| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Department of Radiology, Research Institute of Radiological Science, Yonsei
University College of Medicine, 134, Shinchon-dong, Seodaemun-gu, Seoul
120-752, South Korea.
2 Department of Surgery, Yonsei University College of Medicine, Seoul, South
Korea.
Received October 4, 2007;
accepted after revision November 29, 2007.
Address correspondence to E. K. Kim
(ekkim{at}yuhs.ac).
Abstract
 Top Abstract IntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
SUBJECTS AND METHODS. We prospectively classified 4,668 breast sonograms according to BI-RADS final assessment category. Suspicious sonographic findings were divided into major and minor suspicious findings. Category 1 was normal and category 2 was a benign finding such as cyst or nodule with uniform and intense hyperechogenicity. A nodule neither category 2 nor category 4 or 5 was defined as category 3. A nodule with one or more suspicious findings, not category 5, was defined as category 4. A nodule with two or more major suspicious findings was defined as category 5.
RESULTS. Of the 4,668 cases, 321 cases failed to undergo follow-up of at least 1 year. The PPV was 0.1% in category 1 (3/2,191), 0% in category 2 (0/773), 0.8% in category 3 (6/737), 31.1% in category 4 (161/519), and 96.9% in category 5 (123/127). In palpable lesions (n = 751), the PPV was 2.2% in category 1 (2/93), 0.9% in category 3 (2/217), 54% in category 4 (107/198), and 98% in category 5 (98/100). In nonpalpable lesions (n = 3,596), the PPV was 0.05% in category 1 (1/2,098), 0.8% in category 3 (4/520), 16.8% in category 4 (54/321), and 92.6% in category 5 (25/27).
CONCLUSION. As with mammography, placing sonographic lesions into BI-RADS categories is useful for predicting the presence of malignancy.
Keywords: BI-RADS • breast cancer • mammography • oncologic imaging • sonography • women's imaging
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Since 2000, all sonography examinations performed in our institution have been categorized into five final assessments. There have been several reports of the positive predictive value (PPV) of BI-RADS final assessment for mammography [16–21], but to our knowledge, few studies have been published about the results of BI-RADS final assessment for sonography [22, 23].
The objective of this study was to report our classification of sonographic findings according to BI-RADS and calculate the PPV for each BI-RADS assessment category to determine whether the BI-RADS assessment category for sonography is a good predictor of malignancy.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
The indications for breast sonography examinations were palpable abnormalities in 759 women (16.3%), follow-up of a sonographic lesion in 589 (12.6%), nonpalpable mammographic abnormalities in 482 (10.3%), nipple discharge in 135 (2.9%), and screening sonography in mammographically dense breast (BI-RADS density categories 3 and 4) in 2,703 (57.9%). Thus, 1,965 (42.1%) were diagnostic sonograms, and 2,703 (57.9%) were screening sonograms.
Breast density in mammography was graded according to the following BI-RADS density categories [24]: type 1, breasts composed almost entirely of fat; type 2, scattered fibroglandular densities that could obscure a lesion on a mammogram; type 3, breast tissue is heterogeneously dense, which may lower the sensitivity of mammography; or type 4, extremely dense breasts, which lowers the sensitivity of mammography. Mammographic ab normalities included masses, focal asymmetry, architectural distortion, and microcalcifications with possible mass.
Imaging Methods
All sonography examinations were performed by one of three dedicated breast
imaging radiologists with 4–8 years of experience, and the examiner knew
the results of the clinical examination and mammography when performed at the
time of sonography. Sonography was performed using equipment (HDI 3000 and
5000, Philips-ATL) and electronically focused near-field probes with a
bandwidth of 5-10–MHz or 5-12–MHz, respectively. Compounding
imaging was used in all exam inations performed with the HDI 5000 machine.
Color or power Doppler images were not routinely obtained. All sonography
examinations included bilateral whole-breast imaging and investigation of any
abnormality found on clinical or mammo graphic examination.
A total of 4,288 mammographic examinations were performed before (n = 4,087) or after (n = 201) sonography. Of these, 3,390 women had mammograms obtained with dedicated equipment (DMR, GE Healthcare) in our hospital. Standard craniocaudal and mediolateral oblique views were routinely obtained, and additional mammographic views were obtained as needed. The remaining 898 mammograms were obtained at an outside institution. Mammograms were assessed by one of three experienced radiologists. Mammography was not primarily performed for palpable masses in cases of young age (< 35 years) or pregnancy.
