| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Department of Radiology, Duke University School of Medicine, DUMC 3808, Durham, NC 27710.
Received May 19, 2004;
accepted after revision August 11, 2004.
Supported by National Institutes of Health grant number
R01-CA090561-01A1.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. Sonograms of 403 solid lesions were analyzed by one of three dedicated breast radiologists. Each lesion was described using features from the sonographic BI-RADS lexicon. Lesion description and biopsy results were correlated. PPV and NPV were calculated.
RESULTS. Histologic results showed that 141 (35%) of 403 masses were malignant. Sonographic BI-RADS descriptors showing high predictive value for malignancy include spiculated margin (86%, 19/22), irregular shape (62%, 102/164), and nonparallel orientation (69%, 75/109). Sonographic BI-RADS descriptors highly predictive of benign lesions include circumscribed margin (90%, 160/178), parallel orientation (78%, 228/294), and oval shape (84%, 200/237). For the sonographic BI-RADS features of mass margin, shape, orientation, lesion boundary, echo pattern, and posterior acoustic features, descriptors chosen were significantly (p < 0.001) different for malignant and benign masses.
CONCLUSION. Descriptors from the new sonographic BI-RADS lexicon can be useful in differentiating benign from malignant solid masses.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The positive predictive value (PPV) of mammographic features described in the original mammography BI-RADS lexicon has been investigated [6, 7]. These studies found that the mammography lexicon was useful in differentiating benign and malignant breast lesions. Several studies have assessed the utility of sonographic features in distinguishing benign from malignant lesions [8-10]. However, to our knowledge, no studies to date have assessed the PPV and negative predictive value (NPV) of sonographic features as described in the new sonographic BI-RADS lexicon. The purpose of this study was to evaluate the sonographic features of solid breast lesions with known histologic diagnosis and determine the predictive value of features from the new sonographic BI-RADS lexicon for malignant versus benign diagnosis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Cases for analysis in this study were selected from those recommended for biopsy. Between February 23, 2000, and February 28, 2002, 654 female patients with 738 lesions underwent sonographically guided biopsy at our facility. A lesion was included in this study if it corresponded to a solid mass on sonography and if both mammographic and sonographic films taken before the biopsy were available for review. Forty lesions were excluded from the study because they underwent biopsy before acquisition of the mammographic or sonographic images, because such biopsies can cause tissue changes that distort the appearance of the mass. In addition, 181 lesions were excluded because either mammographic or sonographic films were unavailable for review, 82 lesions were excluded because no solid mass was found at histology (including 68 cysts, six ducts, one abscess, and seven cases of normal breast tissue), 13 lesions were excluded because they were found to be lymph nodes at histologic diagnosis, and 10 lesions were excluded because final pathologic diagnoses were obtained at other facilities and were unavailable. Nine lesions were excluded because only nonpalpable calcifications were observable on mammography, and such lesions would not typically be sent for sonography and sonographically guided biopsy in clinical practice. The remaining 403 lesions in 369 patients constituted the study population, including 202 palpable lesions and 201 nonpalpable ones. The median age for these patients was 50.0 years (age range, 18-89 years).
Mammography
Most patients underwent initial craniocaudal and mediolateral-oblique
mammography, with additional true lateral and spot compression magnification
mammograms available in almost all cases. Young patients under 30 years old
with palpable abnormalities received a single unilateral mediolateral-oblique
view to limit radiation to the developing breast. Mammographic breast
composition revealed BI-RADS type 1 (fatty) in 16 (4%) of the 403 cases, type
2 (scattered fibroglandular densities) in 157 cases (39%), type 3
(heterogeneously dense) in 162 cases (40%), and type 4 (extremely dense) in 68
cases (17%). One hundred ten cases (27%) were occult to mammography but
visible on sonography. Two hundred thirty-three lesions were visible on
mammography as a discrete mass seen on both craniocaudal and
mediolateral-oblique projections. An additional 60 lesions were visible on
mammography as focal asymmetry (n = 24), architectural distortion
(n = 32), and calcifications (n = 4) presenting as a
palpable mass.
