| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Department of Radiology, Kinki University School of Medicine, 377-2
Ohno-Higashi, Osaka-Sayama, Osaka 589-8511, Japan.
2 Department of Surgery, Kinki University School of Medicine, Osaka 589-8511,
Japan.
Received February 3, 2003;
accepted after revision April 1, 2003.
Address correspondence to M. Inoue
(inoue{at}med.kindai.ac.jp).
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. MDCT was performed in 149 women with suspected breast tumors, and 173 breast lesions were detected. These breast lesions were classified as either mass or nonmass enhancing lesions. For mass lesions, the margin, shape, and enhancement patterns were evaluated. For nonmass enhancing lesions, the distribution of enhancement and the types of time-density curve patterns were evaluated. MDCT was compared with mammography and sonography as a method of revealing breast tumors.
RESULTS. Of the 173 breast lesions detected, 150 were mass lesions, 131 (87%) of which were malignant. Of the 23 nonmass enhancing lesions, 21 (91%) were malignant. The most highly predictive features for lesion malignancy were an irregular margin (100%), an irregular shape (99%), and rim enhancement (100%). Similar features were the most accurate signs of malignancy—a spiculated and irregular margin (90%). On time-density curves, the washout and plateau patterns showed high positive predictive value (93%) and sensitivity (91%) for malignancy. However, these patterns had low negative predictive value (42%) and specificity (48%). Seven breast lesions that could not be detected on mammography or sonography were identified on MDCT. MDCT more accurately revealed the margin of the tumor invasion in 11 breast tumors than did mammography or sonography.
CONCLUSION. The features revealed on MDCT can help to distinguish benign lesions from carcinomas. MDCT can add to the data obtained with mammography or sonography in patients with suspected breast tumors.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Multidetector CT (MDCT) scanners can acquire multiple CT data sets with each rotation of the X-ray tube and can scan through large anatomic areas three to seven times faster than can single-detector helical CT scanners [6]. The speed with which MDCT can scan an area and the thinner collimation possible with MDCT improve the spatial and time resolution of images acquired to detect breast lesions. The purpose of our study was to evaluate the potential usefulness of dynamic MDCT in the differentiation of breast cancers from benign lesions. In addition, the detectability of breast tumors on MDCT was compared with the detectability of breast tumors on mammography and sonography.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
MDCT Protocol
On an MDCT scanner (Aquilion, Toshiba Medical, Tokyo, Japan) set for 1-mm
collimation and a pitch of 5.5, we scanned our patients from the level of the
axilla to the lower edge of the breast. Four breath-hold acquisitions were
obtained before and 1, 3, and 8 min after a IV rapid bolus administration of
nonionic contrast material. We infused 100 mL of nonionic contrast material
(Omnipaque [iohexol], Daiichi Pharmaceutical, Tokyo, Japan) at a rate of 3.0
mL/sec. The data were reconstructed at 0.6-mm increments. In the initial phase
of our study, 100 patients underwent MDCT in the prone position. The last 49
patients underwent MDCT in the supine position because it allowed surgical
simulation on three-dimensional data displays.
After reconstruction, the images were transferred to a workstation. Multiplanar reconstructions (axial, coronal, and sagittal) were used for the evaluation of tumors. For evaluation of time-density curve patterns, lesion interpretation was performed using the region-of-interest method for all 173 lesions. Types of time-density curve patterns were categorized as washout (an abrupt decline in density 3-8 min after contrast material injection), plateau (stabilized enhancement without a change in density 3-8 min after the injection), or persistent (an increasing density throughout the 8-mm period) (Fig. 1).
|
Image Interpretation
One radiologist reviewed all detected lesions for morphologic features and
time-density curve patterns. All detected lesions were classified as either
mass or nonmass enhancing lesions. Descriptive terms of morphologic features
are listed in Tables 1 and
2. For mass lesions, margin,
shape, enhancement pattern, and types of time-density curve patterns were
evaluated. For nonmass enhancing lesions, distribution of enhancement and the
types of time-density curve patterns were evaluated. The positive predictive
value of each morphologic feature was calculated. Statistical analysis was
performed using the chi-square test with the Yates correction.
|
|
Comparison with Mamography and Sonography
Two radiologists reviewed the mammograms, and one radiologist reviewed the
sonograms obtained by two surgeons. Although sonography parameters were not
optimized for this comparison, the two surgeons who acquired the sonograms are
specialists in breast surgery and are experienced in breast sonography.
