| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Breast Imaging |
1 From the Departments of Clinical Radiology (A.K.K., M.S., M.H.B., R.S.V.), Pathology (V.J.K.), Oncology (V.K.), and Surgery (P.K.M.), Kuopio University Hospital and Kuopio University, Puijonlaaksontie 2, FIN-70210 Kuopio, Finland. Received April 30, 2004; revision requested July 14; revision received October 13; accepted November 15. Supported by Kuopio University Hospital (EVO funding 406023). Address correspondence to A.K.K. (e-mail: Anna.Koskela{at}kuh.fi).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
MATERIALS AND METHODS: The study was approved by the ethics committee of the hospital; informed consent was obtained. Biopsy was performed in 197 patients with 205 lesions (97 masses, 108 microcalcifications). The first sample (from the center) was collected in container A; second and third samples (2 mm from center), in container B; and additional samples, in container C. Malignancies, atypical ductal hyperplasia (ADH), and radial scars were excised. Benign lesions were followed up mammographically (mean, 24 months). Strict sensitivity and working sensitivity were calculated separately. Stereotactic biopsy with diagnosis of a nonmalignant lesion that, after surgery, proved to be malignant was considered false-negative when strict sensitivity was calculated. Stereotactic biopsy with diagnosis of ADH or radial scar was considered true-positive if the findings at surgery corresponded to the results at biopsy or indicated malignancy and was considered false-positive if the findings at surgery were benign when working sensitivity was calculated. Sensitivity, specificity, and overall accuracy of stereotactic biopsy were determined for masses and microcalcifications in all three containers by using surgical samples and findings at mammographic follow-up as reference. At 
2 analysis, P < .05 was considered to indicate significant difference.
RESULTS: Strict sensitivity of the first sample was 77% (66 of 86) (90% [35 of 39] for masses, 66% [31 of 47] for microcalcifications). Results of the first sample were false-negative significantly more often in microcalcifications (n = 16) than in masses (n = 4) (P = .010). Combined results of containers A and B (ie, three samples) yielded higher strict sensitivity than those with first sample alone (95% [37 of 39] for masses [P = .196], 91% [43 of 47] for microcalcifications [P < .001]). With multiple samples, strict and working sensitivity were both 100% (39 of 39) for masses and 91% (43 of 47) and 98% (46 of 47), respectively, for microcalcifications. Four false-negative diagnoses (ADH, three cases; lesion with discordant mammographic and stereotactic biopsy findings, one case) were microcalcifications.
CONCLUSION: More than three samples are needed (a minimum number was not determined) for a histologic diagnosis of a mass lesion by using an add-on stereotactic biopsy device.
© RSNA, 2005
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
Investigators in previous studies attempted to define the minimum number of cores required for a reliable histologic diagnosis (3–5). It seems that masses are accurately diagnosed with five cores (3,4), but an additional number of samples may be necessary in the investigation of clusters of microcalcifications (3–5). All these studies were performed by using a dedicated prone biopsy table. Such a table offers several advantages. With the patient in a prone position, vasovagal reactions (6) and patient motion (2) are eliminated. A dedicated biopsy table, however, is expensive and occupies much space.
An add-on unit with the patient in a sitting position and conventional mammographic equipment costs less, uses less space and, furthermore, can be used for mammography in addition to breast biopsy. It may, therefore, be the first choice for most of the smaller centers where breast cancer is diagnosed. As far as we know, there are no published studies in which the number of cores required for reliable histopathlogic diagnosis with an add-on unit is described.
The aim of this study was to prospectively determine whether there is a minimum number of cores required for histopathologic diagnosis of mammographically detected nonpalpable breast lesions with an add-on 14-gauge stereotactic core-needle biopsy device.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
The biopsy material obtained by using an add-on stereotactic biopsy device was collected as follows: The first sample was collected in container A; the second and the third, in container B; and all additional samples, in container C. A histologic evaluation and report were performed for each container separately. After core-needle biopsy, all the women either underwent surgical excision or, if the diagnosis from stereotactic core-needle biopsy was benign, were followed up with mammography. The study was approved by the ethics committee of the hospital. Informed consent was obtained.
