Stereotactic Biopsy of the Breast Using an Upright Unit, a Vacuum-Suction Needle, and a Lateral Arm-Support System

Transcription

1 Dianne Georgian-Smith 1,2 Carl D Orsi 3 Ellen Morris 4 Crandon F. Clark, Jr. 5 Ellen Liberty 3 Constance D. Lehman 1 Received January 8, 2001; accepted after revision October 8, Presented at the annual meeting of the American Roentgen Ray Society, Seattle, April May Department of Radiology, University of Washington Medical Center, Box , 1959 N. E. Pacific, Seattle, WA Present address: Department of Radiology, Massachusetts General Hospital, Wang ACC 219-Q, 15 Parkman St., Boston, MA Address correspondence to D. Georgian-Smith. 3 Department of Radiology, University of Massachusetts Memorial Medical Center, Rm. 2144, 55 Lake Ave. N., Worcester, MA Department of Radiology, South Shore Hospital, 55 Fogg Rd., South Weymouth, MA Department of Radiology, Travis Air Force Base, CA AJR 2002;178: X/02/ American Roentgen Ray Society Stereotactic Biopsy of the Breast Using an Upright Unit, a Vacuum-Suction Needle, and a Lateral Arm-Support System OBJECTIVE. This study evaluated the vacuum-suction needle (8- to 11-gauge) on an upright stereotactic machine with a lateral arm-support system. SUBJECTS AND METHODS. From July 1999 to August 2000, 185 core biopsies of the breast in 179 consecutive patients were planned in four institutions using 8- to 11-gauge vacuumsuction probes on an upright stereotactic unit. Needles were stabilized and attached to the x, y, and z coordinates of the machines via a lateral arm-support system. Needle entry was made in the x-axis. RESULTS. Five patients were canceled, and 180 biopsies were performed in 174 patients while the patients were in seated (n = 171) and lateral decubitus (n = 9) positions. An average of 9.5 cores were taken (range, 5 26 cores). Targeting was successful in 176 (98%) of 180 cores. Lesions were missed because of movement associated with Parkinson s disease (n = 1), or because the mass was obscured (n = 1) or calcifications were not in the core specimen (n = 2). Findings in 152 (84%) of the biopsies were benign and 28 (16%) were malignant. Forty-one lesions underwent surgical excision and 106 underwent mammographic follow-up. Discordance was 4% (6 /147). Complications included vasovagal reactions (n = 10, 5.6%), bleeding (n = 5, 3%), hematomas (n = 3, 1.7%), vomiting (n = 1, 0.6%), and technical failure (n = 1, 0.6%). CONCLUSION. Vacuum-suction needle core biopsies can be successfully performed on an upright stereotactic machine with a lateral arm attachment. Thinly compressed breasts and lesions located near the chest wall are well sampled. The vasovagal rate is higher than that on a prone table but is acceptable. S tudies have validated the accuracy of stereotactic biopsy by several methods: concordance of core biopsy and surgical results for malignancy and atypical ductal hyperplasia, surgical results versus core biopsy in cases of mammographic discordance [1 20], and the false-negative rates of benign core biopsies with surgical results or surveillance [21]. However, these studies have primarily used prone stereotactic tables. Moreover, the advantages of vacuum-suction needles over spring-loaded core biopsy needles have also been evaluated, mostly on prone tables [8 20, 22, 23]. In August 1999, General Electric Medical Systems (Milwaukee, WI) released a lateral arm-support system for vacuum-suction needles to be used with upright, add-on stereotactic units. The purpose of our study was to evaluate this system. Subjects and Methods Between June 1999 and August 2000, 185 stereotactic core biopsies in 179 consecutive patients were scheduled at four institutions using the lateral armsupport system for vacuum-suction needles on upright, add-on units, consisting of a Senographe DMR with a Senovision or Stereotix 2 stereotactic unit (General Electric Medical Systems). The study set consisted of 180 completed core biopsies in 174 consecutive patients. Five patients had their procedures canceled for the following reasons: one patient in the lateral decubitus position experienced a vasovagal reaction as a result of severe neck pain before targeting and skin incision. Two patients were canceled because of excessive bleeding such that calcifications were not sampled; subsequent excisional biopsies yielded radial scar and ductal carcinoma in situ. One procedure was aborted after only two core samples were obtained because the patient had two vasovagal reactions. Because calcium was present in one core sample that showed fibrocystic change, the patient underwent mammographic follow-up; the calcifications were stable at 1 year. Finally, one patient was canceled because of equipment failure. No cores were obtained, and surgical biopsy showed a small hyalinized fibroadenoma. Biopsies were performed with the 11-gauge Mammotome probe (Ethicon Endo-Surgery, Cincinnati, OH) (n = 156) and with the 8-gauge Minimally Invasive Breast Biopsy needle (United States Surgical, Norwalk, CT) (n = 24). Core nee- AJR:178, April