Using histologic findings as the gold standard, one radiologist compared MDCT with mammography and sonography in the detection of breast lesions and the depiction of tumor invasion.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Masses.—Table 1 shows the frequency of morphologic features (descriptors) for mass lesions and the positive predictive values for these features. Among mass lesions, the features of malignancy with high positive predictive value were an irregular margin (100%), an irregular shape (99%), and rim enhancement (100%). Figures 2A, 2B, and 2C shows a patient with an invasive ductal carcinoma with an irregular shape and a spiculated margin. We found a significantly higher frequency of carcinomas among masses with a spiculated or irregular margin than among those with a smooth margin (99% vs 45%, p < 0.0001). Of the 15 malignant lesions with smooth margins, eight were invasive ductal carcinomas, three were DCIS, three were mucinous carcinomas, and one was a medullary carcinoma. Figures 3A, 3B, and 3C shows a patient with a mucinous carcinoma with a smooth margin.
|
|
|
|
|
|
The frequency of malignancy among masses with an irregular shape was significantly higher than that of masses with the other shapes (99% vs 67%, p < 0.0001). The frequency of carcinoma among masses with rim or homogeneous enhancement was significantly higher than that of heterogeneous enhancement (95% vs 63%, p < 0.0001).
Nonmass lesions.—Table 2 shows the frequency and positive predictive value of nonmass morphologic features (descriptors). Twenty-one of 23 nonmass lesions were malignant. Nine (39%) of 23 nonmass lesions were DCIS. Figures 4A, 4B, and 4C shows a patient with DCIS displaying segmental enhancement. Approximately two thirds of all DCIS lesions were nonmass enhancing lesions. The incidence of nonmass enhancement in invasive ductal carcinomas was significantly lower than that in DCIS (6% vs 64%, p < 0.0001).
|
|
|
Assessment of Time-Density Curve Patterns
Table 3 shows the frequency
and positive predictive value of time-density curve patterns. Of the 173
lesions, 110 (64%) showed a washout pattern, 39 (23%) showed a plateau
pattern, and 24 (14%) showed a persistent pattern. The incidence of carcinoma
was significantly higher in lesions showing the washout pattern than in those
showing the persistent pattern (95% vs 58%, p < 0.0001), although
24% of benign tumors also showed the washout pattern. The difference between
the incidence of the persistent pattern in carcinomas (14/152 or 9%) and in
benign tumors (10/21 or 48%) was statistically significant (p <
0.0001). Four mucinous carcinomas showed the persistent pattern, and the
remaining two showed the plateau pattern. One of our more interesting findings
was that none of the six mucinous carcinomas showed the washout pattern.
|
Diagnostic Accuracy
Table 4 shows the diagnostic
reliability of four lesion features: margin, shape, internal enhancement, and
time-density curve pattern. A spiculated or irregular margin, an irregular
shape, a rim or homogeneous enhancement, and the washout or plateau pattern on
time-density curves were associated with malignancy. Time-density curve
pattern showed the highest sensitivity (91%) among the four features for
predicting malignancy. Margin and shape had both high specificity and high
positive predictive value (both 99%) for predicting malignancy. A spiculated
or irregular margin had the highest negative predictive value (55%) for
malignancy. Similarly, a spiculated or irregular margin showed the highest
accuracy (90%) for malignancy.