Patients
Of the 212 consecutive patients with 220 nonpalpable, mammographically detected breast lesions who were eligible for this study, 15 patients were excluded. In seven of the excluded patients, the location of the lesion was high, near the axillary fossa, or so close to the thoracic wall and pectoral muscle that it could not be reached by using the stereotactic equipment. Such lesions were excised surgically. In two (1%) patients, the biopsy had to be terminated after the first pass because of a vasovagal reaction. In five patients, the samples had all been placed in one container instead of in three containers, so separate analysis could not be performed. In addition, one patient died from unrelated causes before any follow-up examination. Thus, 197 patients (mean age, 56 years; range, 32–88 years) with 205 lesions were included in this study. Eight patients had two lesions: Seven patients had two lesions in the same breast, and one patient had a lesion in each breast.
Mammographic Findings
A complete mammographic imaging work-up, which included magnification views, was performed for all lesions. Comparison with prior mammograms, if available, was performed. For 24 (12%) patients, this was the first mammographic examination; no mammograms existed for comparison. Mammographic findings were categorized either as masses, with or without microcalcifications, or as clusters of microcalcifications. The size of the lesion was determined as the greatest diameter of the lesion measured directly on the mammogram and was reported in millimeters. The classification of the mammographic findings was independently performed by two radiologists (M.S. and M.H.B., with 8 and 9 years of experience in breast imaging, respectively) who were blinded to the histologic results. All lesions were retrospectively categorized according to the standardized Breast Imaging Reporting and Data System (BI-RADS) recommended by the American College of Radiology (7), which is not in routine use in our country. The BI-RADS category was assigned as follows: category 2 for lesions classified as benign, category 3 for lesions classified as probably benign, category 4 for lesions classified as a suspicious abnormality, and category 5 for lesions classified as highly suggestive of malignancy. In cases of discrepancy, a consensus reading was conducted. The number and size of malignant lesions in each BI-RADS category were reported.
Stereotactic Biopsy
All biopsies were performed by using a regular mammographic machine (Sophie; Planmed, Helsinki, Finland) and an add-on stereotactic biopsy device (Cytoguide; Planmed) with the patient in an upright seated position. An automated biopsy gun (Biopty; Bard, Covington, Ga), with a 23-mm excursion and a 14-gauge needle, was used.
The biopsy procedures were performed by one of five radiologists (including M.S., M.H.B.), all with 4–6 years of experience in performance of breast biopsy. For mass lesions, after the lesion was localized, the first needle pass was targeted to the center of the lesion. Routine prefire stereotactic views were obtained to confirm the position of the needle. Needle-tip location was modified, if needed, to ensure the central position. The following three to four needle passes were performed in a clockwise direction at 2 mm from the center of the lesion. Additional needle passes were performed after individual consideration, depending on the lesion distribution. For microcalcifications, there was more variability in needle placement. The intent was to target the most suspicious area for core-needle biopsy. If the calcifications were tightly clustered, the first pass was performed either by targeting the center of the lesion or by selecting a particularly distinctive calcification that could be reliably discerned on the two stereotactic images. Subsequent passes were planned according to the geography of the calcifications. A minimum of four biopsies (range, 4–15; mean, 7) were performed in each patient. The overall biopsy procedure was accomplished, as previously described (8). Infections and hematomas that required further treatment were not noted in our patient population.
Specimens removed at core-needle biopsy were collected separately in marked containers: the first sample (obtained from the center of the lesion) was collected in container A; the second and the third samples (obtained 2 mm from the center of the lesion), in container B; and all additional samples, in container C.
For lesions evident as microcalcifications without a mass, the presence of microcalcifications in the biopsy material was confirmed with findings from radiography of the cores or with those from the histopathologic analysis report. Samples were considered sufficient (a) for microcalcifications if microcalcifications were detected either at radiography of the cores or at histopathologic analysis of clusters of microcalcifications and (b) for mass lesions if a histopathologically determined diagnosis was in concordance with the mammographic appearance.
