| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Radiology Department, Palo Alto Medical Clinic, 795 El Camino Real, Palo Alto,
CA 94301.
2 Surgery Department, Palo Alto Medical Clinic, Palo Alto, CA 94301.
Received June 6, 2002;
accepted after revision August 27, 2002.
Address correspondence to R. J. Jackman.
Abstract
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
MATERIALS AND METHODS. From July 1991 through June 2001, 1687 consecutive patients (age range, 29-94 years; median age, 58 years) with 1894 lesions were referred by clinicians in a multispecialty clinic. The patients underwent stereotactic, prone, histologic biopsy of 1851 lesions (98%) and needle-localized breast biopsy of 43 lesions (2%). We performed stereotactic biopsies successively with 14-gauge automated large-core devices and 14- or 11-gauge vacuum-assisted devices. We evaluated lesions by patient, breast, lesion, and procedural variables to determine why stereotactic biopsy was not performed.
RESULTS. Of 1851 lesions referred for stereotactic biopsy, biopsies were canceled in 42 lesions (2%) not considered suspicious enough to warrant biopsy. Of 1809 lesions in which stereotactic biopsy was considered to be warranted, stereotactic biopsy was canceled for technical reasons in 29 lesions (2%). Of 43 lesions referred for surgical biopsy, stereotactic biopsy was thought to be technically problematic in five (12%). Inability to accomplish a stereotactic biopsy in 34 (2%) of 1852 lesions needing a biopsy was due to proximity to the chest wall (n = 10, 29%), inadequate lesion visualization unrelated to lesion depth (n = 19, 56%), and patient factors (n = 5, 15%).
CONCLUSION. Stereotactic biopsy had a technical success rate of 98% (1780/1809) and was used for histologic diagnosis in 96% (1780/1852) of mammographically detected lesions. Inadequate lesion visualization accounted for 85% (29/34) of stereotactic biopsy failures.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In July 1991, we changed from performing needle-localized breast biopsies to performing stereotactic prone biopsies for mammographically detected lesions. This retrospective study was undertaken to determine what factors during our first 10 years of performing stereotactic biopsies were associated with completion or cancellation of the biopsy.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
From July 1991 through June 1992, all patients decided whether to undergo a percutaneous biopsy or a surgical biopsy after consultation with a breast surgeon. After the first year, a preponderance of patients made a decision about the type of biopsy after consultation with their referring physician (before the biopsy day) and a diagnostic radiologist (immediately before the scheduled biopsy). Patients continued to have a consultation with a surgeon before biopsy if a discretely palpable mass was present or if the patient or referring clinician made such a request. Most discretely palpable masses were biopsied by the surgeons using either histologic needle biopsy or excisional biopsy, both performed using palpation guidance. Informed consent was obtained from each patient before biopsy.
Patients whose weight exceeded the biopsy table limit of 300 lb (135 kg), who had a bleeding diathesis, or whose anticoagulant therapy could not be temporarily stopped were restricted from scheduling a stereotactic biopsy. We had no standard prebiopsy restrictions based on presumed inability of the patient to cooperate with the procedure, breast size, or lesion size, position, or conspicuity.
Patient, breast, lesion, and procedural variables were recorded at the time of biopsy completion or cancellation. The only patient variable was age, and the only breast variable was tissue density (extremely dense, heterogeneously dense, scattered fibroglandular densities, or almost entirely fat). Lesions were categorized before biopsy by referring clinicians for palpability (discrete, vague, or impalpable) and by the radiologist who performed the biopsy by mammographic type (microcalcifications without a mass; masses, including asymmetric densities and areas of architectural distortion, with or without associated calcifications; or miscellaneous), maximum mammographic lesion diameter, and Breast Imaging Reporting and Data System (BI-RADS) lexicon of the American College of Radiology [9] (categories 2-5). Starting February 1995, we recorded breast tissue density and the BI-RADS classification before biopsy for all lesions; those items were recorded retrospectively for all lesions without a completed stereotactic biopsy before that time. Other variables were recorded for all lesions.
We performed stereotactic biopsies using a prone-biopsy table (Mammotest or Mammotest Plus/S; Fischer Imaging, Denver, CO) with three successive techniques. Biopsies were performed from July 1991 to mid April 1995 by large-core biopsy with a long-throw (2.3-cm excursion) automated biopsy gun (Biopty; Bard Urological, Covington, GA) using a variety of 14-gauge cutting needles. From mid March 1995 to mid June 1996, we used a vacuum-assisted technique with a Mammotome device (Biopsys Medical/Ethicon Endo-Surgery, Cincinnati, OH) with 14-gauge probes. From mid May 1996 through June 2001, we used a vacuum-assisted technique with 11-gauge probes. During the 1-month overlap between the various biopsy techniques, availability of the new probe determined which method was used. No lesion or patient variables were considered in determining the method. Data from the vacuum-assisted biopsies with the 14- and 11-gauge probes were combined for analysis.
