Women s Imaging Original Research

Women s Imaging Original Research Villa et al. Biopsy of Atypical Ductal Hyperplasia Women s Imaging Original Research WOMEN S IMAGING Alessandro Villa 1 Alberto Tagliafico 2 Fabio Chiesa 1 Maurizio Chiaramondia 3 Daniele Friedman 4 Massimo Calabrese 2 Villa A, Tagliafico A, Chiesa F, Chiaramondia M, Friedman D, Calabrese M Keywords: atypical ductal hyperplasia, follow-up, microcalcifications, underestimation rate, vacuumassisted breast biopsy DOI:10.2214/AJR.11.6588 Received January 23, 2011; accepted after revision March 25, 2011. 1 Department of Radiology, DICMI-University of Genova, Largo Rosanna Benzi 5, Genova 16132, Italy. Address correspondence to A. Villa (dr.willa@libero.it). 2 Department of Radiology, National Institute for Cancer research (IST), Genova, Italy. 3 Department of Pathologic Anatomy, University of Genova, Genova, Italy. 4 Department of Breast Surgery, University of Genova, Genova, Italy. AJR 2011; 197:1012 1018 0361 803X/11/1974 1012 American Roentgen Ray Society Atypical Ductal Hyperplasia Diagnosed at 11-Gauge Vacuum- Assisted Breast Biopsy Performed on Suspicious Clustered Microcalcifications: Could Patients Without Residual Microcalcifications Be Managed Conservatively? OBJECTIVE. The purpose of our study was to establish whether it might be safe for women with a diagnosis of atypical ductal hyperplasia (ADH) at stereotactically guided vacuum-assisted breast biopsy without any residual microcalcification after the procedure to undergo mammographic follow-up instead of surgical biopsy. MATERIALS AND METHODS. From October 2003 to January 2009, 1173 consecutive 11-gauge vacuum-assisted breast biopsy procedures were performed. ADH was found in the specimens of 114 patients who underwent vacuum-assisted breast biopsy for a single cluster of suspicious microcalcifications smaller than 15 mm; 49 had residual microcalcifications, and 65 had microcalcifications completely removed by the procedure. Of 49 patients with residual microcalcifications, 41 underwent surgical biopsy. Of 65 patients without residual microcalcifications, 26 underwent surgical biopsy, 35 were not surgically treated and were managed conservatively with mammographic follow-up, and 4 had follow-up of less than 24 months. RESULTS. In 41 patients with residual microcalcifications who underwent surgical biopsy, 8 malignant lesions were found at surgery. The underestimation rate was 20% (8/41). In 26 patients without residual microcalcifications who underwent surgical biopsy, no malignant lesions were found. One malignant lesion was found in the 35 patients managed conservatively at follow-up. The underestimation rate in patients without residual microcalcifications using surgical biopsy or mammographic follow-up as the reference standard was 1.6% (1/61). CONCLUSION. Patients without residual microcalcifications after vacuum-assisted breast biopsy could possibly be managed in a conservative way with mammographic follow-up. A typical ductal hyperplasia (ADH) is a proliferative breast lesion considered to increase a woman s lifetime risk of developing breast carcinoma [1]. Increased risk of breast cancer in women with biopsy-proved ADH has been reported to be 4 5 times that of the general population at 10 years and even higher up to 11 times at 10 years in women with a family history of breast cancer [1]; however, this concept is still controversial [2]. Pathologically, ADH can be defined as a ductal proliferation with some but not all of the features of a low or intermediate grade of ductal carcinoma in situ (DCIS) [3]. The diagnostic criteria commonly used to establish a diagnosis of ADH are the following (Fig. 1): ducts completely filled and exhibit sharp punched-out spaces or micropapillae but lack uniform cytologic features, ducts are filled by a uniform population of cells with cytologic features of low-grade DCIS but lack architectural features, and cytologic and architectural features met but failure to meet size criteria (fewer than two ducts involved or less than 2 mm in aggregate dimension) [1, 4, 5]. Many authors have sought possible pathologic or clinical radiologic predictors to reduce the ADH underestimation rate to under 2% using 14-gauge core biopsy or 14-, 11-, or 9-gauge vacuum-assisted breast biopsy [6 15]. Several studies have found subgroups of patients that might be treated conservatively when ADH involves fewer than three foci and microcalcifications are completely removed at needle biopsy [13, 15]. Other studies, although showing a correlation between radiopathologic predictors and the underestimation rate of ADH, found that the underestimation rate remained at more than the 2% threshold and for these reasons surgical excision was recommended for biopsy-proved ADH [7, 8, 16]. Previously reported data are 1012 AJR:197, October 2011