Histopathologic Analysis of Core-Needle Biopsy Specimens
One of five pathologists (including V.J.K.), with 7–20 years (mean, 12 years) of experience with breast biopsy analysis, analyzed the contents of each container separately for the presence of (a) invasive carcinoma, (b) ductal carcinoma in situ (DCIS), (c) radial scar, (d) atypical ductal hyperplasia (ADH), or (e) a specific benign diagnosis. The pathologist, however, was not blinded to the study design.
Treatment and Follow-up
In accordance with our clinical practice, the histopathologic findings in each lesion were reviewed in conjunction with the mammographic findings and clinical history in a multidisciplinary meeting to plan further treatment of the patient. A radiologist, a pathologist, a plastic surgeon, an oncologist, and a surgical breast nurse who acted as coordinator, all of whom had wide experience in breast disease, met on a weekly basis.
All women whose biopsy results indicated the presence of invasive carcinoma or DCIS underwent surgical treatment. Those with high-risk lesions, such as ADH and radial scars, also underwent surgical resection (9–12), as did women who had benign lesions with discordant mammographic findings. Mammographic surveillance was planned to last a minimum of 2 years (13–15), and mammographic follow-up at 1 year and 2 years was recommended for women who had all other lesions with benign biopsy results.
Statistical Analysis
Calculations were performed separately for mass lesions and microcalcifications in all three containers. The sensitivity and specificity values with 95% confidence intervals and the overall accuracy of stereotactic core-needle biopsy were determined by using the results at analysis of surgical samples and at mammographic follow-up as the reference standard.
Sensitivity was separately calculated in two categories, strict analysis and working analysis. For strict analysis, strict sensitivity was defined such that any diagnosis of a nonmalignant lesion at stereotactic core-needle biopsy that proved to be malignant after surgical excision was considered false-negative. For working analysis, working sensitivity was defined such that a diagnosis of ADH or radial scar (high-risk lesions) at stereotactic core-needle biopsy was considered true-positive if the findings in the final surgical sample corresponded to the results at biopsy or indicated malignancy (DCIS or invasive carcinoma), and the diagnosis was considered false-positive if the pathology report at surgical excision indicated benign results (16).
The strict false-negative rate was determined by dividing the number of lesions that were originally considered benign at core-needle biopsy (false-negative results at core-needle biopsy) by the total number of surgically confirmed malignancies. The working false-negative rate was determined by dividing the number of lesions that were originally considered benign (high-risk lesions were calculated as true-positive lesions if the final surgical sample corresponded to the results at biopsy or indicated malignancy) by the number of surgically confirmed malignancies and high-risk lesions.
Malignancy rate was determined by dividing the number of malignancies in each BI-RADS category by the number of lesions in each category. Calculations were performed separately for mass lesions and for microcalcifications.
The numbers of underestimated ADH and DCIS lesions were separately calculated for both masses and lesions with microcalcifications. Underestimated ADH lesions were defined as lesions that yielded ADH at core-needle biopsy and carcinoma at surgery (17). Underestimated DCIS lesions were defined as lesions that yielded DCIS at core-needle biopsy and invasive carcinoma at surgery (17).
The 2 analysis was applied for group comparisons of dichotomized discrete variables. Differences were considered statistically significant if the P value was less than .05. The 95% confidence intervals are approximations of the confidence intervals (preferred when percentages are near 100 or zero) that are determined on the basis of formulas given in Fleiss et al (18). All other analyses were performed with statistical software (SPSS for Windows, version 11.5; SPSS, Chicago, Ill).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
BI-RADS Classification
According to the BI-RADS classification, there were two (1.0%) category 2 lesions, 35 (17.0%) category 3 lesions, 121 (59.0%) category 4 lesions, and 47 (22.9%) category 5 lesions. Altogether, 149 (72.7%) of the lesions were found on screening mammograms. Forty-nine (23.9%) lesions were found on mammograms obtained because of a symptom (lump, pain, or eczema) somewhere in the breasts), and seven (3.4%) lesions, on mammograms obtained because of a history of hormone replacement therapy.