Image visualization techniques changed during the study; the dates of change are not available. In the first few years, images were visualized on mammographic film. In the last several years, we began using digital images with a variety of charge-coupled devices and monitors. Apart from the first several months of 2000, when a color monitor was used exclusively, digital images were viewed on a black-and-white monitor. In the past few months, after completion of this study, we used a higher resolution charge-coupled device and an associated black-and-white monitor (Mammo Vision Elite; Fischer Imaging).
A surgical skin hook (Converse Skin Hook, large N4717; Storz Surgical Instruments, St. Louis, MO) was used to stabilize the breast tissue during needle insertion during most biopsies and to increase breast thickness during biopsy of lesions close to the biopsy needle skin entry site, including in thin breasts.
All lesions in patients referred for stereotactic biopsy were evaluated immediately before biopsy on mammographic images made on the biopsy table or on a mammography machine to determine whether a biopsy was indicated and technically feasible. If the stereotactic biopsy was canceled, one of the authors retrospectively assigned a revised BI-RADS category (which could be either the same or different from the initial BI-RADS category) after reviewing the prebiopsy mammographic reports, the report of the radiologist canceling the biopsy, and (when available) the mammographic images. Reasons for cancellation were categorized as inadequate lesion suspicion to warrant a biopsy (with a revised BI-RADS category 2 or 3) or as various difficulties precluding biopsy of a worrisome lesion (with a revised BI-RADS category 4 or 5). The same author reviewed the medical record to determine whether the patient underwent a completed biopsy of this lesion at a later date or whether imaging follow-up occurred.
One of the authors reviewed the medical records of all patients who were referred for a needle-localized breast biopsy instead of a stereotactic biopsy to determine the reason for that referral. The other author reviewed the prebiopsy mammography reports and the mammographic images to make a subjective determination about whether a stereotactic biopsy would be technically feasible on those lesions using our current biopsy technique.
Histologic diagnoses were recorded for all biopsied lesions as malignant, high-risk, or benign. Malignant lesions included invasive carcinoma, ductal carcinoma in situ (DCIS), lymphoma, and sarcoma. We considered atypical ductal hyperplasia, atypical lobular hyperplasia, lobular carcinoma in situ, radial scar, phyllodes tumor, and papilloma to be high-risk lesions, in which the associated presence of carcinoma can be underestimated at percutaneous biopsy. We recommended that most high-risk lesions undergo surgical excision. Lesions not categorized as histologically malignant or high-risk were classified as benign.
During most of the study period, we recommended that the patient undergo bilateral mammography at 12 months after biopsy if the benign diagnosis was either fibroadenoma or lymph node. For all other benign diagnoses thought to be concordant with the mammographic findings, we recommended mammography of the affected breast (or breasts) 6 months after biopsy and a bilateral mammogram at 12 months after biopsy. In the later stages of the study period, we omitted the 6-month postbiopsy mammogram and recommended a bilateral mammogram 12 months after biopsy for almost all patients whose lesions had a benign concordant diagnosis. Subsequently, all patients with benign lesions not thought to need a repeated biopsy were asked to have annual diagnostic mammographic follow-up for at least 3 years after biopsy. A repeated biopsy, by a more aggressive percutaneous biopsy or by surgical excision, was recommended if discordance was found between the mammographic and histologic findings [10] or if substantial lesion progression was found at mammographic follow-up [11,12,13,14].
We wanted to evaluate lesions within a multispecialty clinic whose staff had a common philosophic approach to biopsy of lesions requiring imaging guidance for completion of the biopsy. Lesions in patients referred by physicians outside our multispecialty clinic were excluded. We also excluded lesions from within our multispecialty clinic that were biopsied by a surgeon who briefly worked using a different philosophic approach. During a 2-month period, that surgeon evaluated 24 lesions with discretely or vaguely palpable masses by a combination of palpation, breast sonography, and fine-needle aspiration biopsy (n = 23) or 14-gauge histologic biopsy (n = 1). The biopsies were performed with palpation guidance, with or without sonographic guidance, and were often performed before a mammogram had been obtained.