Biopsy of Atypical Ductal Hyperplasia Fig. 1 Atypical ductal hyperplasia. A, Photomicrograph of lesion involving single ductlike structure in 56-year-old woman shows architectural features typical of ductal carcinoma in situ (cribriform with rigid epithelial bridges) but lacks cytologic characteristics (monomorphous cell population). (H and E, 200) B, Photomicrograph of lesion involving multiple ductules of single terminal ductal-lobular unit in 52-year-old woman shows heterogeneous proliferation of cells arranged with micropapillary architecture. Lumen of few ductules is occupied by detached cell clusters. (H and E, 150) not concordant; therefore, our aim was to evaluate the underestimation rate of ADH in those patients who did not have residual calcifications after 11-gauge vacuum-assisted breast biopsy and, specifically, we attempted to establish whether in this subgroup of the population the underestimation rate was lower than 2% the risk of malignancy that is widely accepted to define lesions as probably benign and for which a short-term imaging follow-up would be appropriate as an alternative to tissue sampling. Materials and Methods Patients and Procedures Our institutional review board approved this retrospective study. The electronic database of the pathologic anatomy unit of our institution (University of Genova) was accessed and histologic reports of high-risk lesions were reviewed to search for biopsy-proved ADH. From October 2003 to January 2009, 1173 consecutive stereotactic procedures were performed with an 11-gauge directional vacuum-assisted breast biopsy device (Mammotome, Ethicon Endo-Surgery). A total of 167 high-risk lesions were found: 118 were classified as ADH, 25 as flat epithelial atypia, 10 as lobular neoplasia, 4 as radial scar, and 10 as papillary lesions (Fig. 2). All diagnoses of ADH were made by a pathologist with 20 years of experience in breast pathology using established criteria as defined by Page et al. [1], Page and Rogers [3], and Tavassoli and Norris [4]. A lesion was designated as ADH when the stereotactic vacuum-assisted breast biopsy pathology report indicated the presence of ADH, isolated or accompanied by other high-risk lesions such as lobular neoplasia, flat epithelial atypia, or radial scar, excluding lesions with concomitant DCIS foci or invasive carcinoma. This was in agreement with criteria previously used by other authors [1, 3, 8] Only patients who underwent stereotactic biopsy for a single cluster of suspicious microcalcifications were included in the study. The mean patient age was 57 years (age range, 38 76 years). Of 118 ADH lesions, 114 corresponded to a single cluster of microcalcifications. Four cases of ADH were excluded from the study because of a multiple cluster of microcalcifications. Mammographic and Biopsy Evaluation Preliminary image acquisitions and successive spot magnifications were performed by two fullfield digital mammography units (Giotto Image SD Full-Field Digital Mammography Unit, IMS Italy). Each microcalcification cluster was retrospectively reviewed on a high-resolution digital mammographic screen by a radiologist with 15 years of experience in breast radiology, and the Fig. 2 Graph illustrates pathologic results of lesions categorized as high-risk lesions. ADH = atypical ductal hyperplasia, FEA = flat epithelial atypia, LN = lobular neoplasia, RAD SCAR = radial scar, PAP = papillary lesion. No. of Lesions at Pathology 120 100 80 60 40 20 0 ADH 118 radiologic findings were classified according to the BI-RADS category system [17] for patients in which this information was missing in the original reports. In addition, the size of each calcification cluster was recorded and the mean was calculated. Stereotactic vacuum-assisted breast biopsy procedures