Breast Cancer (2013) 20:312

DOI 10.1007/s12282-011-0326-x

SPECIAL FEATURE

Preoperative evaluation for intraductal spread

of breast cancer through current imaging tests:
their strengths and limitations

Comprehensive ultrasound diagnosis for intraductal spread

of primary breast cancer
Kazutaka Nakashima Takuya Moriya

Received: 30 August 2011 / Accepted: 13 December 2011 / Published online: 31 January 2012
The Japanese Breast Cancer Society 2012

Abstract
Background The current capacity of preoperative ultrasound examinations to estimate the intraductal spread of
primary breast cancer was evaluated. The current ultrasound examination has three modes, B-mode, vascularity,
and elastography, and this study evaluated this comprehensive ultrasound examination.
Methods Of the patients undergoing breast cancer surgery
for a breast mass at our department between April 2010 and
March 2011, we evaluated 99 in whom breast-conserving
surgery was scheduled, and performed preoperative ultrasounds and stored the digital ultrasound images and
pathology data. ACR BI-RADS-US lexicon elastography
for the main mass and peri-tumoral ductal lesions (PTDL,
defined in main sentence), and the vascularity and elasticity
pattern (defined in main sentence) were evaluated.
Results A cut end-negative resection at the first procedure
after ultrasound estimation was achieved in 76.7%, and the
breast was conserved by an additional resection with a
negative margin in 90.6%. Shadowing of mass posterior
acoustic features tends to indicate cut end-negative results
significantly. There was a statistical difference between
vascularity in PTDL and elasticity pattern of PTDL.
Conclusion Using the recent comprehensive ultrasound
examination with vascularity and elastography was useful
for the preoperative estimation of intraductal spread of
primary breast cancer in our retrospective study.
K. Nakashima (&)
Department of Breast and Thyroid Surgery, Kawasaki Medical
School, 577 Matsushima, Kurashiki, Okayama 701-0192, Japan
e-mail: urbandoc@med.kawasaki-m.ac.jp
T. Moriya
Pathology 2, Kawasaki Medical School, Kurashiki,
Okayama, Japan

Keywords Comprehensive ultrasound
Elastography

Vascularity
Intraductal spread
Breast-conserving surgery

Introduction
Ultrasound (US) is the most frequently used and important
modality in clinical practice for breast cancer. In particular,
in determining the resection area in breast-conserving surgery, the accuracy in evaluating the localization by US, by
which images can be scanned in the same position as in
surgery, is higher than that by MRI and CT; using the latter
two, evaluation of accurate localization is difficult because
of the different positions sometimes needed for scanning
(e.g., arm position). It is possible to have the same position
during the operation as during the preoperative US, as in
intraoperative ultrasound examinations. Moreover, the resolution of the current high-resolution and high-contrast US is
less than 0.2 mm (1418 MHz Probe), and marking for
localization of US is possible directly on the body surface;
US has a variety of advantages for preoperative examination.
Recent development of high-resolution and high-contrast US equipment and examination techniques, as well as
intensive education, has enabled observation of morphological findings at the millimeter level. In addition, due to
the prevalence of highly sensitive Doppler and high-contrast elastography, US that used to provide information
only for morphological diagnosis has provided other
qualitative information.
Comprehensive diagnosis based on such morphological
and qualitative information has suggested the possibility of
a new diagnostic method in breast US. In this study, we
defined comprehensive ultrasound examination as the
new diagnotic method with ultrasound B mode, elastography, and vascularity.

123

In our department, when breast-conserving surgery is

considered in cases with non-palpable breast cancer or
difficult preoperative diagnosis of the extension of the
lesion, preoperative US diagnosis for intraductal spreading
is carried out, and marking on the skin surface is done with
a permanent marker. When breast-conserving surgery is
considered difficult at this step, mastectomy plus immediate breast reconstruction is indicated.
During the resection, a gap from the skin to surface is
taken into account, and the tumor is resected with a 10-mm
margin from the marking. Intraoperative pathology is
always carried out, and cross sections of the margins on the
nipple, and inner and lateral sides are submitted. When the
cut end is positive, another resection with a distance of
10 mm is added. Intraoperative pathology is re-examined,
and when the cut end is positive, mastectomy is carried out.
The accuracy of preoperative US diagnosis for intraductal spreading was investigated in our department by
examining the process of preoperative US diagnosis for
intraductal spreading and breast-conserving surgery, and a
novel method was devised based on the assumption that
evaluation of intraductal spreading was possible by pathological diagnosis of the cut end of the resected tissue.