Sonographic Imaging Technique
Sonographic examinations were performed to evaluate mammographically
identified masses or densities in 161 (40%) of 403 cases, palpable lesions in
66 cases (16%), and lesions that were both palpable and mammographically
visible in 132 cases (33%). Incidental lesions in 44 cases (11%) were noted
during sonography of other mammographic or palpable masses. Sonography was
performed by one of five dedicated breast imaging radiologists with 6-18 years
of experience, and mammographic views were available to the radiologist at the
time the sonogram was performed. Sonography was performed using a
variable-frequency linear transducer set at 12 MHz (Sonoline Elegra, Siemens
Medical Solutions). For lesions in the lateral aspect of the breast, the
patient was imaged in the supine-oblique position, and for other lesions, the
patient was supine. Images were acquired in both radial and antiradial
projections with and without caliper measurements. Additional gray-scale
images were obtained in almost all cases to better show the lesion. Doppler,
color Doppler, and power Doppler images were not part of the routine imaging
protocol.
Imaging Interpretation and Data Analysis
For this study, each case was evaluated by one of three dedicated breast
radiologists with 6-11 years of experience. Each case was analyzed by only one
observer. Information about the patient's age, physical examination findings,
family history of breast cancer, and personal history of breast malignancy was
available to each radiologist to best reproduce a realistic clinical
situation. The radiologist was blinded to the histologic diagnosis during the
evaluation.
The interpreting radiologist first evaluated the mammograms. The observer determined the breast parenchyma density and then determined if the lesion in question was visible on mammography. If visible, the lesion was described using BI-RADS mammographic descriptors of mass margin (circumscribed, obscured, microlobulated, ill-defined/indistinct, or spiculated), shape (oval, round, lobular, or irregular), and density (fat-containing, low, equal, or high). The presence or absence of calcifications was noted, along with any associated findings (e.g., architectural distortion) or special cases (e.g., axillary lymphadenopathy).
After initial interpretation of mammograms, prior mammographic views were presented for comparison if available. Prior mammograms were available for 195 (48%) of the 403 lesions.
After reviewing the mammographic images, the observer was presented with static sonographic images. The observer was first asked to assess each mass using the lexicon popularized by Stavros et al. [8]. Subsequently, the observer was asked to assess the same images using features of the new sonographic BI-RADS lexicon [4].
For each category from the mammographic BI-RADS lexicon, the sonographic lexicon described by Stavros et al. [8], and the sonographic BI-RADS lexicon, the radiologist was limited to selecting one best feature descriptor. If the lesion could be described by more than one term, the observer was instructed to choose the term that described a lesion that was most suspicious for malignancy.
Sonographically guided core needle biopsy was performed for each lesion using a 14-gauge needle (Achieve, Allegiance Health Care; C. R. Bard). The histologic diagnosis was recorded for each case and correlated with the imaging feature analysis. Lesions initially diagnosed by core needle biopsy as atypia, malignant, and having discordant imaging and histology underwent surgical excision. For discordant and atypical lesions, final histologic diagnoses were based on excisional biopsy results.
Statistical software (SAS software system, version 8.2, SAS Institute) was used to calculate the Pearson's chi-square test when assessing statistical significance.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Of 233 lesions visible as a mass on mammography, 87 (37%) were malignant. PPVs for mammographic BI-RADS features are shown in Table 1. Descriptors chosen for features of mass shape, margin, and density are significantly (p < 0.0003) different for malignant versus benign masses.
|
Table 2 shows the predictive
value of morphology descriptors defined by Stavros et al.
