miRNA profiling of primary lung and head and neck squamous cell carcinomas: Addressing a diagnostic dilemma.

OBJECTIVE: To determine whether microRNA (miRNA) profiling of primary lung and head and neck squamous cell carcinomas could be useful to identify a specific miRNA signature that can be used to further discriminate between primary lung squamous carcinomas and metastatic lesions in patients with a history of head and neck squamous cell cancer. METHODS: Specimens of resected primary head and neck and lung squamous cell carcinomas were obtained from formalin-fixed, paraffin-embedded blocks. Paraffin blocks were sectioned and deparaffinized, and total RNA was isolated and profiled. Quantitative polymerase chain reaction was performed to verify array results. RESULTS: Twelve head and neck and 16 lung squamous cell carcinoma samples met quality control metrics and were included for analysis. Forty-eight miRNAs were differentially expressed (P < .05) between the 2 groups. Of these, 30 were also significantly associated (q < .25) with tumor type in 2 independent sets of primary head and neck and lung squamous carcinomas profiled by The Cancer Genome Atlas consortium, including miR-34a and miR-10a. The ratio of miR-10a and miR-10b was especially predictive of primary cancer site in all 3 data sets, with area under the (receiver operating characteristics) curve values ranging from 0.922 to 0.982. Quantitative polymerase chain reaction confirmed the association of miR-34a expression and the miR-10:miR-10b ratio with tumor type. CONCLUSIONS: MicroRNA expression may be useful for discriminating between head and neck and lung squamous cell carcinomas, including miR-34a and the miR-10a:miR-10b ratio. This differentiation has clinical importance because it could help determine the appropriate therapeutic approach.