Specimens
The total number of core samples obtained was 1449. In the specimens of clusters of microcalcifications (108 lesions), histopathologic evidence of calcium was detected in 57 (52.8%) lesions in container A, in 81 (75.0%) lesions in containers A and B combined, and in 95 (88.0%) lesions in at least one of the three containers.
Radiography was performed in specimens taken from 48 (44.4%) lesions that were evident as clusters of microcalcifications. Results revealed calcifications in at least one specimen in 46 (96%) of 48 lesions. In four lesions with calcifications seen on the radiographs of specimens, there was no calcium seen at histopathologic analysis.
In 13 clusters of microcalcifications (12 patients), there was no calcium detected at histopathologic analysis. In four of these clusters (three cases of DCIS and one case of fibrocystic disease), microcalcifications were seen on radiographs of the specimen.
In nine clusters of microcalcifications (eight patients), there were no calcifications seen either on radiographs of the specimen or at histopathologic analysis. In three of these lesions (two patients), the histologic finding at core-needle biopsy was malignancy (DCIS or invasive ductal carcinoma). In addition, surgical excision revealed fibrosis with microcalcifications in two lesions. In the remaining four clusters of microcalcifications, the diagnosis at core-needle biopsy was fibrosis in two lesions (BI-RADS category 3), liponecrosis in one lesion (BI-RADS category 3), and adipose tissue and hematoma in one lesion (BI-RADS category 4). These lesions were not surgically excised. On the follow-up mammograms, the lesions remained unchanged. In one mass lesion (BI-RADS category 3), the findings at core-needle biopsy led to a nondiagnostic histologic result. The patient with this lesion was followed up with mammography, with no change detected during 28 months.
Strict and Working Analyses
The overall strict and working analyses for the three containers are shown in Tables 1 and 2. As described earlier, sensitivity was separately calculated in two categories (strict and working sensitivities). The strict sensitivity of the first sample was 77% (66 of 86) (90% [35 of 39] for masses and 66% [31 of 47] for microcalcifications). The results with the first sample were false-negative significantly more often for microcalcifications (n = 16) than they were for mass lesions (n = 4), with P = .010. Combined results with containers A and B, that is, three samples, yielded higher strict sensitivity than did results with the first sample alone: The strict sensitivity increased to 95% (37 of 39) for mass lesions (P = .196) and to 91% (43 of 47) for microcalcifications (P < .001). For multiple samples, a strict sensitivity of 100% (39 of 39) for mass lesions was reached; however, our data do not permit determination of the number of samples that are needed beyond three. For microcalcifications, multiple samples did not improve the strict sensitivity (91% [43 of 47]).
|
|
The working specificity of container A, containers A and B combined, and all three containers together was 98% (117 of 119) (97% [56 of 58] for mass lesions and 100% [61 of 61] for microcalcifications). The working overall accuracy of container A was 90% (185 of 205) (94% [91 of 97] for mass lesions and 87% [94 of 108] for microcalcifications); of containers A and B, 97% (198 of 205) (96% [93 of 97] for mass lesions and 97% [105 of 108] for microcalcifications); and of containers A, B, and C, 99% (202 of 205) (98% [95 of 97] for mass lesions and 99% [107 of 108] for microcalcifications).
Histopathologic Analysis of Stereotactic Core-Needle Biopsy Specimens
Histopathologic analysis of diagnostic specimens removed at core-needle biopsy revealed carcinoma in 83 (40.5%) of the 205 lesions, ADH lesions in three (1.5%), a radial scar in one (0.5%), and a specific benign diagnosis in 113 (55.1%). Altogether, 54 (26.3%) invasive carcinomas, 32 (15.6%) cases of DCIS, and two (1.0%) radial scars were found in the surgical specimens, so at surgery, malignancy was confirmed in all but one lesion with a malignant result at core-needle biopsy. In this particular case, the main diagnosis was a radial scar at core-needle biopsy, but there was a mild suspicion of tubular carcinoma. The final surgical diagnosis was a radial scar. The results of a repeated retrospective analysis of the core specimens from this lesion also indicated a radial scar; the pathologists who performed the repeated analysis were blinded to the results of the previous histologic analysis of the core specimens. Among the 33 lesions (diameter range, 4–14 mm; mean, 14 mm) for which a case of DCIS was indicated at core-needle biopsy, findings at surgery revealed invasive ductal carcinoma in seven (21%) and confirmed cases of DCIS in 26 (79%). The number of malignancies and the malignancy rate according to the mammographic findings are shown in Table 3.
