Introduction
Branchioma is a rare site-specific benign tumor with an adult male predominance, typically occurring in the lower neck and combining bland squamoid epithelial with variable mesenchymal elements, reminiscent of thymic tissue, hence the widely used historical terminology “ectopic hamartomatous thymoma” [
1]. Since its first description by Smith and McClure in 1982 [
2] and later by Rosai et al. [
3], branchioma was reported in many case series and single case reports under different names such as ectopic hamartomatous thymoma, branchial anlage mixed tumor or thymic anlage tumor, and biphenotypic branchioma [
4‐
6]. In the upcoming WHO classification of the head and neck tumors that is currently in beta version, the name “branchioma” is adopted for this neoplasm [
7].
Branchioma is composed of endodermal and mesodermal lineage derivatives, particularly, of epithelial islands, spindle cells, and mature adipose tissue [
7], but no tissue of genuine thymic origin or thymic differentiation was described so far [
4‐
6]. Epithelial cells are arranged in non-keratinizing epithelial islands, cysts, or glandular structures, sometimes with a peripheral rim of residual myoepithelial cells. The spindle cells are plump and arranged in haphazard, storiform, or fascicular patterns with intermingled thick collagen. Sometimes, the spindle-shaped cells grow in solid sheets with interspersed adipose tissue, imitating a pleomorphic adenoma-like morphology [
8]. Both the epithelial and spindled components most likely originate from the same progenitor cell [
4] and both show positivity for pan-keratins as well as frequent nuclear androgen receptor (AR) expression, which might explain its prevailing occurrence in postpubertal men [
9]. However, the spindle cells also show CD34, p63, and SMA positivity. The latter 2 markers also decorate the myoepithelial cell component, when present [
10].
The herein presented case expands the morphological, immunohistochemical, and molecular genetic spectrum of this rare tumor entity.
Discussion
Branchioma is a rare lower neck tumor with 85 reported cases to date. The origin of branchioma is controversial; postulated origin from ectopic thymic remnants in the lower neck justified the original terminology “ectopic hamartomatous thymoma” [
3]. However, recent investigations point rather toward the branchial cleft apparatus as the most probable origin [
4,
5]. Branchioma is not a hamartoma but a true neoplasm most likely derived from the rudimental embryological structures of endoderm and mesoderm, which are responsible for its triphasic morphology including epithelial cells, spindle cells, and adipose tissue [
4]. The proportion of each of these components varies from case to case. The spindle cells have haphazard to fascicular arrangement, the epithelial cells grow in cystic, solid, or pseudoglandular patterns, and the adipose component is dispersed throughout the tumor. Both spindle and epithelial cells are immunoreactive with high-molecular-weight cytokeratins, myoepithelial markers (p63 and p40), and the androgen receptor [
9]. The spindle cells show reactivity with CD34 and SMA [
6,
15] and partially resemble the stroma of spindle cell lipoma [
14]. Similar to the latter, immunohistochemical RB1 loss was observed in our case, but no
RB1 genetic alteration was detected. This suggests point mutations and/or other alternate molecular or epigenetic mechanisms responsible for RB1 loss that are not detectable by the FISH method.
Molecular findings in branchioma have been reported in only two studies [
8,
10]. One study has looked for
PLAG1 rearrangements to assess distinctness of branchioma from pleomorphic adenoma; no rearrangements have been detected in the 4 analyzed branchiomas [
8]. Another study investigated 3 branchiomas using a custom, targeted NGS panel including 1385 pan-cancer-related genes. A hotspot
HRAS (pGln61Lys) mutation was found in one case of branchioma with intraductal type carcinoma, whereas no definitive oncogenic drivers or copy number alterations were found in the other two cases [
10]. We herein expand on the spectrum of molecular findings in branchioma, in which we detected 5 pathogenic/likely pathogenic gene mutations, particularly two
MSH6 mutations, two
PTEN mutations, and one
KRAS alteration, indicating molecular heterogeneity in branchioma. Admittedly, we did not perform microsdissection and separate molecular testing of the different tumor components. However, lack of atypia and proliferative features in the nested/organoid (neuroendocrine-like) component suggests it probably represents a morphological variant of the epithelial component. The presence of the retained MSH6 expression in the context of its molecular genetic alteration is not surprising. Loss of MMR proteins precludes heterodimerization of MLH1-PMS2 and MSH2-MSH6. The loss of MSH6 immunoexpression is related to gene mutation with loss of the epitope for the MSH6 antibody. In our case, there was an altered
MSH6 gene but the antibody epitope for MSH6 defective protein was probably retained which resulted in positive nuclear immunoexpression.