BI-RADS Classification
Categorization of the sonography examination into categories BI-RADS
1–5 was performed by one of the three radiologists who performed the
sonography examinations. All interpretations were made in conjunction with
mammograms when mammograms were available, and the decision of category was
based on the highest level of mammographic or sonography findings; thus, in
the case of a lesion that was morphologically probably benign on mammography
but suspicious on sonography, the final assessment was based on the sonography
features.
Suspicious sonographic findings were based on our experiences and the report of Stavros et al. [1]. We tried to further divide suspicious findings into major and minor suspicious findings to distinguish category 4 and 5 lesions. Major suspicious findings were irregular shape, spiculated margin, and microcalcifications. Minor suspicious findings included microlobulated or angular margin, nonparallel orientation, duct extension, complex echogenicity, and posterior shadowing (Table 1).
|
The final assessment category made for BI-RADS on a scale from 1 to 5 is summarized in Table 2 (Figs. 1, 2, 3). Because we excluded patients with known malignancies, category 6 was not used. In addition, we did not use category 0 as a final radiologic conclusion in this study. When more than one mass was found in both breasts, a single final assessment was made based on the mass with the most suspicious features.
|
|
|
|
Biopsy and Follow-Up
We recommended a 1-year mammographic follow-up examination in women with a
category 1 or 2 lesion, 6-month sonography follow-up in women with a category
3 lesion, and biopsy in those with a category 4 or 5 lesion; however, tissue
sampling was also performed when the patient or clinician wanted to confirm
the breast lesion dia gnosis in lesions characterized as category 1, 2, or
3.
Biopsies were performed in 1,057 cases, with 884 (83.6%) undergoing sonography-guided core needle biopsy, 88 (8.3%) under going excisional biopsy, and 85 (8.0%) under going fine-needle aspiration with a 23-gauge needle. The choice of excisional biopsy rather than sonography-guided needle biopsy was based of the preference of the surgeon. Fine-needle aspiration was indicated for probable complicated cysts and symptomatic cysts. Sonography-guided core biopsies of masses were performed with a 14-gauge automated core biopsy needle (SACN Biopsy Needle, Medical Device Technologies) and a spring-loaded biopsy gun (Promac 2.2L, Manan Medical Products). Three hundred sixty-one of 884 core needle biopsies were followed by surgical excision because of the presence of malignancy (n = 270), patient's desire (n = 58), atypical pathologic results (n = 10), imaging–pathologic discordance (n = 13), or increase in size (n = 10). Some lesions originally diagnosed as atypical ductal hyperplasia (ADH) or benign were found to be malignant on further surgical excision; these lesions were recorded in this study by their excisional biopsy result. All patients with breast lesions with a benign biopsy result and imaging–pathologic concordance and patients with lesions without tissue confirmation that were categorized BI-RADS 3 were recom mended to receive follow-up by sonography at 6 months, mammography and sono graphy at 1 year, and one yearly mammography or sonography examination (or both) thereafter. Patients with lesions that were categorized BI-RADS 1 and 2 were recommended to have a follow-up mammography at 1 year.
In cases in which biopsy was not performed, we analyzed the follow-up mammogram or sonography results for at least 12 months as registered in our hospital information system.
Statistical Analysis
For all examinations, we calculated the PPV (the number of breast cancers
divided by the total number of examinations per category x 100) of the
five BI-RADS assessment categories. BI-RADS category 1, 2, and 3 lesions were
regarded as benign and BI-RADS category 4 and 5 lesions as malignant.
False-negative examinations were defined as the pathologically confirmed
malignant lesions in BI-RADS category 1, 2, and 3. Statistical comparisons of
the quantitative data were performed using a Student's t test.
Statistical significance was assigned a p value of less than 0.05.
Statistical analysis was performed with the SAS system (Magree SAS Macro
program, SAS Institute).
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
|
|
|
Among the 3,125 women, 510 women (16.3%) had palpable lesions and 2,615 (83.7%) had nonpalpable lesions. The age of the women with the palpable lesions was significantly younger than that of the women with nonpalpable lesions (40.2 vs 48.5 years, respectively; p < 0.01). There were 209 malignancies in the palpable group: 191 invasive, 15 noninvasive, two metastases, and one lymphoma (mean size, 1.8 cm; SD, 0.9 cm). There were 84 malignancies in the nonpalpable group: 72 invasive and 12 noninvasive (mean size, 0.8 mm; SD, 0.4 mm).