2 Georgian-Smith et al. Fig. 1. Upright stereotactic unit with phantom. A, Photograph shows lateral arm-support device (black arrow) and vacuum-suction needle (open arrow) (Minimally Invasive Breast Biopsy; United States Surgical, Norwalk, CT) suspended upside down. Needle entry (white arrow) lies in x-axis parallel to image receptor and perpendicular to X-ray beam. B and C, Photographs show that lateral arm rotates around central point, allowing different approaches for skin insertion. A B dle biopsies were performed with the patients either sitting in a chair (n = 171) or reclining in a lateral decubitus position (n = 9) with the abnormal breast in the nondependent position. The x-, y-, and z-axes were determined by the standard techniques of stereo pair images and selection of the target by the radiologist. The lateral arm was attached to the stereotactic unit and had full mobility in each of the three axes, as well as axial rotation around a central point analogous to the mechanism of a compass (Fig. 1). Consequently, skin entrance could be varied. The lateral arm system supported the casing of the driver mechanism 1018 by suspending it upside down from a metal straight arm and by supporting the needle shaft with a sterile needle holder. Once the lesion s coordinates were determined, it was clamped into a fixed position. Needle insertion was made parallel to the image receptor in the x-axis of compression (Fig. 1). In contrast to the prone or standard upright units in which the X-ray beam is parallel to the direction of the needle, the X-ray beam is perpendicular to the longitudinal axis of the needle with this technique. The patient s skin was disinfected at the site of the determined skin entry, and local anesthetic was administered along the expected track of the nee- C dle. An incision was made with a scalpel blade parallel to the compression paddle. After needle insertion was complete, a pair of stereo views was obtained to document that the needle was either in a prefire position with the needle tip aiming at the target or adjacent to the lesion as in a postfire position (Figs. 2 and 3). To verify lesion sampling at core biopsy, first the sampling of calcifications was determined by core specimen radiographs showing calcifications (or a decrease in the number of calcifications between the radiographs obtained before and after biopsy) and concordance of pathology of the core biopsy results AJR:178, April 2002

3 Stereotactic Biopsy of the Breast A D B E C F Fig year-old woman with solid round ill-defined mass that was not visible on sonography. This patient underwent stereotactic core biopsy with lateral arm-support device. A, Preliminary scout mammogram shows mass (arrow) within confines of biopsy area. B and C, Prefire mammographic stereo pair views. Orientation of needle tip (short arrows) in relation to mass (long arrows) is similar because X-ray beam is perpendicular to needle path in both images. D and E, Postfire stereo pair views. Mass (long arrows) is at sampling notch (short arrows). F, Mammogram obtained after sampling shows biopsy cavity (short arrows) at mass (long arrows). Mass is now smaller. Pathology revealed infiltrating ductal carcinoma (not shown). with the mammographic findings. Second, sampling of masses, architectural distortion, or asymmetric densities were made by ghost tracks of air density in the lesions on postbiopsy radiographs, or postbiopsy clip placement within the lesion, and concordance of pathology and mammography results. The pathologic results of all core biopsies were reviewed. Surgical follow-up was recommended for patients in whom mammographic pathologic discordance was seen, patients in whom targeting was unsuccessful, and those with atypical ductal hyperplasia and malignant core results, ductal carcinoma in situ, or invasive ductal or lobular carcinoma core results. Comparison was made between the pathologic results of core and excisional tissues. Complications of infection, vasovagal reactions, hematomas, and bleeding were documented. Results We completed 180 biopsies in 174 consecutive patients whose ages ranged from 35 to 85 years (mean age, 54.6 years). We found 155 cal- AJR:178, April 2002 cifications, 16 masses, four masses with calcifications, three asymmetric densities, and two findings of architectural distortion. An average of 9.5 core samples (range, 5 26 samples) were taken from each patient; an average of 12 cores were taken from calcified lesions and 7.5 cores from those that were not calcified. Because needle sampling was taken in the plane parallel to compression, patients with thinly compressed breasts could still be biopsied. The smallest compression (smallest zaxis) recorded was 1.7 cm. Targeting and sampling of the lesion were successful in 176 (98%) of 180 cases. The four cases that were not successfully sampled were as follows. One of the cases counted as unsuccessful did not show calcifications in the core specimen radiographs. The histology was benign, and calcifications were present. Follow-up was with imaging rather than repeated biopsy, and the calcifications were sta- ble. Another case, terminated early because of bleeding, did not show calcifications at core specimen radiograph and pathology; the patient went on to surgical biopsy. One obese patient with Parkinson s disease and hemiplegia could not be immobilized in spite of body-wrapping the patient to the chair. A prone table biopsy was not an option for this patient because she could not lie on her side or her stomach without feeling asphyxiated. Core biopsies were taken, but calcifications were not obtained. This patient was followed up with imaging because of her surgical risk. At 12 months, the calcifications were stable. In the fourth case in which targeting could not be confirmed, the mass was obscured by lidocaine. Imaging at 6 months showed a stable residual mass. One hundred fifty-two (84%) of the core biopsies were benign (141 fibrocystic change, nine atypical ductal hyperplasia, one atypical 1019

4 Georgian-Smith et al. A B D C E F Fig year-old woman with cluster of pleomorphic calcifications who underwent stereotactic core biopsy with lateral arm-support device. A, Preliminary scout mammogram shows calcifications (arrow) in center of biopsy area. B and C, Prefire mammographic stereo pair views show calcification (long arrows) and needle tip (short arrows) relationship. D and E, Postfire mammographic stereo pair views show calcifications (long arrows) within sampling notch (short arrows). F, Mammogram obtained after sampling shows that calcifications have been completely removed and replaced by marking clip (long arrow) at edge of biopsy cavity (short arrows). Pathology revealed ductal carcinoma in situ (not shown). ductal hyperplasia and atypical lobular hyperplasia, and one lobular carcinoma in situ) and 28 (16%) were malignant (18 ductal carcinoma in situ, nine invasive ductal, and one invasive lobular carcinoma) (Table 1). Forty-one patients underwent surgical follow-up. Of the 28 malignant core results, 27 (96%) underwent surgical follow-up and one patient with ductal carcinoma in situ was lost to followup. Seventeen of the 27 cases were concordant pathologies. Two cases of invasive ductal were in situ at surgery (Table 2). Two cases of ductal carcinoma in situ were upgraded to infiltrating TABLE 1 ductal carcinoma, and six cases of ductal carcinoma in situ were benign at surgical pathology (four atypical ductal hyperplasia and two fibrocystic change) (Table 2). Thirteen cases that were benign after stereotactic biopsy went on to surgical follow-up. There was concordance in five cases (one lobular carcinoma in situ and four fibrocystic change). The cases of fibrocystic change included three cases of mammographic pathologic discordance and one case of papillomatosis, initially at core biopsy as well as surgical excision. Seven cases were atypical ductal hyperplasia after stereotac- Histology by Core Biopsy in 180 Procedures in 174 Patients Benign Findings No. Fibrocystic change Atypical ductal hyperplasia Atypical ductal and lobular hyperplasia Lobular carcinoma in situ Total 152 Malignant Findings Ductal carcinoma in situ Invasive ductal carcinoma Invasive lobular carcinoma No Note. Core biopsies of the breast were performed in four institutions using 8- to 11-gauge vacuum-suction probes on an upright stereotactic unit with a lateral arm-support system tic biopsy, of which two were upgraded to in situ ductal carcinoma and five were fibrocystic change after surgery (Table 2). The one case of mixed atypical ductal hyperplasia and atypical lobular hyperplasia was lobular carcinoma in situ at surgical pathology (Table 2). Of the 180 core biopsies, 139 were not followed up surgically. One patient with in situ ductal carcinoma was lost to follow-up. Two patients with atypical ductal hyperplasia did not have surgical follow-up. One was lost to follow-up because she was homeless; she had not returned to the hospital for medical care at 18 months after her needle biopsy. The second patient without follow-up was a poor surgical risk because of cardiac disease. The stereotactic biopsy in this patient was a follow-up core biopsy for architectural distortion 2 years after she had undergone a core stereotactic biopsy obtained with a 14-gauge spring-loaded core biopsy needle (Bard Magnum; Bard Urological, Covington, GA) that had also yielded atypical ductal hyperplasia. Mammographic follow-up continued, and findings were stable at 6 months after the vacuum-suction biopsy. AJR:178, April 2002

