Radiology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online before print August 9, 2007, 10.1148/radiol.2443061588
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2443061588v1
244/3/718    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Forsberg, F.
Right arrow Articles by Hall, A. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Forsberg, F.
Right arrow Articles by Hall, A. L.
(Radiology 2007;244:718-726.)
© RSNA, 2007

Breast Imaging

Breast Lesions: Imaging with Contrast-enhanced Subharmonic US—Initial Experience1

Flemming Forsberg, PhD, Catherine W. Piccoli, MD 2, Daniel A. Merton, BS, RDMS, Juan J. Palazzo, MD, and Anne L. Hall, PhD

1 From the Departments of Radiology (F.F., C.W.P., D.A.M.) and Pathology (J.J.P.), Thomas Jefferson University, Suite 763J, Main Building, 132 S 10th St, Philadelphia, PA 19107; and GE Healthcare, Milwaukee, Wis (A.L.H.). Received September 13, 2006; revision requested November 9; revision received January 10, 2007; final version accepted February 1. Supported in part by the U.S. Army Medical Research Material Command under DAMD17-00-1-0464 and by GE Healthcare, Princeton, NJ. Address correspondence to F.F. (e-mail: flemming.forsberg{at}jefferson.edu).


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Purpose: To prospectively compare accuracy of gray-scale subharmonic imaging (SHI) with that of standard gray-scale ultrasonography (US), power Doppler US (with and without contrast material), and mammography for the diagnosis of breast cancer, with histopathologic or clinical follow-up results as the reference standard.

Materials and Methods: This HIPAA-compliant pilot study had institutional review board approval; all subjects gave written informed consent. Fourteen women (age range, 37–66 years) had 16 biopsy-proved breast lesions. In SHI, pulses are transmitted at one frequency, but only echoes at half that frequency (the subharmonic) are received. A US scanner was modified to perform gray-scale SHI (transmitting at 4.4 and receiving at 2.2 MHz). Precontrast imaging (gray-scale US and power Doppler) was followed by contrast material–enhanced power Doppler and gray-scale SHI. A reader blinded to mammographic and pathologic findings assessed diagnosis on a six-point scale. Sensitivity, specificity, accuracy, and receiver operating characteristic (ROC) curves were computed for mammography, gray-scale and power Doppler imaging (pre- and postcontrast), and SHI.

Results: Of the 16 lesions, four (25%) were malignant. Mammography had 100% sensitivity and 20% specificity. Sensitivity and specificity, respectively, were 50% and 92% for precontrast imaging and 75% and 75% for contrast-enhanced power Doppler. SHI had 75% sensitivity and 83% specificity. Specificity was higher for all US modes than for mammography (P < .04). There were no significant differences in specificity among US modes or in sensitivity (P ≥ .50). Area under the ROC curve for the diagnosis of breast cancer was 0.64 for standard gray-scale US and power Doppler US, 0.67 for contrast-enhanced power Doppler US, 0.76 for mammography, and 0.78 for SHI (P > .20). Contrast enhancement was better with SHI than with power Doppler (100% vs 44% of lesions with good or excellent enhancement; P = .004).

Conclusion: SHI appears to improve the diagnosis of breast cancer relative to conventional US and mammography, albeit on the basis of results in a very limited number of subjects.

© RSNA, 2007


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Mammography is the first-line imaging test for screening, detecting, and diagnosing breast lesions (1,2). However, in current clinical practice, 65%–90% of all breast biopsy samples are found to be benign (2,3). There is, therefore, a need for an imaging technique to improve the diagnosis of breast cancer by enabling reliable differentiation between malignant and benign masses. Ultrasonographic (US) imaging may be one such technique (3–5). It enables easy differentiation between cystic and solid lesions (unlike mammography) and can also be used to guide breast biopsies in real time. Moreover, US has been shown to improve the characterization of solid breast lesions as benign or malignant (3–5). Diagnosis of breast cancer on the basis of Doppler US flow imaging findings has produced mixed results owing to overlap between flow measurements in benign and malignant tumors (5–7). One problem may be the lack of sensitivity of US for depicting flow in small tumor neovessels (8). Because angiogenic vascular morphology is an independent predictor of malignant breast disease (9–11), it should therefore be an important marker for breast US flow imaging.

Microbubble-based US contrast agents combined with nonlinear contrast agent–specific imaging techniques markedly improve both the sensitivity and the specificity of diagnostic US (8,12). At higher acoustic pressures (>0.5 MPa), US contrast agents produce marked nonlinear signal components in the received echoes that span the range from subharmonic to ultraharmonic frequencies and can be exploited for imaging (12). One particular nonlinear contrast agent imaging technique is second harmonic imaging (HI), which is commercially available most commonly in combination with pulse inversion imaging (12–16). In pulse inversion techniques, a pair of pulses 180° out of phase with one another is transmitted, and the received signals are summed to cancel first (and other odd) harmonics. However, HI (even in combination with pulse inversion imaging) suffers from reduced blood-to-tissue contrast that results from second harmonic signal generation and accumulation in tissue. Subharmonic imaging (SHI), in which sound pulses are transmitted at one frequency (f0) but echoes are only received at the subharmonic frequency (f0/2), may be an attractive alternative to HI owing to the lack of subharmonic generation in tissue and the marked subharmonic scattering produced by some contrast agents (16). Feasibility studies of contrast material–enhanced imaging in the SHI mode have been conducted in vitro and in vivo by our group (17–22) and by others (23–27).