There were nine false-negative results: Five cases (cases 1–5) were symptomatic and four (cases 6–9) were asymptomatic; these are summarized in Table 4. Six were mammographically negative and four were sonographically negative. Six cases were classified as category 3 and three as category 1. Retrospectively, two of the six false-negative cases in category 3 had a noncircumscribed margin sonographically (Fig. 5).
|
|
Table 5 shows the PPV for the five BI-RADS assessment categories for all examinations. In category 1, three lesions were malignant, two of which were palpable. These lesions showed heterogeneous fibroglandular tissues without abnormalities on mammography, and sonography failed to depict any abnormalities at the palpable area. They looked like prominent parenchymal tissue (Fig. 6). One case (case 9) that showed negative findings on both screening mammography and sonography was found to be malignant 8 months later when the patient complained of a palpable mass, and subsequent mammography and sonography show ed an ill-defined mass that was confirmed as T2N1 invasive ductal carcinoma. There were no malignant lesions in category 2. In category 3, six of 737 lesions (0.8%) were malignant. Within category 3, there were two malignancies of 217 palpable lesions. In category 4, 31.1% of lesions were malignant, and 96.9% were malignant in category 5. Four of 127 (3.1%) category 5 lesions turned out to be benign: their pathologic diagnoses were granular cell tumor, fat necrosis, fibrocystic change, and adenosis.
|
|
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
Recently, Hong et al. [31] reported PPV and negative predictive value (NPV) of features described in the sonographic BI-RADS lexicon and concluded that descriptors from the sonographic BI-RADS lexicon could be useful in differentiating benign from malignant solid masses. However, they focused only on the lexicon, not on the final assessment. The final assessment is more important in the management of patients with breast disease, and our study indicated that sonographic final assessment is comparable to screening mammography. Most of the published data on medical audits involve screening mammography [17, 18, 20, 24], and few data concerning diagnostic mammography and sonography are available. Zonderland et al. [21] reported the PPV of BI-RADS for mammography and sonography in 2,762 cases, of which 2,108 (76.3%) were mammographic evaluation alone and 654 (23.7%) were joint mammography and sonography evaluation. Recently, Costantini et al. [22] reported sonography audit results in 178 cases, but these were limited to biopsyproven cases only. The results from our study population containing more than 4,000 cases and categories 1–5 were generally comparable to those of previous studies [16–18, 21, 22]. In the reports of Costantini et al. [22] and Zonderland et al. [21], the PPV of category 3 was 7.7% and 3.9%, respectively, whereas it was 0.8% in our study. In the study by Costantini and colleagues, only lesions that underwent cytologic or histologic examination were analyzed and only a small number of women with probably benign lesions and proved diagnoses (n = 26) were included. In the study by Zonderland and coworkers, the PPV of category 3 was six of 154 (3.9%), and three of six false-negative cases showed microcalcifications on mammography. In contrast to those two studies, our study contained a relatively large number of category 3 lesions (n = 737), and we did not include mammographic microcalcification that results in low false-negative results.
A category 3 lesion is judged to have a 2% or lower probability of malignancy, and several studies have suggested that probably benign lesions can be followed up with short-term mammography [24, 32–34]. In our study, the PPV of category 3 lesions was 0.8% overall: 0.9% for palpable lesions and 0.7% for nonpalpable lesions. Recently, Graf et al. [35] described 112 lesions that were palpable circumscribed solid breast masses according to mammography and sonography and were followed up for at least 2 years with no diagnoses of breast carcinoma. They concluded that palpable noncalcified solid breast masses with benign morphology according to mammography and sonography could be managed similarly to nonpalpable BI-RADS category 3 lesions, with short-term follow-up. In our study, two palpable cancers were miscategorized as category 3, one of which had minor suspicious sonographic findings retrospectively. In the study by Zonderland and colleagues [21], two palpable category 3 lesions that turned out to be malignant showed irregular margins retrospectively. These observations indicate that meticulous sonographic evaluation is necessary. However, an emphasis on the detection of subtle sonographic suspicious features may lead to an increase in the classification of lesions into category 4, resulting in increased numbers of unnecessary biopsies. In our study, we applied strict sonographic criteria to classify category 3 lesions and had low false-negative rates in category 3; however, the PPV of category 4 was 31.1%, lower than in other studies (Table 5).
