Molecular analysis of pancreaticobiliary fine needle aspirate/duct brushing cytocentrifugation supernatant fluid.

Abstract

e21121 Background: Reliance on minimally invasive sampling techniques (fine needle aspiration (FNA), duct brushing) continues to increase for pancreaticobiliary (PB) pre-cancer/cancer diagnosis. Fundamental limitations include low cellularity, sampling variability, and indeterminate microscopic diagnosis. To address these issues, we developed a novel approach involving DNA mutational profiling of the soluble (or free) DNA in the cytocentrifugation supernatant fluid (supernatant) currently discarded during cytology preparation. Methods: 77 PB specimens underwent combined microscopic/molecular analysis using free DNA in cytology supernatant fluid. Pancreatic FNA and biliary cytobrushes had either surgical pathology or positive cytology as the gold standard. DNA was extracted from 2 mL of the supernatant and from corresponding microdissected cytology cells. Profiling included DNA quantity and quality, and analyses for KRAS mutations and 16 LOH (loss of heterozygosity) mutations. Results: The cell-free supernatant showed significantly higher levels of more intact amplifiable DNA than typical microdissection specimens. All non-neoplastic specimens (n=38) showed no false positive mutations. 37/39 malignant specimens showed mutations (sensitivity 95%, specificity 100%). Mutational profiles of the microdissected cells and the supernatant demonstrated high concordance. However, mutation detection was consistently superior in the supernatant fluid (mean 6.7 mutations) compared to the microdissected cells (mean 4.2). Mutation clonality was also significantly higher in the supernatant. Conclusions: Despite the lack of cells in many PB samples, the post-centrifugation supernatant fluid provides an abundant, robust, and highly representative source of DNA for molecular analysis. Of note, the greater extent and clonality of detected mutations in the supernatant fluid supports it being enriched with neoplastic DNA when cancer is present. This novel approach complements standard cytology practice by increasing sensitivity and specificity of cancer detection while reducing sampling variation.