[8]. The descriptors of
spiculated margin (91%, 21/23) and taller-than-wide shape (71%, 76/107) were
highly predictive of malignancy. The presence of a thin echogenic
pseudocapsule and a circumscribed margin was highly predictive of a benign
diagnosis, with NPVs for malignancy of 95% (63/66) and 90% (173/193),
respectively. Descriptors for the Stavros features shown are significantly
(p 
0.0001) different between malignant and benign masses.
|
The PPV of malignancy for the sonographic BI-RADS lexicon features were
also calculated (Table 3).
Predictive values for sonographic mass shape and mass margin in palpable and
nonpalpable masses only are also shown
(Table 4). A significant
difference (p 0.0005) between descriptors of benign masses and
those of malignant masses is shown. Spiculated margin (PPV = 86%, 19/22),
irregular shape (PPV = 62%, 102/164), and nonparallel orientation (PPV = 69%,
75/109) display high predictive value for malignancy. Circumscribed margin,
oval shape, and parallel orientation are predictive of a benign lesion with
NPVs of 90% (160/178), 84% (200/237), and 78% (228/294), respectively.
|
|
Surrounding tissue effects were defined in only 56 (14%) of 403 masses. These included one mass with three surrounding tissue effects (Cooper's ligament changes, skin thickening, and architectural distortion), two masses with Cooper's ligament changes and architectural distortion, one mass with edema and architectural distortion, and one mass with Cooper's ligament changes and skin retraction. The other 51 cases had 24 instances of Cooper's ligament change, 15 examples of architectural distortion, seven instances of duct change, and five cases of edema. Lesions in 42 (75%) of the 56 cases were malignant. Similarly, 76% (16/21) of masses with calcifications were malignant, although only 21 masses were described as having either macrocalcifications or microcalcifications visible on sonography.
Special cases were infrequently identified, which is understandable because only solid masses that had undergone biopsy were included in this study. Foreign bodies, clusters of microcysts, and masses in or on the skin are unlikely to be recommended for biopsy, and complicated cysts are not solid lesions and thus were excluded from this study. Three lesions (1%) in 403 cases were described as special cases (two intramammary lymph nodes and one axillary lymph node) but were not found to represent lymph nodes at histology.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although mammography is recognized as the best method of screening for breast cancer, breast sonography has become well established as a valuable imaging technique. Although there has been some controversy regarding the utility of sonography when evaluating solid breast masses for the likelihood of malignancy [12, 13], several studies have suggested that sonographic appearance can be useful in differentiating benign from malignant solid breast masses [8-10]. Rahbar et al. [9] found that the features most likely to predict a benign diagnosis in solid masses were round or oval shape, had a circumscribed margin, and had a width-to-anteroposterior ratio greater than 1.4. Features most predictive of malignancy were irregular shape, microlobulated or spiculated margin, and width-to-anteroposterior ratio of less than or equal to 1.4.
Stavros et al. [8] developed a classification scheme for solid breast masses that has a 98.4% sensitivity and 99.5% NPV for malignancy. Of particular note, Stavros et al. specifically described and gave pictorial examples of lesions illustrating each sonographic descriptor in the lexicon used to build the classification system, thus enabling the continued use of these descriptors in clinical practice. Despite the reported high sensitivity and NPV of the lexicon popularized by Stavros et al. [8], Baker et al. [14] found that interobserver agreement was at best moderate for six of seven sonographic features. This lack of consistency suggested that a more global standardized terminology was necessary for general clinical use.
A standardized lexicon for sonography was developed in 2003 by the ACR in light of the increasing use of sonography in clinical practice. Like its mammographic counterpart, the sonographic BI-RADS lexicon was intended to provide a unified language for sonographic reporting and research and to avoid ambiguity in the communication and teaching of sonographic interpretation [3-5]. The study presented here was undertaken to explore the PPV and NPV values of these new sonographic BI-RADS features.
