Hidradenoma in Axillary Lymph Nodes

Abstract

Axillary lymph nodes are a common site of metastatic carcinoma, particularly from the breast. Epithelial inclusions within axillary lymph nodes are uncommon but well recognized, most frequently encountered as small, incidental clusters of mammary-type epithelium; squamous, Müllerian, and mixed types are also sporadically identified. Hidradenoma is a benign adnexal neoplasm classically composed of epidermoid/squamoid, intermediate/basaloid, and mucinous cells that most commonly originates in the skin but can arise in the breast parenchyma, with morphologic and genetic features that overlap with mucoepidermoid carcinoma. Rare instances of hidradenomas within pelvic and axillary lymph nodes have been reported, raising the specter of metastasis. We report 3 cases of axillary lymph node hidradenoma presenting without a concurrent breast or cutaneous lesion. Morphology, immunohistochemistry, and molecular studies were evaluated. Two cases were found to harbor CRTC1::MAML2 rearrangements, while a rarer CRTC3::MAML2 fusion was identified in a third case. Following excision, the 3 patients were without evidence of residual disease, with clinical follow-up of ∼2 years each. Recognition of this entity is essential to avoid misclassification as metastatic carcinoma and prevent overtreatment.

Uncommonly, axillary lymph nodes harbor epithelial inclusions that may be misinterpreted as metastatic breast carcinoma.¹ Hidradenoma is a benign adnexal neoplasm classically composed of epidermoid/squamoid, intermediate/basaloid, and mucinous cells that most commonly originate in the skin but can arise in breast parenchyma, with morphologic and genetic features that overlap with mucoepidermoid carcinoma (MEC).² Further complicating this distinction, hidradenomas have been identified within lymph nodes, both in cases with synchronous cutaneous hidradenoma and in isolation, raising the specter of metastasis.^3–6 Alternatively, these isolated nodal lesions may be primary to the lymph node, perhaps arising from a preexisting epithelial inclusion.^3–5 Most reported nodal hidradenomas in the literature have involved inguinal nodes (n=6), with 3 of those cases in isolation, without a concurrent or prior cutaneous lesion.^3–5 Less frequently, axillary nodes are involved (n=2), with 1 reported in isolation.³ All reported cases demonstrated benign clinical behavior.^3–6

Breast MEC, salivary gland MEC, and breast and cutaneous hidradenomas harbor CRTC1::MAML2 and rarer CRTC3::MAML2 rearrangements.^2,7–12 One isolated inguinal lymph node hidradenoma was reported positive for MAML2 break-apart fluorescence in situ hybridization (FISH),⁴ but the specific fusion partner was not identified. Additional genetic studies of nodal hidradenomas are lacking. Herein, we report 3 cases of axillary nodal hidradenoma without concurrent breast or cutaneous lesions, expanding upon the sole prior report of this phenomenon.³ Moreover, MAML2 rearrangements and corresponding fusion partners were identified in each case using in situ hybridization (ISH) and/or next-generation sequencing (NGS), including both CRTC1 and rarer CRTC3; this is the first report of fusion partners in nodal hidradenomas at any site.

MATERIALS AND METHODS

Case Selection

Three patients with axillary nodal hidradenoma (NH1-NH3) were identified from consultation cases at the authors’ institutions. Morphology, immunohistochemistry, and molecular studies were evaluated to establish the diagnosis. All patients underwent excision of the involved node; 1 patient (NH2) underwent a core biopsy preceding excision.

Immunohistochemistry and In Situ Hybridization

The following antibodies were used for immunohistochemistry: CK5/6 (D5/16B4, 1:50; Cell Marque, Rocklin, CA), ER (SP1, undiluted; Ventana, Tucson, AZ), GATA3 (L50-823, 1:50; Biocare Medical, Concord, CA), Ki67 (MIB1, 1:200; DAKO, Santa Clara, CA), MUC4 (8G7, 1:200; Cell Marque), p63 (BC4A4, 1:50; Biocare Medical), S100 protein (polyclonal, 1:1000; DAKO), SOX10 (BC34, 1:50; Biocare Medical), and TRPS1 (PAS-845874, 1:500; ThermoFisher, Waltham, MA). Antigen retrieval was as follows: for CK5/6, GATA3, Ki67, MUC4, p63, and SOX10, BOND ER2 (Leica Biosystems, Richmond, VA); for ER and TRPS1, Ventana CC1 (Ventana); and for S100 protein, enzyme (Leica Biosystems). The BaseScope RNA ISH assay (Advanced Cell Diagnostics, Newark, CA) targeting the recurrent breakpoint of either CRTC1::MAML2 (# 1195241) or CRTC3::MAML2 (# 1240501) was performed on NH1 and NH2 as previously described.⁷ Due to limited tissue, CRTC1::MAML2-specific FISH was alternatively performed on NH3 as previously described.⁷ RNAscope RNA ISH (Advanced Cell Diagnostics) targeting LINC00473 (# 442558) was performed on NH1 and NH2 as previously described (here, on the Leica platform).⁷

