Intraductal carcinoma (IDC) of the salivary glands is an uncommon and enigmatic tumor, our understanding of which is rapidly evolving. Recent studies have demonstrated multiple IDC subtypes and consistent gene fusions, most frequently involving RET. Because IDC is a ductal proliferation surrounded by flattened myoepithelial cells, it was previously presumed to be analogous to breast ductal carcinoma in situ, but recent evidence has shown that the myoepithelial cells of fusion-positive IDC harbor the same genetic alterations of the ductal cells and are therefore neoplastic. In addition, there are rare reports of fusion-positive IDC with overt areas of irregular invasion lacking myoepithelial cells, but this phenomenon is not well documented or understood. This study aims to better characterize these frankly invasive carcinoma ex-IDC. All cases of frankly invasive carcinoma ex-IDC were obtained from the authors' files. Inclusion criteria included a component of concurrent or antecedent IDC and/or a fusion known to be associated with IDC. Immunohistochemistry (S100, SOX10, mammaglobin, androgen receptor, p63, p40) and molecular analysis (targeted RNA sequencing or large panel DNA next generation sequencing) was performed. Clinical follow-up was obtained from medical records. Ten cases of frankly invasive carcinoma ex-IDC were identified. The tumors occurred in 8 men and 2 women ranging from 33 to 82 years (mean, 66.3). All but one case arose in the parotid gland. In 4 cases, the IDC component was intercalated duct type. It was mixed apocrine/intercalated duct in two, and in the remaining 4 cases, no residual IDC was identified. The frankly invasive carcinomas were remarkably heterogeneous, ranging from minimally to widely invasive beyond the confines of the IDC, low-grade to high-grade, with morphologies that varied from duct-forming to those having clear cell or sarcomatoid features, to frankly apocrine. The original diagnoses for these cases were (adeno) carcinoma, not otherwise specified (n = 6), salivary duct carcinoma (n = 3), and secretory carcinoma (n = 1). All cases harbored fusions: NCOA4::RET (n = 6), TRIM33::RET (n = 2), TRIM27::RET (n = 1), and STRN::ALK (n = 1). Clinically, one tumor recurred locally, cervical lymph node metastases occurred in five patients, and distant metastasis later developed in four of these patients. Our findings highlight striking diversity in frankly invasive carcinomas that arise from fusion-positive IDC, a tumor which may serve as a precursor neoplasm like pleomorphic adenoma. These carcinomas vary in their extent of invasion, grade, histologic appearances, and clinical behavior. Importantly, in contrast to pure IDC, which is believed to be indolent, many frankly invasive cases were aggressive. Because RET and ALK fusions are targetable, it is important to recognize the broad spectrum of frankly invasive carcinomas that can arise from IDC, particularly because some cases are completely overrun or recur without any recognizable IDC component. These results suggest fusion analysis may be of clinical benefit on any salivary gland (adeno) carcinoma, not otherwise specified or salivary duct carcinoma.
Read full abstract