|
In the 108 lesions evident as a cluster of microcalcifications without a mass, stereotactic core-needle biopsy revealed cases of invasive carcinoma in 12 (11.1%) lesions, cases of DCIS in 31 (28.7%) lesions, cases of ADH in three (2.8%) lesions, and a benign diagnosis in 58 (53.7%) lesions. Results of histopathologic analysis of four (3.7%) lesions remained nondiagnostic. The mammographic appearance of the specimens from malignancies obtained at core-needle biopsy was classified as BI-RADS category 4 in 24 (56%) of 43 lesions and BI-RADS category 5 in 19 (44%) of 43 lesions. All three cases with a diagnosis of ADH at core-needle biopsy proved to be malignant (two cases of DCIS and one case of invasive carcinoma) at surgery (strict false-negative cases) (Fig 1). In the fourth case with a strict false-negative diagnosis, surgery was performed on the lesion because of a discordance between the mammographic finding and the histopathologic diagnosis of fibrocystic disease at core-needle biopsy. Invasive lobular carcinoma was found. This was the single working false-negative case in the data (Fig 2 ). The strict false-negative rate was 5% (four of 86), and the working false-negative rate was 1% (one of 88). Among the 108 clusters of microcalcifications, results at stereotactic core-needle biopsy led to underestimation for three (100%) of three cases of ADH and seven (23%) of 31 cases of DCIS.
|
|
|
|
The remaining 58 clusters of microcalcifications and 53 mass lesions (105 patients) that were considered benign at core-needle biopsy (including five lesions with nondiagnostic histopathologic findings at core-needle biopsy) were thus followed up with clinical examination and mammography. Follow-up was performed for all (100%) of these patients. The mean mammographic follow-up time was 24 months (range, 6–39 months). Of the patients who were followed up, 65.7% (69 of 105) were followed up for 24 months and 90.5% (95 of 105) were followed up for more than 20 months. No patients were unavailable for follow-up, but one patient died and two patients moved to another district after adjunctive first follow-up mammography at 6 months.
Mammographic change occurred in 2.9% (three of 105) of lesions. One mass lesion with a diagnosis of fibroadenoma at core-needle biopsy had grown when the 1-year mammogram was obtained. The diagnosis of fibroadenoma was confirmed at surgery. One cluster of microcalcifications was noted to have grown at 1-year follow-up mammography, and one mass lesion, at 2-year follow-up mammography. At repeat biopsy, the initial diagnosis of fibrocystic disease was confirmed in both cases. No change was noted in the remaining 108 lesions (102 patients).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
In the present study, only three (1.5%) of 205 biopsy procedures were terminated because of a vasovagal reaction. Two of the patients involved were excluded from the study because the procedure had to be terminated after only one pass, but the cores that were removed were diagnostic. The third patient who had a vasovagal reaction underwent biopsy twice, but neither the first biopsy, which was terminated after three passes, nor the second biopsy, which included six passes, was diagnostic. Our results are consistent with those of Caines et al (20), who reported vasovagal attacks in 1.6% of the procedures, and of Wunderbaldinger et al (21), who reported vasovagal collapses in 2% of the procedures that led to termination of the biopsy procedure. Vasovagal reactions can be averted by using premedication with atropine, which, however, was not used in our study.
Prone tables have also been advocated over add-on units because the prone position is thought to minimize patient movement (2). Patient movement may make the biopsy procedure more time consuming and increase the radiation dose. Only one biopsy in our study failed because of patient movement. The repeat biopsy was accomplished successfully.
For the majority of our patients, the reason for biopsy was a suspicious lesion found at screening mammography. This finding explains the relatively small diameter (mean, 11 mm) of the lesions. Despite the small size of the lesions, the sensitivity and false-negative rates acquired with an add-on unit are comparable with those in previous reports with mean lesion diameters of 13–28 mm (3,16,21).
