristics of scirrhous carcinoma may explain the high prevalence of
lesions with serrated borders on contrast-enhanced MR images. Identification
of scirrhous carcinoma is important for several reasons. First, scirrhous
tendency has been reported to be closely associated with patient poor
prognosis [17]. Second,
scirrhous carcinomas tend to be missed more frequently at
fine-needle-aspiration biopsy than other subtypes of invasive ductal carcinoma
[18].
Solid tubular carcinoma is characterized histopathologically by a large
solid cluster of cancer cells with expansive growth forming distinct
boundaries [10]. An absence of
estrogen and progesterone receptors has been found to occur more frequently in
solid tubular carcinomas than in the other two subtypes
[12]. Also, the frequencies of
DNA aneuploidy, epidermal growth factor receptor positivity, and p53 protein
positivity have been shown to be significantly higher in solid tubular
carcinomas than in other subtypes
[12]. In our study, the
expansive growth pattern of solid tubular carcinoma was seen as delayed rim
enhancement on contrast-enhanced MR images. Delayed rim enhancement on
contrast-enhanced MR images has been previously discussed in the literature
[1921].
Buadu et al. [19] reported
that delayed peripheral enhancement with centrifugal progression was seen in
carcinomas that had an expansive growth pattern and a high marginal vessel
density with or without a vascularized rim of connective tissue. Matsubayashi
et al. [20] reported that a
high degree of surrounding fibrosis and inflammatory changes correlated
significantly with depiction of delayed rim enhancement. These previous
reports indicate that delayed rim enhancement, which was frequently observed
in solid tubular carcinoma, may represent compressed or proliferated
connective tissue surrounding the tumor.
Papillotubular carcinoma is defined as a carcinoma characterized by
papillary projection and tubule formation. Part of the tumor may contain a
solid pattern of tumor cell proliferation
[10]. Although papillotubular
carcinomas displayed linear or bandlike enhancement more frequently than other
histologic subtypes, no single MRI finding specifically corresponded to this
subtype. In fact, as shown in Figure
4, papillotubular carcinomas were included in all five groups
(AE), indicating substantial overlap between its MRI appearance and the
MRI appearance of other subtypes.
Study Limitations
Several limitations of this study should be acknowledged. We studied a
relatively small number of patients. Only invasive ductal carcinoma of no
special type was retrospectively evaluated in our study. Additional studies
that include noninvasive breast cancer, special types of invasive breast
cancer, and benign breast lesions will be useful in determining the diagnostic
significance of MRI findings. Another limitation is that the mean maximum
diameter of the tumors that we evaluated was approximately 2025 mm. The
analysis may not apply to smaller lesions because enhancement in smaller
lesions is generally more homogeneous. Finally, we also acknowledge that
predicting the histopathologic subtype does not necessarily affect clinical
management. All these patients still underwent conventional therapy, including
surgery, radiation, and medication.
In conclusion, we found that MRI features showed a close relationship with
histopathologic subtypes of invasive ductal carcinomas of no special type.
High-resolution contrast-enhanced MRI can be a noninvasive diagnostic tool for
histopathologic subtypes of invasive breast cancer.
Acknowledgments
We thank Ichiro Kadoya and Chikai Shoji for their cooperation and
encouragement.
References
- Orel SG, Schnall MD, Powell CM, et al. Staging of suspected breast
cancer: effect of MR imaging and MR-guided biopsy.
Radiology1995; 196:115
122[Abstract/Free Full Text]
- Brown J, Buckley D, Coulthard A, et al. Magnetic resonance imaging
screening in women at genetic risk of breast cancer: imaging and analysis
protocol for the UK multicentre studyUK MRI Breast Screening Study
Advisory Group. Magn Reson Imaging2000; 18:765
776[Medline]
- Liu PF, Debatin JF, Caduff RF, Kacl G, Garzoli E, Krestin GP.