The 1894 lesions were divided into four groups for data analysis. In three groups of lesions, the stereotactic biopsy was scheduled and then either completed (n = 1780), canceled for a worrisome lesion because of inability to complete the biopsy (n = 29), or canceled because of inadequate concern about the lesion to proceed with the biopsy in a patient who agreed to mammographic follow-up (n = 42). Patients in the fourth group of lesions were not referred for a stereotactic biopsy and went directly to an imaging-guided surgical biopsy (n = 43). Age was a variable per patient, and breast density was a variable per breast. All other variables were per lesion; therefore, the lesion was used as the unit of analysis.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Stereotactic Biopsy Canceled Because Lesion Was Reassessed as Benign
or Probably Benign
Stereotactic biopsy was canceled in 42 (2.3%) of 1851 lesions referred for
stereotactic biopsy and given a revised BI-RADS category of 2 (n = 5)
or 3 (n = 37) in patients who agreed to mammographic follow-up rather
than a stereotactic biopsy after they reviewed the mammographic images with
the radiologist. The original BI-RADS category in those lesions was 3
(n = 13) or 4 (n = 29). Follow-up analysis of those 42
lesions revealed a later biopsy of three lesions (all noncalcified masses),
with the histologic diagnosis being malignant in two lesions (both invasive
ductal carcinoma with associated DCIS) and benign in one. One of those lesions
had a successful stereotactic biopsy at our facility 8 months after the
canceled biopsy and was included in the 1780 lesions with a successful biopsy.
Two lesions had surgical biopsies performed later: one, 2 weeks and one, 24
weeks after the canceled biopsy. Thirty-six lesions had stable mammographic
follow-up, but no biopsy, and were assumed to be benign. Follow-up was
long-term (range, 23-109 months; median, 44 months) in 30 lesions and
short-term (range, 11-20 months; median, 16 months) in six lesions. The final
three (7%) of the 42 lesions had neither biopsy nor mammographic follow-up.
Eliminating those 42 lesions left 1852 lesions in need of a biopsy.
Lesions Referred for Surgical Rather Than Stereotactic Biopsy
The 43 lesions sent for surgical biopsy rather than stereotactic biopsy
constituted 2.3% of the total 1894 lesions referred for biopsy under imaging
guidance and 2.3% of the 1852 lesions in need of a biopsy. The reasons for
referral of 43 lesions for a surgical biopsy instead of a stereotactic biopsy
were divided into three groups in which the patient was told a surgical biopsy
was needed, and a fourth group in which the patient was offered a choice
between a stereotactic biopsy and a surgical biopsy.
In one group, the breast surgeon made the decision concerning 15 lesions. Nine patients with palpable lesions that were going to be surgically excised without imaging guidance had second impalpable lesions in need of an imaging-guided biopsy in the same breast (n = 8) or the opposite breast (n = 1), and the surgeon decided to remove both lesions surgically. Two lesions that the surgeon wanted to remove were discretely palpable, but the surgeon wanted imaging guidance before undertaking the lumpectomy to ensure an accurate biopsy. Single lesions were surgically biopsied because of recurrent or residual calcifications 12 months after a lumpectomy for carcinoma, an impalpable mass that looked like a complex cyst on sonography, an impalpable mass that the surgeon wanted to remove because of the patient's age, and a wire suture (placed during a prior surgical biopsy) that the patient thought might be the cause of breast pain.
The radiologist reviewing the prebiopsy mammogram requested the surgical biopsy in a second group of patients (11 lesions). Five of the 11 lesions were subjectively still thought to be technically problematic for completion of a stereotactic biopsy: three were calcific clusters in proximity to the chest wall, and two were faint calcific clusters in very small breasts. The other six of the 11 lesions would currently have a vacuum-assisted stereotactic biopsy: three were recurrent calcifications at a previous lumpectomy site, two were spiculated masses suggestive of radial scar [15], and one was a calcified mass with significant mammographic progression 4 years after a benign stereotactic biopsy.
The third group of patients (four lesions) referred for a surgical biopsy presented during an initial 2-month transition period before the decision to do stereotactic biopsy in almost all breast lesions needing imaging-guided biopsy was firmly established.
In a fourth group of patients (13 lesions), the patient was offered a choice of biopsy type and requested the surgical biopsy because she wanted the lesion surgically removed. Two of the 13 lesions were discretely palpable, and 11 were impalpable.
Thus, using our current technology, vacuum-assisted stereotactic biopsy was thought to be technically feasible in 88% (38/43) of lesions referred for a surgical biopsy.
