CME. False-Positive Sentinel Lymph Nodes in Breast Cancer Patients Caused by Benign Glandular Inclusions

Transcription

1 Anatomic Pathology / Glandular Inclusions in Sentinel Nodes False-Positive Sentinel Lymph Nodes in Breast Cancer Patients Caused by Benign Glandular Inclusions Report of Three Cases and Review of the Literature Yan Peng, MD, PhD, 1 Raheela Ashfaq, MD, 1 Gene Ewing, MD, 1 A. Marilyn Leitch, MD, 2 and Kyle H. Molberg, MD 1 Key Words: Axillary sentinel nodes; False-positive; Benign glandular inclusions; Breast carcinoma; Immunostains for myoepithelial markers DOI: /JVB8QFQNW5HBN7UJ Upon completion of this activity you will be able to: provide a differential diagnosis for glandular groups of epithelial cells in axillary lymph nodes. define histologic and immunohistochemical features of benign glandular inclusions in axillary lymph nodes. list the immunohistochemical stains that can be applied to lymph nodes to differentiate benign glandular inclusions from metastatic carcinoma of breast. The ASCP is accredited by the Accreditation Council for Continuing Medical Education to provide continuing medical education for physicians. The ASCP designates this educational activity for a maximum of 1 AMA PRA Category 1 Credit per article. The authors of this article and the planning committee members and staff have no relevant financial relationships with commercial interests to disclose. Questions appear on p 146. Exam is located at CME Abstract We report 3 cases of sentinel lymph nodes (SLNs) containing benign glandular inclusions (BGIs) in patients with breast carcinoma that were initially misdiagnosed as metastatic carcinoma. The first case had an SLN with glandular elements adjacent to a squamous inclusion cyst, the second had an SLN with a single complex gland showing apocrine features, and the third had 2 SLNs, each containing rare glands lined by bland columnar cells and surrounded by thin, fibrous bands. All glandular elements were distinctly different from the corresponding invasive carcinoma. Immunostains for myoepithelial markers revealed smooth muscle myosin reactivity and scattered p63+ nuclei, indicating the presence of myoepithelial cells. Based on morphologic and immunohistochemical findings, a diagnosis of BGIs was established. Our case series report indicates that comparison with the morphologic features of primary breast carcinoma and using immunohistochemical analysis for myoepithelial markers are important ancillary tools in distinguishing BGIs from metastatic carcinoma. Biopsy of axillary sentinel lymph nodes (SLNs) has become the standard of care in the management of breast cancer. Pathologic evaluation of SLNs provides a reliable means of predicting the status of the remaining axillary lymph nodes, thus avoiding the complications associated with a complete axillary node dissection (CAND) in patients with negative SLNs. Recently, the concept of intranodal benign glandular inclusions (BGIs) in SLNs has gained the attention of breast surgeons and pathologists because their presence may be mistaken for metastatic carcinoma. BGIs in lymph nodes are not uncommon and have been well documented in the literature. 1-3 Although well recognized, 4-9 the occurrence of this phenomenon in axillary lymph nodes is extremely rare. Most reported examples of BGIs have been in patients without breast carcinoma, but there are a few reports of BGIs in patients with breast carcinoma. 5,6 Previously, BGIs in patients with breast carcinoma have been reported in nonsentinel axillary lymph nodes. 7,8 Recently, a single case report demonstrated the presence of BGIs in an axillary SLN in a patient with ipsilateral breast carcinoma. 6 We report 3 cases with 4 axillary SLNs containing BGIs with different histologic appearances in patients with ipsilateral invasive ductal carcinoma of the breast. Initially, the BGIs in all 3 cases were misdiagnosed as metastatic carcinoma. On second pathologic review for quality assurance purposes and with the assistance of immunohistochemical stains for myoepithelial cells in 2 cases, corrected diagnoses of BGIs were made, fortunately before further surgical intervention. Thus, unnecessary CANDs were avoided in all 3 patients. To the best of our knowledge, this is the largest case series in the literature of BGIs in axillary SLNs in patients with ipsilateral breast carcinoma. Am J Clin Pathol 2008;130: DOI: /JVB8QFQNW5HBN7UJ 21