5 Stereotactic Biopsy of the Breast TABLE 2 Core Biopsy Versus Surgical Pathology Discordance in 20 Patients Core Biopsy Surgical Biopsy Findings Findings FCC ADH DCIS IDC LCIS FCC a 0 0 ADH DCIS IDC ADH and ALH Note. FCC = fibrocystic change, ADH = atypical ductal hyperplasia, DCIS = ductal carcinoma in situ, IDC = invasive ductal carcinoma, LCIS = lobular carcinoma in situ, ALH = atypical lobular hyperplasia. a Includes one case of calcifications that was FCC at core biopsy and DCIS at 6-month follow-up, and one case of missed calcifications and early termination at core biopsy because of excessive bleeding that was DCIS at surgery. TABLE 3 Study Summary of Stereotactic Core Biopsy Studies Using Upright Units Needle Gauge Note. N/A = not available. a Does not include insufficient samples at core biopsy. b Cases include all malignant and benign cases with surgical follow-up. Concordance is defined as matching benign and malignant histologies at core biopsy and surgery. Atypical ductal hyperplasia and lobular carcinoma in situ were considered as benign histologies in calculation. Malignant histologies did not distinguish in situ ductal from invasive ductal. c Cases of calcifications only, because results also included sonographic core biopsies of masses. d Cases include surgical follow-up for atypical ductal hyperplasia or cancer at core biopsy or mammographic pathologic discordance. e Needles were all spring-loaded except for the 11-gauge, which was vacuum-suction and was used in 16% (36/225) of cases. The remaining 136 cases were fibrocystic change at stereotactic biopsy; 63 had followup at 6 months, and two cases showed changes at imaging. One case was ductal carcinoma in situ (Table 2), and the second showed fibrocystic changes at surgical follow-up. None of the 43 cases that had 1-year follow-up needed surgical intervention. Of the 30 cases of fibrocystic change without follow-up (30/180,17%), 15 were lost to follow-up, 11 missed follow-up appointments, two cases of co-morbidity thwarted mammographic follow-up, and two cases were scheduled for follow-up at the time this article was written. In summary, of 180 core biopsies, follow-up occurred in 147 (82%). The stereotactic core biopsy discordance rate was 4% (6/147) (Table 2): four cases were benign at core biopsy and malignant at surgical pathology (two atypical ductal hyperplasia, one missed calcifications, and one in situ ductal at 6-month follow-up), Core Biopsies (No.) Sampling Success (%) Surgical Concordance (%) Caines et al. [1] 18, (96/116) a,b Doyle et al. [2] c 96 N/A Bolivar et al. [3] (108/124) b,c Hirst and Davis [4] (27/30) b Blue and Harman [5] (24/25) b Welle and Clark [6] (10/15) d Welle et al. [7] 14, 11 e d and two malignant cases were underestimated (two in situ ductal upgraded to invasive). Complications that were encountered included vasovagal reactions (n = 10, 5.6%), bleeding (n = 5, 3%), hematoma (n = 3, 1.7%), vomiting and dizziness (n = 1, 0.6%), and technical failure (n = 1, 0.6%). All of the vasovagal reactions occurred when the patients were in the sitting position. In one case, the patient developed a 6-cm hematoma 5 days after the completion of the study, when she resumed her anticoagulant therapy. She had only scant bleeding during the procedure, and hemostasis was quickly achieved at the completion of the study. The hematoma then became infected, and she underwent a 3-week course of antibiotics. Follow-up sonography at 6 months showed a seroma that was half as large as the initial one. The remaining two hematomas had occurred immediately at the time of the procedure and were treated with compression only. These cases were not included in the five cases of bleeding. Discussion The evolution of the lateral arm attachment began in Germany and was led by Joachim Teubner, who made a preliminary horizontal type in 1989 (Teubner J, personal communication). This attachment was introduced at the annual Radiological Society of North America meeting in November 1990 and was patented by Teubner in 1993 [24]. Further modifications and improvements resulted in the accommodation of larger and heavier vacuum-suction probes that weighed approximately 4.5 kg. The attachment received approval from the United States Food and Drug Administration in August The evolution of stereotactic core biopsy has progressively moved to larger samples to increase accuracy. This has been measured by the concordance of core pathology with surgical results, mostly the underestimation of in situ malignancy and invasive disease by core biopsy [1 20, 25, 26] and, recently, by the false-negative rate of benign core biopsies [21]. Concordance has increased as sample sizes have increased with the needle gauge and with the exchange of the springloaded for the vacuum-suction needle [22, 23, 27]. Until now, these advantages