| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Original Research |
1 Breast Imaging Section, Department of Radiology, Memorial Sloan-Kettering
Cancer Center, 1275 York Ave., New York, NY 10021
2 Department of Pathology, Memorial Sloan-Kettering Cancer Center, New York,
NY.
Received October 26, 2006;
accepted after revision May 18, 2007.
Supported by a grant from the New York State Empire Clinical Investigator
Program.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. With institutional review board approval, retrospective review was performed of a database of 342 lesions that had MRI-guided vacuum-assisted biopsy during a 39-month period. Biopsy samples were obtained in a 1.5-T magnet using a 9-gauge MRI-compatible vacuum-assisted biopsy device. Medical and pathology records were reviewed to determine the number of discordant lesions and surgical outcome. Statistical analysis was performed.
RESULTS. Among 342 lesions that had MRI-guided vacuum-assisted biopsy, results were discordant in 24 (7% [95% CI, 3–14%]) lesions. The discordance rate was significantly (p < 0.001) higher among MRI target lesions that were possibly missed rather than sampled. A trend was seen (p < 0.06) toward a higher discordance rate in MRI lesions that were sampled rather than excised at MRI-guided vacuum-assisted biopsy. Subsequent surgery in 20 discordant lesions yielded cancer in six (30% [12–54%]), including ductal carcinoma in situ (DCIS) in two and invasive carcinoma in four (three ductal and one lobular, all with DCIS). The cancer rate among discordant lesions was significantly higher in postmenopausal women than in premenopausal women (p <0.05).
CONCLUSION. Imaging–histologic discordance was found in 7% of lesions that had MRI-guided vacuum-assisted biopsy. Among discordant lesions, surgical excision revealed cancer in 30%. Imaging–histologic correlation is essential after MRI-guided vacuum-assisted biopsy to avoid delay in the diagnosis of breast cancer.
Keywords: biopsy • breast • breast cancer • MRI • women's imaging
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Although MRI-guided 9-gauge vacuum-assisted biopsy is being increasingly used for diagnosis of indeterminate or suspicious lesions identified at breast MRI, few data address imaging–histologic discordance with this technique [6–9]. This study was undertaken to determine the frequency of discordance at MRI-guided 9-gauge vacuum-assisted biopsy and to assess the rate and histology of cancer identified in these discordant lesions.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The 342 lesions in this study did not include 27 lesions that had MRI-guided vacuum-assisted biopsy and subsequent MRI-guided needle localization on the same day as part of a validation study at our institution [6] but did include 95 lesions reported in our first clinical experience with MRI vacuum-assisted biopsy as an alternative to surgical biopsy [8]. An additional 52 lesions scheduled for MRI-guided vacuum-assisted biopsy during this 39-month period did not undergo vacuum-assisted biopsy, because no suspicious lesion was identified on the day of biopsy (n = 46), the patient was unable to cooperate with the procedure (n = 4), or technical factors precluded biopsy (n = 2, both extremely posterior lesions, one of which was in a thin breast).
Indications for MRI that led to the detection of these 342 lesions included high-risk screening (e.g., in women with genetic predisposition, strong family history of breast cancer, prior breast cancer, biopsy-proven diagnosis of atypical hyperplasia or lobular carcinoma in situ, or prior mantle irradiation for Hodgkin's disease) in 165 (48%), extent of disease assessment in the ipsilateral or contralateral breast in a woman with proven breast cancer in 83 (24%), problem solving in 61 (18%), and follow-up after previous breast MRI for enhancement considered "probably benign" in 33 (10%) [10].
For MRI-detected lesions referred for biopsy at our center, correlative sonography is recommended at the discretion of the interpreting radiologist and the referring clinician. Lesions referred for sonography are generally masses larger than 0.5 cm. If a sonographic correlate is confidently identified, biopsy is usually performed under sonographic guidance. For MRI-detected lesions that have no sonographic correlates, that have vague sonographic findings that do not definitely correlate with the MRI lesion, or that are better seen on MRI than sonography, biopsy is usually performed under MRI guidance.
The decision regarding MRI-guided vacuum-assisted biopsy versus MRI-guided localization is made in consultation among the radiologist, the clinician, and the patient. In general, MRI-guided vacuum-assisted biopsy is considered the method of choice for solitary lesions that are accessible to the procedure. MRI-guided needle localization is often performed if multiple lesions are present, if the lesion is considered inaccessible to MRI-guided vacuum-assisted biopsy (e.g., due to extreme posterior location), or if the patient or referring clinician prefers surgical excision.