2 Peng et al / Glandular Inclusions in Sentinel No d e s Materials and Methods Three cases were identified from the surgical pathology files at our institution between November 2006 and May Clinical information was obtained from electronic medical records and the referring physicians. SLN identification was performed by using a dual injection technique for mapping. Technetium sulfur colloid was injected in the subdermal peritumoral region at least 2 hours before surgery and then blue dye, lymphazurin blue or methylene blue, was injected in the operating room in the peritumoral and often subareolar location. Any node that was radioactive or stained blue with dye was removed. The SLNs were fixed in 10% buffered formalin and entirely submitted for histologic evaluation. Sections were prepared for routine H&E and immunohistochemical staining. The latter was performed on a TechMate 1000 automated immunostainer (Ventana Medical Systems, Tucson, AZ). Monoclonal antibodies to the following antigens were used: cytokeratin AE1/AE3, smooth muscle myosin heavy chain, p63, and HER-2/neu (all from DakoCytomation, Carpinteria, CA); CD10 (Biocare, Concord, CA); and estrogen and progesterone receptors (Ventana Medical Systems). Results Clinical Features The clinicopathologic characteristics of the 3 cases are summarized in ztable 1z. All 3 patients underwent breast core biopsy before excision. Excisional specimens showed predominantly ductal carcinoma in situ (high grade) along with relatively small amounts of invasive ductal carcinoma (low to intermediate grade), and no lymphovascular invasion was identified zimage 1Az, zimage 2Az, and zimage 3Az. Ipsilateral SLN biopsies were performed at the time of primary excision. There were a total of 9 SLNs identified; 4 of these SLNs contained BGIs on which immunohistochemical analysis for cytokeratins and myoepithelial markers (smooth muscle myosin, p63, and CD10) was performed. All other SLNs were negative for malignancy and without epithelial elements confirmed by cytokeratin AE1/AE3 stains. Histologic and Immunohistochemical Features Case 1 The SLN was submitted for frozen section and revealed intranodal glandular elements with microcalcifications (1.8 mm) adjacent to a benign-appearing squamous inclusion cyst containing keratin debris zimage 1Bz. The frozen section diagnosis was deferred. The permanent sections also showed the intranodal glandular elements zimage 1Cz and zimage 1Dz that were positive for cytokeratin but had a bland appearance. A diagnosis of metastatic carcinoma was made on the permanent sections. On a second pathologic review of the SLN before further surgical management, immunostains for myoepithelial markers were performed to explore the possibility of BGIs. The presence of strongly positive smooth muscle myosin and p63 immunostains surrounding the glands indicated the presence of myoepithelial cells zimage 1Ez and zimage 1Fz and supported their benign nature. The glandular components were also positive for estrogen and progesterone receptors. The patient was free of disease 13 months after the SLN biopsy. Case 2 The second case had an SLN with a single, complex, cytokeratin-positive gland with partial apocrine features embedded in the lymphoid stroma zimage 2Bz and zimage 2Cz. The gland ztable 1z Clinicopathologic Features of the Three Cases With SLNs Containing BGIs in Patients With Breast Carcinoma Case 1 Case 2 Case 3 Age (y)/sex 65/F 46/F 56/F Laterality of invasive ductal carcinoma Left Left Right Invasive ductal carcinoma Greatest dimension (cm) Grade I II II Estrogen receptor Positive (2+) ND * Positive (2+) Progesterone receptor Positive (2+) ND * Positive (1+) HER-2/neu Negative by immunohisto- ND * Negative by immunohistochemical analysis chemical analysis No. of SLNs identified No. of SLNs with BGIs Size of SLN with BGIs (cm) , 1.5 BGIs, benign glandular inclusions; SLNs, sentinel lymph nodes. * Not done because there was too little material for further study. 22 Am J Clin Pathol 2008;130: DOI: /JVB8QFQNW5HBN7UJ

