| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Breast Imaging |
1 From the Departments of Radiology (M.C.H.) and Surgery (D.Y.N.) and the Clinical Research Institute (W.K.M, J.G.I.), Seoul National University Hospital and the Institute of Radiation Medicine, Seoul National University Medical Research Center, 28 Yongon-Dong, Chongno-Gu, Seoul 110-744, South Korea. From the 1999 RSNA scientific assembly. Received October 11, 1999; revision requested November 6; revision received February 14, 2000; accepted February 28. W.K.M., J.G.I., and D.Y.N. supported by grant no. 03-99-031 from the Seoul National University Hospital Research Fund. Address correspondence to W.K.M. (e-mail: moonwk@radcom.snu.ac.kr).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: Fifty nonpalpable breast lesions in 50 patients were prospectively evaluated with power Doppler US before and after injection of the contrast agent SH U 508A. Lesion vascularity and the morphology of vessels on US scans were analyzed and were correlated with histologic results.
RESULTS: Surgical excision revealed 22 cancers and 28 benign lesions. At nonenhanced power Doppler US, eight (36%) of 22 cancers and four (14%) of 28 benign lesions were vascular. At contrast agent–enhanced power Doppler US, 21 (95%) cancers and six (21%) benign lesions were vascular (P < .001). Irregular vessels were seen in three cancers and one benign lesion at nonenhanced power Doppler US and in 11 cancers and one benign lesion at contrast-enhanced power Doppler US. By using the presence of vascularity in the mass as the diagnostic criterion for malignancy, the sensitivity, specificity, and positive and negative predictive values of power Doppler US changed from 36%, 86%, 67%, and 63%, respectively, to 95%, 79%, 78%, and 96% after contrast agent injection.
CONCLUSION: Contrast-enhanced power Doppler US was superior to nonenhanced power Doppler US in the demonstration and characterization of tumor vascularity in nonpalpable breast lesions. Contrast-enhanced power Doppler US may be useful for the differentiation between nonpalpable breast cancers and benign tumors.
Index terms: Breast neoplasms, diagnosis, 00.30 • Breast neoplasms, US, 00.12983 • Ultrasound (US), contrast media, 00.12988 • Ultrasound (US), power Doppler studies, 00.12983
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Specific to Doppler US imaging is the assumption that neoangiogenesis within a malignant mass may enable differentiation from benign disease (9,10). Power Doppler US based on total integrated power of the Doppler spectrum is now considered superior to color Doppler US in the demonstration of vascular flow because of such advantages as high sensitivity to slow flow, no angle dependency, and no aliasing (11,12).
Recently developed microbubble US contrast agents also improve the detection of characteristic neovascular morphologic features by enhancing the signal strength in small vessels (13). Investigators in previous studies of Doppler US found that echo enhancement by a contrast agent could increase the diagnostic accuracy in breast tumors (14–16). In a study of 34 patients with breast tumors (14), injection of a microbubble agent increased the sensitivity from 89% (16 of 18) to 100% (18 of 18) and the specificity from 88% (14 of 16) to 100% (16 of 16). In that study, however, most of the tumors were palpable, and tumor vascularity was analyzed at conventional color Doppler US imaging. To our knowledge, the usefulness of a contrast agent in power Doppler US has not been evaluated in nonpalpable breast lesions.
We performed a prospective study to investigate the value of power Doppler US and a microbubble US contrast agent in the differentiation of nonpalpable breast lesions. The primary objective was to assess the sensitivity, specificity, and predictive values of power Doppler US before and after the administration of an US contrast agent.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
On mammograms, the lesions were seen as a mass in 18 cases, a mass with microcalcifications in 25 cases, microcalcifications in four cases, an asymmetric density in one case, and architectural distortion in two cases. The size of the lesion was 4–9 mm in 19 cases, 10–19 mm in 27 cases, and 20–23 mm in four cases. According to the American College of Radiology Breast Imaging Reporting and Data System (BI-RADS) (17), the final assessment category was category 3, probably benign lesions, in three cases; category 4, suspicious lesions, in 34 cases; and category 5, highly suspicious lesions, in 13 cases. Biopsy was performed in three probably benign lesions because of patient or referring clinician preference on clinical grounds. In four lesions that manifested only as clustered microcalcifications at mammography, the finding and location of the lesion at gray-scale US were confirmed with repeat gray-scale US examination after mammographically guided needle localization.