MAML2 Fusion Analysis

NH2 was subjected to the Archer FusionPlex Pan Solid Tumor v2 panel (Integrated DNA Technologies, Boulder, CO), with methods previously described.² NH1 was subjected to whole transcriptome sequencing. Briefly, tumor RNA was extracted from formalin-fixed, paraffin-embedded tissue blocks and used for complementary DNA synthesis. Library preparation was performed using the KAPA RNA HyperPrep Kit with RiboErase (Roche, Indianapolis, IN) according to the manufacturer’s specifications, and resulting libraries were sequenced on a NovaSeq 6000 (Illumina, San Diego, CA). Sequence reads were mapped to the reference human genome build GRCh37 (hg19). Fusion variant calling was performed using a custom pipeline including STAR-Fusion, Pizzly, Arriba, CICERO, EricScript, and FusionCatcher.

RESULTS

Three patients with axillary nodal hidradenoma were identified: a 57-year-old man with a 3.8 cm lesional node (NH1), a 65-year-old woman with a 2 cm node (NH2), and a 41-year-old man with a 2.1 cm node (NH3). No patients reported concurrent cutaneous or breast lesions; NH1 had a history of a benign ipsilateral axillary skin lesion removed several years prior (not available for review). All underwent excision of the involved node, with histopathology demonstrating hidradenoma within the lymph node parenchyma (Fig. 1). Clinical follow-up demonstrated no evidence of residual disease for each patient, at 20, 28, and 24 months after surgery, respectively.

FIGURE 1.

Hidradenomas within axillary lymph nodes. A, Ultrasound of NH1 revealed a 3.8 cm predominantly cystic mass with irregular hypoechoic peripheral soft tissue (arrow) in the right axilla. The radiologic impression was that of a necrotic mass or lymph node. B and C, Low-power views demonstrate lesional tissue within lymph node parenchyma: NH1 (B) and NH3 (C). D and E, Intermediate and epidermoid cells are present in variable proportions: NH3 (D) and NH1 (E). F, A rare mucinous cell (arrow) was identified in 1 case. G, Clear cell change can be prominent. A minor population of characteristic glandular/lining cells (arrows) is usually seen. H, Hyalinized stroma was variably present (NH1 shown in F to H).

Morphology, immunohistochemistry, and molecular studies were evaluated (Figs. 2, 3). Each case was characterized by a multinodular, solid proliferation with variable papillary architecture, often embedded in a hypocellular, hyalinized stroma. Epidermoid and intermediate cells predominated in various proportions; nuclear pleomorphism was absent. A rare mucinous cell was only identified in 1 case. All exhibited variable clear cell change, and 2 demonstrated a characteristic lining/glandular population. Mitotic activity was inconspicuous (0-1 mitotic figure per 10 high-power fields), and necrosis was absent. Immunohistochemistry for high-molecular-weight keratin and p63 typically highlighted the epidermoid and intermediate cell populations, and the lining/glandular cells were positive for GATA3. Estrogen receptor, SOX10, S100 protein, and MUC4 were negative, and the Ki67 proliferation index was universally low (<10%).

FIGURE 2.

Immunohistochemistry and in situ hybridization on axillary lymph node hidradenomas. A, p63 immunohistochemistry diffusely highlights the epidermoid and intermediate cells of nodal hidradenoma (H&E inset). B, GATA3 conversely highlights the lining/glandular cells. C, Fluorescence in situ hybridization with green- and orange-labeled probes spanning the CRTC1 (19p13) and MAML2 (11q21) loci, respectively, confirmed a CRTC1::MAML2 fusion indicated by the yellow signals. D, LINC00473 RNAscope was positive in CRTC1/3::MAML2-harboring hidradenomas (NH2 shown in D, and NH3 shown in A to C).

