Oral mucosal scrapes capture cancer associated microRNA expression consistent with histopathology

Oral mucosal abnormalities considered to have malignant potential may involve a large area of mucosa. Standard care histopathological investigation is invasive, limited to site selection and subject to variation in assessment between pathologists. A quantitative, minimally invasive, rapidly collected, oral mucosal site-specific assessment will assist in decision making and increase diagnostic precision. This study aimed to validate a workflow to analyse oral scrape derived cancer-associated microRNA analysis for mucosal site-specific assessment as a surrogate biomarker to histopathological diagnosis. Forty-one oral scrapes were collected from 33 patients undergoing investigation at the Royal Dental Hospital of Melbourne before mucosal biopsy. RNA from oral scrapes was used to investigate ten cancer-associated microRNAs. An algorithm for categorical high or low-risk based on histopathological diagnosis was developed. The novel risk stratification algorithm utilised two microRNAs and categorised all cases of carcinoma and severe dysplasia as high-risk and accurately distinguished all non-potentially malignant disorders as low-risk lesions. This study provides proof-of-concept that oral scrapes can be predictably collected in a clinical workflow to assess the expression of cancer-associated microRNA specific to an oral mucosal site with minimal invasiveness.