were performed in the prone position on a dedicated biopsy table (Mammotome, Ethicon Endo-Surgery) by the same radiologist who categorized the microcalcification clusters. Written informed consent was obtained from all patients. Each microcalcification cluster was stereotactically centered and a small amount of local anesthetic was injected. Tissue sampling of single microcalcification clusters was performed by 11-gauge vacuum-assisted breast biopsy and a mean of 12 specimens (range, 9 16) were obtained. Radiographs of every pathologic specimen were obtained to confirm the presence of microcalcifications. After biopsy, each specimen was placed in standard biocassettes for tissue sample embedding, formalin fixed, and sent to the pathologist for histologic examination (Fig. 3). Three levels were cut in each core specimen. If calcifications were not detected in any of the core samples in which they were visualized by the radiograph of the specimens or for other reasons, such as suspicious areas, additional deeper-level specimens (up to five levels) were obtained by the original pathologist. Upgrade occurred every time the biopsy pathologic reports yielded ADH and DCIS or infiltrating ductal carcinoma (IDC) was found at surgery. After the vacuum-assisted breast biopsy procedure, mammography in upright position was performed to ascertain the presence or absence of residual microcalcifications and to assess the correct placement of a metallic marker in the biopsy site (Fig. 4). The absence of residual microcalcifications was established when not even a single microcalcification was detected when comparing the prebiopsy mammogram with the postbiopsy mammogram. FEA 25 LN 10 RAD 4 PAP 10 AJR:197, October 2011 1013

Villa et al. Fig. 3 Specimens for radiography check to detect microcalcifications. A, Twelve biopsy specimens were placed like clock hours to correlate with core specimen radiographic examinations. B, Magnification radiograph of biopsy specimens was obtained to assess presence of microcalcification. Study Design The patients were subdivided into three groups: patients surgically treated without (group A) and with (group B) residual microcalcifications and patients with completely removed microcalcifications that were managed conservatively with mammographic follow-up (group C) as an alternative to surgical excision. The underestimation rate of ADH was calculated for the subgroup of patients with no residual microcalcifications after vacuum-assisted breast biopsy and surgically treated (group A) and was compared with the ADH underestimation rate of patients still with microcalcifications after vacuumassisted breast biopsy and surgically treated (group B). The underestimation rate of ADH was also evaluated for the aggregated subgroup of patients without residual microcalcifications both surgically treated and managed conservatively (groups A and C) and was compared with the underestimation rate Fig. 4 Mammogram obtained after vacuumassisted breast biopsy in 49-year-old woman to assess presence of some residual calcification and establish correct position of metallic marker. TABLE 1: Comparison Between Groups A and B All ADH (n = 114) Group A vs Group B of the patients in group B. Furhermore, we tried to establish whether in the subgroup of the population with no residual microcalcifications after vacuumassisted breast biopsy procedures (group A and aggregated groups A and C) the underestimation rate was lower than 2%. In group B, eight patients were excluded because they were surgically treated in other institutions and pathologic reports could not be retrieved; in group C the length of follow-up ranged from 6 to 72 months, with a median length of 53 months. Four patients had a follow-up period of less than 24 months (6, 12, 16, and 17 months, respectively), which is the minimum imaging follow-up deemed adequate to define a lesion as benign or not evolutive. Thus, these patients were excluded. In group A, all patients underwent surgery. Pathologic reports after surgery and follow-up information for patients who did not undergo surgical excision were used as the reference standard. The relationship between BI-RADS category and underestimation rate of ADH was also evaluated. Statistical Analysis Statistical analysis was performed with commercially