Abstract 694: Brush biopsy miRNA based oral squamous cell carcinoma detection and localization-overcoming barriers to classifier development

Abstract Cancers of the oral cavity arise from the surface tissues and, despite prominence of these sites, over 50% are not diagnosed until they are in advanced stage. Unfortunately, patients with advanced stage oral cancer have high mortality rates because treatments are usually not successful, complex and have high complication rates. Among the main reasons for delays in diagnosis of squamous cell carcinoma, which make up over 90% of oral cancers, is the resemblance of these lesions to benign lesions. Specifically, oral squamous cell carcinoma (OSCC) is often misdiagnosed as a benign process, most commonly benign leukoplakia (BL) or oral lichen planus (OLP) and is not interrogated for diagnosis confirmation. On the other extreme, patients with BL or OLP that have no malignant potential are subjected to unnecessary stress and surgical interventions when these lesions are mistaken as cancers. Unlike cancers of other organs, no screening protocols or non-invasive diagnostic tests are currently available to reliably detect these cancers early when treatments can save lives. We have developed a noninvasive brush-based tissue collection and miRNA (and not mRNA) analysis method that can differentiate OSCC from other mucosal lesions. This brush based sampling method showed 93% accuracy in differentiating OLP from OSCC based on the expression of as few as 6 miRNAs. We will report on the validation of this classifier with a set of 40 OSCC and OLP samples, and the training of a classifier to differentiate BL from OSCC with over 35 additional samples. We will also discuss the role of technical replicates and clinician calibration in efficient miRNA based classifier development. Citation Format: Madeleine W. Fine, Samira Cheruku, Antonia Kolokythas, Ashleigh Weyh, Joel Schwartz, Beth Miloro, Guy R. Adami. Brush biopsy miRNA based oral squamous cell carcinoma detection and localization-overcoming barriers to classifier development [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 694.