The sonographic BI-RADS descriptors of spiculated margin, irregular shape, and nonparallel orientation showed high predictive value for malignancy in this study, whereas circumscribed margin, oval shape, and parallel orientation were highly predictive of a benign diagnosis. Similar results were found whether features were assessed for palpable or nonpalpable masses. This stratification can be understood in view of the abnormal processes that these descriptors represent [4]. As in mammography, sonographic evidence of spiculated margin suggests infiltrating growth of the lesion into the surrounding tissue, whereas an irregular shape can indicate inconsistent growth and advancement of the lesion edge. Nonparallel orientation on sonography can suggest spread of the lesion through tissue-plane boundaries. All of these characteristics are more likely to be associated with malignant lesions. In contrast, circumscribed margin and oval shape representing smooth uniform growth without involvement of surrounding tissue are associated more with a benign lesion. Similarly, parallel orientation suggesting containment in one tissue plane is more indicative of a benign process.
Sixteen (9%) of 172 masses described as both oval and circumscribed on sonography were malignant at histology. This relatively high malignancy rate in the setting of apparently benign sonographic features is likely due to concomitant mammographic features that appeared more worrisome and prompted biopsy, including interval enlargement of a lesion compared with prior mammographic findings. In addition, the benign sonographic features described by Stavros et al. [8] had a slightly lower predictive value in our study than was reported in that article. This is likely due to the fact that the observing radiologists in our study were interpreting static images, whereas real-time evaluation might have provided more information, leading to the choice of a more suspicious descriptor. Whereas static image interpretation is not typical of our routine practice, viewing static images represents a common method for interpreting clinical breast sonograms.
Identification of surrounding tissue effects such as edema, architectural distortion, or changes to the Cooper's ligaments was infrequent. However, identification of surrounding tissue effects had a high predictive value for malignancy, suggesting that recognition of such features could be helpful in the final assessment of a sonogram. In addition, the observers in this study occasionally encountered some difficulty in describing a sonographic feature with the terms offered by the sonographic BI-RADS lexicon. For example, in describing the lesion boundary, the radiologist may have believed that neither of the two available options—abrupt interface or echogenic halo—accurately reflected what he or she observed. These experiences may reflect some unfamiliarity with the relatively new sonographic BI-RADS lexicon, and further instruction and experience in the usage of the BI-RADS descriptors may be warranted.
Each case in this study was interpreted by one of three radiologists, all of whom were dedicated breast radiologists working at the same academic institution. No case was assessed by more than one observer. To our knowledge, no studies have yet been published showing the degree of intraobserver or interobserver variability with the sonographic BI-RADS criteria. Such studies would be desirable for a more complete assessment of the utility of the sonographic BI-RADS lexicon.
This study was limited to solid masses. In particular, for lesions such as simple cysts, which are often clearly benign on sonography, the NPV of benign descriptors (e.g., circumscribed margin) is likely to be much higher.
This study has three additional weaknesses. First, biopsy had already been performed for all patients included in this study. Therefore, although they were blinded to biopsy results, the observers were aware that their descriptions and assessments did not directly affect patient care, which could theoretically have affected their choices. Second, when evaluating the sonogram, the radiologist was not blinded to the mammogram and indeed was asked to analyze the mammogram immediately before analyzing the sonogram. Although the mammographic appearance of the lesion could influence the radiologist's assessment of the sonographic image, mammograms and sonograms are routinely interpreted in tandem in actual clinical practice. This design was therefore intended to combat the artificial quality imposed by the retrospective nature of this study and is in keeping with the practice recommendations of tandem interpretation as described in the ACR BI-RADS manual [4]. Third, only biopsy-proven lesions were included in this study. Thus, this study did not address the predictive value of BI-RADS features in more benign-appearing lesions that were interpreted as definitely benign or were recommended for follow-up only. During the time covered by this study, however, follow-up assessments were rarely used in clinical practice at this institution: that is, any solid mass that was not determined to be a benign finding (i.e., BI-RADS category 2) virtually always underwent biopsy. Therefore, few such cases would be excluded from the study on this basis. Regardless, it would be useful for future studies to have a broader patient population for analyses, incorporating lesions with long-term follow-up results in addition to those with histologic diagnoses.