Stereotactic Biopsy Canceled for Technical Reasons
Eliminating the 42 lesions not biopsied because of a revised BI-RADS
category and the 43 lesions with surgical biopsy left 1809 lesions with an
attempted stereotactic biopsy. Biopsy was canceled in 29 (1.6%) of those 1809
lesions for technical reasons; the revised BI-RADS category in those 29
lesions was 4 (n = 26) or 5 (n = 3). Three factors accounted
for these technical cancellations. Seven (24%) of the technical cancellations
were for calcification lesions too close to the chest wall; 17 (59%) (14
noncalcified masses and three calcific clusters) were for lesions inadequately
seen, unrelated to lesion depth, to proceed with the biopsy; and five (17%)
were in four patients who were unable to lie prone for a sufficient time for
the biopsy to be completed.
No cancellations occurred because the patient exceeded the biopsy table weight limit, the breast was too large, the patient had a bleeding diathesis, anticoagulant therapy could not be temporarily stopped, an implant was present in the breast [16], or the lesion was too small. We do not have data about biopsies that were temporarily postponed because of equipment failure, uncorrected anticoagulation problems, late arrival of the patient, temporary health problems (e.g., cough), health insurance problems, or other transient reasons.
The 29 lesions with biopsies canceled for technical reasons had three outcomes. First, 17 lesions were adequately seen for a successful later biopsy, without interval imaging. One lesion underwent a stereotactic biopsy at our facility 2 weeks after the canceled biopsy (when a radiologist with more biopsy experience performed the procedure) and was included in the 1780 lesions with a completed stereotactic biopsy. One lesion that was more evident on mammograms obtained with the patient in an upright position than in a prone position was successfully stereotactically biopsied with the patient in an upright position at another facility. Fifteen lesions had surgical biopsy after needle localization under mammographic (n = 14) or sonographic (n = 1) guidance at a median of 5 weeks (range, 1-33 weeks) after the canceled biopsy.
Second, 10 patients with lesions that were not seen adequately to proceed with biopsy using mammographic or sonographic guidance were recommended to have a mammogram 6 months later. Four of those 10 lesions underwent stereotactic biopsy at our facility because mammography performed at 7, 25, 46, or 84 months after the canceled biopsy showed lesion progression; all four were included in the 1780 lesions with a completed stereotactic biopsy. Six lesions not having a biopsy had stable mammographic follow-up (one each at 12, 29, 36, and 63 months; two at 37 months) and were assumed to be benign.
In the 21 lesions that underwent a subsequent biopsy, the final histologic diagnosis was malignant in 12 lesions, high risk in two lesions, and benign in seven lesions. The malignant lesions were invasive ductal carcinoma (n = 5, four with associated DCIS), invasive lobular carcinoma with associated DCIS (n = 1), and DCIS alone (n = 6). Thus, malignancy was found in 44% (12/27) of lesions with later successful biopsy (n = 21) or mammographic follow-up (n = 6). The malignant lesions were evenly split between masses (n = 6) and calcifications (n = 6).
The last two of the 29 lesions for which biopsies were canceled for technical reasons had neither biopsy nor mammographic follow-up. One lesion was in a patient whom we determined to be too ill to require further evaluation and the other, in a patient lost to follow-up. The relationship between the follow-up method and the reason for cancellation of the biopsy is shown in Table 2.
|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Large-core biopsies performed under sonographic guidance have several reported advantages over stereotactic biopsies. If appropriate equipment and expertise are available, those lesions visible on both sonography and mammography might be preferentially biopsied under sonographic guidance because, compared with stereotactic biopsy, the procedure costs less, requires less time to perform, is better tolerated by the patient, provides real-time needle visualization, and does not use ionizing radiation [17, 18]. Our sonographic machines were fully occupied with diagnostic imaging examinations during this study, with a scheduling backlog of several weeks. We did no sonographically guided biopsies of the breast or any other organs during this study.
Of the 1894 mammographically detected lesions referred for an imaging-guided biopsy, 43 (2.3%) were referred for a diagnostic needle-localized surgical breast biopsy and not a stereotactic biopsy. With our philosophy in 2002, which evolved during this study, a percutaneous stereotactic or sonographically guided biopsy is performed whenever possible for a mammographically detected lesion in need of an imaging-guided biopsy. An imaging-guided surgical breast biopsy is now performed without an antecedent percutaneous biopsy attempt only if specifically requested by the patient. In January 2002, after acquisition of an additional sonography scanner, we started doing sonographically guided biopsies with a handheld 11-gauge vacuum-assisted device [19]. This new strategy includes performing sonographically guided biopsies of some discretely palpable lesions of the type that underwent palpation-guided biopsies by surgeons during the study period.