FIGURE 3.

Cytomorphologic features and immunohistochemical and molecular results of axillary lymph node hidradenomas.

FISH, RNA ISH, and/or fusion testing by NGS was performed to identify and characterize MAML2 rearrangements. CRTC1 exon 1 fused to MAML2 exons 2 to 5 was identified in 2 cases (1 by NGS and BaseScope-ISH; another by CRTC1::MAML2 fusion-specific FISH); a third case was found to harbor the alternative partner CRTC3 (by NGS and BaseScope-ISH). LINC00473 RNA ISH (RNAscope) was positive in 2/2 cases (1 CRTC1::MAML2, 1 CRTC3::MAML2).

DISCUSSION

The identification of epithelial proliferations within axillary lymph nodes can carry significant clinical implications. In routine practice, such findings prompt consideration of metastatic breast carcinoma, potentially resulting in pathologic upstaging. If detected in a core needle biopsy (CNB), neoadjuvant systemic therapy and altered surgical management may be considered. More commonly reported in inguinal lymph nodes, hidradenoma involving axillary lymph nodes poses a significant diagnostic challenge, as it combines a nodal location, epithelial differentiation, and an immunohistochemical profile that overlaps with bona fide breast carcinoma.

In addition to breast malignancies, cutaneous carcinomas of the trunk and upper extremities, and lung carcinoma may also metastasize to axillary lymph nodes. Primary breast and lung mucoepidermoid carcinomas are rare but well-documented, and these sites proximal to the axilla are the most common extrasalivary sites of MAML2-rearranged MEC. Metastasis and/or mucoepidermoid carcinoma from the breast or other site represents a genuine consideration when evaluating a MAML2-rearranged epithelial lesion in the axilla, and clinical correlation is essential in this context. Furthermore, the distinction between hidradenoma and hidradenocarcinoma can be challenging; features such as nuclear pleomorphism, elevated mitotic activity and Ki67 labeling index, and necrosis (absent in this case series) are more frequently identified in hidradenocarcinomas.¹³ If there is no evidence of a concurrent or prior separate lesion, hidradenoma primary to the axillary lymph node should be considered. MAML2 rearrangement, while strongly associated with MEC, should not be automatically interpreted as evidence of malignancy in isolation. Detection by ISH or NGS can confirm a diagnosis of hidradenoma or MEC, but importantly does not distinguish between them. This distinction is likely not possible in the limited tissue of a CNB; we recommend a diagnosis of “MAML2-rearranged neoplasm” in this setting, deferring final classification to excision with clinical and imaging correlation.²

We recently reported on the use of RNA ISH to detect the specific breakpoint of CRTC1::MAML2 or CRTC3::MAML2 by BaseScope, or the long noncoding RNA LINC00473 by RNAscope, as a downstream surrogate of MAML2 fusion.⁷ These chromogenic markers demonstrated 92% to 93% sensitivity and 99% to 100% specificity in the context of salivary gland, breast, and cutaneous CRTC1/3::MAML2-rearranged neoplasms.⁷ Anecdotally, we have had cases with scant tumor in a CNB that did not allow for extensive ancillary workup, and on which only LINC00473 RNA ISH was possible, given limited tissue. A positive result was strongly suggestive, and the final diagnosis of MEC was confirmed on excision. These findings underscore the importance of integrating molecular and surrogate marker results within an appropriate morphologic and clinicopathologic framework.

CONCLUSION

Hidradenoma should be recognized as a rare but important benign epithelial neoplasm that may involve axillary lymph nodes, either arising from epithelial inclusions or representing limited lymphatic dissemination without malignant behavior. Awareness of this entity is particularly critical in the contemporary era of routine molecular testing, where detection of a MAML2 rearrangement may otherwise prompt a diagnosis of metastatic carcinoma. To our knowledge, this series adds 3 cases to only 1 previous report of axillary hidradenoma without a concurrent cutaneous lesion.³ Moreover, this is the first report to identify MAML2 fusion partners in nodal hidradenomas of any site, including both CRTC1 and the rarer CRTC3. This series adds to the evidence of the seemingly benign clinical behavior of isolated nodal hidradenomas; recognition of this adnexal neoplasm and appropriate workup can prevent misdiagnosis as malignancy and overly aggressive therapeutic intervention.