|
|
||||||||
1 Department of Radiology, University of Washington Medical Center, 1959 NE
Pacific, Seattle, WA 98195.
2 Breast Imaging, Seattle Cancer Care Alliance, 825 Eastlake Ave. E, G4-830,
Seattle, WA 98109-1023.
3 Mayo Clinic, 4500 San Pablo Rd., Jacksonville, FL 32224.
Received June 28, 2004;
accepted after revision September 22, 2004.
Presented at the 2004 annual meeting of the American Roentgen Ray Society,
Miami Beach, FL.
Abstract
|
|
|---|
MATERIALS AND METHODS. Consecutive MRI-guided vacuum-assisted breast biopsies of 38 lesions in 28 women performed between May and September 2003 at two practice sites in the United States were retrospectively reviewed. Lesion characteristics including size, morphology, and enhancement were recorded. Times to perform each procedure, defined as the time from the start of the first localizing scan to the final scan after biopsy, were recorded. Histologic results for all lesions were obtained, and surgical, imaging, or clinical follow-up was performed.
RESULTS. Enhancing masses and foci ranged from 2.5 to 19 mm. Nonmasslike enhancements ranged from 6 to 70 mm. All 38 biopsies (100%) were technically successful, and no complications were associated with any of the biopsy procedures. The average time to perform the 19 single-site MRI-guided procedures was 38 min (range, 23-57 min). The 11 multiple-site biopsies performed in a single breast averaged 59 min (range, 51-68 min), and eight bilateral biopsies averaged 64 min (range, 46-80 min). Histologic results from vacuum-assisted breast biopsy revealed malignancy in 14 lesions (37%), atypical ductal hyperplasia in two lesions (5%), and benign findings in 22 lesions (58%). One of two lesions with atypical ductal hyperplasia was upgraded to ductal carcinoma in situ after surgery, for an overall cancer yield of 40% (15/38).
CONCLUSION. This new method of MRI-guided vacuum-assisted breast biopsy is a safe, effective, and time-efficient means of MRI-guided tissue sampling.
|
|
|---|
Because of the increasing use of MRI and its high rate of both true- and false-positive lesions occult on other forms of imaging, a method is needed for accurate tissue sampling of lesions seen only on MRI. The most common methods of MRI biopsy currently in use are wire localization and surgical excision or core needle biopsy [8-17].
Under stereotactic mammographic guidance, vacuum-assisted biopsy has proven to be more effective than core needle biopsy and less invasive than surgical excision of breast lesions [18, 19]. The advantages of vacuum-assisted breast biopsy include a single probe insertion with directional sampling and rapid collection of larger samples. MRI-compatible devices have been developed that allow vacuum-assisted needle biopsy to be performed under MR guidance [20, 21]. Clinical studies in Europe have shown that vacuum-assisted breast biopsy was successful in 98% of 341 lesions [22] and have identified important barriers to more widespread use of this technology, including needle artifact, tissue shift during probe insertion, and washout of contrast enhancement during the procedure. The challenges of this early technology have limited application to lesions larger than 10 mm in diameter, and the devices to date are not commercially available in the United States [22, 23].
Recently, a new commercially available method of MRI-guided vacuum-assisted breast biopsy that addresses many of the limitations identified by earlier prototypes has been developed. The purpose of this study was to evaluate this method of vacuum-assisted breast biopsy in women undergoing MRI-guided breast biopsy of lesions identified only on MRI.
|
|
|---|
Lesion Characteristics
MRI interpretation criteria, assessments, and recommendations were based on
the breast MRI lexicon established by the American College of Radiology
[24]. All lesions identified
on diagnostic MRI were assessed by the radiologist as category 4 (suspicious
abnormality) or category 5 (highly suggestive of malignancy). A targeted
second-look sonography examination was performed on all lesions. If the lesion
was clearly visible, a sonographically guided core biopsy was performed. If
the lesion was not visible on sonography, as was the case in the 38 lesions
included in this study, an MRI-guided vacuum-assisted breast biopsy was
performed.