Classification of Mammographic Findings
In the present study, stereotactic core-needle biopsy was performed in all patients who were referred to our hospital for biopsy. Core-needle biopsy was also performed for two lesions classified as BI-RADS category 2 and 35 lesions classified as BI-RADS category 3, with no malignancies detected in them. Our results are similar to those in previous reports with a very low incidence of malignancy (0.5%–4%) detected among lesions categorized as probably benign (13–15).
Number of Cores
A recommendation of five cores for masses and a minimum of five cores for microcalcifications has been determined for biopsy of nonpalpable breast lesions with dedicated prone equipment (3,4). A minimum of four specimens should be obtained with US-guided core-needle breast biopsy and a 14-gauge needle (22). Our experience with an add-on device is similar. Strict accuracy after one sample was removed was 90% (184 of 205); after three samples were removed, 97% (198 of 205); and after more than three (multiple) samples were removed, 98% (200 of 205). The strict sensitivity of multiple samples reached 100% (39 of 39) for masses; however, our results do not provide data to indicate how many more than three samples are sufficient. For microcalcifications, use of multiple samples did not contribute to improvement in the strict sensitivity (91% [43 of 47]), but an acceptable working sensitivity of 98% (46 of 47) was reached. The 95% confidence intervals shown in Tables 1 and 2 precluded defining the optimum or minimum number of samples needed, even for mass lesions.
In the present study, all false-negative cases, as well as underestimation of cases of DCIS, occurred at the stereotactic core-needle biopsy of microcalcifications. For masses, it was not possible to evaluate the underestimation rate for cases of DCIS (only two lesions) or that of ADH (no lesions), which is one limitation of our study. It is well known that, in clusters of microcalcifications, the diagnosis of ADH, as well as of DCIS, at core-needle biopsy may lead to underestimation of the presence of invasive carcinoma, and the lesions should be surgically removed. With this practice, the removal of multiple samples helped to reach a high working accuracy also for microcalcifications.
Seven of 33 cases of DCIS diagnosed by using stereotactic core-needle biopsy proved to be invasive ductal carcinoma at surgery. The underestimation rate of 21% (seven of 33) is comparable to results in previous investigations (23–25). Results at core-needle biopsy caused underestimation of the presence of carcinoma in all (three of three) of the ADH lesions in the present study. In a series of 1032 lesions for which a diagnosis was determined by using core-needle biopsy, Meyer et al (27) detected 18 ADH lesions, 10 of which were malignant at surgery. Similar results have been reported by others (9,10,26,27).
Limitations
The findings in five lesions remained nondiagnostic at core-needle biopsy in this study. For one mass, the results of histopathologic analysis of the core were nondiagnostic. For four clusters of microcalcifications, no microcalcifications were detected at radiography of the core or at histopathologic analysis. For one of these clusters of microcalcifications, the results of histopathologic analysis of the core were nondiagnostic. The lesion remained stable at 12- and 20-month follow-up mammography, despite the suspicious mammographic appearance (classification of BI-RADS category 4). On images obtained before and after stereotactic biopsy, these lesions were correctly targeted. One limitation of the present study was that these five lesions were not surgically removed. The decision not to remove them was made in the multidisciplinary meeting and was not in accordance with the original study protocol. One patient died of unrelated causes. A 7-month follow-up study revealed that her mammographic findings were stable. In the other cases, no changes were noted in the follow-up studies at 14 months, 25 months, and 28 months.