Although in vivo SHI has been reported in animal studies (17,21), to date, no one has, to our knowledge, produced in vivo SHI images in humans. We hypothesized that SHI can be performed in the breast at clinically relevant contrast agent doses (approximately half the dose used in animal studies). Thus, the purpose of our study was to prospectively compare the accuracy of gray-scale SHI with that of gray-scale US, power Doppler US (with and without contrast material), and mammography for the diagnosis of breast cancer, with histopathologic or clinical follow-up results as the reference standard.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
The study was approved by the institutional review board of Thomas Jefferson University and was compliant with the Health Insurance Portability and Accountability Act. All subjects who met the inclusion criteria over the study period and who were willing to participate gave written informed consent prior to enrollment in the study. The US contrast agents studied in this trial are approved by the U.S. Food and Drug Administration for use in echocardiography and were employed off-label for breast imaging (this was explicitly stated in the consent forms).

This pilot study was sponsored in part by GE Healthcare, Princeton, NJ. The sponsor provided the US microbubble contrast agent Optison (perflutren protein-type microspheres). The authors of this article (except A.L.H., who is an employee of GE Healthcare) had sole control of the data generated by this trial and the information provided for publication.

Subjects
Fourteen women with 16 breast lesions were enrolled in this prospective study between January 2005 and December 2005. The subjects were women older than 21 years of age with a breast mass or a parenchymal abnormality other than a mass that resulted in a breast biopsy (core or surgical) being scheduled. The mean age was 54 years (range, 37–66 years). Eleven of the women (nine of whom had one lesion each and two of whom had two lesions each) were referred after mammography revealed a solid mass or a suspicious area (eg, microcalcifications). The remaining three women (with one lesion each) were included on the basis of findings at palpation, US, or magnetic resonance imaging. Participants had to be in medically stable condition; pregnant or nursing women were excluded. Demographic characteristics (ie, age and race) were recorded.

US Examinations
First, a baseline gray-scale US examination was performed by a sonographer (D.A.M., with 20 years of experience) with a scanner (Logiq 9; GE Healthcare, Milwaukee, Wis) and a broad-bandwidth linear array transducer (model 7L; bandwidth, 3–7 MHz) to identify the largest section of the mass or abnormality seen with mammography. Second, baseline power Doppler images of the target region were obtained. Image parameters such as gain, pulse repetition frequency, and filter settings were adjusted for each woman to optimize flow visualization and were then kept constant across all US modes. Digital cine clips of the mass or area of abnormality were stored on the scanner hard drive for each baseline imaging mode. Next, a US contrast agent was administered intravenously via a peripheral vein, typically the antecubital vein. The first six subjects evaluated received Optison in a 0.5-mL dose. After the voluntary recall of Optison in November 2005 (28), the remaining eight participants received 0.25 mL of perflutren lipid microspheres (Definity; Bristol-Myers Squibb Imaging, North Billerica, Mass). A digital cine clip was obtained with the gain and power Doppler settings unaltered from the precontrast settings, except that the gain could be lowered if excessive color blooming occurred because of the contrast agent (29). This allowed for side-by-side comparison of the pre- and postcontrast power Doppler studies.

After a waiting period of 15 minutes (to avoid any cumulative effect of the contrast agent and to ensure a return to baseline conditions), a second contrast agent injection was administered for gray-scale SHI, and a digital cine clip was acquired for later offline analysis. The contrast doses employed were 4.0 and 1.4 mL for Optison and Definity, respectively (equivalent to approximately 0.05 and 0.02 mL per kilogram of body weight). The Logiq 9 scanner was modified (by A.L.H., with 22 years of experience as a US engineer) to operate in gray-scale subharmonic mode with a transmit frequency of 4.4 MHz and a receive frequency of 2.2 MHz. This corresponds to the resonance frequency of Optison, which is approximately 2 MHz and is the frequency range used with the US scanners employed for our previous SHI work (18,22). This is an important design criterion for SHI, because the threshold above which subharmonic generation occurs reaches a minimum when the insonation frequency is twice the resonance frequency of the bubbles imaged (17). Moreover, SHI was implemented as a pulse inversion technique. Although the experimental software used for SHI did not provide a calibrated indication of the acoustic field, we have previously shown that the incident acoustic pressure amplitude with SHI is less than 1.5 MPa (peak negative pressure, 0.65 MPa) when measured in water (22). Moreover, the software was tested at GE Healthcare and was found to comply with U.S. Food and Drug Administration power output requirements.