Canceled Stereotactic Biopsies
Stereotactic breast biopsies were excluded ahead of time and were canceled
on the day of the scheduled biopsy for a variety of reasons. In
Table 3, we have eliminated
some excluded and canceled lesions from different series to make cancellation
criteria more comparable.
|
Biopsies were performed using a 14-gauge automated large-core technique in two series [6, 7]. Philpotts et al. [6] reported cancellation in 16% (89/572) of lesions scheduled for biopsy. Not counting the 23 lesions diagnosed as cysts and the 17 lesions reassessed as benign (other than a cyst), biopsies were canceled in 9% (49/532) of the remaining lesions in their series. Verkooijen et al. [7] did not include 14 lesions that were reassessed as benign before starting the biopsy procedure; they canceled biopsies in 13% (64/476) of the remaining lesions.
Biopsies were performed using an 11-gauge directional vacuum-assisted technique in a series by Philpotts et al. [8] that reported cancellations in 3% (36/1246) of lesions. Not counting the eight lesions reassessed as benign, the three lesions with equipment failures, and the three lesions in patients undergoing anticoagulant therapy, biopsies were canceled in 2% (22/1232) of the remaining lesions in that study.
We performed biopsies using 14-gauge automated large-core and 14- and 11-gauge directional vacuum-assisted techniques. We would have restricted patients from scheduling a stereotactic biopsy if they had weight exceeding the table limit, a bleeding diathesis, or anticoagulant use that could not be temporarily stopped, but we had no such patients. Excluding the 43 lesions referred for imaging-guided surgical biopsy, we canceled biopsies in 4% (71/1851) of lesions referred for stereotactic biopsy. Eliminating the 42 lesions reassessed as benign or probably benign, biopsies were canceled in 2% (29/1809) of the remaining lesions (Table 3).
When we combined the actual success rate in the 1809 lesions in which a stereotactic biopsy was thought to be necessary with the subjective feasibility rate in the 43 lesions referred for a needle-localized biopsy, a stereotactic biopsy was performed or thought to be feasible in 98% (1818/1852) of all lesions. Inability to accomplish a stereotactic biopsy in 2% (34/1852) of lesions was due to inadequate lesion visualization because of proximity to the chest wall (n = 10, 29%), inadequate lesion visualization unrelated to lesion depth (n = 19, 56%), and inability of the patient to lie prone long enough for the biopsy to be completed (n = 5, 15%). The 34 lesions were masses (n = 16, 47%), which represents 2.0% (16/803) of masses needing a biopsy, and calcifications (n = 18, 53%) which represents 1.7% (18/1048) of calcifications needing a biopsy. The remaining lesion referred for a surgical excision was a small wire suture that we would currently remove by vacuum-assisted biopsy [20].
Stereotactic Technical Problems and Advances
During the course of the 10-year period reviewed in our study, a variety of
stereotactic biopsy technical advances (in addition to the previously
mentioned change from performing automated large-core biopsies to performing
vacuum-assisted biopsies) have helped us complete and not cancel biopsies of
some lesions. Most of the techniques are thoroughly discussed and illustrated
in an article by Burbank [2].
We have opinions but no data on the effect of those changes.
Lesions in very thin breasts can be problematic, leading to what is termed a "negative stroke margin" (that is, the tip of the fired needle penetrates the skin on the far side of the breast and can potentially strike the image receptor). For a biopsy to be completed, the collection chamber near the tip of the needle must be completely inside the breast tissue. We use a variety of maneuvers to increase the breast thickness in the direction of needle penetration (z-axis direction). First, peripheral pressure on the breast tissue is applied to increase breast thickness by forcing breast tissue to protrude through the aperture on the front compression paddle which has been described as being performed with a woodworking C clamp [2, 21] or with a compressible elongated plastic sponge [22]. We have accomplished the same result by having a radiologist or technologist apply hand pressure to the periphery of the breast while a technologist compresses the breast between the front compression paddle and the image receptor. Second, use of a reversed compression paddle behind the breast (the air-gap technique) is also helpful in increasing the thickness of the breast tissue [2, 3]. Combining the air-gap technique with hand pressure to the periphery of the breast (between the two compression paddles) is extremely helpful in increasing breast thickness (Fig. 1A,1B).
|
|
In a very thin breast, it is helpful to advance the vacuum-assisted needle slowly into the breast in an uncocked (postfire) position until the patient feels pain deep in the breast opposite the needle entrance side or until the needle tip moves the skin on the deep side of the breast. A long, thin needle is then inserted deep into the breast, immediately adjacent to the biopsy needle, and local anesthetic is injected into the deep skin (from the inside of the breast rather than the outside). Tissue sampling can then begin.