of the larger needles have been almost exclusively limited to their use on dedicated prone stereotactic tables because the large size of the vacuum-suction needle devices does not allow them to fit between the tube head and image receptor on some upright units. The prone tables are effective but limited, in that they can only be used for core biopsy and not for standard mammography, they require more space than standard mammography machines, and they are more costly than add-on stereotactic units. Therefore, smaller practices have found upright units to be more cost-effective. Unfortunately for these practices, the vacuum-suction needles were not easily adaptable to the upright units or could not fit between the tube head and the breast when needle entry was in the z-axis. This study was performed to evaluate the performance of the new lateral arm with vacuum-suction needles on add-on upright stereotactic units. Few studies in the literature used add-on upright machines. In 1989, Evans and Cade [28] and Fajardo et al. [29] described their initial investigations of stereotactically guided fine- AJR:178, April

6 Georgian-Smith et al. needle aspirations. Since then, seven studies using primarily spring-loaded core biopsy needles have described successful targeting in % of their biopsies and 67 97% agreement between core biopsy and surgical pathologic results (Table 3). The largest series of core biopsies on only upright units was by Hirst and Davis [4], which included 297 cases of calcifications. These researchers found no false-positive cases by surgical follow-up in 22 malignancies at core biopsy. At the time that Hirst and Davis submitted their data, all 275 benign cases remained benign at follow-up and at surgery (n = 6), although the time interval of follow-up was not specifically noted for the entire cohort. However, an addendum to the study [4] stated that at 2 years after its completion, 215 of 275 benign core biopsies remained benign. The average time and range in months for follow-up were not noted. These investigators did note two cases of benign core results that were subsequently found to be malignant, but they did not mention in how many additional needle biopsies were performed. In a recent study with add-on upright equipment, investigators successfully used the 11-gauge Mammotome probe in 36 of their total 225 procedures, but the patients were studied in the lateral decubitus position [7]. The authors noticed that there was improved concordance with the Mammotome compared with the 14- gauge tru-cut needle: none of seven cases of atypical ductal hyperplasia or in situ ductal carcinoma were upgraded at surgery. However, the sample size of only 36 patients was small. Moreover, their images show that there is little clearance space for the Mammotome and tube head. We question whether this fact limits its application for large-breasted women. The study by Parker et al. [8] of 103 patients included a mixture of upright (n = 30) and prone (n = 73) biopsies and used different needle gauges, from 14- to 18-gauge [8]. For the entire study, these authors found a core versus surgical agreement of 87%. However, when the 14-gauge cases performed on the prone table were singled out, there was 97% agreement in 29 cases that underwent surgery. Many similar studies followed [9 15]. Jackman et al. [26] had one of the largest studies 450 cases. There was concordance in 85% (99/116); 69% (11/16) of core biopsies with atypical ductal hyperplasia were discordant with surgical excision using a 14-gauge spring-loaded needle. The introduction of vacuum-suction needles with 14-gauge needles and then 11-gauge needles improved on concordance as the size of tissue sampled increased [14 20]. Burbank [14] took an average of 26 core samples and found no underestimation of disease. Burbank found no cases of atypical ductal hyperplasia (n = 8) at core biopsy and cancer at surgery, and no cases of ductal carcinoma in situ (n = 32) at core biopsy and invasion at surgery. Other investigators [15, 19, 20] using 11- gauge vacuum-suction needles showed increased concordance over 14-gauge springloaded core biopsy needles for atypical ductal hyperplasia and ductal carcinoma in situ but found enough underestimation of disease to continue to recommend surgical excision. Our results, obtained by using both 11- gauge and 8-gauge vacuum-suction needles, were in keeping with those studies using the same needles and prone tables. Successful sampling of the lesions occurred in 98% of cases. We found six cases (4%, 6/147) of underestimation of malignancy or invasion by benign (includes atypical ductal hyperplasia) and in situ ductal disease at core biopsy. Liberman et al. [27] reported 4% discordance using 14-gauge vacuum-suction needles and 2% discordance using 11-gauge vacuum-suction needles. Alternatively, the 96% concordance