Patients and methods

Of the cases undergoing breast cancer surgery for a breast
mass at our department between April 2010 and March
2011, 99 patients for whom breast-conserving survey was
scheduled and rapid intraoperative pathology samples of
the resected margin were submitted were enrolled in this
study. When patients received open biopsy or primary
systemic treatment and US images of a mass were not
recorded digitally, they were excluded from the subjects.
US images and glass slides for pathology were collected
retrospectively. The accuracy of preoperative marking for
resection areas based on US images and factors influential
for the accuracy were retrospectively investigated by
comparison of the pathological and US characteristics of
tumors with the results of rapid intraoperative pathology on
the surgical margin and the final pathological diagnosis.
Preoperative US diagnosis for intraductal spreading was
carried out in 67 of 99 cases. In 57 cases, Doppler scanning
and real-time tissue elastographyTM (elastography) in
addition to B-mode US were employed for diagnosis, and
further investigations were carried out on these 57 cases.
The US equipment and probes used were Hitachi Aloka
Medical EUB 7500 and L74M, respectively. This study
was approved in advance by the ethics committee, Kawasaki Medical School, on 21 July 2011 (approval no. 866).
A preoperative US diagnosis for intraductal spreading
was made in almost the same position as in surgery, and

123

Breast Cancer (2013) 20:312

marking was, in principle, carried out along the tumor

boundary, but it was done including a halo in cases of a
mass with a halo.
In our department, we have long focused on the finding
that mammary duct-like structures are often observed
around a mass and employ this as a reference in the
examination. In this study, this structure was focused on and
termed the peri-tumoral ductal lesion (PTDL) (Fig. 1). It
was defined as a tubular low-echoic lesion extending to the
adjacent tissue beyond 5 mm from the tumor. Presence or
absence of PTDL by B-mode, Doppler images, and findings
by elastography were examined.
With regard to the results of preoperative US diagnosis for
intraductal spreading, cases were categorized as follows
according to the intraoperative pathology on the tumor
margin: cut end-negative cases in all directions, first
resection negative (FRN); cut end-positive cases that were
later found to be cut end-negative by re-submission, second
resection negative (SRN); and cases that were finally
diagnosed as cut end-positive and received mastectomy,
final positive (FP). True positive in preoperative comprehensive US diagnosis for intraductal spreading was
defined as FRN, and it was compared with the following
pathological findings and US findings.
According to the General Rules for Clinical and Pathological Recording of Breast Cancer [1], ductal carcinoma
in situ (DCIS), papillotubular carcinoma (PAP), solidtubular carcinoma (SOL), scirrhous carcinoma (SCI),
invasive lobular carcinoma (ILC), multiple metastases, ER
expression, PgR expression, HER2 status, pT, pN, Ki-67,
grade, and presence or absence of EIC were examined
pathologically.
According to the ACR BI-RADS-US Lexicon Classification [2], US findings were classified according to the
following categories for evaluation: mass shape (oval,
round, lobulated, irregular with angles, and irregular),
internal echo (anechoic, hyperechoic, complex, hypoechoic, and isoechoic), mass margin (circumscribed, indistinct,
angular, microlobulated, and speculated), mass echo pattern (anechoic, hyperechoic, complex, hypoechoic, and
isoechoic), and mass posterior acoustic features (normal,
enhancement, shadowing, and combined pattern). Presence
or absence of architectural distortion in surrounding tissue
was also examined.
In this study, attention was paid to the patterns of tumor
invasion to adjacent tissue and lactiferous duct-like structures indicating tumor extension to the adjacent tissue. It
was hypothesized that elastic areas around the tumor were
wider and extended the boundary of the tumor by images
[3], and the elasticity score of elastography [4], which
served as the criteria for evaluation of tumors, was also
examined for judgment of degrees of tumor invasion to
adjacent tissue (Fig. 2). The elasticity score comprised