|
Comparison of MDCT with Mammography and Sonography
Of the 173 breast lesions detected on MDCT, 20 lesions (13 invasive
carcinomas, four cases of fibrocystic disease, and three fibroadenomas) were
not revealed on mammography, and 11 lesions (five invasive carcinomas, three
DCIS, one case of fibrocystic disease, and two fibroadenomas) were not
revealed on sonography. Both mammography and sonography failed to reveal seven
multifocal or multi-centric lesions that were shown on MDCT in seven patients
(four invasive carcinomas, two fibroadenomas, and one case of fibrocystic
disease) (Figs. 5A, and
5B). The primary tumors in the
seven patients had been detected on mammography or sonography.
|
|
MDCT was superior to mammography and sonography in the depiction of the margin of the tumor invasion in 11 breast lesions (six invasive ductal carcinomas, three invasive lobular carcinomas, one DCIS, and one fibroadenoma) (Figs. 6A, and 6B). For the depiction of the tumor margins of the remaining 155 breast tumors, MDCT was equal to mammography or sonography.
|
|
Change of Surgical Methods
MDCT data were the basis for a change in surgical methods in six of the 147
patients. Breast conservation surgery was changed to mastectomy for four
patients. For the other two patients, additional lesions detected on MDCT
required local excision.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
We undertook this study to describe the features of MDCT-detected breast lesions with terms from the breast CT lexicon that we developed, using the lexicons of MR imaging-detected breast lesions as our references [12, 13]. We also sought to assess the positive predictive value of these features. In our study, 87% of lesions were categorized as masses and 13% were nonmass enhancing lesions. We found the frequency of DCIS to be significantly higher in nonmass enhancing lesions than in mass lesions. Nine of 14 DCIS were detected as nonmass lesions. These findings are analogous to those of previous studies of lesions detected on mammography and MRI [7, 9, 14-16].
We found that a spiculated margin had a positive predictive value for malignancy of 99%, which was substantially higher than the 70-80% frequency of carcinoma among masses with spiculated margins detected mammographically [9]. This result may be explained by the fact that MDCT can depict the mass margin accurately without overlapping other tissues. In contrast, 45% of smoothly marginated masses that we studied were malignant. However, we cannot report that a smooth margin is a reliable indicator of benignity on MDCT. First, the proportion of malignant masses in our study was higher than the proportion of such masses in other studies [7, 11]. Second, three mucinous carcinomas and one medullary carcinoma had smooth margins. The results are consistent with histologic findings for mucinous and medullary carcinomas [17].
For nonmass enhancing lesions, the incidence of malignancy was 91%, higher than the incidence of 67-86% for lesions detected on MRI reported in previous studies [18, 19]. Perhaps this result is related to the fact that the proportion of cases of fibrocystic disease was low in our study.
Reliable features for breast cancer on MDCT were nonmass enhancing lesions and mass lesions with a spiculated margin, irregular shape, and rim enhancement. Spiculated or irregular margins showed the highest accuracy as features with which to differentiate malignant from benign lesions. Although MDCT of the breast can be used to predict the presence of breast cancer with an accuracy of 90%, biopsy remains the final diagnostic procedure.
In terms of the time-density curve, the positive predictive values of the washout, plateau, and persistent patterns were 95%, 85%, and 58%, respectively. We found that the washout or plateau pattern was a predictor of likelihood of carcinoma. The sensitivity and the positive predictive value of the washout or plateau pattern on the time-density curve were similar to those of the spiculated or irregular margin. However, specificity, negative predictive value, and accuracy of the washout or plateau pattern were lower than those of the spiculated or irregular margin. We believe that a spiculated or irregular margin of a lesion on MDCT is the most reliable indicator of malignancy. Thus, the morphologic analysis is more accurate than is the time-density curve analysis. In addition, the time-density curve analysis exposes the patient to increased radiation dose, and we currently do not perform a time-density curve analysis as a routine examination in patients with breast lesions at our institution.