rate in our study is in keeping with many of the studies that used vacuum-suction probes, whose results ranged from 75% to 100% [14 20]. Therefore, stereotactic vacuum-assisted biopsy can be successfully performed with patients in the upright position. One advantage of the lateral arm approach for vacuum-suction needles over the one described by Welle et al. [7] is that very thin breasts can still be sampled. In this study, a cluster of calcifications was completely removed in a patient who compressed to 1.7 cm. Compression of 1.3 cm or less requires placement of a Plexiglas block between the breast and the image receptor. This is done to elevate the breast, because there is distance in the z-axis between the needle and the edge of the metal casing. Without elevation of a very thin breast to a z-compression of at least 1.3 cm, the needle would not be able to be lowered in the z-axis without the casing hitting the image receptor. Welle et al. [7] used a standard z-axis approach perpendicular to the image receptor with vacuum-suction probes and did not comment on whether they encountered cases that could not be performed because of compression that was too thin. Unlike the Mammotome probe, the Minimally Invasive Breast Biopsy probe has a plastic outer cannula that can be advanced to cover the skin in cases in which the compression of the breast is smaller than the biopsy notch or in cases in which the lesion is so superficial that there may be excision of skin with biopsy of a lesion when using a standard z-axis approach. The drawback, however, is that this outer cannula then renders the biopsy notch smaller than usual. This limitation is rarely encountered when using the lateral approach. With the lateral approach, biopsies can be performed without decreasing the size of the biopsy notch, unless the lesions are very superficial, because the biopsy access is in the x-axis and the breast is therefore maximally expanded. An advantage of the lateral approach is that the administration of lidocaine for the skin incision is made to the side of the lesion and not superimposed on the target, as occurs in the z-axis approach, resulting in obscuration. Our study found only one case, a mass, in which sampling could not be confirmed at the time of biopsy because of lidocaine administration. Short-term followup at 6 months did confirm stability, however. In general, obscuration by lidocaine may be less a factor with the lateral approach than in the standard, z-axis approach. Another benefit of the upright unit with the lateral arm is that it allows access to lesions near the chest wall. One patient in this series had an 8-mm nodule juxtaposed to the pectoral muscle in the craniocaudal view and superimposed on the muscle in the mediolateral oblique view. The nodule was not visible on sonography and was not accessible on the prone table. This nodule was removed by the vacuum-suction needle on the upright unit without complication; it was also confirmed on the short-term follow-up mammogram. The histology report indicated benign ductal hyperplasia. Use of the vacuum-suction needle over a spring-loaded core biopsy needle is indicated when clip placement is important for possible surgical follow-up and when confirmation of sampling of a mass or asymmetric density is desired. Being able to place clips now with upright units is a considerable improvement when performing stereotactic biopsies. One disadvantage of the upright units may be a higher frequency of complications, particularly vasovagal reactions. Of the 180 biopsies in our study, 5.6% (10/180) involved vasovagal reactions, and all of them occurred with the patient in the sitting position. However, we did have one vasovagal reaction in a patient in the lateral decubitus position, but because it occurred in association with neck pain before the start of the procedure, this 1022 AJR:178, April 2002

7 Stereotactic Biopsy of the Breast was not included as one of the ten vasovagal reactions. Reports of two early studies in which patients were positioned on prone tables specifically mentioned that there were no vasovagal episodes [9, 10]. In 1990, Parker et al. [8] noted two (2/30, 6.7%) episodes of vasovagal reaction that occurred when the patient was in the upright position and none in the prone position. Similarly, Liberman et al. [18] used an 11-gauge vacuum-suction needle and described one complication each of bleeding, vasovagal reaction, and nausea in 112 patients. In contrast, far more incidences of vasovagal reactions have been reported in the literature when upright add-on units were used. Welle and Clark [6] and Welle et al. [7] reported this complication in 37.5% (3/8) and in 20% (4/20) of patients, respectively, who were biopsied in the sitting position. Similarly, Blue and Harman [5] reported a 20% rate of vasovagal reactions (5/25), but Hirst