Biopsy Radiologists
Biopsies are performed at our center with a 9-gauge MRI-compatible
vacuum-assisted biopsy device (Automated Tissue Excision and Collection
[ATEC], Suros Surgical Systems)
[6]. In this study, biopsies
were performed by one of 14 radiologists specializing in breast imaging. All
radiologists had prior experience in breast MRI, percutaneous stereotactic,
and sonographically guided breast biopsy. Three radiologists had performed a
median of five (range, 1–22) MRI-guided vacuum-assisted biopsies during
the previous validation study at our institution
[6]. The rest of the
radiologists observed or assisted one or more experienced radiologists before
assuming responsibility for MRI-guided vacuum-assisted biopsy.
Positioning and Biopsy
After giving informed consent for MRI-guided biopsies, the patient was
positioned prone in the 1.5-T magnet (Signa, GE Healthcare). A dedicated
breast surface coil and breast biopsy device were used, either the Biopsy
Breast Array Coil (Model BBC, MRI Devices; now In vivo, Intermagnetics) or the
Open Breast Coil (Model OBC-63, MRI Devices; now In vivo) with a
grid-localizing system (Biopsy Positioning Device, Model MR-BI-160, MRI
Devices; now In vivo). MRI biopsy technique has been previously described
[6,
11]. The biopsy site was
marked with a titanium marker (TriMark, Suros).
After biopsy, the breast was compressed with ice, sterile strips were applied, a two-view digital mammogram was obtained to document clip placement, and a sterile gauze bandage was applied. Postbiopsy instructions were given verbally and in writing, and the patient was told when she would be contacted with biopsy results.
Imaging–Histologic Correlation: Definition of Discordance
Imaging and histologic findings are "concordant" when
histologic findings provide a sufficient explanation for the imaging features
and "discordant" when they do not
[1]. At stereotactic or
sonographically guided biopsy, discordance occurs when the biopsy was
performed for calcifications, but no calcifications are identified at specimen
radiography; when lesions highly suggestive of malignancy (BI-RADS category 5)
yield benign results; and when lesions that are suspicious (BI-RADS category
4) yield histology that does not match the imaging pattern (i.e., a discrete
mass yielding "benign breast tissue")
[10]. At our institution, the
radiologist performing percutaneous biopsy does not finalize the report until
the histologic review is complete; the radiologist then correlates histology
and imaging findings and dictates an addendum to the report indicating whether
concordance exists. For discordant cases, rebiopsy (usually surgical excision)
is recommended.
At MRI-guided biopsy, as at percutaneous biopsy with other methods, discordance occurs when the histology does not provide a sufficient explanation for the imaging findings: for example, a BI-RADS category 5 lesion yielding benign results, or a BI-RADS category 4 lesion in which the histology and imaging findings do not match. For the purposes of this study, the impression of discordance was formulated on the basis of the histologic findings and images obtained on the day of the biopsy (both before and after tissue acquisition), interpreted in conjunction with the MRI performed before the date of biopsy.
Data Collection and Analysis
In a protocol approved by our institutional review board, data collected
for this study included MRI indication, patient age, menopausal status, MRI
lesion parameters (type, size, and location), biopsy parameters, and
histologic results. MRI examinations performed before and on the day of biopsy
(including images obtained before and after tissue acquisition) and postbiopsy
mammograms were reviewed.
Each MRI lesion was described by one of 14 radiologists before biopsy according to lesion type (mass, nonmass, or focus of enhancement) and final BI-RADS assessment categories (category 1, negative; 2, benign; 3, probably benign; 4, suspicious; 5, highly suggestive of malignancy) as per the BI-RADS breast MRI lexicon [10]. Lesion location was recorded as retroareolar (within 2 cm of the nipple), upper outer quadrant, upper inner quadrant, lower outer quadrant, or lower inner quadrant. Mammographic parenchymal pattern was classified as class 1 (fatty), 2 (scattered fibroglandular densities), 3 (heterogeneously dense), or 4 (dense) [12].
Images from MRI examinations performed after completing tissue acquisition on the day of biopsy were reviewed to determine whether the imaging target was excised, sampled, or possibly missed. Additional IV contrast material was injected after tissue acquisition but before postbiopsy imaging on the day of biopsy at the discretion of the biopsy radiologist to assess for residual target lesion. The target was considered to have been excised if the images obtained after tissue acquisition showed a postbiopsy change, such as air at the biopsy site, with interval removal of the target lesion. The target was considered to have been sampled if some of the lesion was excised but portions of the lesion remained. The target was judged to be "possibly missed" if the lesion appeared to persist or if there was uncertainty as to whether the target was sampled.