Reference Standard
After the contrast-enhanced US study, the subjects underwent core or surgical biopsy within 2 hours. Biopsy was followed by a final histopathologic assessment by a pathologist (J.J.P., with 20 years of experience), which served as the reference standard (except in one lesion, for which the reference standard was provided by results of 9 months of clinical follow-up because the scheduled biopsy was changed to watchful waiting). The presence or absence of various lesion components (eg, microcalcifications) was also recorded. A number of other histopathologic variables were also recorded as part of the participants' clinical care, but, because none of these variables demonstrated significant differences compared with the US imaging results in this pilot study (P > .12), data regarding these variables have been omitted for the sake of brevity.

Data Assessment
Evaluation of the real-time US cine clips was performed offline by a radiologist (C.W.P., with 17 years of experience in breast imaging) who was blinded to the mammographic and pathologic findings. The diagnoses at pre- and postcontrast US were made on the basis of the radiologist's experience and were expressed by using a six-point scale, in which 0 indicated no lesion seen (no findings); 1, benign findings; 2, probably benign findings; 3, indeterminate findings; 4, probably malignant findings; and 5, malignant findings. The pre- and postcontrast US diagnostic criteria (detailed below) were evaluated for each subject in one consecutive reading as follows: Standard gray-scale images were reviewed first, gray-scale and power Doppler (baseline) US images were reviewed next, baseline and contrast-enhanced power Doppler US images were then reviewed, and, finally, gray-scale and SHI images were reviewed (one after the other in all cases). Although this may have introduced some bias by comparing pre- with postcontrast results, it was considered the more realistic approach to how contrast-enhanced US may be used in clinical practice. Finally, each mammogram was read by a radiologist as part of the participant's standard clinical assessment and was assessed by using the Breast Imaging Reporting and Data System (30).

All imaging variables collected were compared with all histopathologic variables (F.F., with 16 years of experience in US research). The pre- and postcontrast US diagnostic criteria included the following: overall diagnosis, the size of the lesion, degree of enhancement (none, mild, good, excellent, or excessive), degree of vascularity (avascular, mild, moderate, highly focused, highly diffuse, or no lesion seen), blood vessel morphology, and vessel anastomoses.

Statistical Analysis
Because this was a pilot study, no statistical power analysis could be performed a priori. Moreover, for the statistical analyses detailed below, lesions were considered to be independent and contrast agents were assumed to perform equivalently to allow the very limited data set to be analyzed as a whole.

Sensitivity, specificity, and accuracy for lesion characterization were calculated for all five imaging modalities (the four US modes and mammography). Lesions that were diagnosed as being in the indeterminate category were classified as malignant, as in our prior study of contrast-enhanced US imaging of the breast (31), because in practice any lesion characterized as indeterminate would, like those in higher categories, have to undergo biopsy. Comparisons of sensitivities, specificities, and accuracies for diagnosing cancer with the different imaging modalities were performed (F.F.) with the McNemar test for correlated proportions (against the histopathologic reference standard) and software (Stata 8.0; Stata Corporation, College Station, Tex) and by considering exact binomial P values of less than .05 to indicate significance.

The incremental value of combining the US modes (individually or together) with mammography was analyzed by using logistic regression and receiver operating characteristic (ROC) analyses (32,33). Differences between ROC curves were tested by computing Mann-Whitney statistics with software (Stata 8.0). All histopathologic variables were compared (by using two-way tests) with imaging judgments of diagnosis and with vascularity characteristics for the different modalities. When both sets of variables were nominal, {chi}2 tests were performed. When both types of variables were ordinal or continuous, correlations were calculated. When one type of variable was nominal and one was continuous, nonparametric rank order tests, such as Mann-Whitney U tests or Kruskal-Wallis tests, were performed (34).


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 
Subject and Lesion Characteristics
A total of 90 subjects were screened for enrollment from January to June and from October to December of 2005. Of these subjects, 76 were excluded (Fig 1). The 14 women enrolled in this study were predominantly Caucasian (10 women [71%]), while four (29%) were African American. Among the 16 lesions, there were 12 benign lesions and four cancers (25%) with average diameters of 12 mm ± 7 (standard deviation) and 18 mm ± 11, respectively (P = .22). A significant difference in age was found between subjects with benign lesions and those with malignant lesions (52 years ± 8 and 61 years ± 5, respectively; P = .04). The majority (three of four) of cancers were invasive, but there was one lesion of ductal carcinoma in situ. One mass was considered benign after 9 months of follow-up, while the histopathologic classification of the other 11 benign lesions included one with fibrocystic changes and five fibroadenomas. The remaining five masses were classified as "other" and included three complex cysts, one lymph node, and a dermal scar. Microcalcifications were observed in five participants (three with benign and two with malignant disease). Only three women presented with no associated mass; two of these women had benign disease.


Figure 1
View larger version (19K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 1: Flowchart of procedures performed in this pilot study.