The contrast agent was SH U 508A (Levovist; Schering, Berlin, Germany), which is a suspension of monosaccharide microparticles (galactose) in sterile water. Microbubbles (2–8 µm in average diameter) are stabilized in the microparticle suspension; they can traverse the pulmonary capillary bed and enhance the signal of blood. Before US, the agent was prepared for injection by shaking it with 5 mL of water for 5–10 seconds. A milky suspension of galactose microparticles and microbubbles was created with the disaggregation of the granules. After standing for 2 minutes for equilibration, 6.0 mL of a 400 mg/mL suspension was injected manually at 1 mL/sec via an 18–20-gauge cannula placed in an antecubital vein. An additional 10 mL of physiologic saline solution was then injected at the same rate to flush the cannula. The intravenous cannula was placed prior to the commencement of the US examinations.
All US examinations were performed by one experienced radiologist (W.K.M.) with power map US equipment (HDI 3000; Advanced Technology Laboratories, Bothell, Wash). A 10–5-MHz, 38-mm linear transducer was used in all cases; the transducer automatically shifts to 6 MHz when in the power Doppler interrogation mode. The following settings were used: low transmit power, mechanical index of 0.7, and 55-dB log compression. The color box was enlarged to include the lesion and a margin of normal breast tissue. The scanning plane was selected for optimal visualization of vascularity. Color sensitivity was set at a level where the background color was just suppressed and a small vessel could be detected. The resultant power Doppler US gains ranged from 75% to 85%. Care was taken not to apply pressure with the probe, as this has previously been found to obliterate the small, low-pressure vessels associated with breast lesions.
Dynamic power Doppler US data were recorded on videotape and were stored in the US unit until the signal enhancement had completely diminished. Static images were obtained before and 30, 60, 90, 120, 180, 210, and 240 seconds after injection of the contrast agent. At each time frame, one representative hard-copy image was obtained in all cases.
All patients underwent hook wire–guided surgery within 24 hours of the US examinations. Needle localization was performed with US guidance in 18 lesions and mammographic guidance in 32 lesions. Radiography of the specimen was performed in all patients, and microcalcifications were confirmed in all 29 lesions. Surgical excision of 50 nonpalpable lesions revealed 22 cancers and 28 benign lesions. Malignant lesions were 14 cases of ductal carcinoma in situ (DCIS) and eight cases of invasive carcinoma. DCIS lesions were histologically comedocarcinoma in six lesions and noncomedocarcinoma in eight lesions. The nuclear grade of DCIS was classified as low grade in one lesion, intermediate grade in nine lesions, and high grade in four lesions. Eight invasive carcinomas included six ductal carcinomas, not otherwise specified; one 15-mm lobular carcinoma; and one 8-mm tubular carcinoma. The histologic diameter of the lesions was 5–30 mm (mean, 9.8 mm) for DCIS and 6–22 mm (mean, 13.1 mm) for invasive cancer.
Benign lesions were 13 cases of nonproliferative fibrocystic disease, nine cases of proliferative fibrocystic disease without atypia, one case of atypical ductal hyperplasia, and five fibroadenomas. The histologic diameter of the fibroadenomas was 5–15 mm (mean, 7.8 mm).
On mammograms, the mean lesion size was 11 mm (range, 4–21 mm) in cancers and 10 mm (range, 4–23 mm) in benign disease.
All US images were preoperatively analyzed and assessed in consensus by two radiologists (W.K.M., J.G.I.). At gray-scale US, the lesions were described according to the shape as oval, round, lobular, or irregular; according to the orientation as wider than tall or taller than wide; according to the echogenicity as hyperechoic, isoechoic, mildly hypoechoic, or markedly hypoechoic; according to the echotexture as homogeneous or heterogeneous; according to the margin as well defined, microlobulated, ill defined, or spiculated; according to the acoustic transmission as shadowing, enhancement, or normal; and according to the presence of a pseudocapsule (5).
At nonenhanced and contrast agent–enhanced power Doppler US, lesion vascularity and the distribution and morphology of the vessels were analyzed subjectively by two radiologists, one who actually performed the examination (W.K.M.) and one who evaluated the video and static images (J.G.I.). The US images obtained at the time of maximal enhancement were analyzed to determine the morphology of the tumor vessels. The vascularity of the lesion, determined according to the number and the estimated area of vessels within the mass, was classified into three categories: avascular, defined as without flow; hypovascular, defined as one or two vessels or less than 20% color flow; and hypervascular, defined as more than three vessels or 20% or greater color flow.
The amount of enhancement after contrast agent injection was classified into three categories: minimal, defined as no or equivocal additional vessels; moderate, defined as one or two additional vessels; and marked, defined as more than three additional vessels.