For all cases, the following lesion characteristics were recorded: type (focus, mass, or nonmasslike enhancement), size, location, and enhancement kinetics (persistent, plateau, or washout).
Biopsy Procedure
At both institutions, biopsies were performed following standard clinical
protocol. The time to perform the procedure, defined as time from the first
MRI sequence (localizing) to the last scan sequence (after clip deployment),
was recorded for each procedure.
|
|
|
Components of the ATEC Introducer Set (Suros Surgical Systems) include a needle guide, a coaxial introducer sheath, a sharp nonferrous inner stylet, and a plastic localizing obturator (Fig. 2). These components provide access to the target lesion and verification of sheath placement. The combined introducer sheath and stylet allows access to the biopsy site while minimizing fluid egress. The non-artifact-producing localizing obturator appears as a signal void, a "black dot," on the image screen in the sagittal view and identifies the position of the sheath in reference to the target lesion. The tip of the obturator corresponds to the center of the sampling chamber once the obturator is removed and the sampling device is inserted through the sheath. Placement of the biopsy site marker is accomplished with the introducer sheath still in place by guiding the marker through the sheath to the biopsy target site.
|
|
|
|
|
MRI-Guided Biopsy Procedure
After written informed consent was obtained from the patient and IV access
was established, the patient was placed prone with the affected breast or
breasts (for bilateral biopsies) positioned in the dedicated breast coil
(model OBC-63 Breast Array Coil, MRI Devices). The breast was stabilized in a
biopsy guidance grid (ATEC, Suros Surgical Systems) by providing moderate
compression (Fig. 4). A vitamin
E capsule was taped to the skin in the center of the grid for reference. The
biopsy guidance grid also allows determination of the x and
y coordinates of the target lesion.
After localizing sequences had been performed, unenhanced and contrast-enhanced sagittal images were obtained to confirm the persistence of the enhancing lesion noted on the prior diagnostic MRI (Figs. 5A, 5B, 5C, and 5D). The images were reviewed on the monitor, and the locations of the vitamin E capsule placed on the skin surface and the lesion were marked on a reference diagram. These markings correspond to the x and y coordinates of the lesion and target. In addition, the depths (z coordinates) of the lesion and vitamin E capsule were recorded. The depth (z) difference between the lesion and the vitamin E skin marker was calculated and recorded on the diagram.
The MRI table was moved out of the bore of the magnet, the patient's skin was then sterilized with commercially available povidone-iodine and local anesthetic (1% lidocaine mixed with bicarbonate solution [10:1]) was injected. A sterile plastic coaxial sheath with a sharp inner nonferrous metallic stylet was inserted to the calculated depth using the centimeter markers on the exterior of the sheath. Once the coaxial sheath was in place, the inner cutting stylet was removed and replaced with a plastic MRI-visible obturator. With the coaxial sheath and obturator in place, the patient was returned to the bore of the magnet.
A sagittal sequence was performed through the region of the obturator and the lesion to confirm accurate targeting (Figs. 5A, 5B, 5C, and 5D). Once correct targeting was confirmed, the patient was moved out of the bore of the magnet and tissue sampling was performed using the vacuum-assisted breast biopsy device through the coaxial sheath. Sampling was performed preferentially in the direction of the lesion. For example, if the lesion was noted to be inferior in relation to the probe on the sequence performed after obturator insertion, samples were obtained from areas between the 4- and 8-o'clock positions. When tissue sampling was complete, the probe was retracted 5 mm and an MRI-safe site marker was placed through the coaxial sheath (ATEC Mark Biopsy Site Identifier, Artemis Medical). A final sagittal sequence through the biopsied region was performed to verify the location of the sampling defect and of the site marker (Figs. 5A, 5B, 5C, and 5D).
At the completion of the biopsy, the coaxial system was removed. Manual compression was applied to achieve hemostasis, and a sterile dressing was applied.