 
US Findings
All lesions imaged with mammography (13 of the 16 lesions) were scored as Breast Imaging Reporting and Data System category 3 or 4 (two and 11 lesions, respectively). In vivo breast SHI resulted in an almost complete suppression of tissue signals, allowing better visualization of lesion vascularity relative to that provided by baseline power Doppler (Fig 2). This suppression of tissue signals was even more marked on the baseline images (the ones obtained before administration of Optison), which were almost completely black (no subharmonic flow signals were detectable before the microbubbles arrived, only noise). The internal morphology of the vascularity associated with the breast masses was also visualized better with SHI than with contrast-enhanced power Doppler imaging (Fig 3). The enhancement in SHI mode typically lasted 1.0–1.5 minutes (ie, a bit shorter than that with contrast-enhanced power Doppler imaging).


Figure 2A
View larger version (81K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2a: Fibroadenoma (arrows) imaged in its largest section by using (a) baseline power Doppler mode and (b) SHI mode after injection of 4.0 mL of Optison. Although the correct diagnosis (probably benign) was obtained with both modalities, note the improved display in SHI mode of small intratumoral vessels (indicating that only minimal bubble destruction is occurring) and the excellent suppression of tissue echoes.

 

Figure 2B
View larger version (96K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 2b: Fibroadenoma (arrows) imaged in its largest section by using (a) baseline power Doppler mode and (b) SHI mode after injection of 4.0 mL of Optison. Although the correct diagnosis (probably benign) was obtained with both modalities, note the improved display in SHI mode of small intratumoral vessels (indicating that only minimal bubble destruction is occurring) and the excellent suppression of tissue echoes.

 

Figure 3A
View larger version (52K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3a: Benign ductal microcalcifications and fibrocystic changes, as well as areas of hyperplasia and adenosis (arrows in a), are depicted along the longest axis with (a) contrast-enhanced power Doppler imaging and (b) SHI. Note the excessive color blooming in a compared with the improved depiction of small branching vessels within and around the lesion (arrows) in b. However, the SHI image is dominated by excessive noise outside the focal region, and a false-positive assessment was rendered with both imaging modes.

 

Figure 3B
View larger version (72K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 3b: Benign ductal microcalcifications and fibrocystic changes, as well as areas of hyperplasia and adenosis (arrows in a), are depicted along the longest axis with (a) contrast-enhanced power Doppler imaging and (b) SHI. Note the excessive color blooming in a compared with the improved depiction of small branching vessels within and around the lesion (arrows) in b. However, the SHI image is dominated by excessive noise outside the focal region, and a false-positive assessment was rendered with both imaging modes.

 
Although the Logiq 9 US scanner was optimized for SHI performed with Optison, it also worked well when Definity was used (Fig 4). In this example, two lesions were visualized at gray-scale US (Fig 4a), and both were considered cysts in the radiologist's evaluation. However, after contrast agent administration in power Doppler mode (Fig 4b), the lesion on the subject's right was upgraded to the indeterminate category (because of the spotty peripheral flow visualized), and the one on the left was changed to probably malignant (because of the more complete vascular filling observed). Once gray-scale SHI was performed, the former lesion was downgraded to probably benign because it was essentially avascular in this imaging mode (Fig 4c), while the latter lesion remained an indeterminate mass, given the extensive vascularity demonstrated. Histopathologic examination revealed that the lesion on the right was ductal carcinoma in situ, while the lesion on the left was a cyst (as correctly predicted at SHI, where indeterminate lesions were treated as malignant).


Figure 4A
View larger version (88K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4a: (a) Gray-scale US image shows ductal carcinoma in situ (*) and cyst ({otimes}) visualized in their largest cross sections. (b) Contrast-enhanced power Doppler image shows spotty peripheral flow in the cancer and more complete filling (color blooming artifact) in the cyst. (c) Gray-scale SHI image shows both peripheral and intratumoral flow in the cancer (*), while the cyst ({otimes}) is essentially avascular.

 

Figure 4B
View larger version (72K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4b: (a) Gray-scale US image shows ductal carcinoma in situ (*) and cyst ({otimes}) visualized in their largest cross sections. (b) Contrast-enhanced power Doppler image shows spotty peripheral flow in the cancer and more complete filling (color blooming artifact) in the cyst. (c) Gray-scale SHI image shows both peripheral and intratumoral flow in the cancer (*), while the cyst ({otimes}) is essentially avascular.

 

Figure 4C
View larger version (78K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 4c: (a) Gray-scale US image shows ductal carcinoma in situ (*) and cyst ({otimes}) visualized in their largest cross sections. (b) Contrast-enhanced power Doppler image shows spotty peripheral flow in the cancer and more complete filling (color blooming artifact) in the cyst. (c) Gray-scale SHI image shows both peripheral and intratumoral flow in the cancer (*), while the cyst ({otimes}) is essentially avascular.

 
Comparison of Imaging Modalities
Mammography achieved a sensitivity of 100% and a specificity of 20%, for an overall accuracy of 38% (Table). Standard gray-scale US and baseline power Doppler performed identically, with sensitivities of 50%, specificities of 92%, and accuracies of 81%. The combination of baseline US and contrast-enhanced power Doppler imaging had a sensitivity, specificity, and accuracy of 75%; these values increased with SHI to 75%, 83%, and 81%, respectively. Contrast-enhanced power Doppler US had a significantly higher specificity than mammography (P = .03), as did the other US modes (P = .008). There were no significant differences among the specificities or sensitivities achieved with the US modes (P ≥ .50 for both) in this data set. Among the accuracy values calculated, only that for SHI showed a trend toward a significant improvement over the accuracy of mammography (P = .07), while differences for all other comparisons were nonsignificant (P ≥ .11).