Breast Cancer (2013) 20:312

Fig. 1 Peri-tumoral ductal lesions (PTDL): PTDL means hypoechoic ductal lesion beyond 5 mm from the tumor. Red arrows indicate the parts
of PTDLs on B-mode images

scores of 1, 2, 3, 4, and 5. Since all cases had a malignant

tumor, scores of only 3, 4, or 5 were evaluated.
In terms of PTDL, the elasticity pattern of PTDL was
examined in addition to the presence of PTDL, microcalcification in PTDL, and vascularity in PTDL. With regard to the
elasticity pattern of PTDL, green pattern, red line pattern, blue line pattern, blue to red line pattern, and blue
line skip pattern were defined according to the color of PTDL
(Table 1). Figure 3 shows typical cases with these patterns.
Findings and the results of marking for prediction of the
resection area by preoperative US were subjected to the
chi-square test, and important and effective factors for
preoperative US diagnosis for intraductal spreading were
investigated.

Results
Of 67 patients that were subjected to preoperative US
diagnosis for intraductal spreading, there were 49 FRN, 16
SRN, and 2 FP cases. With regard to preoperative US
marking for prediction of the resection area revealed that
cut end-negative resection was achieved in 73.1% and

breast was conserved by additional cut end-negative

resection in 97.0%. On the other hand, in 32 cases that
were not subjected to preoperative US marking, there were
23 FRN, 6 SRN, and 3 FP cases, and cut end-negative
resection by the first procedure was achieved in 76.7%. The
breast was conserved by additional resection with a negative margin in 90.6%. Considering that indications of
preoperative US marking for prediction of the resection
area were dependent on the surgeon in charge and a large
number of cases had lesions relatively hard to identify
because they were hardly palpable, they were thought to be
excellent results.
When compared with pathological factors, none of
tumor types, multiple/solitary occurrence, status of ER,
PgR, and HER2, pT, pN, Ki-67, grade, and presence or
absence of EIC, according to the General Rules for Clinical
and Pathological Recording of Breast Cancer, were statistically significant. Meanwhile, cases that were finally cut
end-positive and subjected to mastectomy had only PAP
according to the General Rules for Clinical and Pathological Recording of Breast Cancer (Table 2).
With regard to ACR BI-RADS-US findings, neither
mass shape nor mass margin yielded a significant

123

Breast Cancer (2013) 20:312

Fig. 2 Elasticity scores: a score 3, b score 4, and c score 5. Score 3

means the blue area is smaller than the tumor outline on the B-mode
image (d). Score 4 means the blue area almost corresponds with the

tumor outline on the B-mode image (d). Score 5 means the blue area
is larger than the tumor outline on the B-mode image (d)

difference. As to mass echo pattern, complex cases

tended to be cut end-positive at the first resection (Table 3,
p \ 0.05). In terms of mass posterior acoustic features,
shadowing cases were cut end-negative at the first
resection to a significant extent (Table 4, p \ 0.01).
Patients with presence of architectural distortion in
surrounding tissues tended to be cut end-negative, but there
was no statistical significance (p = 0.07).
Despite no significant difference (Table 5, p = 0.06),
mass elastography revealed a score of 4, which indicated
few invasive findings with clear delineation from the surrounding tissue, and there was a tendency of FRN. With
regard to PTDL focusing mostly on the wide expansion of

the lesion, there was, unfortunately, no significant correlation with the presence or absence of PTDL by B-mode
US or microcalcification in PTDL (p = 0.05). Comedotype microcalcification was identified with relative ease,
and breast-conserving surgery by quadrantectomy may be
recommended in cases with a certain level of tumor
expansion.
Meanwhile, there was a statistical difference in vascularity in PTDL and elasticity pattern of PTDL. With
regard to vascularity in PTDL, positive cases revealed
SRN and FP significantly more frequently, and they
were cut end-positive more frequently at the first resection
(Table 6, p \ 0.05).