To the best of our knowledge, there have been few reports on large numbers of breast sonography audit results including category 1 and 2 lesions. Geller et al. [36] reported the use of BI-RADS for mammographic evaluation of women with signs and symptoms of breast disease. They reported inconsistencies between the assessment category and management recommendations, indicating that the reporting system for diagnostic mammograms was unreliable at that time. In our study, BI-RADS was applied to sonography of palpable breast masses and nonpalpable lesions with a comparable diagnostic index (Table 5). However, placing palpable lesions in category 1 requires great caution. In the current study, two of 93 (2.2%) palpable category 1 lesions were proved to be malignancies and their physical examinations were all suspicious. The NPV of sonography with mammography is high enough (typically 95–100%) to exclude breast malignancy in a palpable lump [37, 38]; however, if the physical examination is suspicious, a biopsy should not be delayed. Durfee et al. [39] reported that they performed sonography on clinically palpable breast cancers that were undetectable on mammography and 9% (3/35) were not revealed by sonography either. These lesions were all suspicious at clinical examination. In our study, two cases of palpable breast cancer were not visualized by either mammography or sonography (n = 209), and three cases were interpreted as probably benign. Thus, the false-negative interpretation rate in palpable breast cancer was 1.9% (4/209).
Differentiation between category 4 and 5 may seem to be unnecessary because all of these lesions require biopsy; however, it is important because the proper designation of category 5 lesions leads to imaging–pathologic discordance if the core biopsy result is benign. Therefore, we tried to discriminate category 4 and 5 lesions according to our sonographic interpretation criteria, and the PPVs of category 4 and 5 lesions in our study were 31.1% and 96.9%, respectively. Four cases of 127 category 5 lesions were proved to be benign, so our results indicate that preoperative histopathologic diagnosis is necessary before definite treatment in category 5, although the probability of malignancy was very high [40]. ACR BI-RADS encourages the subdivision of final assessment category 4 into subcategory 4a, 4b, or 4c to communicate the level of suspicion to referring physicians and patients [15], and recently Lazarus et al. [23] reported that the PPVs of such subcategorization (6% for 4a, 15% for 4b, and 53% for 4c) were good enough to predict the likelihood of malignancy, although their study was limited by sample size. Further studies with a larger sample size are required.
There are limitations to our study. First, we have used mammography when interpreting sonography, so our interpretation of sonography was probably biased by mammography in the cases in which the level of suspicion was higher in mammography. It could be one of the reasons that explain the differences from the results of other studies. However, practically, this influence does actually occur when breast sonography is performed and interpreted. Second, we followed mammographic or sonographic stability for at least 1 year in cases that did not undergo biopsy. Because some cancers may progress slowly after this time, such cancers may have been missed in this study. However, the follow-up of negative examinations (categories 1, 2, and 3 lesions) in large data sets is difficult and the diagnostic index would be minimally affected if some carcinomas developed later. Third, sonography is an operator-dependent examination technique. Although the three radiologists performing and interpreting breast sonography in this study were experienced breast imagers, interobserver variability in breast sonography remains a concern [23, 41, 42]. Finally, Berg et al. [43] recently reported that lesion detection and characterization can be problematic in physician-performed whole-breast sonography. In particular, fewer than 50% of lesions 5 mm or smaller were detected, and detection of lesions 5.1–11 mm in mean diameter was variable. Considering that 57.9% of our study population received breast sonography to screen mammographically dense breasts, long-term follow-up may be required in such cases.
In conclusion, our study results indicate that the clinical application of BI-RADS final assessment to breast sonography was successful as a predictor of malignancy. Proper classification of BI-RADS final assessment will help referring physicians, radiologists, and patients to understand their management options and implications.
|
TopAbstractIntroductionSubjects and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  J. H. Shin, B.-K. Han, E. Y. Ko, Y. H. Choe, and S.-J. Nam Probably Benign Breast Masses Diagnosed by Sonography: Is There a Difference in the Cancer Rate According to Palpability? Am. J. Roentgenol., April 1, 2009; 192(4): W187 - W191. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