Data were entered in a computerized spreadsheet (Excel, Microsoft). We calculated the frequency of discordance and the likelihood of cancer in discordant lesions in our study, and we compared these frequencies with the discordance rate and likelihood of cancer in discordant lesions in prior studies of stereotactic and sonographically guided breast biopsy [1–5]. Tests for statistical significance were performed using computerized statistical software (Epi-Info, Centers for Disease Control and Prevention) and using the chi-square and Fisher's exact tests, with p value of less than 0.05 considered significant. The 95% CIs were calculated using the Geigy scientific tables [13].
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Factors Associated with Discordance
Correlations between discordance and specific patient, lesion, or biopsy
factors are shown in Table 1.
Discordance was present in six (43%) of 14 lesions in which the imaging target
was possibly missed, in 17 (7%) of 236 lesions in which the target was
sampled, and in one (1%) of 92 lesions in which the target was excised at
MRI-guided vacuum-assisted biopsy. Discordance was significantly (p
< 0.001) more likely if the MRI target lesion was possibly missed rather
than sampled. A trend was seen (p < 0.06) toward a higher
discordance rate if the MRI lesion was sampled rather than excised at
MRI-guided vacuum-assisted biopsy. Other factors evaluated had no significant
impact on the discordance rate (Table
1).
|
Surgical Outcome in Discordant Lesions
Surgical outcomes in 20 discordant lesions are shown in
Table 2. Surgery revealed
benign histology in 13 (65%) of 20 discordant lesions and high-risk findings
(atypical ductal hyperplasia and lobular carcinoma in situ) in one discordant
lesion. Surgery yielded cancer in six (30%) of 20 discordant lesions,
including ductal carcinoma in situ (DCIS) in two and invasive carcinoma in
four (three ductal and one lobular, all with DCIS)
(Table 3). The median size of
invasive cancer was 0.4 cm (range, 0.1–0.5 cm). The median interval from
vacuum-assisted biopsy to surgery in these six cancers was 32 days (range,
16–35 days). Cancer was significantly (p < 0.05) more likely
in discordant lesions in postmenopausal than in premenopausal women. No other
factors evaluated had a significant association with the cancer rate in
discordant lesions (Table
1).
|
|
Surgical outcomes were unavailable for four discordant lesions. Two women with BI-RADS category 4 lesions sampled at MRI vacuum-assisted biopsy sought treatment elsewhere. One woman with a BI-RADS category 4 lesion in which the target was excised at MRI-guided vacuum-assisted biopsy chose MRI follow-up rather than surgery; follow-up MRI 4 months after the initial biopsy confirmed target lesion removal, and an additional 6-month follow-up MRI is planned. Another woman with a BI-RADS category 4 lesion sampled at MRI-guided vacuum-assisted biopsy had repeated MRI-guided vacuum-assisted biopsy that yielded benign concordant results; no suspicious findings were identified on follow-up MRI examinations, most recently 16 months after the initial biopsy.
Our Results Versus Results of Prior Studies of Percutaneous Breast Biopsy
Our results at MRI-guided vacuum-assisted biopsy can be compared with
published investigations of percutaneous stereotactic and sonographically
guided biopsy
[1–5]
(Table 4). In our study,
MRI-guided vacuum-assisted biopsy was significantly more likely to yield
discordant results than percutaneous biopsies performed under stereotactic or
sonographic guidance in the published literature (24 /342 = 7% vs 123 /4,002 =
3%; p < 0.001). The proportion of discordant lesions having
subsequent surgical excision in our study of MRI-guided vacuum-assisted biopsy
was comparable with prior reports of stereotactic or sonographically guided
biopsy (20 /24 = 83% vs 101 /123 = 82%, p = 1.0). Among discordant
lesions that had repeat biopsy, a trend was seen toward a higher cancer rate
in lesions that had MRI-guided vacuum-assisted biopsy in our study than in
prior reports of stereotactic or sonographically guided biopsy (6 /20 = 30% vs
14 /101 = 14%; p <0.1).
