Searching for Mammary Analog Secretory Carcinoma of Salivary Gland Among Its Mimics

Andre Pinto; Vania Nosé; Claudia Rojas; Yao-Shan Fan; Carmen Gomez-Fernandez


Mod Pathol. 2014;27(1):30-37. 

In This Article


The recent identification of mammary analog secretory carcinoma as a primary salivary gland tumor with a characteristic molecular alteration, but with morphologic features that overlap primarily with those of acinic cell carcinoma has prompted several groups to review their archival cases. Acinic cell carcinomas have a broad spectrum of histologic patterns, including solid, microcystic, papillary-cystic, and follicular. Thus, salivary gland tumors with these architectural features, have been historically classified as acinic cell carcinomas, and to a lesser extent, as adenocarcinomas NOS.[3,4,7] Other entities like cystadenocarcinoma and mucoepidermoid carcinoma have also been used as potential diagnoses for what we now understand may represent mammary analog secretory carcinoma.[3]

In determining the diagnostic categories for the retrospective analysis of our archival cases, acinic cell tumors were selected as most of the newly diagnosed mammary analog secretory carcinomas have been shown to be initially classified as such. Cribriform cystadenocarcinomas have also been cited as possible diagnoses for mammary analog secretory carcinomas. We also decided to include cases diagnosed as adenocarcinoma, NOS, as 'unclassifiable' salivary gland adenocarcinomas may be inadvertently placed into this category as a diagnosis of exclusion. We chose not to review other tumors types (polymorphous low-grade adenocarcinoma, mucoepidermoid carcinoma, among others) as these have specific histomorphologic characteristics, which we believe are separate and distinct from those in the classic description of mammary analog secretory carcinoma.

In the review of our cases, we found that the overlap between mammary analog secretory carcinoma and acinic cell carcinoma, in particular, may be quite striking. As an example, one of our cases diagnosed as acinic cell carcinoma was morphologically and immunophenotypically indistinguishable from mammary analog secretory carcinoma but failed to show the characteristic ETV6-NTRK3 gene rearrangement by FISH, underscoring the importance of this ancillary technique in this differential (Figure 4).

Reclassifying tumors is becoming a regular exercise for the modern pathologist. Supported by the increasing sophistication and feasibility of molecular techniques, this practice is becoming more and more common on a routine basis, and this may or may not have significant clinical implications for the patient. When, for instance, two neoplasms that have distinct genetic profiles, but similar therapeutic options and outcomes, are separated, no major impact is produced clinically. However, if tumors are biologically distinct enough to the point that the treatment will differ, we as pathologists have a major role in patient care.

This phenomenon still seems unclear when analyzing the behavior of mammary analog secretory carcinoma because of the paucity of cases reported. A male predilection is undoubtedly noted,[1,3–5] and differs from acinic cell carcinoma, which has a slight female predominance. Parotid is the salivary gland most commonly involved,[1,3,5] although these tumors may occur in other major and minor salivary glands. A recent study has suggested that 'acinic cell carcinomas' that occur outside of the parotid are actually mammary analog secretory carcinomas.[15] A higher incidence of nodal metastasis and poorer disease-free survival seem to occur when compared with acinic cell carcinoma.[3,4]

Our data revealed a higher predominance of these tumors in men, which correlates with the current literature. Immunohistochemistry for S-100 seems to be a very useful technique for distinguishing mammary analog secretory carcinoma, typically positive for this marker, from acinic cell carcinoma, reported to be positive in only 10% of cases.[16,17] The S-100 staining pattern for mammary analog secretory carcinoma can range from partial to diffuse and from weak to intense. However, it is important to remember that this marker has to be used in conjunction with morphology, as many other salivary gland tumors like monomorphic and pleomorphic adenomas, polymorphous low-grade adenocarcinomas, and adenoid cystic carcinomas can be reactive for this immunostain.[18]

Similar to its mammary counterpart, mammary analog secretory carcinomas are positive for mammaglobin.[1,6,19] Although reportedly negative in most cases of acinic cell carcinoma,[1,15] this marker has not been widely tested among other salivary gland tumors. Table 3 contains a review of studies about the expression of S-100 and mammaglobin in mammary analog secretory carcinomas.

ANO1, also known as DOG1, was recently shown to be an indicator of salivary acinar cell and intercalated duct differentiation.[9] It tends to be positive in acinic cell carcinoma, with a characteristic apical-luminal membranous and intercalated duct staining, and negative in mammary analog secretory carcinoma. Therefore, this stain may be used as a negative marker in an immunohistochemical panel for distinguishing cases with morphologic overlap between mammary analog secretory carcinoma and acinic cell carcinoma. All the cases initially classified as acinic cell tumors in our series demonstrated negative immunoreactivity for this marker, and were re-classified as mammary analog secretory carcinomas after FISH studies showed a positive ETV6 translocation.

Of note, an unusual finding in our study was the presence of ETV6 amplification in tumor cells seen in one case, and deletion of this gene in the majority of cells in another. Both tumors were present in the parotid gland of male patients. The significance of these genetic alterations remain unclear, however, they may indicate that alternative pathways involving the ETV6-NTRK3 gene could be related to the pathogenesis of this tumor, assuming that those cases represent mammary analog secretory carcinoma.