Alena Skálová’s research while affiliated with Charles University in Prague and other places

Background: Primary oncocytic salivary gland tumors and oncocytic subtypes of traditionally non-oncocytic salivary gland neoplasms are occasionally encountered in fine needle aspiration specimens, biopsies, and resections. Oncocytes are cells, either non-neoplastic or neoplastic, containing increased numbers of mitochondria resulting in cells with abundant eosinophilic cytoplasm and a low N/C ratio. Summary: A broad range of salivary gland tumors can be oncocytic including oncocytoma, Warthin tumor, mucoepidermoid carcinoma, salivary duct carcinoma, and others, especially those tumors where the oncocytic pattern represents a subtype of neoplasm; the oncocytic pattern can create a diagnostic challenge due to marked similarities in the oncocytic pattern of cells. Key messages: While their microscopic cytologic and histologic features may be similar, these tumors differ intrinsically at the molecular level. Ancillary studies such as immunologic (e.g., androgen receptor for salivary duct carcinoma) and molecular analysis, e.g., FISH for detecting the MAML2 or PLAG1/HMGA2 gene alterations in mucoepidermoid carcinoma and pleomorphic adenoma, respectively, can be used to classify these oncocytic tumors in difficult cases.

Canalicular tumors of the salivary glands have recently emerged as an entity characterized by distinct morphology and recurrent HMGA2 gene rearrangement. In this study, we analyzed 40 cases intending to elucidate their features further. The monophasic or biphasic tumors exhibited a growth pattern of interconnected anastomosing trabeculae and canaliculi, accompanied by a classical pleomorphic adenoma in one-third of the cases. Invasive growth into surrounding adipose tissue was revealed in one case which was, therefore, diagnosed as epithelial-myoepithelial carcinoma. Although the tumor cells uniformly expressed HMGA2 protein in all cases, cytokeratin 7, S100 protein, and SOX10 displayed either diffuse positivity or highlighted the luminal and abluminal cell populations, respectively. Areas with morphological oncocytoid change and AR-immunopositivity of luminal cells were seen in 13/14 (93%) of tested biphasic cases. HMGA2 rearrangement was detected by RNA-sequencing in 30 cases. The most common alteration was an HMGA1::WIF1 fusion, but several novel or rare fusion partners were identified, including ARID2, FHIT, MSRB3 and its antisense variant MSRB3-AS1, IFNG-AS1, and the long intergenic region LINC02389. In addition, FISH revealed HGMA2 break-apart in the remaining 10 cases where targeted sequencing failed to detect any alteration or where RNA sequencing could not be performed. Notably, the loss of the 3'-untranslated region of HMGA2 emerges as the common denominator for the described rearrangements, possibly disrupting its negative regulation by small regulatory RNAs. Awareness of this lesion ensures appropriate diagnosis and clinical management, especially with regard to the possibility of malignant transformation described in this and previous studies.

The pathology of reactive, dysplastic, and neoplastic sinonasal seromucinous glands is complex, and their contribution to tumorigenesis of sinonasal carcinomas remains controversial. In our practice, we have observed the presence of respiratory epithelial adenomatoid hamartomas (REAH) and seromucinous hamartomas (SH) associated with adenoid cystic carcinomas (AdCC) in a subset of cases. In many of these cases, genuine atypical features and dysplastic characteristics of the glands were noted at the interface of SH and AdCC. To investigate this phenomenon further, 88 sinonasal AdCC cases were selected from the authors’ files and analyzed histologically, immunohistochemically, and genetically searching for MYB/MYBL1 and NFIB gene fusions. HPV testing was also performed. Univariate statistical analysis was conducted on our cohort. Thirty-one cases (35%) showed features of atypical sinonasal glands arising in SH (ASGSH) at the SH-AdCC interface, characterized by bilayered epithelium, architectural disarray, mild nuclear polymorphism, and atypia, sometimes with colloid-like material in the lumen. The MYB immunomarker was negative in 14 ASGSHs (with a positive internal control in AdCC cells), while only two cases showed faint and moderate to weak expression of the antibody in ASGSH glands. In 12 cases, the immunostaining of ASGSH could not be properly assessed, while AdCC cells were negative. The immunostaining was not performed in five cases. Our findings suggest that a subset of sinonasal AdCC may originate in a multistep dysplastic process within SH, consistent with an SH-ASGSH-AdCC progression sequence.