For the past several years, one of the authors has used a surgical skin hook (Fig. 2A,2B) during each stereotactic biopsy. The skin hook is inserted into the small skin nick created with the tip of a scalpel blade. The hook is used in each breast to pull the skin of the breast in the direction of the biopsy needle being inserted into the breast to prevent the needle from indenting the breast tissue and potentially moving the lesion inside the breast. Continued traction of the skin along the shaft of the biopsy needle is used to effectively increase the z-axis thickness of the breast during some biopsies. No images are obtained with the hook in place, but the traction maneuver presumably slightly changes the z-axis depth of the lesion. This maneuver helps, rather than hinders, successful completion of the biopsy because it is used in breasts, including thin breasts, in which the lesion is close to the skin entry point of the probe to facilitate completely covering the collection chamber of the biopsy probe with breast tissue.
|
|
Despite these maneuvers, air—rather than tissue—can be suctioned when vacuum is applied to the probe. A final step of vigorously twisting the anesthetized, retracted skin around the shaft of the probe to create an airtight seal has eliminated the problem of any breast being too thin for a stereotactic biopsy to be completed. We do not know whether this technique originated in our facility or whether it was adopted at the suggestion of someone else. We are unaware of the technique being previously reported in the literature.
Lesions close to the chest wall are one of the major reasons for a canceled biopsy. The Fischer table positions the needle using polar coordinates [2]. Because the needle is angled up toward the chest wall, the needle tip can be positioned in any lesion that is visualized on images obtained with the patient in prone position on the biopsy table. The Lorad table (Danbury, CT) uses Cartesian coordinates for needle positioning. Because the needle is inserted parallel to the undersurface of the biopsy table, the needle tip cannot be positioned in lesions located in the extreme upper aspect of images acquired with the patient prone.
With either table, some techniques have been found that can assist in performing biopsies of lesions in close proximity to the chest wall. It is critical that the patient be positioned with a maximal amount of breast tissue protruding down through the aperture of the table. The patient must be in a relatively comfortable position with the muscles relaxed. The breast tissue must be firmly but relatively comfortably pulled downward from the chest wall and held in position by firm compression. An arm-through-the-hole technique [2, 3, 23] drops more breast tissue through the table aperture, especially for lesions positioned in the upper outer quadrant of the breast. We often use a target-on-scout technique with that maneuver because the lesion can be obscured by the protruding arm on one of the two 15° stereotactic images [2, 3]. A small compression plate allows one to compress the breast tissue without as much potential compression of adjacent nonbreast tissue; this plate is particularly helpful for lesions in the axillary tail of the breast, where the lesion is close to the humeral head. Rarely, we have removed the padding on the biopsy table and placed just a towel between the patient and the metal top of the table (Chavez J, personal communication). Doing so is less comfortable for the patient, but it has allowed us to complete biopsies in patients willing to tolerate the discomfort when other maneuvers have failed to visualize an extremely deep lesion.
Very faint calcifications can be difficult to visualize. A combination of
firm breast compression (to prevent motion) and low kilovoltage (
25 kVp,
to increase the conspicuity of the calcifications) is crucial. The quality of
the imaging system is important. We were able to visualize faint
calcifications with approximately equal ability using mammographic film or
using our original digital system with a black-and-white monitor. The period
of a few months in which we exclusively used a color monitor was the only
period in which we were unable to adequately visualize some of the faintest
calcifications seen on nonmagnification mammographic films obtained on an
upright mammography unit. We next worked with color and black-and-white
monitors next to each other for more than a year; it was subjectively quite
evident that very faint calcifications were seen only on the black-and-white
monitor. With the higher-resolution digital system we have used since the
completion of this study, we are able to visualize some faint calcifications
not evident even on magnification mammographic films. Although we biopsy only
those calcifications seen on nonmagnification mammographic films, the newest
digital system (and elimination of the color monitor) allows us to visualize
the calcifications needing biopsy with more certainty.
The main advantage of a digital receptor rather than a film receptor is the almost immediate display of the image, which decreases the time needed to complete each biopsy. The main disadvantage of the digital receptor is the small image size of 5 x 5 cm. We occasionally obtain a larger film image with the patient in the prone position, mainly if we are attempting to biopsy a small lesion in the periphery of a large breast. Once the lesion is identified, we reposition the breast and proceed with the biopsy using digital imaging.