In conclusion, the ACR BI-RADS lexicon provides standardized terminology to facilitate accurate and consistent breast sonography and mammography reporting. This study shows that features from the standardized sonographic lexicon can be helpful in distinguishing benign from malignant solid masses. Further evaluation of the sonographic BI-RADS lexicon would be useful, including studies using prospective real-time analyses of sonograms and interobserver variability studies of the descriptors. Such studies would help further to assess the utility of the sonographic BI-RADS lexicon in clinical breast radiology.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  S. J. Kim, E. Y. Ko, J. H. Shin, S. S. Kang, S. H. Mun, B.-K. Han, and E. Y. Cho Application of Sonographic BI-RADS to Synchronous Breast Nodules Detected in Patients with Breast Cancer Am. J. Roentgenol., September 1, 2008; 191(3): 653 - 658. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E.-K. Kim, K. H. Ko, K. K. Oh, J. Y. Kwak, J. K. You, M. J. Kim, and B.-W. Park Clinical Application of the BI-RADS Final Assessment to Breast Sonography in Conjunction with Mammography Am. J. Roentgenol., May 1, 2008; 190(5): 1209 - 1215. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Wang, D. M. Ikeda, B. Narasimhan, T. A. Longacre, R. J. Bleicher, S. Pal, R. J. Jackman, and S. S. Jeffrey Estrogen Receptor-Negative Invasive Breast Cancer: Imaging Features of Tumors with and without Human Epidermal Growth Factor Receptor Type 2 Overexpression Radiology, February 1, 2008; 246(2): 367 - 375. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. Jesneck, J. Y. Lo, and J. A. Baker Breast Mass Lesions: Computer-aided Diagnosis Models with Mammographic and Sonographic Descriptors Radiology, August 1, 2007; 244(2): 390 - 398. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
O. Graf, T. H. Helbich, G. Hopf, C. Graf, and E. A. Sickles Probably Benign Breast Masses at US: Is Follow-up an Acceptable Alternative to Biopsy? Radiology, July 1, 2007; 244(1): 87 - 93. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. Mesurolle, L. Kadoch, M. El-Khoury, A. Lisbona, N. Dendukuri, and W. D. Foulkes Sonographic Features of Breast Carcinoma Presenting as Masses in BRCA Gene Mutation Carriers J. Ultrasound Med., June 1, 2007; 26(6): 817 - 824. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
Y.-W. Chang, K. H. Kwon, D. E. Goo, D. L. Choi, H. K. Lee, and S. B. Yang Sonographic Differentiation of Benign and Malignant Cystic Lesions of the Breast J. Ultrasound Med., January 1, 2007; 26(1): 47 - 53. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. G. ODLE Breast Ultrasound Radiol. Technol., January 1, 2007; 78(3): 222M - 242M. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Costantini, P. Belli, R. Lombardi, G. Franceschini, A. Mule, and L. Bonomo Characterization of solid breast masses: use of the sonographic breast imaging reporting and data system lexicon. J. Ultrasound Med., May 1, 2006; 25(5): 649 - 659. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. Rhiem, U. Flucke, and R. K. Schmutzler BRCA1-Associated Breast Carcinomas Frequently Present with Benign Sonographic Features. Am. J. Roentgenol., May 1, 2006; 186(5): E11 - E12. [Full Text] [PDF]
|
|
|
|
|
|
J. A. Baker, A. Hong, E. L. Rosen, and M. S. Soo Reply. Am. J. Roentgenol., May 1, 2006; 186(5): E12 - E13. [Full Text] [PDF]
|
|
|
|
|
|
W. A. Berg Sonographically Depicted Breast Clustered Microcysts: Is Follow-Up Appropriate? Am. J. Roentgenol., October 1, 2005; 185(4): 952 - 959. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