Improved diagnostic accuracy in dynamic contrast-enhanced MRI of the breast by
combined quantitative and qualitative analysis. Br J
Radiol 1998;71:501
509[Abstract]
- Malich A, Boehm T, Facius M, et al. Differentiation of
mammographically suspicious lesions: evaluation of breast ultrasound, MRI
mammography and electrical impedance scanning as adjunctive technologies in
breast cancer detection. Clin Radiol2001; 56:278
283[Medline]
- Solomon B, Orel S, Reynolds C, Schnall M. Delayed development of
enhancement in fat necrosis after breast conservation therapy: a potential
pitfall of MR imaging of the breast. AJR1998; 170:966
968[Free Full Text]
- Turetschek K, Huber S, Floyd E, et al. MR imaging characterization
of microvessels in experimental breast tumors by using a particulate contrast
agent with histopathologic correlation. Radiology2001; 218:562
569[Abstract/Free Full Text]
- Neubauer H, Li M, Kuehne-Heid R, Schneider A, Kaiser WA. High grade
and nonhigh grade ductal carcinoma in situ on dynamic MR mammography:
characteristic findings for signal increase and morphological pattern of
enhancement. Br J Radiol2003; 76:3
12[Abstract/Free Full Text]
- Japanese Breast Cancer Society. General rules for
clinical and pathological recording of breast cancer, 14th ed.
Tokyo, Japan: Kanehara, 2000:19
56
- Page DL, Anderson TJ. Diagnostic histopathology of the
breast. New York, NY: Churchhill Livingstone, 1987:193
197
- Sakamoto G. Histological classification of breast cancer [in
Japanese]. Gan No Rinsho 1985; [suppl]:105
113
- Kurosumi M, Tabei T, Inoue K, et al. Prognostic significance of
scoring system based on histological heterogeneity of invasive ductal
carcinoma for node-negative breast cancer patients. Oncol
Rep 2003;10:833
837[Medline]
- Tsutsui S, Ohno S, Murakami S, Kataoka A, Kinoshita J, Hachitanda
Y. Histological classification of invasive ductal carcinoma and the biological
parameters in breast cancer. Breast Cancer2003; 10:149
152[Medline]
- Samejima N, Yamazaki K. A study on the vascular proliferation in
tissues around the tumor in breast cancer. Jpn J Surg1988; 18:235
242[Medline]
- Kobayashi S, Iwase H, Ito Y, et al. Clinical significance of bcl-2
gene expression in human breast cancer tissues. Breast Cancer Res
Treat 1997;42:173
181[Medline]
- Weinstein D, Strano S, Cohen P, Fields S, Gomori JM, Degani H.
Breast fibroadenoma: mapping of pathophysiologic features with
three-time-point, contrast-enhanced MR imagingpilot study.
Radiology1999; 210:233
240[Abstract/Free Full Text]
- Lemon SC, Roy J, Clark MA, Friedmann PD, Rakowski W. Classification
and regression tree analysis in public health: methodological review and
comparison with logistic regression. Ann Behav Med2003; 26:172
181[Medline]
- Kodama H. Reevaluation of the histological classification of breast
cancer: a presentation of a new histological classification considering the
relationship between scirrhous tendency and postoperative prognosis [in
Japanese]. Nippon Geka Gakkai Zasshi1985; 86:853
862[Medline]
- Park IA, Ham EK. Fine needle aspiration cytology of palpable breast
lesions: histologic subtype in false-negative cases. Acta
Cytol 1997;41:1131
1138[Medline]
- Buadu LD, Murakami J, Murayama S, et al. Patterns of peripheral
enhancement in breast masses: correlation of findings on contrast medium
enhanced MRI with histologic features and tumor angiogenesis. J
Comput Assist Tomogr 1997;21:421
430[Medline]
- Matsubayashi R, Matsuo Y, Edakuni G, Satoh T, Tokunaga O, Kudo S.
Breast masses with peripheral rim enhancement on dynamic contrast-enhanced MR
images: correlation of MR findings with histologic features and expression of
growth factors. Radiology2000; 217:841
848[Abstract/Free Full Text]
- Sherif H, Mahfouz AE, Oellinger H, et al. Peripheral washout sign
on contrast-enhanced MR images of the breast.
Radiology1997; 205:209
213[Abstract/Free Full Text]

CiteULike
Complore
Connotea
Del.icio.us
Digg
Reddit
Technorati What's this?