We do not know if any difference exists in visualization of faint calcifications imaged on a Fischer table versus a Lorad table. Faint amorphous calcifications are the most difficult calcifications to visualize. Berg et al. [24] found a 20% (30/150) malignancy rate in clustered amorphous calcifications and reported on attempted stereotactic biopsies of 123 such lesions using a Lorad table. Those authors aborted biopsies for 4% (5/123) of lesions because of poor calcification visualization. Apart from our brief experience with a color monitor, we had no canceled biopsies due to poor visualization of calcifications.
Malignancy in Lesions with Canceled Biopsies
A moderate risk of malignancy is present in lesions for which a scheduled
stereotactic biopsy is canceled. Philpotts et al.
[6] reported that in 21%
(19/89) of the canceled biopsies, the appearance of the lesions warranted
biopsy with another method. Malignancy was found at immediate biopsy of 16%
(3/19) of those lesions. All malignancies were masses. Excluding the 19 cases
with immediate biopsy and 23 cysts, follow-up was accomplished in 38 of the
other 47 lesions. Mammographic findings of progression led to delayed biopsy
in three of the 38 lesions, with malignancy found in one lesion that was also
a mass.
Verkooijen et al. [7] found malignancy in 58% of 64 lesions for which stereotactic biopsies were canceled. This is comparable to the malignancy rate in the successfully completed biopsies. Data about lesion type were not reported.
In our practice, if a stereotactic biopsy is canceled, a choice is made between attempted biopsy by other means and follow-up. If the patient is unable to lie prone or the lesion is inaccessible by prone stereotactic biopsy, the BI-RADS category does not change and biopsy should be attempted by another method. If the lesion is inadequately visualized to complete a prone stereotactic biopsy, unrelated to chest wall proximity, all images are reviewed to decide which revised BI-RADS category to assign. A BI-RADS category of 4 or 5 is assigned, and biopsy is attempted by another method if the lesion is still suspicious enough to warrant a biopsy. BI-RADS category 1 is assigned if the area of concern is definitely proven to be a pseudolesion (such as superimposition of normal structures), and BI-RADS category 2 is assigned if the lesion is proven to be definitely benign (such as skin calcifications). If the lesion is not suspicious enough to warrant a biopsy attempt by other means but does not qualify as BI-RADS category 1 or 2, we advise assigning a BI-RADS category 3 and obtaining a 6-month follow-up mammogram.
If the choice of a revised BI-RADS category is not clear from the mammographic images alone, use of other imaging modalities may be helpful. We advise sonography for mass lesions with a canceled stereotactic biopsy, but we had only one such lesion that was adequately seen on sonography to complete a biopsy and no such lesions for which sonography was helpful in assigning a revised BI-RADS category. We do not know if MR imaging or radionuclide imaging of minimally evident mammographic lesions or sonographic imaging of minimally evident calcific clusters would help to complete a biopsy or to guide revised BI-RADS coding.
We also found malignancies in lesions with canceled biopsies. In 27 of 29 lesions with a canceled biopsy and a revised BI-RADS category 4 or 5, a later successful biopsy (n = 21) or mammographic follow-up (n = 6) was performed. Malignancy was found in 44% (12/27) of those lesions, which is higher than the 36% (688/1899) malignancy rate of all lesions in the study having either biopsy or follow-up.
Of 42 lesions with a canceled biopsy and a revised BI-RADS category of 2 or 3, a later successful biopsy (n = 3) or mammographic follow-up (n = 36) was performed in 39 lesions. Malignancy was found in 5% (2/39) of those lesions. The one lesion with a delayed successful stereotactic biopsy is shown in Figure 3A,3B,3C.
|
|
|
Miscellaneous Information
Excluding patients with the 30 lesions for which a surgical biopsy was
suggested because of preference of the surgeon (n = 15) or the
radiologist (n = 11) or during a 2-month transition period
(n = 4), patients with the other 1864 lesions were offered the choice
between a stereotactic biopsy and a needle-localized surgical biopsy. The
clinicians and radiologists were strongly supportive of avoiding a diagnostic
operation for lesions needing an imaging-guided biopsy. In that setting,
patients chose a stereotactic biopsy for 99.3% (1851/1864) of lesions and a
surgical biopsy for 0.7% (13/1864) of lesions. When the lesion was discretely
palpable and the patient was offered a choice of biopsy type, patients chose a
stereotactic biopsy for 93.3% (28/30) of lesions and a surgical biopsy for
6.7% (2/30) of lesions. We do not know what the patient choice would have been
if all patients with discretely palpable lesions, with or without need for
imaging guidance during biopsy, had been offered a choice of biopsy type.