Moreover, our case adds to the morphological (prominent neuroendocrine-like nested/organoid features) and immunophenotypic (CD34 expression combined with RB1 loss) heterogeneity/pitfalls related to the differential diagnosis of branchioma. The nested/organoid component retained the epithelial immunophenotype and did not express any of neuroendocrine markers, consistent with a morphological variant of the epithelial part of the tumor and not a dedifferentiation or transdifferentiation. Weissferd and Moran described a series of thymomas with histological neuroendocrine-like differentiation and pancytokeratin positivity but none of the neuroendocrine markers was positive [
16], a feature analogous to our current case. Differentiation between thymomas and branchiomas is based on morphology and IHC level. Thymomas are positive for cytokeratins and p63 but show admixture of immature T cell lymphocytes according to subtype and are positive for PAX8 in 55% of cases. However, they are negative with CD34 and SMA.
The lateral neck location together with biphasic tumor cell morphology requires a careful diagnostic approach, especially in fine-needle aspiration biopsies or small biopsy samples. The clinical and radiological differential diagnoses of the adult lateral or anterior neck masses must consider processes of developmental, infectious, or neoplastic nature [
17], keeping in mind that over 75% of lower neck masses are likely malignant, mostly metastasis of squamous cell carcinoma or lymphoma [
17]. The histological and immunohistochemical differential diagnosis also includes primary or secondary tumors of the epithelial, mesenchymal, neuroepithelial, or neuroendocrine tumor origin.
The combination of cervical localization, variable spindle cell histology, CD34 positivity, and loss of RB1 immunoexpression places branchioma, spindle cell lipoma (SCL), and spindle cell–predominant trichodiscoma (SCPT) in the most common differential diagnosis. SCL is usually localized in the subcutis of the nuchal area of elderly men and is characterized by frequent chromosome 13 and/or 16 deletions, CD34 expression, and RB1 immunonegativity [
18]. Androgen receptor expression is a feature shared by both branchioma and SCL, possibly explaining the male predominance in both [
19]. Spindle cell–predominant trichodiscoma is mainly face-localized hamartoma of the mantle zone of hair follicle (AKA mantleoma) [
20]. In a study by Michalová et al., the authors described 6 cases of SCPT with heterozygous deletion of
RB1 gene, while 18 of 19 cases showed loss of RB1 staining in spindle cells; the morphology of SCPT with
RB1 deletion was indistinguishable from tumors without this genetic alteration [
21].
The loss of 13q14 especially in the RB1 gene is also common in other soft tissue tumors (e.g., cellular angiofibroma or myofibroblastoma of the breast), which together with the overlapping morphology and IHC (loss of RB1 and CD34 positivity) supports the hypothesis of a spectrum of genetically related 13q/RB1 family tumors. Based on our results, branchioma might represent another potential member of this group. However, studies on larger cohort of branchiomas are needed to investigate this possibility.
Finally, the admixture of epithelial elements and spindle cells might suggest biphasic synovia sarcoma (BSS). Indeed, rare branchioma cases in our experience have been initially judged as low-grade BSS. However, the triphasic morphology, the characteristic location of the tumor, and the immunophenotype are distinctive and rule out BSS. The SS18 immunohistochemistry and molecular testing can help to resolve the issue in equivocal cases or in cases with unusual morphology and lack of lipomatous component.
In summary, we reported an unusual case of branchioma with neuroendocrine-like morphology lacking nuclear RB1 expression and harboring several pathogenic mutations. All these findings underline the wider differential diagnosis of this unusual variant of a rare tumor entity. Additional studies of larger cohorts of branchiomas are needed to investigate whether this immunophenotype and molecular background represent a recurrent feature of these tumors.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.