The distribution of vessels was classified as central, peripheral, or both. Central vessels had persistent focal color flow signal reproducibly detected within the lesion and not extending to the margin. Peripheral vessels were blood vessels at the edge of the mass or within 2 mm of the margin of the tumor.
The shape of the vessel was classified as regular or irregular. Irregular vessels were blood vessels that lacked the normal tapering, vessels with tortuosity, or vessels that penetrated the tumor radially. If there was a penetrating vessel (7), it was reported.
The radiologists provided a BI-RADS final assessment category to indicate the probability of malignancy. Each lesion was categorized as benign (benign or probably benign) or malignant (suspicious or highly suggestive of malignancy).
The vascularity of the lesions and the distribution and morphology of the vessels before and after contrast agent administration were correlated with histologic results. To see whether the findings at power Doppler US before and after contrast agent enhancement were different between malignant and benign lesions, the 
2 and Fisher exact tests were performed with a statistical software system (SAS for Windows, version 6.12; SAS Institute, Cary, NC). For practical purposes, grouping of hypervascular and hypovascular lesions versus avascular lesions was done retrospectively, and the statistical analysis was performed for a comparison of vascular versus avascular lesion findings with the Fisher exact test and for a comparison of hypervascular, hypovascular, and avascular lesion findings with the Mantel-Haenszel 2 test. The findings with a P value less than .05 were considered to be statistically significant.
By using the presence of vascularity in the mass or the presence of irregular vessels as the diagnostic criterion for malignancy, diagnostic accuracy, sensitivity, specificity, and positive and negative predictive values were calculated for nonenhanced and contrast-enhanced power Doppler US findings.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
After injection of the contrast agent, 21 (95%) of the 22 cancers were vascular: hypovascular in eight and hypervascular in 13. The increase in power Doppler signal reached a maximum at 30 seconds after injection in 12 cases and at 60 seconds after injection in nine cases. Eighteen cancers showed an increase in vascularity with the contrast agent (Fig 1): moderate in seven and marked in 11. The distribution of the vessels was peripheral in four cases, central in four cases, and peripheral and central in 13 cases. Irregular vessels were seen in 11 cases, and penetrating vessels were seen in five of these 11 cases (Fig 2). Tumors with irregular or penetrating vessels were classified as hypervascular in all 11 cases.
|
|
|
|
Benign Lesions
At nonenhanced power Doppler US, four (14%) of 28 benign lesions were vascular: hypovascular in three and hypervascular in one. The distribution of vessels in benign lesions was peripheral in three cases and peripheral and central in one case. Irregular vessels were seen in one case. A penetrating vessel was not seen in any case.
After injection of the contrast agent, six (21%) of 28 benign lesions were vascular: hypovascular in four and hypervascular in two. No vessels were seen in 22 cases (Fig 3). The increase in power Doppler signal reached a maximum at 60 seconds after injection in four cases and at 90 seconds after injection in two cases. Three benign lesions showed an increase in vascularity with the contrast agent: moderate in one fibroadenoma and one case of atypical ductal hyperplasia and marked in one proliferative fibrocystic lesion without atypia. The distribution of the vessels was peripheral in four cases, central in one case, and peripheral and central in one case. Irregular vessels were seen in one lesion at nonenhanced and contrast-enhanced power Doppler US. A penetrating vessel was not seen in any case.
|
|
Statistical Analysis
The difference in the vascularity between malignant and benign lesions was statistically significant only at contrast-enhanced power Doppler US with analysis of vascular and avascular lesions (P < .001, Fisher exact test) and of hypervascular, hypovascular, and avascular lesions (P < .01, Mantel-Haenszel 2 test). A moderate or marked increase in vascularity with administration of the contrast agent was predominantly seen in malignant lesions (P < .001, Mantel-Haenszel 2 test). The vascularity of invasive cancers and DCIS was significantly different only at nonenhanced power Doppler US (P < .01, Mantel-Haenszel 2 test). The difference in the distribution and morphology of vessels between malignant and benign lesions was not statistically significant at either nonenhanced or contrast-enhanced power Doppler US (P > .05, 2 test for distribution of vessels and Fisher exact test for morphology of vessels).