3 Anatomic Pathology / Original Article A B C D E F zimage 1z (Case 1) Benign glandular inclusions in a sentinel lymph node (SLN). A, Primary invasive ductal carcinoma of the breast (H&E, 200). B, Frozen section of glandular inclusions in the SLN (H&E, 200). C and D, Permanent section of the SLN. C, The glandular inclusions formed a 1.8-mm aggregate with microcalcifications (H&E, 40). D, Higher magnification of the glandular inclusions showing 2 distinct cell layers and a basement membrane (H&E, 200). E, With the immunoperoxidase method, smooth muscle myosin immunostain strongly highlighted a myoepithelial cell layer surrounding the glands ( 40). F, With the immunoperoxidase method, p63 immunostain was positive for the myoepithelial cells with nuclear staining ( 100). 23 Am J Clin Pathol 2008;130:21-27 DOI: /JVB8QFQNW5HBN7UJ 23 23

4 Peng et al / Glandular Inclusions in Sentinel Nodes A B C D E F zimage 2z (Case 2) Benign glandular inclusion in a sentinel lymph node (SLN). A, Primary microinvasive carcinoma of the breast (H&E, 400). B, The SLN contained a single complex gland with a basement membrane embedded in the lymphoid stroma; part of the glandular epithelium showed apocrine features (H&E, 400). C, With the immunoperoxidase method, the gland was positive for cytokeratin AE1/AE3 immunostains ( 400). Also with the immunoperoxidase method, the gland was encircled by smooth muscle myosin reactivity (D, 400), scattered p63+ nuclei (E, 400), and partial CD10 staining (F, 400). 24 Am J Clin Pathol 2008;130: DOI: /JVB8QFQNW5HBN7UJ

5 Anatomic Pathology / Original Article A B C D E zimage 3z (Case 3) Benign glandular inclusions in sentinel lymph nodes (SLNs). A, Primary invasive ductal carcinoma of the breast (H&E, 200). B and C, The first SLN contained a single small gland lined by bland columnar cells that were positive for cytokeratin AE1/AE3 by the immunoperoxidase method (B, 100; C, 200; courtesy of Vilkesh Jaiswal, MD, Harris Methodist Hospital, Fort Worth, TX). D and E, The second SLN had 2 separate small glands with a bland columnar cell lining and a basement membrane (D, H&E, 40; E, H&E, 100). 25 Am J Clin Pathol 2008;130:21-27 DOI: /JVB8QFQNW5HBN7UJ 25 25

6 Peng et al / Glandular Inclusions in Sentinel No d e s was present only on deeper levels, not the initial sections of the node. Per quality assurance review by a second pathologist before surgical intervention, examination with myoepithelial markers was performed, which revealed that the gland was encircled by smooth muscle myosin reactivity and scattered p63+ nuclei zimage 2Dz and zimage 2Ez. A CD10 immunostain was also positive in half of the gland s myoepithelial layer zimage 2Fz. Based on histologic and immunophenotypic findings, a diagnosis of BGI was established. The patient was free of disease 9 months after the SLN biopsy. Case 3 The third case had 2 SLNs, each with rare, small, cytokeratin-positive glands lined by bland, tall columnar cells and surrounded by thin, fibrous bands zimage 3Bz, zimage 3Cz, zimage 3Dz, and zimage 3Ez. Because these glands were not present on cryosections of frozen sections, these 2 nodes were called negative for metastatic carcinoma intraoperatively. A diagnosis of metastatic carcinoma was made on permanent sections of the SLNs (deeper sections of the frozen sections), and the patient was referred to our institution for further surgical management. We reviewed the surgical pathology materials and requested hold slides and the paraffin tissue blocks from both SLNs for immunostains for myoepithelial markers. Unfortunately, no glands were found on the hold slides or deeper sections. We attempted to destain an original H&E-stained slide from 1 of the SLNs to perform a smooth muscle myosin immunostain, but the immunohistochemical staining with appropriate positive controls failed to work on the destained material. The histologic findings indicated that each of the 2 SLNs exhibited benign endosalpingiosis-like glandular inclusions, which were significantly different from the primary breast carcinoma (Image 3A). The patient was free of disease 6 months after the SLN biopsy. Discussion Although benign epithelial inclusions in axillary lymph nodes are extremely rare in patients with breast carcinoma, differentiating them from metastatic carcinoma can be diagnostically challenging. Our case series report indicates that comparison with the morphologic features of the primary breast carcinoma and using immunohistochemical analysis for myoepithelial markers are important tools