For the detection of breast tumors and depiction of tumor invasion, we found MDCT to be superior to mammography and sonography. Therefore, we suggest MDCT as an efficient technique for the preoperative examination and staging of breast tumors. Surgical plans for six patients were changed because of the additional information provided by MDCT. MDCT depicted the tumor invasion of invasive lobular carcinoma more accurately than did the conventional modalities. Invasive lobular carcinoma has a propensity to be multifocal. Thus, when one lesion is detected on mammography, additional lesions may be occult [20, 21]. MDCT may be better suited for the detection and staging of invasive lobular carcinoma than are conventional techniques.
MR mammography is reported to be more accurate than mammography and sonography for assessing the extent of breast cancers [4, 5]. Although one study comparing single-detector helical CT and MR mammography has been reported [22], we know of no reported comparison between MDCT and MR mammography. In the study comparing helical CT and MR mammography, helical CT was found to perform almost as well as MR mammography for assessment of the extent of breast cancer [22]. We plan to perform a study comparing MDCT and MR mammography in assessing the extent of breast cancer using MR equipment with a high field strength.
MDCT has several advantages over MR mammography. With MDCT, the breast can be scanned faster and with thinner collimation than is possible with MR mammography and the patient can be supine, allowing a surgical simulation for breast conservation using a three-dimensional display and an easy approach for CT-guided breast biopsy [23, 24]. In addition, MDCT allows one to evaluate axilla lymph nodes and lung metastases simultaneously.
One of the disadvantages of MDCT is that the breast is exposed to radiation. The measured radiation dose at the skin surface of the breast for a single breath-hold acquisition of MDCT at our institution was 26 mGy, similar to the dose reported for conventional CT and approximately 10 times higher than the dose received during standard mammography [25, 26]. This disadvantage may not be a serious problem for patients who are thought to have breast cancer on the basis of findings on other diagnostic techniques.
MDCT is now used as a complementary tool to mammography and sonography. In clinical practice, breast MDCT is indicated for preoperative patients with breast cancer to detect multicentric or multifocal lesions and to assess the extent of breast cancer and for patients with indeterminate findings for breast lesions on conventional modalities to differentiate malignancy from benign lesions.
In conclusion, reliable features for breast cancer on MDCT were nonmass enhancing lesions and mass lesions with a spiculated margin, an irregular shape, and rim enhancement. These features revealed on MDCT can help to distinguish benign tumors from carcinomas. Time-density curve patterns were not reliable predictors of the benignity of breast tumors. MDCT was superior to mammography and sonography in the detection of breast tumors and depiction of the tumor invasion.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  A. Perrone, L. Lo Mele, S. Sassi, M. Marini, L. Testaverde, L. Izzo, and M. Marini MDCT of the Breast Am. J. Roentgenol., June 1, 2008; 190(6): 1644 - 1651. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
D. K. Kang, E. J. Kim, H. S. Kim, J. S. Sun, and Y. S. Jung Correlation of Whole-Breast Vascularity with Ipsilateral Breast Cancers Using Contrast-Enhanced MDCT Am. J. Roentgenol., February 1, 2008; 190(2): 496 - 504. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. G. Harish, S. D. Konda, H. MacMahon, and G. M. Newstead Breast Lesions Incidentally Detected with CT: What the General Radiologist Needs to Know RadioGraphics, October 1, 2007; 27(suppl_1): S37 - S51. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Shimauchi, T. Yamada, A. Sato, K. Takase, S. Usami, T. Ishida, T. Moriya, and S. Takahashi Comparison of MDCT and MRI for evaluating the intraductal component of breast cancer. Am. J. Roentgenol., August 1, 2006; 187(2): 322 - 329. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Tozaki, T. Kobayashi, S. Uno, K. Aiba, H. Takeyama, H. Shioya, I. Tabei, Y. Toriumi, M. Suzuki, and K. Fukuda Breast-Conserving Surgery After Chemotherapy: Value of MDCT for Determining Tumor Distribution and Shrinkage Pattern Am. J. Roentgenol., February 1, 2006; 186(2): 431 - 439. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
