P04.03.B Rapid DNA methylation-based classification of pediatric brain tumors from ultrasonic aspirate specimens

Abstract Background Cavitating ultrasonic aspirator (CUSA) devices are commonly used in neurosurgical procedures to carefully debulk tumor from adjacent healthy brain tissue. Here, we explore the feasibility of using minced tissue classify otherwise discarded sample material by DNA methylation according respective World Health Organization (WHO) Classification Tumors Central Nervous System (CNS) low pass nanopore whole genome sequencing. Material and Methods 21 aspirated specimens patients undergoing surgery department pediatric neurosurgery at Charité-Universitätsmedizin Berlin with either newly diagnosed cerebral lesions or pre-treated were processed sequencing generate copy number profiles ad-hoc random forest classification. Results compared microarray-based routine profiling. Tumor purity was assessed. In 19/21 (90.5 %) samples minimum amount 1,000 CpG sites sequenced. 20/21 (95.2 cases variation could be generated matched microarray derived profiles, allowing for identification diagnostically therapeutically relevant pathognomonic alterations. 12/17 (70.6 concordantly classified corresponding diagnosis neuropathological workup. Applying recently defined thresholds nanopore-based classification resulted sensitivity 64.7 % specificity 100 %. Conclusion CUSA referred tumors allows methylation-based WHO CNS acceptable high specificity. Hereby, a promising opportunity accurate time- cost-efficient advanced molecular technique is offered