in distinguishing BGIs from metastatic carcinoma in SLNs. Common histopathologic features of BGIs in our cases included a lack of cytologic atypia, usually small size (<2.0 mm), presence of squamous or apocrine differentiation, and presence of myoepithelial components and a basement membrane. A false-positive SLN is of particular concern because it can lead to an unnecessary CAND with the risk of associated complications. Complications of CAND such as paresthesias, pain syndromes, limitations of physical activity, and lymphedema occur in approximately 26% of patients 10 and subjectively decrease the patient s quality of life. Quality assurance pathologic review for all positive SLNs, especially for cases with micrometastasis, is warranted before further surgical management. Currently, intraoperative evaluation of SLNs has further streamlined the SLN procedure. However, frozen section diagnosis for BGIs should proceed with caution because misinterpretation of this phenomenon may prompt a CAND at the time of surgery. Norton et al 6 reported a case in which BGIs in an SLN in a patient with ipsilateral breast carcinoma were misinterpreted as metastatic adenocarcinoma at frozen section, and a level I and II CAND was performed at the time of surgery, revealing 23 benign lymph nodes. Histologic and cytologic features and the location of the BGIs can be helpful in an intraoperative setting. Metastatic deposits are often found within the parenchyma of the lymph node or the subcapsular sinus, whereas most BGIs are found within the fibrous tissue of the lymph node. We recommend careful consideration of the morphologic features and the location of these lesions within the node to prevent a falsepositive diagnosis on frozen section. If there is any question that an epithelial lesion in a node might represent a BGI, deferring the frozen section until permanent sections and, perhaps, immunohistochemical stains are examined, is justified. Resetkova et al 5 reported 2 cases of benign epithelial inclusions in nonsentinel axillary lymph nodes in patients without ipsilateral breast cancer (one patient without breast cancer and the other with contralateral breast carcinoma), and the BGIs contained a myoepithelial layer confirmed by immunohistochemical analysis. Fisher et al 7 reported 5 cases with benign inclusions in nonsentinel axillary lymph nodes in patients with ipsilateral breast carcinoma. Of 81 axillary nodes from the 5 cases, 6 contained benign inclusions, and 2 of the 6 nodes from 1 case showed coexisting metastatic carcinoma. Even though BGIs and metastatic carcinoma can occur simultaneously within the same axillary lymph node, comparison of the primary tumor with the intranodal glandular elements and identification of the presence of a basement membrane and the presence of a myoepithelial cell layer are very helpful in the distinction. It is noteworthy that axillary nodal metastasis with a revertant ductal carcinoma in situ (DCIS) pattern in patients with breast cancer described by Barsky et al 11 did not show persistence of the surrounding myoepithelial layers, although the revertant DCIS areas were characterized by an intact and circumferential basement membrane. The majority of the revertant DCIS areas were intermediate to poorly differentiated and exhibited complete concordance with the primary DCIS pattern. 26 Am J Clin Pathol 2008;130: DOI: /JVB8QFQNW5HBN7UJ

7 Anatomic Pathology / Original Article The histogenesis of BGIs is unknown, but embryologic and embolic phenomena have been suggested. 5,12 BGIs in the axillary lymph nodes may arise from mammary structures or axillary subcutaneous eccrine and apocrine glands. 8 Mammary gland origin is the most commonly reported example of BGIs in axillary lymph nodes. In case 1 of our report, the BGIs were positive for estrogen and progesterone receptors and had a myoepithelial layer, indicating the glandular elements were likely of mammary origin. The proximity of the BGIs to the subcapsular sinus and the history of breast biopsies in our case series suggest that they arose via the embolic mechanism of BGI formation. Our case series illustrates 3 distinct types of BGIs in axillary lymph nodes: BGIs with squamous inclusion cyst; BGIs with apocrine metaplasia; and BGIs with columnar cell features, which are similar to predominant types of intranodal benign glandular inclusions previously described in the literature. 