Histologic Data Collection
Histologic results for all lesions were obtained, and surgical, imaging, or
clinical follow-up was recommended. Biopsy histologic results were classified
into the following categories: invasive carcinoma, in situ carcinoma, atypical
ductal hyperplasia (ADH), or benign.
Statistical Analysis
The average time to complete an examination and SD were computed. The
average lesion size and SD were computed. The positive predictive value was
computed by considering ductal carcinoma in situ (DCIS) or invasive carcinoma
as positive cases.
|
|
|---|
Table 1 summarizes the types and sizes of lesions. Targeted masses and enhancing foci ranged from 2.5 to 19 mm (mean, 9.3 mm). Nonmasslike enhancement ranged from 6 to 70 mm (mean, 23.6 mm).
|
Table 2 presents the type of biopsies performed and the average time to complete each type. Half of the study biopsies (n = 19) were on single lesions, 29% (n = 11) of the biopsies involved multiple lesions in a single breast, and the remaining 21% (n = 8) involved multiple lesions in both breasts. The average time to perform the 19 single-biopsy MRI-guided procedures was 38 min (range, 23-57 min). The 11 multiple biopsies performed in a single breast averaged 59 min (range, 51-68 min), and the eight bilateral biopsies averaged 64 min (range, 46-80 min).
|
All 38 biopsies (100%) were technically successful, and no complications were associated with any of the biopsy procedures. Histologic results from vacuum-assisted breast biopsies revealed malignancy in 14 lesions (37%), ADH in two lesions (5%), and benign findings in 22 lesions (58%) (Table 3). Thirteen of the 14 cases of malignancy and the two cases of ADH were followed by surgical excision, lumpectomy, or mastectomy. One patient with infiltrating carcinoma diagnosed by vacuum-assisted breast biopsy elected not to have further surgery when widespread metastases were identified at further imaging evaluation.
|
One of the two lesions diagnosed as ADH by vacuum-assisted breast biopsy was upgraded to DCIS after surgery, and one of four lesions identified as DCIS was upgraded to invasive cancer. One case of invasive carcinoma was downgraded to DCIS on the basis of analysis of the lumpectomy specimen. Two separate lesions originally identified as DCIS or infiltrating ductal carcinoma on vacuum-assisted breast biopsy were downgraded to benign at surgical excision. In both cases, the biopsy cavity from the vacuum-assisted breast biopsy documenting malignancy was identified in the surgical specimen, confirming the malignancies were removed by the initial vacuum-assisted breast biopsy. Two cases with benign histologic findings at vacuum-assisted breast biopsy were followed by mastectomy, one in a patient who opted for bilateral mastectomies with known cancer in the opposite breast and one in a woman participating in a high-risk screening trial who opted for prophylactic mastectomy. Surgical pathology from the mastectomy in this second case revealed 5 mm of DCIS near the area of the prior MRI-guided vacuum-assisted breast biopsy. The other case was confirmed to be benign.
|
|
|---|
|
This method of vacuum-assisted breast biopsy takes less time to acquire tissue compared with prior published reports of MRI-guided core needle biopsy [15, 27, 28]. For example, MRI-guided core needle biopsy of single lesions required 45 and 55 min on average in two prior published studies, compared with 35 and 38 min for single lesions using MRI-guided vacuum-assisted breast biopsy [25, 27, 28]. The time difference appears to be even more dramatic for multiple lesions, with prior reports of 90 min on average to perform multiple core needle biopsies compared with 69 and 61 min on average to perform multiple biopsies with MRI-guided vacuum-assisted breast biopsy (Table 4).
The data in our study represent our early experience with MRI-guided vacuum-assisted breast biopsy. Of the two facilities in our study, one had no prior experience in MRI-guided vacuum-assisted breast biopsy and the other had only 1 month of experience before the study. The first site did have approximately 1 year of experience with MRI-guided core needle biopsy; the second site had no prior experience with any MRI-guided interventions. It may be that MRI-guided vacuum-assisted breast biopsy can be performed in even less time than reported here once experience is gained. We now find the total table time is approximately 20 min for biopsy of a single lesion.