Table 1 Definitions of the elasticity patterns of PTDL

Definition
Green pattern

Elasticity color of PTDL = elasticity color of tissues around PTDL

Red line pattern

Red line elasticity colors on PTDL

Blue to red line pattern

Blue skip pattern

Blue line elasticity color changes to red line elasticity color of PTDL beyond 5 mm from the tumor margin
Blue line elasticity color skipped on PTDL beyond 5 mm from the tumor margin

All pattern images showed in Fig. 3

123

Breast Cancer (2013) 20:312

Fig. 3 Images of the elasticity pattern of PTDL: a Green pattern

PTDL color is almost the same as the surrounding tissue, b red line
pattern red color on PTDL, c blue to red line pattern PTDL color

changes from blue to red with distance from the main tumor, and
d blue skip pattern blue color lines or blue color deposits are seen
beyond 5 mm

Table 2 Preoperative ultrasound evaluation for tumor spread and

final histology of main tumor

Table 4 Preoperative ultrasound evaluation for tumor spread and

posterior acoustic features of the tumor

DCIS
ILC

FRN

SRN

FP
0

Total

FRN

SRN

FP

Total
14

Normal

Enhancement

13

22

24

2
39

4
16

0
2

6
57

PAP

12

19

Shadowing

SCI

18

23

Combined pattern
Total

SOL

Total

39

16

57
Table 5 Preoperative ultrasound evaluation for tumor spread and
elasticity score of the main tumor

Table 3 Preoperative ultrasound evaluation for tumor spread and

internal echoes in the tumor
FRN
Anechoic

SRN
0

FP
0

Total
0

Hyperechoic

Complex

16

28

35

39

16

57

Hypoechoic
Isoechoic
Total

In terms of the elasticity pattern of PTDL, the blue skip

pattern indicated cut end-positive at the first resection
significantly more often (Table 7, p \ 0.05), and only

FRN

SRN

FP

Total

Score 3

18

24

Score 4
Score 5

11
10

1
9

0
2

12
21

Total

39

16

57

Table 6 Preoperative ultrasound evaluation for tumor spread and

vascularity in PTDL
FRN

SRN

FP

Total

Positive for vascularity in PTDL

Negative for vascularity in PTDL

Total

38
39

12
16

0
2

50
57

123

Breast Cancer (2013) 20:312

cases with this pattern were cut end-positive eventually and

subjected to mastectomy. Cases with a blue to red line
pattern showed FRN more frequently, and it was easier to
determine tumor spreading.

Discussion
Most patients with breast cancer exhibit non-infiltrative
cancer components that extend intraductally outside of the
mass that are palpable or recognized by image modalities.
This is not directly related with prognosis, but accurate
Table 7 Preoperative ultrasound evaluation for tumor spread and
elastography pattern of PTDL

Green pattern

FRN

SRN

FP

Total

10

Red line pattern

19

26

Blue to red line pattern

11

12

Blue skip pattern

Total

39

16

57

Fig. 4 Red line pattern of normal breast ultrasound images with

different causes. The upper image is an elastography image of normal
breast; the lower is an elastography image of the non-tumoral breast
area of a breast cancer patient. Red line patterns of elastography are

123

identification and evaluation of this expansion are the most

important factors for breast-conserving surgery, which is
the recent mainstream operation for the first-line surgery on
breast cancer, because cosmetic results after surgery, the
main aim of breast-conserving surgery, clearly have a
negative correlation with the volume of resection. Therefore, it is necessary to evaluate tumor expansion as accurate as possible preoperatively to achieve the resection at
the minimum volume and conserve a beautiful breast with
excellent cosmetic results [5]. MRI, CT, and US have been
employed to this end recently.
The sensitivity of MRI in the diagnosis of breast cancer
expansion is markedly high, and it has been described as
recommended (C1) in the Clinical Practice Guidelines
for Breast Cancer by the Japanese Breast Cancer Society
2011 [6]. However, high-resolution images are not
obtained unless a prone position is used for MRI scanning,
and image findings are greatly different from the findings
in the position during surgery. Therefore, there is a disadvantage in this likelihood of a gap in spatial information
compared with the findings in the position during surgery.
CT provides images that are similar to those obtained in the

seen around large breast ducts and fatty tissues in normal and nontumor areas of the patients breast. It is especially important that the
red line pattern is seen along the duct wall, not in the duct

Breast Cancer (2013) 20:312

position used during surgery, but its sensitivity has been

considered inferior to MRI.
US allows a direct examination of breasts in the same
position as in surgery, and it is possible to evaluate the
location of a lesion most accurately. However, since the
accuracy depends on the technique, certain training is
required. Meanwhile, current US equipment provides resolution at 200 lm or higher, and as long as examiners learn
the proper scanning technique, accurate morphological
evaluation and visualization of microcalcifications are
sufficiently possible [7].
Furthermore, as the sensitivity of the Doppler mode has
improved and elastography has become prevalent, not only
morphological diagnosis but also qualitative diagnosis
have become possible. Concerning the recent status US, the
amount of information available for diagnosis has greatly
increased, and several reports of prospective studies have
demonstrated evidence of their utility [810]. The Japanese

Breast Cancer Society has evaluated both examinations as

C1 (recommended) in the Clinical Practice Guidelines for
Breast Cancer 2011.
In this study, in cases with difficulty in diagnosing
tumor expansion, cut end-negative first resection was
achieved in 73.1%, and cut end-negative additional resection with breast conservation was achieved in 97.0%.
Considering the technical bias at the resection, it was
presumed that satisfactory accuracy in the evaluation of
tumor expansion was achieved in US comprehensive
diagnosis by a combination of B-mode US, Doppler
scanning, and elastography.
PTDL as defined in this study is discussed as follows.
PTDL elastography images of normal breast tissue and
non-cancerous breast tissue, similar to red line pattern, are
shown. Comparison with B-mode images suggested that
linear low-echoic structure bands were visualized, because
some consisted of a mixture of lactiferous ducts and stroma

Fig. 5 Blue skip pattern of PTDL and its histological component.

Blue line reflects the 2-mm-diameter ductal component of the lesion
(dotted circle a), consisting of breast duct with DCIS surrounding

tissues and lymphocytes. The red line pattern is shown with a 1-mm
breast duct (dotted circle b) without tumor. The blue mass is invasive
ductal carcinoma (c)

123

10

Breast Cancer (2013) 20:312

around them, and others were composed of fat present in

the mammary glands because of regression of mammary
gland tissue (Fig. 4).
Since the green pattern reflects the status of almost no
difference in the distortion between PTDL and adjacent
tissue, it is easy to imagine that tumors exhibiting this
pattern have little intraductal expansion or invasion to
adjacent tissue. However, the green pattern may be present
when mammary glands are markedly hardened, as in cases
with severe mastopathy.
When it comes to the cases with the bblue line pattern,
DCIS, adjacent stromal growth, and infiltration of inflammatory cells were observed in the lactiferous duct tissue
200500 lm in size in the pathological samples where blue
lines were visualized as duct-like structure 12 mm in size.
As a result, they formed a tubular structure 2 mm in size
(Fig. 5). A lactiferous duct 200500 lm in size was hard to
recognize by US, but when it was filled with cancer tissue

or something occluded the duct, it seemed that adjacent

stromal growth and infiltration of inflammatory cells were
also there, and tubes 12 mm in size were formed [11];
these were visualized as PTDL and blue lines.
The potential reason why DCIS was exhibited as a skip
lesion is as follows. Originally, breast cancer does not
develop linearly, but in curves. When tumors grow and
expand, it can be imagined that original curves of the
lactiferous ducts are emphasized and large curves are
observed. One sectional US image may visualize the
structure zone of continuous but curvy lactiferous ducts
with intraductal lesions as skip lesions unless lactiferous
ducts are straight (Fig. 6). The structure of lactiferous ducts
visualized as skip lesions seem to reflect biologically
hardened expanding lesions.
In terms of the red line pattern, areas around the walls
of clearly recognized lactiferous ducts are often widely
observed in red. At the beginning of the study, it had been

Fig. 6 Blue skip pattern of PTDL and its reason for the image
skipping. Breast ducts are not straight; all of them are curved. Breast
ducts with DCIS are more complex. So tomographic images, such as

ultrasound images, will not visualize the continuous duct but will
produce skipped duct images

123

Breast Cancer (2013) 20:312

11

Fig. 8 Red line (of the red line pattern) in this case reflects the strain
image of the fatty tissue inside the breast glands

Fig. 7 Red line (of the blue to red line pattern) in this case reflects
strain image of the ligament around the fatty tissue that has sneaked
into the atrophic breast lobes by tumor traction

thought that relatively soft intraductal areas and large

distortion of lactiferous duct walls reflected the tip of lactiferous ducts where growth of tumors stopped.
However, when the image in Fig. 4 was intensively
observed, lactiferous ducts themselves were not exhibited
in red, but the wall structure zone above and below the
lactiferous ducts were shown in red. Part of red line
pattern showed atrophic fat tissue in the mammary
glands and the condition of adjacent fat tissue dragged by
the tumor, and severe distortion of membranous tissue
such as Coopers ligament showing the red line pattern
was observed in a scattered manner. Due to the created
membranous zone, continuous red lines seemed to be
exhibited in some cross sections (Fig. 7). The same image
was observed when fat was present in the mammary
glands (Fig. 8).
When atrophy due to aging is observed, it has been
postulated that lobules and stroma regress with the ductal
tissue as the axis, and they are replaced with fat tissue. Fat
tissue is present in a membranous manner in the mammary
glands according to the atrophy status. When cancer tissue
has a dragging effect over the adjacent structure, fat tissue
and adjacent membranous structures are distorted, and they
are visualized in lines in some cross sections because they
are membranes, and it can be imagined that they are
visualized in red lines in some cross sections. Taken
together, contrary to the blue line visualized as skip lesions

as previously mentioned, continuously visualized lesions

may be ductal tissue or fat tissue surrounded by membranous structure zones consisting of non-tumorous lesions.
In any case, this red line pattern indicates that expansion
of the tumor ends at the tumor side of this point. The reason
why cut end-negative first resection was achieved at a high
rate in cases with a blue to red line pattern was that the area
of tumor expansion was easier to recognize and accurate
marking was possible.

Conclusion
This is a preliminary study to investigate to what degree US
diagnosis for intraductal spreading was possible by a combination of B-mode US, Doppler scanning, and elastography
by comparison of the findings of a mass and its surroundings.
In the prediction of the expansion area of a breast cancer
lesion, evaluation of the expansion area was insufficient
using B-mode US aimed at morphological diagnosis alone,
but it was suggested that addition of new qualitative diagnostic modalities, such as Doppler scanning and elastography, would greatly improve the diagnostic accuracy.
Reliability of the results in this study was considered
low, because it was presumed that there were a variety of
biases. It is necessary to verify the utility of this diagnostic
method by modifying sample slicing so that the consistency between images and pathological evaluations can be
higher and to validate the accuracy of US comprehensive
diagnoses by planning a multi-center prospective study
excluding biases.

123

12

References
1. Japanese Breast Cancer Society (JBCS). General rules for clinical
and pathological recording of breast cancer. 16th ed. 2008.
2. American College of Radiology (ACR). Breast Imaging Reporting and Data System Atlas (BI-RADS Atlas). 2nd ed. 2003.
3. Cho N, Moon WK, Kim YH, Charg JM, Park SH, Lyou CY.
Sonoelastographic strain index for differentiation of benign
and malignant nonpalpable breast masses. J Ultrasound Med.
2010;29(1):17.
4. Itoh A, Ueno E, Tohno E, Shiina T, Yamakawa M, Matsumura T,
et al. Breast disease, clinical application of US elastography for
diagnosis. Radiology. 2006;239(2):34150.
5. Charfare H, MacLatchie E, Cordier C, Bradley M, Eadie C,
Byrtus A, et al. A Comparison of different methods of assessing
cosmetic outcome following breast conserving surgery and
factors influencing cosmetic outcome. BJMP. 2010;3(1):310.

123

Breast Cancer (2013) 20:312

6. Japanese Breast Cancer Society (JBCS). Clinical practice guidelines
for breast cancer: part of epidemiology and diagnosis. 2011 ed. 2011.
7. Tohno E, Takahashi H, Tamada T, Fujimoto Y, Yasuda H,
Ohuchi N. Educational program and testing using images for the
standardization of breast cancer screening by ultrasonography.
Breast Cancer. 2010. doi:10.1007/s12282-010-0221-x.
8. Chang JM, Moon WK, Cho N, Kim SJ. Breast mass evaluation:
factors influencing the quality of US elastography. Radiology.
2011;258(3).
9. Raza S, Odulate A, Ong EMW, Chikarmane S, Harston CW.
Using real-time tissue elastography for breast lesion evaluation.
J Ultrasound Med. 2010;29(4):55163.
10. Leong LCH, Sim LSJ, Lee YS, Ng FC, Wan CM, Fook-Chong
SMC, et al. A prospective study to compare the diagnostic performance of breast elastography versus conventional breast ultrasound. Clin Radiol. 2010;65(11):88794.
11. Tibor T. Breast cancer: a lobar disease, 2010.