By using the presence of vascularity in the mass as the diagnostic criterion for malignancy, the sensitivity and positive and negative predictive values of power Doppler US improved from 36% (eight of 22), 67% (eight of 12), and 63% (24 of 38) to 95% (21 of 22), 78% (21 of 27), and 96% (22 of 23), respectively, after contrast agent injection. Overall accuracy also improved from 64% (32 of 50) to 86% (43 of 50), whereas the specificity decreased from 86% (24 of 28) to 79% (22 of 28).
By using an irregular vessel as the diagnostic criterion for malignancy, the sensitivity and positive and negative predictive values of power Doppler US improved from 14% (three of 22), 75% (three of four), and 59% (27 of 46) to 50% (11 of 22), 92% (11 of 12), and 71% (27 of 38), respectively, after contrast agent injection. Overall accuracy also improved from 60% (30 of 50) to 76% (38 of 50), whereas the specificity remained at 96% (27 of 28).
Surgical biopsy could have been avoided in 22 (79%) of 28 benign lesions if avascularity at contrast-enhanced power Doppler US had been used as the criterion for excluding malignancy (Table 3). However, using this criterion could have led to one missed biopsy for DCIS.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The presence of angiogenesis in invasive and preinvasive breast lesions raises the possibility that a breast imaging modality that shows lesion vascularity, such as Doppler US, may be useful for the identification or characterization of breast lesions in which the angiogenic phenotype is present. Pathologic vessels typically show an irregular course without progressive diminution in caliber (21).
In this study, power Doppler US rather than color Doppler US was used for the evaluation of nonpalpable lesions of the breast for the following reasons: greater sensitivity of power Doppler US for slow and low-volume flow and better display of vessel morphology. With power Doppler US, more vessels are likely to be depicted and vessels are more likely to be displayed as contiguous structures, which allows vessel morphology and characteristics to be visualized (12,15). In palpable breast cancers, power Doppler US has shown tumor vascularity in 73%–84% of cases (7,8,22). The presence of penetrating vessels as a marker for malignancy produced a sensitivity of 68% (17 of 25), specificity of 95% (58 of 61), positive predictive value of 85% (17 of 20), and negative predictive value of 88% (58 of 66) in one study (7).
These encouraging results could not be reproduced in nonpalpable breast tumors. There was a tendency toward a decreased amount of detected color as the lesion size decreased in both benign and malignant lesions. In this study, only 36% of breast cancers were vascular at nonenhanced power Doppler US. The majority (11 of 14) of DCIS lesions and more than one-third (three of eight) of invasive cancers, including one lobular carcinoma and another tubular carcinoma, were seen as avascular tumors. Irregular or penetrating vessels were seen in only three of 22 cancers.
In our study, injection of contrast material enabled better visualization of vessels seen at nonenhanced power Doppler US and of vessels not seen previously. Of 14 avascular cancers at nonenhanced power Doppler US, 13 cancers showed tumor vascularity after injection of the contrast agent. The one false-negative case was a noncomedocarcinoma DCIS lesion. Irregular or penetrating vessels were seen in 11 cancers at contrast-enhanced power Doppler US and in three cancers at nonenhanced power Doppler US. Our results show that Doppler US enhancement can contribute to the differentiation between nonpalpable breast cancers and benign tumors such as fibroadenomas.
In this study, surgical biopsy might have been avoided in 22 (79%) of 28 benign lesions if avascularity at contrast-enhanced power Doppler US had been used as the criterion for excluding malignancy. However, specificity decreased from 86% (24 of 28) to 79% (22 of 28) after contrast material administration. Three probably benign lesions at mammography included in this study might have influenced the high negative predictive value.
The most important function of an US contrast agent is to enhance the power of the backscattered echo signal. Backscattering and sonic attenuation are interrelated and depend on the concentration of the contrast agent. Backscattering increases at a low concentration, whereas the sonic attenuation caused by backscattering increases at a high concentration (23). In addition to the concentration, the injection rate of the agent may influence the duration of vascular enhancement in the tumor and the induction of color blooming artifact.
In previous studies, the recommended concentration and injection rate of SH U 508A for the evaluation of breast tumors with color Doppler US were 300 mg/mL and 2 mL/sec, respectively (14). When we used these parameters to perform power Doppler US, they rarely showed tumor vascularity in small masses. For these reasons, we chose the injection protocol used in this study after performing a pilot study with variable concentrations of SH U 508A in several nonpalpable breast cancers. Motion and color blooming artifacts, reported as limitations of contrast-enhanced power Doppler US in the abdomen (24), were not a substantial problem in this study. Further studies are needed to establish an optimal injection protocol that is suitable for the purpose of each examination.