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In our study, imaging–histologic discordance occurred in 7% of lesions that had MRI-guided 9-gauge vacuum-assisted biopsy. Our 7% discordance rate at MRI-guided vacuum-assisted biopsy is higher than the 3% discordant rate previously reported at stereotactic or sonographically guided biopsy (Table 4). The higher discordance rate at MRI-guided vacuum-assisted biopsy as compared with other biopsy methods may result from several factors, including the higher risk status of women who have MRI-guided vacuum-assisted biopsy as compared with other biopsy methods, the relative newness of both MRI-guided vacuum-assisted biopsy and correlating MRI findings with histology results, and the variability of enhancement during the menstrual cycle in premenopausal women and during the biopsy procedure in all women [16–19].
We found a significantly higher discordance rate at MRI vacuum-assisted biopsy if the imaging target was possibly missed rather than sampled (43% vs 7%, p < 0.001) and a trend toward higher discordance rate if the MRI target was sampled rather than excised (7% vs 1%, p < 0.06). The latter finding is similar to the stereotactic biopsy experience, in which sampling rather than complete excision of the mammographic target correlated with a significantly higher discordance rate (8 /322 = 2.5% vs 1 /466 = 0.2%; p = 0.004) [20]. Although not usually the goal of MRI-guided vacuum-assisted biopsy, complete removal of the MRI target may reduce the likelihood of discordance, potentially decreasing the rebiopsy rate and improving the cost-effectiveness of MRI-guided vacuum-assisted biopsy. It should be remembered, however, that even in cancers in which the MRI target is removed, surgery reveals residual cancer in 33–63%. Complete removal of the MRI target does not ensure complete removal of the histologic process [8, 20].
Several factors associated with discordance at stereotactic or sonographically guided biopsy were not associated with discordance at MRI-guided biopsy. The discordance rate significantly decreases with operator experience for other percutaneous biopsy methods, but not for MRI-guided vacuum-assisted biopsy. The lack of significant association between operator experience and discordance at MRI-guided biopsy has several possible explanations. Perhaps individuals proficient at other percutaneous biopsy methods may have a minimal learning curve for MRI-guided vacuum-assisted biopsy; it may be that the interpretation of discordance is related to nonlearning curve factors (such as the patient's prior probability of breast cancer); alternatively, it is possible that study of a larger number of lesions may yield a significant association between these parameters. For other percutaneous biopsy methods, discordance is significantly more likely in BI-RADS category 5 than in BI-RADS category 4 lesions. We found no significant difference in discordance rate as a function of BI-RADS category at MRI-guided vacuum-assisted biopsy, but the number of BI-RADS category 5 lesions in this series was small.
|
|
|
|
|
|
|
|
|
|
For those embarking on MRI-guided vacuum-assisted biopsy, we offer suggestions to help identify discordance. Radiologists with expertise in breast imaging and MRI–histologic correlation should perform these procedures. On the day of the biopsy, flushing or aspirating the biopsy cavity and injecting additional IV contrast material after tissue acquisition but before obtaining postbiopsy images may help the radiologist assess for residual enhancement. After the biopsy is done, the radiologist should review all images from the biopsy procedure together with the prior diagnostic MRI, ideally in a calm, appropriately dark room at a PACS monitor that allows scrolling and adjustment of window and level; this assessment is invaluable in determining whether the lesion was excised, sampled, or possibly missed. The biopsy report should not be finalized until histologic analysis is complete; the radiologist can then dictate an addendum indicating the histology results, presence of concordance or discordance, and management recommendations. If concordance is uncertain, postbiopsy MRI may be useful to assess lesion sampling; we generally recommend that this study be performed 1 month after biopsy, although further study is needed to formulate evidence-based recommendations in this regard.
In conclusion, we encountered imaging–histologic discordance in 7% of lesions that had MRI-guided vacuum-assisted biopsy. Among discordant lesions, surgical excision revealed cancer in 30%. Imaging–histologic discordance is an indication for surgical excision after MRI-guided vacuum-assisted biopsy because of the high prevalence of carcinoma in discordant lesions. Imaging–histologic correlation is essential after MRI-guided vacuum-assisted biopsy, as it is after all breast biopsy procedures, to avoid a deleterious delay in breast cancer diagnosis. Additional work is necessary to enhance MRI biopsy techniques, to improve methods for assessing the adequacy of target lesion sampling, to correlate breast MRI findings with histology, and to develop guidelines for the use of postbiopsy MRI, so that more women may benefit from the use of MRI and MRI-guided biopsy in breast cancer diagnosis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  P. A. Causer, C. A. Piron, R. A. Jong, and D. B. Plewes Preliminary In Vivo Validation of a Dedicated Breast MRI and Sonographic Coregistration Imaging System Am. J. Roentgenol., October 1, 2008; 191(4): 1203 - 1207. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