available dedicated software (SAS version 9.1.3, SAS Institute). Comparisons were made between group A and group B and between the aggregated groups A and C and group B with respect to the presence of absence of residual calcifications and presence or absence of malignant lesions at surgical biopsy or follow-up (discrete dichotomous variables).the underestimation rate of ADH for each group was calculated. A possible association with BI-RADS categories and underestimation rate of ADH was also assessed. Differences between groups were analyzed with the Fisher exact test for discrete variables. A p value < 0.05 was considered to represent statistical significance. Results Of the 102 patients with ADH, 61 had microcalcifications completely removed by the vacuum-assisted breast biopsy procedure; the remaining 41 had residual microcalcifications visualized on postbiopsy mammograms. Of the 61 patients with no residual microcalcifications, 26 underwent surgical excision (group A); none had a lesion upgrade compared with stereotactic biopsy. Surgery was not performed in 35 patients (group C) because the referring physician did not recommend it to the patients (n = 8), unrelated patient clinical conditions made surgery With Surgery DCIS or IDC No Upgrade Underestimation Rate (%) Group A Without residual microcalcifications 26 0 26 0 Group B With residual microcalcifications 41 8 33 20 Total 67 8 59 Note Group A composed of patients without residual microcalcifications after vacuum-assisted breast biopsy, and group B composed of patients with residual microcalcifications. Both groups were surgically treated. ADH = atypical ductal hyperplasia, DCIS = ductal carcinoma in situ, IDC = infiltrating ductal carcinoma. For group A versus group B, p = 0.019. 1014 AJR:197, October 2011

Biopsy of Atypical Ductal Hyperplasia TABLE 2: Comparison Between Groups A and C Versus Group B With Surgery or Follow-Up All ADH (n = 114) Group A+C vs Group B (median, 52 wk) DCIS or IDC No Upgrade Underestimation Rate (%) Groups A and C Without residual microcalcifications 61 1 60 1.6 Group B With residual microcalcifications 41 8 33 20 Total 102 9 93 Note Group A composed of patients without residual microcalcifications after vacuum-assisted breast biopsy, and group B composed of patients with residual microcalcifications. Both groups were surgically treated. Group C patients were not surgically treated but were observed on follow-up. ADH = atypical ductal hyperplasia, DCIS = ductal carcinoma in situ, IDC = infiltrating ductal carcinoma For groups A and C versus group B, p = 0.0027. contraindicated (n = 2), or the patient declined the recommended surgery (n = 25). These patients, after being adequately informed on the possible risks and benefits, were managed conservatively with mammographic follow-up, initially with a 6-month interval in the first year after biopsy, then with annual mammography. One patient developed a new cluster of suspicious microcalcifications in the ipsilateral breast that was evaluated with vacuum-assisted breast biopsy after a 48-month follow-up interval. The distance of the new cluster of microcalcifications was 1.5 cm from the previous vacuum-assisted breast biopsy site. The cluster was completely removed by the vacuum-assisted breast biopsy procedure that yielded a low-grade DCIS that was confirmed at surgery. The maximum extent of disease evaluated in the surgical specimen was 1.2 cm. In 41 patients, surgical biopsy was performed because of the presence of residual microcalcifications. At surgery, upgrade occurred in 20% (8/41); six pure DCIS and two DCIS with invasive component were found. The underestimation rate of ADH in patients with removal of all mammographic calcifications who were surgically treated (group A) (0%, 0/26) was significantly different from that associated with residual calcifications (group B) (20%, 8/41) (p < 0.05). The underestimation rate of ADH in patients with removal of all mammographic calcifications who were surgically treated or managed with mammographic followup (groups A and C) (1.6%, 1/61) also was significantly different from that associated with residual calcifications (group B) (20%, 8/41) (p < 0.05) and remained under the 2% threshold (Tables 1 and 2). The mean diameter of each calcification cluster was 0.9 cm (range, 0.2 1.4 cm). To observe a correlation between BI-RADS category and underestimation rate of ADH, only patients with diagnosis of ADH confirmed at surgery were evaluated (66%, 67/102). Of 67 microcalcification clusters, 4 were classified as BI-RADS 3, 62 as BI-RADS 4, and 1 as BI-RADS 5. All upgrade lesions except one, which was classified as BI-RADS 5 (IDC by surgical specimen), were found in the BI-RADS 4 category. No significant correlation between BI-RADS categories and underestimation rate of ADH was found. (p = 0.48, 1.00, and 0.12 for BI-RADS 3, 4, and 5, respectively). Discussion Stereotactic vacuum-assisted breast biopsy has become a widely used alternative to open surgical biopsy in patients with suspicious clustered microcalcifications as well as for reducing the number of surgical biopsies performed for benign lesions. In prior studies, the ADH diagnosis frequency in 11-gauge biopsy specimens ranged from 10% to 28% [8 10, 18 27] (Table 3). In our study, the prevalence of ADH was 11.9% (8/67) using 11-gauge stereotactic core vacuum-assisted breast biopsy. TABLE 3: Published Data on Atypical Duct Hyperplasia (ADH) Underestimation Rate of ADH to 2010 Authors Reference No. Year of Publication No./Total ADH Yielded After 11-Gauge Vacuum-Assisted Breast Biopsy No./Total Patients Surgically Treated No./Total Underestimation Rate Liberman et al. 18 1998 12/112 (11) 10/12 (83) 1/10 (10) Brem et al. 19 1999 20/422 (5) 16/20 (80) 4/16 (25) Burak et al. 20 2000 43/851 (5) 40/43 (93) 5/40 (13) Philpotts et al. 9 2000 26/753 (3) 26/26 (100) 6/26 (23) Adrales et al. 10 2000 90/1081 (8] 62/90 (69) 9/62 (15) Cangiarella et al. 21 2001 10/160 (6) 8/10 (80) 2/8 (25) Lai et al. 22 2001 19/673 (3) 12/19 (63) 2/12 (17) Jackman et al. 8 2002 131/1964 (7) 104/131(79) 22/104 (21) Pandelidis et al. 23 2003 37/1341 (3) 35/37 (95) 5/35 (14) Winchester et al. 24 2003 77/1750 (4) 65/77 (84) 11/65 (17) Sohn et al. 25 2007 88/4579 (2) 78/88 (89) 14/78 (18) Lourenco et al. 26 2007 62/828 (8) 46/62 (74) 13/46 (28) Eby PR et al. 27 2008 58/391 (15) 49/58 (85) 10/49 (20) Penco et al. 29 2010 99/4086 (2) 46/99 (46) 13/46 (28.2) Present study 2010 114/1173 (10) 67/102 (67) 8/67 (11.9) Note Data in parentheses are percentages. AJR:197, October 2011 1015

Villa et al. Our underestimation rate of ADH is one of the lowest compared with those reported by other authors. Many factors contribute to a decrease in the underestimation rate of ADH: some are clinical-radiologic factors and others are pathologic findings. We reviewed only radiographic features of ADH diagnosed with 9-, 11-, and 14-gauge vacuum-assisted breast biopsy and focused on the presence or absence of residual microcalcification clusters. The most important factor in reducing the underestimation rate of ADH is complete removal of microcalcification clusters [9], Cluster extension and the number of cores are less important. Sneige et al. [13], Wagoner et al. [15], and others [28, 29] also examined the underestimation rate of ADH when percutaneous biopsy resulted in complete removal of all suspicious calcifications. Sneige et al., using both 14- and 11-gauge vacuum-assisted breast biopsy, observed no underestimation rate when all calcifications were removed (0 of 15 cases). Similarly, Wagoner et al., using 11-gauge vacuum-assisted breast biopsy, observed an underestimation rate of 5% (3 of 57 cases) when all calcifications were removed versus a rate of 24% (19 of 79 cases) when there were residual calcifications. Additionally, Wagoner et al. distinguished determinant calcifications (those associated with ADH) from unrelated calcifications (those not associated with ADH). When all determinant calcifications were removed, however, there was a 3% underestimation rate (1 of 34 cases) versus a 35% underestimation rate (14 of 40 cases) when there were residual determinant calcifications. Finally, there were no underestimations in the group with all determinant calcifications removed and ADH involving fewer than three foci (0 of 25 cases). Other authors observed no difference in the underestimation rate of ADH in patients who had suspicious calcifications completely removed by vacuum-assisted breast biopsy compared with those with residual microcalcifications after biopsy, irrespective of the presence of determinant calcifications [16]. Penco et al. [29] found an underestimation rate for ADH of 6.7% when all microcalcifications were removed (1/15), but neither the median number of vacuum-assisted breast biopsy specimens nor the pathologic diagnosis of ADH (isolated ADH or accompanied by other high-risk lesions, such as flat epithelial atypia or radial scar) was specified. In our study we considered only the radiologic criterion of presence or absence of residual microcalcifications after 11-gauge vacuum-assisted breast biopsy. The retrospective nature of our study did not allow us to evaluate the number of foci, the grade of atypia, and the difference between determinant and indeterminant calcifications. Other previous studies showed that the underestimation rate was reduced when there were no residual calcifications after vacuum-assisted breast biopsy even under the 2% threshold [9, 13, 15, 18, 28]. For this reason, the absence of residual microcalcifications after the vacuum-assisted breast biopsy criterion was used to show that underestimation rate of ADH remained under the 2% benignity threshold (Table 4). We also considered patients who did not undergo surgical excision and were monitored with 6-month-interval mammographic follow-up for the first year and then with annual mammographic follow-up. Even with the length of follow-up in our study (maximum, 72 months; median, 53 months), one of the longest reported in the literature, 4 patients had a follow-up of less than 24 months, which is the minimum imaging follow-up deemed adequate. Thus, the number was probably not sufficient to detect all cases of vacuum-assisted breast biopsy procedure underestimations. Similar to Liberman et al. [18], Adrales et al. [10], and Darling et al. [30], all calcification clusters were evaluated in our institution using 11-gauge vacuum-assisted breast biopsy. The underestimation rate was greatly reduced by using this technique compared with large-core needle biopsy, core needle biopsy, or 14-gauge vacuum-assisted breast biopsy, presumably resulting from more extensive tissue sampling and the greater likelihood of complete removal of the target lesions [10, 18, 26]. Eby et al. [27] found the TABLE 4: Underestimation Rate of Atypical Ductal Hyperplasia (ADH) Using Residual Microcalcification as Criterion Authors Reference No. Year of Publication Needle Gauge No. of Core Samples (range) Underestimation Rate of ADH (%) Results for Lesions Without Residual Microcalcification Liberman et al. 18 1998 11 14 (1 34) 1/10 (10) None with no residual microcalcification Philpotts et al. a 9 2000 11 12 (5 20) 6/26 (23) No underestimation rate when all calcifications were removed (10 clusters) b Jackman et al. 8 2002 11 At least 12 22/104 (21) 8% (3/36) when percentage of lesion removed was 100% Sneige et al. 13 2003 11 10.8 (6 22) 3/42(7.1) Underestimation rate was 0% without residual microcalcifications Forgeard et al. 28 2008 11 16 29/116(25) None when lesion size was < 6 mm and all microcalcifications removed Wagoner et al. 15 2009 11 and 14 10 (2 26) 33/123 (26.8) 5% underestimation rate without residual microcalcification Kohr et al. 16 2010 11 No. of specimens not given 20/101(19.8) 17% (7/41) without residual microcalcification Penco et al. 29 2010 8 and 11 No. of specimens not given 13/46 (28.2) 6.7% (1/15) when all calcifications were removed Present study 11 12 (9 16) 8/67 (11.9) 0% (0/26) with no residual microcalcification; 1.6% (1/61) considering patients in follow-up a Lesions were calcifications and architectural distortions. Five of six underestimated lesions were calcifications cluster. b No underestimation rate occurred even for two mass lesions (data not included). 1016 AJR:197, October 2011

Biopsy of Atypical Ductal Hyperplasia underestimation rate of focal ADH was less for lesions sampled with 11-gauge vacuumassisted breast biopsy (2 of 25, 8.0%) compared with 9-gauge vacuum-assisted breast biopsy (5 of 31, 16.1%), and the difference was not statistically significant. In our institution, microcalcification clusters were evaluated by a mean of 12 core tissue samples as suggested by Lomoschitz et al. [31]. Previous studies showed that underestimation of malignancy decreases as the average number of cores obtained per lesion increases [32 34]. Jackman et al. [34] found that the underestimation at core needle biopsy, large-core needle biopsy, or vacuumassisted breast biopsy diminished as the mean number of cores per lesion increased (from 6.3 to 9.5) and later showed a further reduction in the underestimation rate when more than 10 cores were acquired per lesion. Zografos et al. [33] found no underestimation of malignancy in 25 cases using an extended core needle biopsy protocol with removal of 96 cores per lesion. Other authors found that the underestimation rate is not related to the number of cores obtained, suggesting that numbers might not accurately reflect how thoroughly the lesion is sampled by core needle biopsy [9, 30, 35]. According to another study [8] no significant difference was found between BI-RADS categories and underestimation rate of ADH. Another important issue that must be considered is the average time interval until the subsequent development of carcinoma (intraductal and invasive carcinoma). In a study by Tavassoli et al. [4], the average interval was about 8.3 years, and in a study from the Mayo Clinic, approximately 20% of patients with ADH developed cancer over a follow-up period of 13.7 years [2]. Despite the low underestimation rate observed in our study in the group of patients without residual microcalcifications, women with a prior diagnosed ADH must be considered a very high-risk population for the lifetime development of malignancy and so must be carefully managed by strict mammographic follow-up. In conclusion, our data suggest that patients with a diagnosis of ADH after 11-gauge vacuum-assisted breast biopsy without residual microcalcifications postprocedure in those institutions that can duplicate an underestimation rate less than 2% could possibly be managed conservatively with 6-month mammographic follow-up and then with annual mammographic follow-up, with an overall underestimation rate under the 2% threshold. However, it is strongly recommended that patients with residual microcalcifications after vacuum-assisted breast biopsy undergo surgical excision because of an elevated underestimation rate. References 1. Page DL, Dupont WD, Rogers LW, Rados MS. Atypical hyperplastic lesions of the female breast: a long-term follow-up study. Cancer 1985; 55:2698 2708 2. Degnim AC, Visscher DW, Berman HK, et al. Stratification of breast cancer risk in women with atypia: a Mayo cohort study. J Clin Oncol 2007; 25:2671 2677 3. Page DL, Rogers LW. Combined histologic and cytologic criteria for the diagnosis of mammary atypical ductal hyperplasia. Hum Pathol 1992; 23:1095 1097 4. Tavassoli FA, Norris HJ. A comparison of the results of long-term follow-up for atypical intraductal hyperplasia and intraductal hyperplasia of the breast. Cancer 1990; 65:518 529 5. Tavassoli FA. Atypical hyperplasia: a morphologic risk factor for subsequent development of invasive breast carcinoma. Cancer Invest 1992; 10:433 441 6. Yeh IT, Dimitrov D, Otto P, Miller AR, Kahlenberg MS, Cruz A. Pathologic review of atypical hyperplasia identified by image-guided breast needle core biopsy: correlation with excision specimen. Arch Pathol Lab Med 2003; 127:49 54 7. Bonnett M, Wallis T, Rossmann M, et al. Histopathologic analysis of atypical lesions in imageguided core breast biopsies. Mod Pathol 2003; 16:154 160 8. Jackman RJ, Birdwell RL, Ikeda DM. Atypical ductal hyperplasia: can some lesions be defined as probably benign after stereotactic 11-gauge vacuum-assisted biopsy, eliminating the recommendation for surgical excision? Radiology 2002; 224:548 554 9. Philpotts LE, Lee CH, Horvath LJ, et al. Underestimation of breast cancer with 11-gauge vacuum suction biopsy. AJR 2000; 175:1047 1050 10. Adrales G, Turk P, Wallace T, et al. Is surgical excision necessary for atypical ductal hyperplasia of the breast diagnosed by Mammotome? Am J Surg 2000; 180:313 315 11. O Hea BJ, Tornos C. Mild ductal atypia after large-core needle biopsy of the breast: is surgical excision always necessary? Surgery 2000; 128: 738 743 12. Doren E, Hulvat M, Norton J, et al. Predicting cancer on excision of atypical ductal hyperplasia. Am J Surg 2008; 195:358 362 13. Sneige N, Lim SC, Whitman GJ, et al. Atypical ductal hyperplasia diagnosis by directional vacuum-assisted stereotactic biopsy of breast microcalcifications: considerations for surgical excision. Am J Clin Pathol 2003; 119:248 253 14. Ely KA, Carter BA, Jensen RA, Simpson JF, Page DL. Core biopsy of the breast with atypical ductal hyperplasia: a probabilistic approach to reporting. Am J Surg Pathol 2001; 25:1017 1021 15. Wagoner MJ, Laronga C, Acs G. Extent and histologic pattern of atypical ductal hyperplasia present on core needle biopsy specimens of the breast can predict ductal carcinoma in situ in subsequent excision. Am J Clin Pathol 2009; 131:112 121 16. Kohr JR, Eby PR, Allison KH, et al. Risk of upgrade of atypical ductal hyperplasia after stereotactic breast biopsy: effects of number of foci and complete removal of calcifications. Radiology 2010; 255:723 730 17. American College of Radiology. Breast imaging reporting and data system (BI-RADS), 4th ed. Reston, VA: American College of Radiology, 2003 18. Liberman L, Smolkin JH, Dershaw DD, et al. Calcification retrieval at stereotactic, 11-gauge, directional, vacuum-assisted breast biopsy. Radiology 1998; 208:251 260 19. Brem RF, Behrndt VS, Sanow L, et al. 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; 172:1405 1407 20. Burak WE Jr, Owens KE, Tighe MB, et al. Vacuum-assisted stereotactic breast biopsy: histologic underestimation of malignant lesions. Arch Surg 2000; 135:700 703 21. Cangiarella J, Waisman J, Symmans WF, et al. Mammotome core biopsy for mammary microcalcification: analysis of 160 biopsies from 142 women with surgical and radiologic follow-up. Cancer 2001; 91:173 177 22. Lai JT, Burrowes P, MacGregor JH. Diagnostic accuracy of a stereotaxically guided vacuumassisted large-core breast biopsy program in Canada. Can Assoc Radiol J 2001; 52:223 227 23. Pandelidis S, Heiland D, Jones D, et al. Accuracy of 11-gauge vacuum-assisted core biopsy of mammographic breast lesions. Ann Surg Oncol 2003; 10:43 47 24. Winchester DJ, Bernstein JR, Jeske JM, et al. Upstaging of atypical ductal hyperplasia after vacuum-assisted 11-gauge stereotactic core needle biopsy. Arch Surg 2003; 138:619 622 25. Sohn V, Arthurs Z, Herbert G, et al. Atypical ductal hyperplasia: improved accuracy with the 11-gauge vacuum-assisted versus the 14-gauge core biopsy needle. Ann Surg Oncol 2007; 14: 2497 2501 26. Lourenco AP, Mainiero MB, Lazarus E, et al. Ste- AJR:197, October 2011 1017

Villa et al. reotactic breast biopsy: comparison of histologic vacuum-assisted breast biopsy is not a therapeutic lesions with cancer or atypical hyperplasia. Radi- underestimation rates with 11- and 9-gauge vacu- procedure even when all mammographically ology 1994; 193:91 95 um-assisted breast biopsy. AJR 2007; 189:1164; found calcifications are removed: analysis of 4086 33. Zografos GC, Zagouri F, Sergentanis TN, et al. [web]w275 279 procedures. AJR 2010; 195:1255 1260 Minimizing underestimation rate of microcalcifi- 27. Eby PR, Ochsner JE, DeMartini WB, Allison 30. Darling ML, Smith DN, Lester SC, et al. Atypical cations excised via vacuum-assisted breast biop- KH, Peacock S, Lehman CD. Is surgical excision necessary for focal atypical ductal hyperplasia found at stereotactic vacuum-assisted breast biopsy? Ann Surg Oncol 2008; 15:3232 3238 28. Forgeard C, Benchaib M, Guerin N, et al. Is surgical biopsy mandatory in case of atypical ductal hyperplasia on 11-gauge core needle biopsy? A retrospective study of 300 patients. Am J Surg 2008; 196:339 345 29. Penco S, Rizzo S, Bozzini AC, et al. Stereotactic ductal hyperplasia and ductal carcinoma in situ as revealed by large-core needle breast biopsy: results of surgical excision. AJR 2000; 175:1341 1346 31. Lomoschitz FM, Helbich TH, Rudas M, et al. Stereotactic 11-gauge vacuum-assisted breast biopsy: influence of number of specimens on diagnostic accuracy. Radiology 2004; 232:897 903 32. Jackman RJ, Nowels KW, Shepard MJ, et al. Stereotaxic large-core needle biopsy of 450 nonpalpable breast lesions with surgical correlation in sy: a blind study. Breast Cancer Res Treat 2008; 109:397 402 34. Jackman RJ, Burbank F, Parker SH, et al. Atypical ductal hyperplasia diagnosed at stereotactic breast biopsy: improved reliability with 14-gauge, directional, vacuum-assisted biopsy. Radiology 1997; 204:485 488 35. Liberman L, Dershaw DD, Glassman JR, et al. Analysis of cancers not diagnosed at stereotactic core breast biopsy. Radiology 1997; 203:151 157 1018 AJR:197, October 2011