In breast masses, tumor vascularity has been assessed with magnetic resonance (MR) imaging after injection of an extracellular contrast agent (25,26). Comparative studies of contrast-enhanced MR imaging and contrast-enhanced power Doppler US showed similar sensitivity and specificity in the detection and characterization of breast cancers (27). The potential advantage of MR imaging is the detection of multifocal disease. US is simple to perform, is performed in real time, and is relatively inexpensive. Microbubble contrast agent strictly confined to the vascular bed allows objective quantification of the enhancement dynamics of tumors at power Doppler US that is analogous to the time-activity curves in nuclear medicine isotope tracer studies (16). The assessment of vascularity with contrast-enhanced power Doppler US may also be helpful in guiding therapy, as some treatments specifically target the blood supplies of the tumors (28).
We did not perform quantitative measurements in this study because this was thought to be unduly time-consuming, yielding little additional diagnostic benefit. Quantitative parameters such as flow velocity or calculated indexes appear much less reliable than qualitative morphologic analysis of the vessels (15,29). Nonblinded interpretation and a small number of study cases could be the limitations. The two radiologists were not blinded to the US sequences; they knew which were with and which were without the contrast agent. This study is for characterizing nonpalpable masses that have already been demonstrated at mammography and US. Further studies are needed to determine whether the combination of contrast-enhanced power Doppler and gray-scale sonographic appearances can reduce the frequency of unnecessary biopsy of nonpalpable breast lesions.
In conclusion, our initial experience indicates that contrast-enhanced power Doppler US was superior to conventional power Doppler US in the demonstration and characterization of tumor vascularity in nonpalpable lesions of the breast. Contrast-enhanced power Doppler US may be useful for the differentiation between nonpalpable breast cancers and benign tumors.
|
FOOTNOTES |
|---|
Author contributions: Guarantor of integrity of entire study, W.K.M.; study concepts and design, W.K.M., D.Y.N., M.C.H.; definition of intellectual content, W.K.M., J.G.I., M.C.H.; literature research, W.K.M.; clinical studies, W.K.M., J.G.I., D.Y.N.; experimental studies, W.K.M., J.G.I., D.Y.N.; data acquisition and analysis, W.K.M., J.G.I., D.Y.N.; statistical analysis, W.K.M.; manuscript preparation and editing, W.K.M., J.G.I., M.C.H.; manuscript review, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Du, F.-H. Li, H. Fang, J.-G. Xia, and C.-X. Zhu Microvascular Architecture of Breast Lesions: Evaluation With Contrast-Enhanced Ultrasonographic Micro Flow Imaging J. Ultrasound Med., June 1, 2008; 27(6): 833 - 842. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R.-F. Chang, S.-F. Huang, W. K. Moon, Y.-H. Lee, and D.-R. Chen Solid Breast Masses: Neural Network Analysis of Vascular Features at Three-dimensional Power Doppler US for Benign or Malignant Classification Radiology, April 1, 2007; 243(1): 56 - 62. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. M. Kwee, M. A. A. J. van den Bosch, M. El Ouamari, B. Kemps, L. Arensman, R. van Hillegersberg, and W. P. Th. M. Mali Contrast-enhanced breast ultrasonography reveals an unusual breast tumor in a male patient with gynecomastia. J. Ultrasound Med., October 1, 2006; 25(10): 1347 - 1351. [Full Text] [PDF]
|
|
|
|
 |
|
 J. L. del Cura, E. Elizagaray, R. Zabala, A. Legorburu, and D. Grande The Use of Unenhanced Doppler Sonography in the Evaluation of Solid Breast Lesions Am. J. Roentgenol., June 1, 2005; 184(6): 1788 - 1794. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. Santamaria, M. Velasco, X. Farre, J. A. Vanrell, A. Cardesa, and P. L. Fernandez Power Doppler Sonography of Invasive Breast Carcinoma: Does Tumor Vascularization Contribute to Prediction of Axillary Status? Radiology, February 1, 2005; 234(2): 374 - 380. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Ozdemir, K. Kilic, H. Ozdemir, C. Yucel, S. Andac, and M. Colak Contrast-Enhanced Power Doppler Sonography in Breast Lesions: Effect on Differential Diagnosis After Mammography and Gray Scale Sonography J. Ultrasound Med., February 1, 2004; 23(2): 183 - 195. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. K. Moon, J. S. Myung, Y. J. Lee, I. A. Park, D.-Y. Noh, and J.-G. Im US of Ductal Carcinoma In Situ RadioGraphics, March 1, 2002; 22(2): 269 - 281. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