Biopsy of lesions that are far medial or posterior in the breast can prove challenging. Until recently, access for MRI-guided biopsy was limited to the lateral side of the breast. However, coils that allow medial access are now available and new coils that provide a full range of options for access including superior and inferior approaches are being developed. For far posterior lesions, we have found that removing the pad on the coil and rolling the patient to a position similar to the positioning on the stereotactic table for gaining access to far posterior lesions can improve access.
This study is the largest to date of MRI-guided vacuum-assisted breast biopsy using the ATEC device. However, there are limitations to this study. Benign biopsies were not followed by surgical excision in 20 of the 22 benign cases. In one case in which the vacuum-assisted breast biopsy sample was benign at histology, DCIS was identified at prophylactic mastectomy. Longer term follow-up of patients with benign histologic results from MRI-guided breast biopsy samples is needed to determine the true rate of false-negative biopsies. This information from long-term follow-up of patients and further study of lesion characteristics associated with various forms of benign abnormalities will help in management decisions regarding suspicious MRI lesions with benign histology. At this time, we determine imaging-pathology concordance on a case-by-case basis and recommend surgical excision for all cases of ADH and for any discordant cases. In addition, we recommend 6-month follow-up MRI examinations in all patients with benign histologic results. If the lesion is stable or decreased at the 6-month follow-up examination, we recommend resumption of routine screening.
In conclusion, this method of tissue sampling for suspicious MRI lesions is safe and accurate, faster than core needle biopsy, and less invasive than wire localization and surgical excision. As the use of breast MRI grows, this procedure provides an effective and minimally invasive method of sampling lesions that are suspicious for malignancy.
Acknowledgments
We thank Bonnie Thursten and John McCloskey for their contributions to the
success of this project and Tamara Fernando for her invaluable assistance in
the preparation of the manuscript.
|
|
|---|
This article has been cited by other articles:
![]() |
J.-M. Lee, J. B. Kaplan, M. P. Murray, and L. Liberman Complete Excision of the MRI Target Lesion at MRI-Guided Vacuum-Assisted Biopsy of Breast Cancer Am. J. Roentgenol., October 1, 2008; 191(4): 1198 - 1202. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. P. Lourenco, M. B. Mainiero, E. Lazarus, D. Giri, and B. Schepps Stereotactic Breast Biopsy: Comparison of Histologic Underestimation Rates with 11- and 9-Gauge Vacuum-Assisted Breast Biopsy Am. J. Roentgenol., November 1, 2007; 189(5): W275 - W279. [Abstract] [Full Text] [PDF] |
||||
![]() |
J.-M. Lee, J. B. Kaplan, M. P. Murray, L. Bartella, E. A. Morris, S. Joo, D. D. Dershaw, and L. Liberman Imaging Histologic Discordance at MRI-Guided 9-Gauge Vacuum-Assisted Breast Biopsy Am. J. Roentgenol., October 1, 2007; 189(4): 852 - 859. [Abstract] [Full Text] [PDF] |
||||
![]() |
J.-m. Lee, J. B. Kaplan, M. P. Murray, M. Mazur-Grbec, T. Tadic, D. Stimac, and L. Liberman Underestimation of DCIS at MRI-Guided Vacuum-Assisted Breast Biopsy Am. J. Roentgenol., August 1, 2007; 189(2): 468 - 474. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Kuhl The Current Status of Breast MR Imaging * Part I. Choice of Technique, Image Interpretation, Diagnostic Accuracy, and Transfer to Clinical Practice Radiology, August 1, 2007; 244(2): 356 - 378. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Liberman, A. E. Holland, D. Marjan, M. P. Murray, L. Bartella, E. A. Morris, D. D. Dershaw, and R. T. Wynn Underestimation of Atypical Ductal Hyperplasia at MRI-Guided 9-Gauge Vacuum-Assisted Breast Biopsy Am. J. Roentgenol., March 1, 2007; 188(3): 684 - 690. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. G. ODLE Breast MR Radiol. Technol., September 1, 2006; 78(1): 45M - 66M. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |