Imaging-Guided Core Needle Biopsy of Papillary Lesions of the Breast

Transcription

1 Eric L. Rosen 1 Rex C. Bentley 2 Jay A. Baker 1 Mary Scott Soo 1 Received January 30, 2002; accepted after revision April 12, Department of Radiology, Breast Imaging Division, Duke University Medical Center, Box 3808, Durham, NC Address correspondence to E. L. Rosen. 2 Department of Pathology, Diagnostic Pathology Division, Duke University Medical Center, Box 3712, Durham, NC AJR 2002;179: X/02/ American Roentgen Ray Society Imaging-Guided Core Needle Biopsy of Papillary Lesions of the Breast OBJECTIVE. Our objective was to assess the incidence of papillary lesions of the breast diagnosed at imaging-guided core needle biopsy and the need for surgical excision after a benign diagnosis. MATERIALS AND METHODS. This retrospective study included 1374 patients with consecutive suspicious breast lesions that underwent either mammography or sonographically guided large-core needle breast biopsy. Fifty-seven lesions (4%) were classified as papillary lesions. Eleven of the 57 cases were lost to follow-up (n = 6) or had not yet shown 2 years of stability (n = 5) and were excluded from this study. The remaining 46 papillary lesions constitute our study population. RESULTS. Surgical excision was performed in 17 (37%) of 46 papillary lesions. In the group of patients whose lesions were recommended for excision because carcinoma was identified at core biopsy, surgical excision revealed one false-positive and two true-positive diagnoses. In four cases, histologic diagnoses of the excisional biopsy and the core needle biopsy were discordant. One false-positive finding at core needle biopsy initially was interpreted as invasive ductal carcinoma on the basis of core needle biopsy specimens. In three false-negative findings, the initial diagnosis at core needle biopsy was upgraded after surgical excision. Two cases of papilloma with adjacent atypical ductal hyperplasia and one of atypical papilloma were upgraded to ductal carcinoma in situ after surgical excision. Imaging follow-up was performed in the remaining 29 patients. All lesions were stable or had decreased in size during the 2-year follow-up period. The negative predictive value of core needle biopsy for excluding malignancy among the papillary lesions diagnosed in our study was 93%. CONCLUSION. When the histologic diagnosis is benign, our data suggest that papillary lesions may be safely managed with imaging follow-up rather than with surgical excision. However, atypical papillary lesions or those associated with atypia require surgical excision because histologic underestimation occurs at a frequency similar to that in other atypical lesions undergoing core needle biopsy. P apillary lesions are a heterogeneous group of breast lesions, identified histologically by the presence of a fibrovascular stalk. They include papilloma, papillomatosis, atypical papilloma, noninvasive papillary carcinoma, and invasive papillary carcinoma. Although malignant papillary lesions are rare, papillary lesions as a group account for up to 5% of the lesions undergoing breast biopsy. They may be diagnosed after imaging-guided core needle biopsy for nonpalpable, clinically occult imaging-detected lesions, but the subsequent management of these lesions remains controversial [1 5]. Our study was undertaken to assess the need for surgical excision of papillary lesions diagnosed at core needle biopsy. Materials and Methods This retrospective study included all patients at our institution who underwent imaging-guided core needle biopsy of breast lesions from January 1991 through June During this time, 1374 consecutive suspicious breast lesions underwent either mammography or sonographically guided large-core needle breast biopsy. Stereotactic guidance was used for 882 procedures, and sonographic guidance was used for the remaining 492 biopsies performed during the review period. Core needle biopsy was performed with either mammographic or sonographic guidance, depending on which modality best depicted the lesion and the preference of the radiologist performing the procedure. In general, sonography was used for masses, and mammography, for microcalcifications. All procedures were performed by dedicated breast imaging radiologists. AJR:179, November

2 Rosen et al. All sonographically guided procedures were performed with the patient in the supine or supine oblique position. Imaging was performed with a high-resolution 5-9 or 13-MHz linear array transducer (Elegra; Siemens, Isaquah, WA). Biopsies were performed using a freehand technique with a 14- gauge Tru-Cut automated core biopsy needle (Baxter Healthcare, Valencia, CA) and a spring-loaded biopsy gun (Magnum; Bard, Covington, GA). All mammographically guided biopsies were performed on a dedicated prone stereotactic biopsy unit (Stereoguide; Lorad Medical Systems, Danbury, CT), using either a 14-gauge automated Tru-Cut needle and spring-loaded biopsy gun or a vacuum-assisted biopsy device with either 11- or 14-gauge biopsy needles (Mammotome; Biopsys Medical, Irvine, CA). Postprocedure mammography and imaging to document needle placement were performed for all patients. We attempted to obtain a minimum of five specimens from each lesion. Specimen radiographs were obtained to document retrieval of the targeted calcifications in all patients in whom the biopsy target was microcalcifications. Biopsy specimens were evaluated in the department of pathology; a minimum of three H and E stained sections were examined from each biopsy. Papillary lesions were classified according to the criteria of Tavassoli [6]. For purposes of our study, carcinoma arising in a papilloma was classified with papillary carcinoma in situ. These results were then reviewed by the breast imaging radiologists and entered into our biopsy database. Using an outcomes audit of our biopsy database, we retrospectively identified all cases of papillary lesions including papilloma, papillomatosis, atypical papilloma, noninvasive papillary carcinoma, noninvasive papillary carcinoma, and invasive papillary carcinoma. In our study, the term papillomatosis is synonymous with multiple peripheral intraductal papillomas, as defined by others, and does not include epithelial hyperplasia [6]. From the 1374 core biopsies performed during the study period, 57 papillary lesions (4%) were diagnosed. Of these, 34 were biopsied with mammographic guidance and 23, with sonographic guidance. Six of the 57 cases were lost to follow-up and were excluded from this study. Five of the 57 had yet to show 2 years of mammographic stability after core needle biopsy. The remaining 46 papillary lesions diagnosed at core needle breast biopsy constituted our study population. Surgical excision was performed in 17 (37%) of 51 papillary lesions. Twenty-nine (63%) of the remaining 46 lesions underwent imaging follow-up and were shown to be stable for a minimum of 2 years (mean follow-up time ± SD, 45 ± 17 months). Of the 46 lesions included in this study, 21 were noncalcified masses; seven, masses with associated microcalcifications; and 18, clustered or segmental microcalcifications. All were prospectively interpreted as suspicious for malignancy and placed in the Breast Imaging Reporting and Data System (BI- RADS) [7] assessment category 4. Reasons for this assessment included the following: new or increasing masses or calcifications, incidental findings not fulfilling benign or probably benign criteria, and palpable abnormalities. Seven masses were visible on sonography only, without a mammographic correlate. After the core needle biopsy, all pathology results were reviewed weekly in conjunction with results from pertinent imaging from both the diagnostic study and the biopsy procedure. In our review, the pathology results were compared with the relevant imaging findings to determine concordance or discordance of the biopsy results. All treatment decisions were by consensus of the participating breast imaging radiologists, and follow-up recommendations were determined by the outcome of this consensus session. Either 6-month or 1-year imaging follow-up recommendations were made for those patients in whom the pathology outcome was considered concordant. For masses, a benign result was considered concordant if no imaging features highly suspicious for malignancy were present, accurate targeting of the needle was shown, adequate samples were obtained, and the pathology results suggested a process known to manifest as a mass. For microcalcifications, a benign result was considered concordant if the following characters were present: no imaging features highly suspicious for malignancy, accurate targeting of the needle, representative microcalcifications on specimen radiography, and microcalcifications associated with a benign process at pathology. All pathology results revealing atypia, ductal carcinoma in situ, or invasive carcinoma and all discordant cases received recommendations for surgical excision. This prospective review was carried out independently of our study, which is a retrospective review of the outcomes of the 46 papillary lesions undergoing core needle biopsy in our department. The radiology, pathology, and surgery data were correlated. All images were retrospectively reviewed jointly by consensus of two breast radiologists. Lesions identified on mammography were characterized by these two reviewers using the BI-RADS lexicon [7], whereas sonographic lesions were characterized using the relevant criteria of Stavros et al. [8]. The relative frequency of each type of lesion (e.g., developing density, mass, or calcification) and lesion characteristics were determined, as well as lesion size, mode of detection, and whether the lesion was symptomatic. Stability of each lesion after core needle biopsy was assessed by a review of imaging and clinical follow-up. Only lesions undergoing subsequent surgical excision or showing a minimum of 24 months of imaging stability were included in our study. All lesions undergoing subsequent surgical excision were correlated to their specimen radiographs and the final diagnoses. The negative and positive predictive values of core needle biopsies were calculated. The core needle biopsy result was considered accurate if the lesion was initially diagnosed as benign papilloma, papillomatosis, or sclerosed papilloma and was stable for at least 1 year on follow-up imaging or if the initial benign core needle biopsy resulted in an excisional biopsy and its histology was confirmed by surgical excision. The core needle biopsy result was considered falsenegative if the lesion was classified as benign or atypical on the initial core needle biopsy but was upgraded to either in situ or invasive malignancy after surgical excision. The core biopsy result also was considered false-negative if a benign papillary lesion was rebiopsied at a later date because of interval change and showed either in situ or invasive malignancy. The core needle biopsy was considered false-positive if the lesion was incorrectly classified as atypical or malignant on the initial core biopsy but downgraded after excisional biopsy to a benign papillary lesion. Results Twenty-seven benign papillomas (59%), 11 papillomatosis (24%), six atypical papillomas (13%), and two intraductal papillary carcinomas (4%) were diagnosed at core needle biopsy and composed 3% (46/1374) of our biopsy population. The imaging manifestations of the 46 papillary lesions were as follows: 21 noncalcified masses, seven partially calcified masses, 17 clustered microcalcifications, and one asymmetric density. Twentythree lesions were evaluated with directed sonography, and seven masses were depicted only on sonography. Eight of the 46 lesions corresponded to palpable abnormalities. None of the patients presented with a complaint of unilateral spontaneous clear or bloody nipple discharge. The radiologists performing the prospective histologic review believed that of the 46 BI- RADS category 4 lesions in the study (Table 1), 36 (78%) showed concordance between the imaging features and histologic results. Nine lesions revealing atypia or suspicious histology (one case of possible radial scar) were recommended for excision. One oval mass diagnosed as invasive ductal carcinoma, not otherwise specified, was considered an unusual appearance for this diagnosis, and surgical excision was recommended. Imaging follow-up was performed in 29 patients. All lesions were stable or had decreased in size in the 2-year follow-up period. None have subsequently been excised because of interval progression on follow-up imaging. The findings of the eight palpable masses included in our study were benign at core needle biopsy, and the masses have remained stable on imaging follow-up. Five of the eight findings were benign papillomas, two of the eight were papillomatosis, and one of the eight was a sclerotic papilloma. Surgical excision was performed for 17 (37%) of 46 lesions after core needle biopsy. The following were indications for excisional 1186 AJR:179, November 2002

3 Imaging-Guided Biopsy of Breast Lesions TABLE 1 Core Biopsy Findings, Treatment Imaging Presentation and Treatment of 46 Papillary Lesions in Patients Diagnosed with Core Needle Biopsy Noncalcified Mass (n = 21) Calcified Mass (n = 7) a Diagnoses were changed from invasive ductal carcinoma to benign papilloma and from benign papilloma to atypical papilloma. b Diagnosis was changed from atypical ductal hyperplasia to ductal carcinoma in situ. c Diagnoses were changed from atypical ductal hyperplasia to ductal carcinoma in situ and from atypical papilloma to ductal carcinoma in situ. biopsy: carcinoma identified on core biopsy in three of the 17 patients, cellular atypia or a high-risk lesion (radial scar) in 10 of the 17, and excision recommended by the pathologist in the remaining four. In the group of lesions recommended for excision because carcinoma was identified at core needle biopsy, surgical excision revealed one false-positive and two true-positive diagnoses. In both cases of papillary ductal carcinoma in situ diagnosed at core needle biopsy, surgical excision confirmed the initial diagnosis, and neither was upgraded to invasive carcinoma (Fig. 1). In one case initially diagnosed as invasive ductal carcinoma at core biopsy, a subsequent surgical excision revealed only a benign papilloma with extensive adjacent apocrine metaplasia (Fig. 2). Restaining of the initial core biopsy specimen with a smooth-muscle actin stain showed retrospectively a uniform and normal myoepithelial layer surrounding the lesion, consistent with a papilloma, as well as extensive apocrine metaplasia in the adjacent parenchyma. The initial diagnosis of invasive ductal carcinoma was false-positive and therefore changed to a benign papilloma. Ten lesions were surgically excised because the core needle biopsy specimens revealed cellular atypia either in or adjacent to the papillary lesion or a high-risk lesion such as radial scar. Three of these were upgraded to low-grade papillary ductal carcinoma in situ after surgical excision and were considered false-negative results at core needle biopsy (Fig. 3). Two of the three carcinomas were initially diagnosed at Clustered Microcalcifications (n = 17) Asymmetric Density (n = 1) Total (n = 46) Benign, follow-up only Benign, surgical excision Atypical, surgical excision Malignant, surgical excision Total no. of surgical excisions Discrepant core biopsy and surgical diagnoses 2 a 1 b 2 c 0 5 core needle biopsy as papillomatosis with associated atypical epithelial hyperplasia, and the other was an atypical papilloma. The findings of the remaining seven lesions were benign (truenegative at core biopsy), with three atypical papillomas and four benign papillomas or papillomatosis diagnosed. Of the four lesions with no carcinoma or highrisk lesion identified in the initial core specimen recommended for excision by the pathologist, two were interpreted as papilloma, one was interpreted as a benign papilloma, and one was interpreted as a sclerotic papilloma. In all four cases, the pathologist recommended surgical excision to avoid undersampling of a potentially higher grade lesion. Subsequent surgical excision showed concordant benign lesions in three. The benign papilloma was reclassified as an atypical papilloma after excisional biopsy (Fig. 4). Overall, 12 consistent histologic results were confirmed at core needle biopsy and surgical excision: five diagnoses of benign papilloma, four of atypical papilloma, two of papillary ductal carcinoma in situ, and one of a papilloma with adjacent atypical ductal hyperplasia. Another case of atypical papilloma was shown to be papillomatosis after surgical excision. In five cases, histologic diagnoses of the results of excisional biopsy and the core needle biopsy were inconsistent. Three lesions were sampled with an 11-gauge vacuum-assisted device and two with a 14-gauge multipass gun. One false-positive result of core needle biopsy was initially interpreted as an invasive ductal carcinoma on the basis of core needle biopsy specimens. In three false-negative cases, the initial core needle biopsy diagnosis was upgraded after surgical excision. Two cases of papilloma with adjacent atypical ductal hyperplasia and one of atypical papilloma were upgraded to in situ carcinoma after surgical excision. In one other case, an atypical papilloma diagnosed at core biopsy was shown to be papillomatosis at surgical excision. The negative predictive value of core needle biopsy for excluding malignancy among the papillary lesions diagnosed in our study was 93%. Discussion Distinguishing malignant from benign papillary lesions can pose problems for the pathologist, particularly if only a small portion of the lesion is submitted for review. The most important histologic feature separating a benign papillary lesion from papillary carcinoma in situ is the presence of an atypical epithelial proliferation resembling low-grade ductal carcinoma in situ in which a normal myoepithelial cell layer is typically absent. Absence of this cell layer in a papillary lesion indicates a papillary carcinoma [6]. Other intermediate lesions, such as atypical papilloma, are defined less precisely, and their correct classification depends on the percentage of the lesion revealing an atypical epithelial proliferation with or without an intact myoepithelial cell layer. Sampling of the lesion thus becomes a major concern in correctly classifying papillary lesions at core needle biopsies; therefore, the treatment of papillary lesions diagnosed at core needle biopsy remains controversial [1 5]. In our series, 4% of all findings of imagingguided core needle biopsies performed at our institution were papillary lesions. This frequency is similar to that of other reports and suggests that papillary lesions account for a small percentage of all lesions undergoing imaging-guided core needle biopsy [1, 3, 5]. Therefore, radiologic histologic correlation and outcomes are important to ensure that these lesions are appropriately managed. The papillary lesions diagnosed at core needle biopsy in our series had a different presentation from that of most symptomatic papillomas. Most papillary lesions arise in the major ducts, and most women present with serous or sanguineous nipple discharge. Similarly, in women with papillary carcinoma, a palpable subareolar mass is present 90% of the time, and up to 26% of patients present with nipple discharge [6]. In our series, however, only eight (14%) of 57 lesions corresponded to AJR:179, November

4 Rosen et al. A B C D Fig year-old woman with false-positive diagnosis at core needle biopsy. A, Medial lateral oblique mammogram shows oval mass (arrows) in subareolar breast. B, Sonogram reveals complex cystic mass (arrow) with solid component. C, Sonogram shows core needle biopsy performed through solid component using sonographic guidance. Note biopsy needle traversing lesion. D, Photograph of histologic specimen obtained at core needle biopsy shows irregular infiltrative-appearing small glands in desmoplastic stroma, interpreted initially as invasive ductal adenocarcinoma. Subsequent excision of this lesion revealed only benign intraductal papilloma with extensive sclerosis. (H and E, 200) E, Photograph of immunoperoxidase stain on histologic specimen obtained at core needle biopsy shows circumferential staining around all glands, confirming presence of myoepithelial cell layer. Diagnosis was changed to benign intraductal papilloma. ( 400) E 1188 AJR:179, November 2002

5 Imaging-Guided Biopsy of Breast Lesions palpable masses, and none presented with a history of clear or bloody nipple discharge. Instead, the lesions in our series tended to be small nonpalpable imaging-detected abnormalities. The papillary lesions referred for imaging-guided biopsy were more likely to represent clinically occult imaging-detected abnormalities and, therefore, may include a relatively high frequency of peripheral rather than central papillary lesions. These peripheral papillary lesions develop in the distal ductal epithelium, lack the fibrovascular core present in large duct papillomas, and may produce mammographically visible clustered microcalcifications [9]. In our series, papillary lesions were typically nonpalpable imaging-detected abnormalities that were almost equally likely to manifest as masses or as microcalcifications. Our findings agree with and further support the previous reports that state that lesions diagnosed as benign papillomas at core needle biopsy can be safely managed with clinical and radiographic follow-up and do not necessarily require surgical excision [1, 3]. The negative predictive value of core needle biopsy among lesions included in our series was 93%. This finding is similar to the 92% negative predictive value reported by Mercado et al. [3], who described papillary lesions diagnosed with an 11-gauge vacuum-assisted biopsy device. All 27 BI-RADS assessment category 4 lesions diagnosed as benign papillomas at core needle biopsy in our study were considered concordant at imaging histologic review, and none proved to be malignant by either imaging or surgical follow-up. The 29 benign lesions that were followed up with mammography have all shown imaging stability, with a mean follow-up time of 45 months (SD, ±17 months). These data parallel the results reported by Liberman et al. [1], who similarly found no additional malignancies among those lesions identified as benign papillomas at core needle biopsy. In our series, no papillary lesion A diagnosed as benign at core needle biopsy has been subsequently found to be malignant. Our results, in combination with those from previous reports, suggest that papillary lesions diagnosed as benign without atypia by core needle biopsy can be safely followed-up with imaging and do not require immediate surgical excision. In contrast, the diagnosis of atypical papilloma or atypical features at core needle biopsy clearly requires surgical excision. Three (38%) of the eight atypical cases in our series were shown to be papillary carcinoma in situ after excisional biopsy. Liberman et al. [1] also observed that three (30%) of 10 atypical cases were upgraded to ductal carcinoma in situ after surgical excision. Mercado et al. [3] and Philpotts et al. [4] observed no additional malignancies among atypical papillary lesions surgically excised, likely because of the small number of lesions observed in this category. Regardless, the importance of surgical excision after a core needle biopsy result of atypia B Fig year-old woman diagnosed with papillary carcinoma in situ at core needle biopsy and confirmed at excisional biopsy. A, Coned image of craniocaudal mammogram shows irregular mass with ill-defined margins and pleomorphic microcalcifications. B, Stereotactic mammograms taken after firing biopsy gun and before biopsy confirm proper positioning of 11-gauge vacuum-assisted biopsy needle. C, Photograph of histologic specimen obtained at core needle biopsy shows well-developed cribriform architecture in epithelial component of papillary lesion, diagnostic for papillary carcinoma in situ. (H and E, 100) C AJR:179, November

6 Rosen et al. has been well established, and the data regarding atypia associated with or in papillary lesions suggest that surgical excision after core needle biopsy of these lesions will reveal a similar frequency of histologic underestimation [2, 4, 5, 10]. The finding of papillary carcinoma in situ at core needle biopsy is relatively rare; in both cases in our series, however, this diagnosis was confirmed at subsequent excision. Although neither case in our series was upgraded to an invasive carcinoma after surgical excision, understaging of papillary malignancy has been reported in other articles addressing this topic [1 4, 10]. This consideration is important because women with in situ carcinoma do not usually undergo axillary lymph node sampling, whereas those with invasive disease almost always do. Underestimation of in situ carcinoma diagnosed at core needle biopsy is an inherent limi- tation of this technique and has been well documented [11, 12]. This understaging of malignancy has been shown to decrease with the increasing number of tissue samples, larger biopsy needles, and use of a vacuum-assisted biopsy device [10]. However, Mercado et al. [3] reported that two (25%) of eight lesions were upgraded from papillary in situ to invasive carcinoma, even though the lesions were sampled with an 11-gauge vacuum-assisted biopsy device and an average of 11 tissue samples were obtained. Similarly, Liberman et al. [1] reported that three (43%) of seven papillary lesions initially diagnosed at 14-gauge core needle biopsy were upgraded to invasive papillary carcinoma after surgical excision. In our series, one (11%) of nine lesions sampled with a 14-gauge needle and two (25%) of eight lesions sampled with an 11-gauge probe were upgraded from an atypical lesion to ductal carcinoma in situ. Understaging A of ductal carcinoma in situ diagnosed at core needle biopsy is one of the limitations of the technique and has been documented to occur for both papillary and nonpapillary lesions at a similar rate. More important, we identified one falsepositive diagnosis of invasive ductal adenocarcinoma, not otherwise specified, at a core needle biopsy of a papilloma. This case highlights the well-known potential error in the pathologic evaluation of benign papillary lesions in which entrapped benign epithelium in the fibrous wall can be mistaken for invasive carcinoma; staining with smooth-muscle actin stain to help avoid this pitfall [13]. When papillomas are falsely diagnosed as invasive carcinoma, they are not usually mistaken for invasive papillary carcinoma, which may have a similar mammographic appearance to a papilloma (i.e., an oval or lobulated B Fig year-old woman with atypical papilloma found at core needle biopsy that was upgraded to papillary carcinoma in situ at excisional biopsy. A, Spot compression magnification mammogram shows clustered amorphous microcalcifications in focally dense breast parenchyma. B, Photograph of histologic specimen obtained at core needle biopsy shows atypical epithelial proliferation involving only part of biopsy (right half of field) specimen, leading to diagnosis of atypical papilloma. (H and E, 40) C, Photograph of histologic specimen obtained at excisional biopsy reveals papillary carcinoma in situ. (H and E, 100) C 1190 AJR:179, November 2002

7 Imaging-Guided Biopsy of Breast Lesions mass with mostly circumscribed margins). The misdiagnosed papillomas in this situation are usually mistaken for invasive ductal carcinoma not otherwise specified. Breast pathologists have noted, It is the concurrence of many small spaces containing complex patterns of epithelia hyperplasia enveloped by fibrosis which may lead to a mistaken diagnosis of malignancy [13]. In general, the mammographic appearance of these two lesions is different: invasive ductal carcinoma often has an appearance that is highly suggestive of malignancy (i.e., irregular mass with ill- defined spiculated margins) compared with the more benign-appearing papilloma. Although it cannot definitely be differentiated by the mammographic appearance alone, forewarning the pathologist that a papillary lesion is suspected on the basis of imaging features may assist in avoiding this pathologic pitfall. In conclusion, papillary lesions account for a small percentage of lesions undergoing imaging-guided core needle biopsy. When the histologic diagnosis is benign and there is imaging histologic concordance, our data sug- A C gest that these lesions may be safely managed with imaging follow-up rather than with surgical excision. However, atypical papillary lesions or those associated with atypia require surgical excision because histologic underestimation occurs at a frequency similar to other atypical lesions undergoing core needle biopsy. In addition, careful attention to radiologic histologic concordance may help minimize the potential for a false-positive finding on core biopsy resulting from error in the pathologic evaluation of benign papillary lesions mistaken for invasive carcinoma. B D Fig year-old woman with benign papilloma revealed at core needle biopsy that was upgraded to atypical papilloma at excisional biopsy. A, Spot compression mammogram reveals oval mass with mostly circumscribed margins in inferior subareolar breast. B, Sonogram shows solid oval mass with ill-defined margins. C, Photograph of histologic specimen obtained at core needle biopsy shows well-developed papillary architecture. (H and E, 40) D, Photograph of histologic specimen from excisional biopsy shows intraductal papilloma with atypia. Punched-out spaces and monotonous cytology are shown in this high-power view. Elsewhere on specimen, papilloma shows only epithelial hyperplasia. Because atypical area is only focal, diagnosis is atypical papilloma. (H and E, 100) AJR:179, November

8 Rosen et al. References 1. Liberman L, Bracero N, Vuolo MA, et al. Percutaneous large-core biopsy of papillary breast lesions. AJR 1999;172: Reynolds HE. Core needle biopsy of challenging benign breast conditions: a comprehensive literature review. AJR 2000;174: Mercado CL, Hamele-Bena D, Singer C, et al. Papillary lesions of the breast: evaluation with stereotactic directional vacuum-assisted biopsy. Radiology 2001;221: Philpotts LE, Shaheen NA, Jain KS, Carter D, Lee CH. Uncommon high-risk lesions of the breast diagnosed at stereotactic core-needle biopsy: clinical importance. Radiology 2000;216: Jackman RJ, Nowels KW, Rodriquez-Soto J, Marzoni FA Jr, Finkelstein SI, Shepard MJ. Stereotactic, automated, large-core needle biopsy of non-palpable breast lesions: false-negative and histologic underestimation rates after long-term follow-up. Radiology 1999;210: Tavassoli FA. Papillary lesions. In: Tavasoli FA, ed. Pathology of the breast. Norwalk, CT: Appleton & Lange, 1992: American College of Radiology. Breast imaging reporting and data system (BI-RADS), 3rd ed. Reston, VA: American College of Radiology, Stavros TA, Thickman D, Rapp CL, et al. Solid breast nodules: use of sonography to distinguish between benign and malignant lesions. Radiology 1995;196: Kopans DB. Pathologic, mammographic, and sonographic correlation. In: Kopans DB, ed. Breast imaging, 2nd ed. Philadelphia: Lippincott-Raven, 1998: Burbank F. Stereotactic breast biopsy of atypical ductal hyperplasia and ductal carcinoma in situ lesions: improved accuracy with directional, vacuumassisted biopsy. Radiology 1997;208: Lee CH, Carter D, Philpotts LE, et al. Ductal carcinoma in situ diagnosed with stereotactic core needle biopsy: can invasion be predicted? Radiology 2000;217: Jackman RJ, Burbank F, Parker SH, et al. Stereotactic breast biopsy of non-palpable lesions: determinants of ductal carcinoma in situ underestimation rates. Radiology 2001;218: Page DL. Papilloma and related lesions. In: Page DL, Anderson TJ, eds. Diagnostic histopathology of the breast. Edinburgh: Churchill Livingstone, 1987: For the convenience of AJR authors, a standardized form requesting permission to reprint from other publications is now available via the ARRS Web site at AJR:179, November 2002