A limitation of our study is the subjective nature of some of our data. It is difficult to categorize the reason for a biopsy's cancellation (especially in retrospect), the effect of incremental improved techniques on allowing completion of a difficult biopsy, and what revised BI-RADS category to assign to an inadequately visualized lesion (unrelated to lesion depth) with a canceled biopsy.
Our cancellation rates with the different biopsy devices are counterintuitive. When comparable cancellation criteria are applied (Table 3), Philpotts et al. canceled biopsies in 9.2% (49/532) of lesions with the large-core device [6] and 1.8% (22/1232) of lesions with the vacuum-assisted device [8]. We, however, canceled biopsies in 1.3% (8/607) of lesions with the large-core device and 1.7% (21/1202) of lesions with the vacuum-assisted device (Table 3). We think our unexpectedly higher cancellation rate with the vacuum-assisted device is explained by two factors. First, a higher percentage of patients with lesions needing biopsy bypassed an attempt at stereotactic biopsy and went directly to surgical biopsy during the large-core era (3.0% [19/626]) than during the vacuum-assisted era (2.0% [24/1226]) (Table 1). Second, use of the color monitor leading to canceled biopsies of some of the faint calcification lesions occurred during the vacuum-assisted era.
In conclusion, stereotactic biopsy, performed with the philosophy, equipment, and technical maneuvers we used, was technically feasible in 98% of lesions in which it was attempted and was the method used in 96% of mammographically detected lesions in which biopsy was warranted. Inadequate lesions visualization accounted for 85% of stereotactic biopsy failures. Lesions scheduled for a biopsy that is canceled have a moderate risk of being malignant and must either be biopsied by another method or carefully followed mammographically. In our experience, patients offered a choice between stereotactic biopsy or needle-localized surgical biopsy by a physician strongly supportive of avoiding diagnostic surgery for breast lesions requiring imaging-guided biopsy choose stereotactic biopsy 99.3% of the time.
Acknowledgments
We thank Laura Liberman for critical manuscript review.
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  N. Cho and W. K. Moon Digital Mammography-Guided Skin Marking for Sonographically Guided Biopsy of Suspicious Microcalcifications Am. J. Roentgenol., March 1, 2009; 192(3): W132 - W136. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. J. Jackman, F. A. Marzoni Jr., and J. Rosenberg False-Negative Diagnoses at Stereotactic Vacuum-Assisted Needle Breast Biopsy: Long-Term Follow-Up of 1,280 Lesions and Review of the Literature Am. J. Roentgenol., February 1, 2009; 192(2): 341 - 351. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. Zagouri, T. N. Sergentanis, and G. C. Zografos Lesions Too Close to the Chest Wall: A Relative Contraindication for Vacuum-Assisted Breast Biopsy Am. J. Roentgenol., April 1, 2008; 190(4): W270 - W270. [Full Text] [PDF]
|
|
|
|
|
|
S. V. Ghate, E. L. Rosen, M. S. C. Soo, and J. A. Baker MRI-guided vacuum-assisted breast biopsy with a handheld portable biopsy system. Am. J. Roentgenol., June 1, 2006; 186(6): 1733 - 1736. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. J. Jackman and J. Rodriguez-Soto Breast microcalcifications: retrieval failure at prone stereotactic core and vacuum breast biopsy--frequency, causes, and outcome. Radiology, April 1, 2006; 239(1): 61 - 70. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Liberman, N. Bracero, E. Morris, C. Thornton, and D. D. Dershaw MRI-Guided 9-Gauge Vacuum-Assisted Breast Biopsy: Initial Clinical Experience Am. J. Roentgenol., July 1, 2005; 185(1): 183 - 193. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I S Whitaker, B Elmiyeh, M N Siddiqui, and T C Holme Haemopneumothorax after fine needle aspiration of the breast J R Soc Med, November 1, 2003; 96(11): 555 - 556. [Full Text] [PDF]
|
|
|
|
|
|
L. Liberman, E. A. Morris, D. D. Dershaw, C. M. Thornton, K. J. Van Zee, and L. K. Tan Fast MRI-Guided Vacuum-Assisted Breast Biopsy: Initial Experience Am. J. Roentgenol., November 1, 2003; 181(5): 1283 - 1293. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