View this table:
[in this window]
[in a new window]

  		Sensitivity, Specificity, and Accuracy of the Five Imaging Modalities

 
The area under the individual ROC curve for the diagnosis of breast cancer was higher for SHI (0.78) than for baseline US (0.64), power Doppler US (0.64), contrast-enhanced power Doppler US (0.67), and mammography (0.76) (Fig 5). However, none of these differences were significant (P ≥ .20). Logistic regression techniques were used to combine the four US imaging modes and mammography before the ROC analysis was repeated. The highest area under the ROC curve produced was 0.87 for the four US modes combined. However, this value was not significantly different from any of the other areas under the ROC curve that were obtained individually, as well as combined (P > .101).


Figure 5
View larger version (18K):
[in this window]
[in a new window]
[Download PPT slide]
  		Figure 5: Individual ROC curves for the four imaging modalities: baseline gray-scale US (dotted line), contrast-enhanced power Doppler imaging (dash-and-dot line), SHI (solid line), and mammography (dashed line). Note the increased area under the curve for SHI compared with that for baseline US and contrast-enhanced power Doppler.

 
The ability to characterize vessel morphology increased significantly from baseline power Doppler imaging to SHI (zero vs seven [0% vs 44%] of the 16 lesions, P = .016). Contrast-enhanced power Doppler performed in between the other two vascular imaging modes in this regard (three [19%] of the lesions could be characterized), but these differences were not statistically significant (P > .12). A similar pattern was observed for the ability to visualize anastomoses, which increased from 0% to 6% to 25% (or no, one, and four of 16 lesions) with power Doppler imaging, contrast-enhanced power Doppler imaging, and SHI, respectively (.12 < P < .20). Moreover, linear regression showed that the degree of vascularity observed with power Doppler and that observed with contrast-enhanced power Doppler both correlated with SHI findings (r2 ≥ 0.40, P < .009) but did not correlate with each other (r2 = 0.18, P = .103). This suggests that SHI can provide improved visualization of lesion vascularity compared with baseline power Doppler, while avoiding the blooming artifacts associated with contrast-enhanced power Doppler imaging. However, the clinical importance of these findings (if any) is currently unknown (owing to the limited study population). Finally, the enhancement produced in the two contrast-enhanced modes (contrast-enhanced power Doppler and SHI) was also assessed. The results showed that contrast enhancement was more likely to be rated as good or excellent with SHI than with power Doppler imaging (16 vs seven [100% vs 44%] of 16 lesions, P = .004).


   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
ADVANCES IN KNOWLEDGE
 References
 
In total, 14 subjects with four cancers and 12 benign lesions were evaluated in this pilot study. To our knowledge, we have produced the first examples of in vivo SHI images acquired in humans, thus demonstrating that nonlinear subharmonic microbubble signals can be obtained in the breast at clinically relevant contrast agent dosages. In clinical practice, such an SHI evaluation would add 15–20 minutes to the total examination time. The specificity of each of the four US modes was significantly greater than that of mammography (P < .04). Baseline US (ie, gray-scale US and power Doppler imaging) and SHI achieved the highest accuracy (81%) in this study, but only SHI showed a trend toward a significant improvement over mammography (P = .07). The area under the ROC curve for the diagnosis of breast cancer was higher for SHI than for any of the other techniques tested (0.78 vs 0.64, 0.67, and 0.76). However, none of these differences were statistically significant (P ≥ .20). These results indicate that SHI may increase the ability to diagnose breast cancer compared with current techniques (albeit on the basis of findings in a very small subject group).

The degrees of vascularity observed with pre- and postcontrast power Doppler imaging correlated with those observed with SHI (r2 ≥ 0.40, P < .009) but not with each other (r2 = 0.18, P = .103), presumably because of color blooming. SHI also demonstrated the vascular morphology (eg, vessel smoothness and anastomoses) of more lesions than did power Doppler (although this was not always statistically significant: .01 < P < .20). These results indicate that SHI may provide better visualization of lesion neovascularity relative to baseline power Doppler, while avoiding the blooming artifacts associated with contrast-enhanced power Doppler imaging. Finally, in this limited study, power Doppler imaging findings did not alter the diagnosis of any breast lesions when added to gray-scale US findings. However, in a large study of 761 breast masses (5), the incremental value of adding Doppler flow imaging for the evaluation of breast masses was clearly demonstrated (the area under the ROC curve increased from 0.938 for mammography alone to 0.964 when US and color Doppler findings were included in the analysis; P = .0008).