and Davis [4] noted only 5%, of which 2% resulted in cancellation of the procedure (n = 683, including both stereotactic and sonography-guided core biopsies). Caines et al. [1] reported the lowest frequency of vasovagal reactions, at 0.3% (4/125). The 5.6% (10/179) rate of vasovagal reactions in our study, with all episodes occurring in patients who were sitting (n = 170), is better than that of several studies in which biopsies were performed using upright equipment, but it is higher than what has been noted for the prone table. For this reason, other researchers have advocated performing the procedure with the patient in the lateral decubitus position [6, 7]. In our experience, one limitation of the lateral decubitus position is that, when using the lateral arm system for vacuum-suction needles, access to lesions is best from an inferior approach. An approach that is perpendicular to the table top with the currently designed lateral arm may result in needle puncture and advancement that is not controlled. The vacuum-suction probe combined with the casing is large and heavy, approximately 22 kg, and it is best used when it is hanging in a horizontal position (Fig. 1). Moreover, it is advisable to avoid passing the lateral arm and needle in front of the patient s face. Therefore, one is precluded from using a superior approach, leaving the lateral decubitus position for inferior approaches. Perhaps the next generation vacuum-suction needles will be more versatile if they are smaller and lighter. Our study found infrequent bleeding (3%) and hematomas (1.7%) when taking an average of 9.5 cores samples using 8- to 11-gauge vacuum-suction needles. The literature stresses that there are usually no noteworthy complications requiring medical intervention when 11- to 14- gauge vacuum-suction needles and dedicated tables are used [14, 16, 23]. Liberman et al. [30] and Heywang-Köbrunner et al. [16] described hematomas on post core biopsy mammograms at 60% and 16%, respectively. Clinically evident hematomas were only observed in 4% of 108 cases [30]. In studies performed on add-on units with 14-gauge spring-loaded needles, Hirst and Davis [4] reported a 0.9% (6/683) rate of hematomas. Welle et al. [7] noted only one hematoma of 225 biopsies but did not comment on the needle type because both 14-gauge spring-loaded and 11-gauge Mammotome devices were used in their series. In general, hemostasis problems can occur on upright units as well as on dedicated tables, but in relatively low frequencies. When they do occur, they can be treated with minimal care, including local pressure, ice to the incision, and pressure wraps if needed. Patients are advised to discontinue warfarin and nonsteroidal antiinflammatory medications, with the knowledge of their primary health care providers, and are advised not to take the herbal supplement gingko biloba leaf extract, which is a known antagonist to the platelet-activation factor [31]. In spite of these measures, one patient in our study developed a hematoma 1 week after an uneventful and successful biopsy when she was restarted on her anticoagulant medication. In summary, by the methods previously applied to the evaluation of stereotaxis on prone tables using vacuum-suction needles, this lateral arm attachment for the vacuum-suction needle has been shown to be as effective in targeting and biopsying lesions as the prone method. Assessment of the needle-tip location in relation to the lesion is easy with this attachment because the X-ray beam is perpendicular to the needle. Women with thinly compressed breasts and posteriorly located lesions are well served by this approach. The shortcomings of movement and vasovagal reactions may occur, but they rarely affect outcomes; some episodes might be prevented by using a specially designed table or gurney for a lateral decubitus position. Next generation vacuum-suction needles that are lighter and smaller will improve feasibility of their use with upright units. Acknowledgments We acknowledge Heather Cameron, who assisted in the preparation of the manuscript, and Elena DiPardo and Billie Stober, who provided valuable assistance. References 1. Caines JS, McPhee MD, Konok GP, Wright BA. Stereotaxic needle core biopsy of breast lesions using a regular mammographic table with an adaptable stereotaxic device. AJR 1994;163: Doyle AJ, King AR, Miller MV, Collins JP. Implementation of image-guided large-core needle biopsy of the breast on a limited budget. Australas Radiol 1998;42: Bolivar AV, Garcia EO, Ayensa FG. Stereotaxic core needle aspiration biopsy with multiple passes in nonpalpable breast lesions. Acta Radiol 1998;39: Hirst C, Davis N. Core biopsy for microcalcifications in the breast. Aus N Z J Surg 1997;67: Blue J, Harman J. Experience of the upright breast stereotactic core biopsy method and histological correlation with