Identification of calcifications on radiographs of cores provides proof that the targeted lesion was sampled. The major limitation of this study is that the radiographs of cores were not obtained for every lesion with microcalcifications. The practice of routine radiography of cores was instituted in May 2000 at our hospital. Accordingly, radiographs of cores were available only in 44.4% (48 of 108) of the biopsy specimens, and the findings remained negative in four cases (three cases of DCIS and one of fibrocystic disease). The presence of microcalcifications mainly was confirmed at the histopathologic analysis of the cores. Dahlstrom et al (28) interestingly noted that calcifications of less than 100 µm assessed histologically were not visible at radiography of cores and, thus, may not represent the calcifications seen at mammography. Stomper et al (29) reviewed mammographic and histopathologic features of 27 breast cancers that manifested mammographically as noncalcified masses. In their study, the histopathologic analysis revealed that 41% of these lesions exhibited microscopic calcifications in the tumors or adjacent tissue. Microcalcifications of 50–100 µm are seen histopathologically and microcalcifications of more than 150 µm can be seen at mammography. At histopathologic analysis, the sizes of the calcium particles were not measured in our study.
The pathologists who analyzed the specimens in our study were not blinded to the study design, nor were they required to analyze the specimens in alphabetical order (that is, to record an interpretation for specimen A before looking at specimen B, and then to record an interpretation for specimen B before looking at specimen C). The study design created the opportunity for observer bias in the pathologist, which is a potential limitation of the study.
Recently, stereotactic biopsy has been increasingly performed with devices, such as vacuum-assisted 11-gauge needles, that acquire a larger volume of tissue. Compared with 14-gauge core-needle biopsy, 11-gauge vacuum-assisted biopsy has been shown to be advantageous in retrieval of calcifications (30,31) because of the lower frequency of histologic underestimation (24,32) and the lower repeat biopsy rate (33). Stereotactic biopsies with vacuum-assisted devices, however, are more expensive than core-needle biopsies. Therefore, at least in smaller centers, core-needle biopsies may remain the primary biopsy method for mass lesions (3,4).
False-Negative Results
The strict false-negative rate of 5% (four of 86) for the surgically confirmed malignancies is comparable with results in previous studies with a prone biopsy device (10,19,34,35). The reported strict sensitivity rates according to the numbers of cores acquired with the patient in a prone position or with the patient sitting are shown in Table 4 (1,3–5,19,21,36,37). In the present study, the strict sensitivity rates acquired with an add-on device are similar.
|
In conclusion, an add-on stereotactic core-needle biopsy device is comparable to a prone biopsy table in the diagnostic procedure for biopsy of nonpalpable breast lesions. Three samples were insufficient for an accurate diagnosis of a mass lesion. Multiple samples are needed; however, our study results do not provide data to indicate how many more than three cores are sufficient. With microcalcifications, a high working sensitivity can be reached with multiple samples; however, in microcalcifications, a diagnosis of ADH at core-needle biopsy may cause underestimation of malignancy, and these lesions, as well as lesions with discordant findings, should be surgically excised.
|
FOOTNOTES |
|---|
Abbreviations: ADH = atypical ductal hyperplasia • BI-RADS = Breast Imaging Reporting and Data System • DCIS = ductal carcinoma in situ
Authors stated no financial relationship to disclose.
Author contributions: Guarantors of integrity of entire study, A.K.K., R.S.V.; study concepts and design, M.S., A.K.K., R.S.V.; literature research, A.K.K.; clinical studies, M.S., M.H.B., A.K.K., V.J.K.; data acquisition, A.K.K., V.J.K.; data analysis/interpretation, A.K.K., R.S.V.; statistical analysis, A.K.K., R.S.V.; manuscript preparation, A.K.K.; manuscript definition of intellectual content, R.S.V.; manuscript editing, R.S.V., V.K.; manuscript revision/review, M.S., M.H.B., P.K.M.; manuscript final version approval, all authors
|
References |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. S. Wu, J. D. Goldsmith, P. J. Horwich, S. K. Shetty, and M. G. Hochman Bone and Soft-Tissue Lesions: What Factors Affect Diagnostic Yield of Image-guided Core-Needle Biopsy? Radiology, September 1, 2008; 248(3): 962 - 970. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. JOHNSON Breast Biopsy Techniques And Breast Pathology Radiol. Technol., September 1, 2007; 79(1): 44M - 59M. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 G. Vlastos and H. M. Verkooijen Minimally Invasive Approaches for Diagnosis and Treatment of Early-Stage Breast Cancer Oncologist, January 1, 2007; 12(1): 1 - 10. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