There are relatively few reports in the literature on contrast-enhanced US evaluation of breast masses (31,35–46), and most of these studies employed contrast-enhanced power Doppler imaging. Although initial reports were very encouraging, with 100% sensitivity and specificity obtained after contrast agent administration (35), later studies were less successful (37,38), with sensitivities and specificities ranging from 56% to 90%. These results are similar to the 75% sensitivity and specificity achieved with contrast-enhanced power Doppler in our study. To our knowledge, three studies on contrast-enhanced three-dimensional flow imaging of breast lesions have been published to date (31,41,42). Our group studied 55 women and achieved an area under the ROC curve for the diagnosis of breast cancer with contrast-enhanced three-dimensional power Doppler imaging of 0.76 (31), which was significantly better than that achieved with the other US modes evaluated (P < .03). However, Hochmuth et al (42) reported no additional benefits with contrast-enhanced three-dimensional breast imaging, unlike the conclusions of the other two contrast-enhanced three-dimensional imaging studies (31,41). Most likely, the differences reported above are due to the relatively small patient populations studied (ranging from 34 to 220 women).

The use of newer nonlinear gray-scale contrast-enhanced US imaging techniques, such as pulse inversion HI, for breast imaging has, to our knowledge, been reported in only two studies to date (39,45). Although Jung et al (46) refer to "contrast enhanced harmonic ultrasound" in their breast study, they used only contrast-enhanced power Doppler with a background anatomic image obtained in tissue harmonic mode (no contrast-enhanced gray-scale imaging was performed). Cassano and colleagues (45) studied the characteristics of the microcirculation of breast lesions in 50 patients, but only report that their results were "negative" in the sense that no differences could be found between four types of breast lesions (15 fibroadenomas, 15 lesions classified with cytologic examination as probably malignant, 10 lesions with microcalcifications, and 10 postoperative scars with suspected recurrence). Our group (39) previously found, in a study of 33 patients, an area under the ROC curve for the diagnosis of breast cancer with pulse inversion HI (0.83) that was very similar to the SHI results reported here (0.78). Efforts are underway to conduct a larger study in which gray-scale pulse inversion HI and SHI are directly compared for breast imaging; we hope to report those results in the future.

Limitations of our study were mainly due to the very small number of participants included in this initial trial of breast SHI, which limits the statistical power of the study and, therefore, the clinical conclusions that can be made. Instead, our results provide an indication for a future larger clinical trial. Such a trial should involve more than one reader and should use only one US contrast agent—unlike this pilot study, which was limited by not having these factors. Moreover, our study may have been biased from the pre- to postcontrast assessments, because of the order of the reading. Finally, the small sample size of this study necessitated (on three occasions) the inclusion of multiple lesions from the same woman (which may not be statistically independent) and the inclusion of one woman in whom the final diagnosis was not biopsy proved but was instead based on watchful waiting. These were limitations that a future larger trial must avoid with careful study design and suitable statistical power considerations.

Thus, in this pilot study, the use of a contrast agent–specific US imaging technique, gray-scale SHI, was investigated for in vivo breast imaging. To our knowledge, we present the first-ever subharmonic images acquired in humans. SHI appears to improve the diagnosis of breast cancer relative to conventional US and mammography, albeit on the basis of results in a very limited number of subjects. Further studies are required to substantiate these results in a larger patient population.


   ADVANCES IN KNOWLEDGE
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
References
 


   ACKNOWLEDGMENTS
 
We gratefully acknowledge the assistance provided by Colleen Dascenzo, CRA, in recruiting women for this study, as well as the nursing support provided by Donna George, RN.


   FOOTNOTES
 

Abbreviations: HI = harmonic imaging • ROC = receiver operating characteristic • SHI = subharmonic imaging

2 Current address: South Jersey Radiology Associates, Voorhees, NJ. Back

Guarantor of integrity of entire study, F.F.; study concepts/study design or data acquisition or data analysis/interpretation, all authors; manuscript drafting or manuscript revision for important intellectual content, all authors; manuscript final version approval, all authors; literature research, F.F.; clinical studies, F.F., C.W.P., J.J.P.; experimental studies, F.F., C.W.P., D.A.M., A.L.H.; statistical analysis, F.F.; and manuscript editing, all authors


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 ADVANCES IN KNOWLEDGE
 References
 

