Salivary gland tumors include numerous subtypes that vary from benign to highly aggressive, with many showing overlapping histopathological features that can make diagnosis challenging. Most subtypes express driver fusion genes that are tumor specific, and detection of such fusions is useful for differentiating amongst specific diagnoses, determining appropriate tumor grading, and guiding effective treatment. Currently, fusions can be detected by FISH, RT-PCR or through next-generation sequencing approaches, all of which are highly effective methodologies but can be costly or time consuming. We developed a rapid NanoString nCounter platform-based assay to detect salivary gland tumor fusions using a combination of fusion junction-specific probes and an approach through differential exon expression analysis. The assay includes 68 junction-specific probes and analysis of exon expression across 9 fusion-associated genes in a single multiplex assay. Out of 55 retrospective and 171 prospective cases assayed, we accurately detected the majority of cases of pleomorphic adenoma, adenoid cystic carcinoma, cribriform adenocarcinoma, clear cell carcinoma, secretory carcinoma and NUT-rearranged carcinoma, including cases of these tumor types arising in non-salivary gland sites, with the major drawback being an inability to detect MAML2-containing mucoepidermoid samples. With mucoepidermoid carcinoma excluded, the assay shows an overall sensitivity of 96.1% and specificity of 100%. We show that the majority of salivary gland tumor fusions can be effectively detected with a single rapid NanoString based assay, which can serve as a useful adjunctive tool for routine diagnostic head and neck pathology. The assay is low cost with a rapid turnaround time (30h total assay time per sample batch, with minimal technician input required) compared to alternate detection methods.
Read full abstract