4,6-8,10,13 The third case revealed endosalpingiosislike glandular inclusions with a basement membrane that were morphologically similar to those described in a recently published case report. 6 Molecular techniques such as the polymerase chain reaction assay probing for cytokeratin offer even more sensitive methods than immunohistochemical analysis for detecting occult metastasis in SLNs. 14,15 However, false-positive SLNs caused by BGIs can be a particular problem with this method because it does not permit morphologic correlation of a positive result. Therefore, although the polymerase chain reaction assay may become a powerful ancillary tool for the detection of minimal residual disease in SLNs, the possibility of false-positive tests owing to the presence of BGIs needs to be considered. Our 3 cases provide further documentation of the rare occurrence of BGIs in axillary SLNs in patients with breast carcinoma. We hope this report will expand the literature on BGIs in axillary lymph nodes in patients with breast cancer and heighten awareness of its existence and importance among pathologists. It is important to point out that, in the absence of myoepithelial cells and a basement membrane, glands in lymph nodes should be regarded as metastatic carcinoma even if the growth pattern is exceptionally well differentiated. BGIs in axillary SLNs can mimic metastatic carcinoma in patients with breast cancer, leading to a false-positive diagnosis. A conservative approach to the management of such cases is recommended to avoid unnecessary CAND and its associated morbidity. Address reprint requests to Dr Peng: Dept of Pathology, University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Blvd, Dallas, TX References 1. Kempson RL. Consultant case: benign glandular inclusions in iliac lymph nodes. Am J Surg Pathol. 1978;2: Karp LA, Czernobilsky B. Glandular inclusions in pelvic and abdominal para-aortic lymph nodes. Am J Clin Pathol. 1969;52: Schnurr RC, Delgado G, Chun B. Benign glandular inclusions in para-aortic lymph nodes in women undergoing lymphadenectomies. Am J Obstet Gynecol. 1978;130: Edlow DW, Carter D. Heterotopic epithelium in axillary lymph nodes: report of a case and review of the literature. Am J Clin Pathol. 1973;59: Resetkova E, Hoda SA, Clarke JL, et al. Benign heterotopic epithelial inclusions in axillary lymph nodes: histological and immunohistochemical patterns. Arch Pathol Lab Med. 2003;127:e25-e Norton LE, Komenaka IK, Emerson RE, et al. Benign glandular inclusions a rare cause of a false positive sentinel node. J Surg Oncol. 2007;95: Fisher CJ, Hill S, Millis RR. Benign lymph node inclusions mimicking metastatic carcinoma. J Clin Pathol. 1994;47: Fraggetta F, Vasquez E. Epithelial inclusions in axillary lymph node associated with a breast carcinoma: report of a case with a review of the literature. Pathol Res Pract. 1999;195: Holdsworth PJ, Hopkinson JM, Leveson SH. Benign axillary epithelial lymph node inclusions: a histological pitfall. Histopathology. 1988;13: Voogd AC, Ververs JM, Vingerhoets AJ, et al. Lymphoedema and reduced shoulder function as indicators of quality of life after axillary lymph node dissection for invasive breast cancer. Br J Surg. 2003;90: Barsky SH, Doberneck SA, Sternlicht MD, et al. Revertant DCIS in human axillary breast carcinoma metastases. J Pathol. 1997;183: Zhang C, Sung CJ, Gass J, et al. Squamous inclusions cysts with evidence of focal glandular differentiation in an axillary lymph node [letter]. Histopathology. 2005;47: Layfield LJ, Mooney E. Heterotopic epithelium in an intramammary lymph node. Breast J. 2000;6: Nissan A, Jager D, Roystacher M, et al. Multimarker RT-PCR assay for the detection of minimal residual disease in sentinel lymph nodes of breast cancer patients. Br J Cancer. 2006;94: Kurosumi M, Takei H. Significance and problems of histopathological examination and utility of real-time reverse transcriptase polymerase chain reaction method for the detection of sentinel lymph node metastasis in breast cancer. Breast Cancer. 2007;14: From the 1 Department of Pathology and 2 Division of Surgical Oncology, University of Texas Southwestern Medical Center at Dallas. Am J Clin Pathol 2008;130: DOI: /JVB8QFQNW5HBN7UJ 27