  1. Tabár L, Dean PB. Mammography and breast cancer: the new era. Int J Gynaecol Obstet 2003;82:319–326.[CrossRef][Medline]
  2. Kopans DB. Breast imaging. 2nd ed. Philadelphia, Pa: Lippincott Williams & Wilkins, 1998.
  3. Zonderland HM, Coerkamp EG, Hermans J, van de Vijver MJ, van Voorthuisen AE. Diagnosis of breast cancer: contribution of US as an adjunct to mammography. Radiology 1999;213:413–422.[Abstract/Free Full Text]
  4. Stavros AT, Thickman D, Rapp CL, Dennis MA, Parker SH, Sisney GA. Solid breast nodules: use of sonography to distinguish between benign and malignant lesions. Radiology 1995;196:123–134.[Abstract/Free Full Text]
  5. Taylor KJ, Merritt C, Piccoli C, et al. Ultrasound as a complement to mammography and breast examination to characterize breast masses. Ultrasound Med Biol 2002;28:19–26.[CrossRef][Medline]
  6. Bohm-Velez M, Mendelson EB. Computed tomography, duplex Doppler ultrasound and magnetic resonance imaging in evaluating the breast. Semin Ultrasound CT MR 1989;10:171–176.[Medline]
  7. Adler DD, Carson PL, Rubin JM, Quinn-Reid D. Doppler ultrasound color flow imaging in the study of breast cancer: preliminary findings. Ultrasound Med Biol 1990;16:553–559.[CrossRef][Medline]
  8. Ferrara KW, Merritt CR, Burns PN, Foster FS, Mattrey RF, Wickline SA. Evaluation of tumor angiogenesis with US: imaging, Doppler and contrast agents. Acad Radiol 2000;7:824–839.[CrossRef][Medline]
  9. Weidner N, Folkman J, Pozza F, et al. Tumor angiogenesis: a new significant and independent prognostic indicator in early-stage breast cancer. J Natl Cancer Inst 1992;84:1875–1887.[Abstract/Free Full Text]
  10. Gasparini G, Harris AL. Clinical importance of the determination of tumor angiogenesis in breast carcinoma: much more than a new prognostic tool. J Clin Oncol 1995;13:765–782.[Abstract/Free Full Text]
  11. Relf M, LeJeune S, Scott PA, et al. Expression of the angiogenic factors vascular endothelial cell growth factor, acidic and basic fibroblast growth factor, tumor growth factor beta-1, platelet-derived endothelial cell growth factor, placenta growth factor, and pleiotrophin in human primary breast cancer and its relation to angiogenesis. Cancer Res 1997;57:963–969.[Abstract/Free Full Text]
  12. Goldberg BB. Raichlen JS, Forsberg F. Ultrasound contrast agents: basic principles and clinical applications. 2nd ed. London, England: Dunitz, 2001.
  13. Albrecht T, Hoffman CW, Schettler S, et al. B-mode enhancement using phase-inversion US with air-based microbubble contrast agent: initial experience in humans. Radiology 2000;216:273–278.[Abstract/Free Full Text]
  14. Forsberg F, Liu JB, Chiou HJ, Rawool NM, Parker L, Goldberg BB. Comparison of fundamental and wideband harmonic contrast imaging of liver tumors. Ultrasonics 2000;38:110–113.[CrossRef][Medline]
  15. Wilson SR, Burns PN, Muradali D, Wilson JA, Lai X. Harmonic hepatic US with microbubble contrast agent: initial experience showing improved characterization of hemangioma, hepatocellular carcinoma and metastasis. Radiology 2000;215:153–161.[Abstract/Free Full Text]
  16. Kim TK, Choi BI, Han JK, Park SH, Moon SG. Hepatic tumors: contrast agent-enhancement patterns with pulse-inversion harmonic US. Radiology 2000;216:411–417.[Abstract/Free Full Text]
  17. Forsberg F, Shi WT, Goldberg BB. Subharmonic imaging of contrast agents. Ultrasonics 2000;38:93–98.[CrossRef][Medline]
  18. Shi WT, Forsberg F, Hall AL, et al. Subharmonic imaging with microbubble contrast agents: initial results. Ultrason Imaging 1999;21:79–94.[Medline]
  19. Shi WT, Forsberg F, Raichlen JS, Needleman L, Goldberg BB. Pressure dependence of subharmonic signals from contrast microbubbles. Ultrasound Med Biol 1999;25:275–283.[CrossRef][Medline]
  20. Bhagavatheeshwaran G, Shi WT, Forsberg F, Shankar PM. Subharmonic generation from contrast agents in simulated neovessels. Ultrasound Med Biol 2004;30:199–203.[CrossRef][Medline]
  21. Forsberg F, Liu JB, Shi WT, Furuse J, Shimizu M, Goldberg BB. In vivo pressure estimation using subharmonic contrast microbubble signals: proof of concept. IEEE Trans Ultrason Ferroelectr Freq Control 2005;52:581–583.[Medline]
  22. Forsberg F, Liu JB, Shi WT, et al. In vivo perfusion estimation using subharmonic contrast microbubble signals. J Ultrasound Med 2006;25:15–21.[Abstract/Free Full Text]
  23. Shankar PM, Krishna PD, Newhouse VL. Advantage of subharmonic over second harmonic backscatter for contrast-to-tissue echo enhancement. Ultrasound Med Biol 1998;24:395–399.[CrossRef][Medline]
  24. Chomas J, Dayton P, May D, Ferrara K. Nondestructive subharmonic imaging. IEEE Trans Ultrason Ferroelectr Freq Control 2002;49:883–892.[CrossRef][Medline]