surgical biopsy. N Z Med J 1998;111: Welle GJ, Clark ML. Adaptation of an add-on stereotaxic breast biopsy unit: use of a dedicated reclinable mammography chair. AJR 1997;169: Welle GJ, Clark M, Loos S, et al. Stereotactic breast biopsy: recumbent biopsy using add-on upright equipment. AJR 2000;175: Parker SH, Lovin JD, Jobe WE, et al. Stereotactic breast biopsy with a biopsy gun. Radiology 1990;176: Elvecrog EL, Lechner MC, Nelson MT. Nonpalpable breast lesions: correlation of stereotaxic large-core needle biopsy and surgical biopsy results. Radiology 1993;188: Gisvold JJ, Goeliner JR, Grant CS, et al. Breast biopsy: a comparative study of stereotaxically guided core and excisional techniques. AJR 1994;162: Mikhail RA, Nathan RC, Weiss M, et al. Stereotactic core needle biopsy of mammographic breast lesions as a viable alternative to surgical biopsy. Ann Surg Oncol 1994;1: Brenner RJ, Fajardo L, Fisher PR, et al. Percutaneous core biopsy of the breast: effect of operator experience and number of samples on diagnostic accuracy. AJR 1996;166: Meyer JE, Christian RL, Lester SC, et al. Evaluation of nonpalpable solid breast masses with stereotaxic large-needle core biopsy using a dedicated unit. AJR 1996;167: Burbank F. Stereotactic breast biopsy of atypical ductal hyperplasia and ductal carcinoma in situ lesions: improved accuracy with directional, vacuumassisted biopsy. Radiology 1997;202: Darling ML, Smith DN, Lester SC, et al. Atypical ductal hyperplasia and ductal carcinoma in situ as revealed by large-core needle breast biopsy: results AJR:178, April

8 Georgian-Smith et al. of surgical excision. AJR 2000;175: Heywang-Köbrunner SH, Schaumlöffel U, Viehweg P, Höfer H, Buchmann J, Lampe D. Minimally invasive stereotaxic vacuum core breast biopsy. Eur Radiol 1998;8: Seoudi H, Mortier J, Basile R, Curletti E. Stereotactic core needle biopsy of nonpalpable breast lesions: initial experience with a promising technique. Arch Surg 1998;133: Liberman L, Smolkin JH, Dershaw DD, Morris EA, Abramson AF, Rosen PP. Calcification retrieval at stereotactic, 11-gauge, directional, vacuum-assisted breast biopsy. Radiology 1998;208: Brem RF, Behrndt VS, Sanow L, Gatewood OMB. Atypical ductal hyperplasia: histologic underestimation of carcinoma in tissue harvested from impalpable breast lesions using 11-gauge stereotactically guided directional vacuum-assisted biopsy. AJR 1999;172: Won B, Reynolds HE, Lazaridis CL, Jackson VP. Stereotactic biopsy of ductal carcinoma in situ of the breast using an 11-gauge vacuum-assisted device: persistent underestimation of disease. AJR 1999;173: Jackman RJ, Nowels KW, Rodriguez-Soto J, Marzoni FA Jr., Finkelstein SI, Shepard MJ. Stereotactic, automated, large-core needle biopsy of nonpalpable breast lesions: false-negative and histologic underestimation rates after long-term follow-up. Radiology 1999;210: Philpotts LE, Shaheen NA, Carter D, Lange RC, Lee CH. Comparison of rebiopsy rates after stereotactic core needle biopsy of the breast with 11-gauge vacuum suction probe versus 14-gauge needle and automatic gun. AJR 1999;172: Jackman RJ, Marzoni FA Jr, Nowels KW. Percutaneous removal of benign mammographic lesions: comparison of automated large-core and directional vacuum-assisted stereotactic biopsy techniques. AJR 1998;171: Teubner J, DeBartolomeo JY, Rouchy D, inventors. General Electric CGR S.A., France, assignee. Mammograph provided with an improved needle carrier. U.S. patent 5,219,351. Washington, DC: U.S. Patent Office, Liberman L, Cohen MA, Dershaw DD, Abramson AF, Hann LE, Rosen PP. Atypical ductal hyperplasia diagnosed at stereotaxic core biopsy of breast lesions: an indication for surgical biopsy. AJR 1995;164: Jackman RJ, Nowels KW, Shepard MJ, Finkelstein SI, Marzoni FA Jr. Stereotaxic large-core needle biopsy of 450 nonpalpable breast lesions with surgical correlation in lesions with cancer or atypical hyperplasia. Radiology 1994;193: Liberman L, Drotman M, Morris EA, et al. Imaging-histologic discordance at percutaneous breast biopsy: an indicator of missed cancer. Cancer 2000;89: Evans WP, Cade SH. Needle localization and fine-needle aspiration biopsy of nonpalpable breast lesions with use of standard and stereotactic equipment. Radiology 1989;173: Fajardo LL, Davis JR, Wiens JL, Trego DC. Mammography-guided stereotactic fine-needle aspiration cytology of nonpalpable breast lesions: prospective comparison with surgical biopsy results. AJR 1990;155: Liberman L, Hann LE, Dershaw DD, Morris FA, Abramson AF, Rosen PP. Mammographic findings after stereotactic 14-gauge vacuum biopsy. Radiology 1997;203: Blumenthal M. Ginkgo biloba leaf extract. In: Blumenthal M, ed. The complete German commission E monographs: therapeutic guide to herbal medicines. Boston: American Botanical Council, Austin, Texas; Integrative Medicine Communications, 1998: AJR:178, April 2002