  25. Goertz DE, Cherin E, Needles A, et al. High frequency nonlinear B-scan imaging of microbubble contrast agents. IEEE Trans Ultrason Ferroelectr Freq Control 2005;52:65–79.[CrossRef][Medline]
  26. Goertz DE, Needles A, Burns PN, Foster FS. High-frequency, nonlinear flow imaging of microbubble contrast agents. IEEE Trans Ultrason Ferroelectr Freq Control 2005;52:495–502.[CrossRef][Medline]
  27. Ganor Y, Adam D, Kimmel E. Time and pressure dependence of acoustic signals radiated from microbubbles. Ultrasound Med Biol 2005;31:1367–1374.[CrossRef][Medline]
  28. Drug Alert Class 1 Medicines Recall-Amersham Health AS (Part of GE Healthcare)-Optison Suspension for Injection 3ml Vial. Medicines and Healthcare Products Regulatory Agency. November 23rd, 2005. http://www.info.doh.gov.uk/doh/EmBroadcast.nsf/0/641a8cc5d80ba800802570c2003f54fd?OpenDocument. Accessed March 29, 2006.
  29. Forsberg F, Liu JB, Burns PN, Merton DA, Goldberg BB. Artifacts in ultrasonic contrast agent studies. J Ultrasound Med 1994;13:357–365.[Abstract]
  30. American College of Radiology. Breast imaging reporting and data system. 3rd ed. Reston, Va: American College of Radiology, 1998.
  31. Forsberg F, Goldberg BB, Merritt CR, et al. Diagnosing breast lesions with contrast-enhanced 3-dimensional power Doppler imaging. J Ultrasound Med 2004;23:173–182.[Abstract/Free Full Text]
  32. Metz CE. ROC methodology in radiologic imaging. Invest Radiol 1986;21:720–733.[Medline]
  33. DeLong ER, DeLong DM, Clarke-Pearson DL. Comparing the areas under two or more correlated receiver operating characteristic curves: a nonparametric approach. Biometrics 1988;44:837–845.[CrossRef][Medline]
  34. Rosner B. Fundamentals of biostatistics. 3rd ed. Boston, Mass: PWS-Kent, 1990.
  35. Kedar RP, Cosgrove D, McCready VR, Bamber JC, Carter ER. Microbubbble contrast agent for color Doppler US: effect on breast masses. Radiology 1996;198:679–686.[Abstract/Free Full Text]
  36. Madjar H, Prompeler HJ, Del Favero C, Hackeloer BJ, Llull JB. A new Doppler signal enhancing agent for flow assessment in breast lesions. Eur J Ultrasound 2000;12:123–130.[CrossRef][Medline]
  37. Stuhrmann M, Aronius R, Schietzel M. Tumor vascularity of breast lesions: potentials and limits of contrast-enhanced Doppler sonography. AJR Am J Roentgenol 2000;175:1585–1589.[Abstract/Free Full Text]
  38. Goldberg BB, Merritt CRB, Needleman L, et al. Contrast enhanced ultrasound and mammography for breast cancer detection [abstr]. J Ultrasound Med 2001;20:S95.
  39. Furuse F, Forsberg F, Goldberg BB, et al. Pulse inversion harmonic imaging for breast cancer diagnosis [abstr]. Ultrasound Med Biol 2003;29:S93–S94.
  40. Kim SW, Park SS, Ahn SJ, et al. Identification of angiogenesis in primary breast carcinoma according to the image analysis. Breast Cancer Res Treat 2002;74:121–129.[CrossRef][Medline]
  41. Yang WT, Tse GM, Lam PK, Metreweli C, Chang J. Correlation between color power Doppler sonographic measurement of tumor vasculature and immunohistochemical analysis of microvessel density for the quantitation of angiogenesis. J Ultrasound Med 2002;21:1227–1235.[Abstract/Free Full Text]
  42. Hochmuth A, Boehm T, Bitzer C, Fleck M, Schneider A, Kaiser WA. Differentiation of breast masses using 3-D sonographic and echo-enhancer-based evaluation of the vascular pattern: initial experiences. Ultrasound Med Biol 2002;28:845–851.[CrossRef][Medline]
  43. Zdemir A, Kilic K, Ozdemir H, Yucel C, Andac S, Colak M. Contrast-enhanced power Doppler sonography in breast lesions: effect on differential diagnosis after mammography and gray scale sonography. J Ultrasound Med 2004;23:183–195.[Abstract/Free Full Text]
  44. Kettenbach J, Helbich TH, Huber S, Zuna I, Dock W. Computer-assisted quantitative assessment of power Doppler US: effects of microbubble contrast agent in the differentiation of breast tumors. Eur J Radiol 2005;53:238–244.[CrossRef][Medline]
  45. Cassano E, Rizzo S, Bozzini A, Menna S, Bellomi M. Contrast enhanced ultrasound of breast cancer. Cancer Imaging 2006;6:4–6.[CrossRef][Medline]
  46. Jung EM, Jungius KP, Rupp N, et al. Contrast enhanced harmonic ultrasound for differentiating breast tumors: first results. Clin Hemorheol Microcirc 2005;33:109–120.[Medline]




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
2443061588v1
244/3/718    most recent
Right arrow Submit a response
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Forsberg, F.
Right arrow Articles by Hall, A. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Forsberg, F.
Right arrow Articles by Hall, A. L.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
RADIOLOGY RADIOGRAPHICS RSNA JOURNALS ONLINE