Detection of Diagnostic and Prognostic Methylation-based Signatures in Liquid Biopsy Specimens from Patients with Meningiomas

A preeminent and currently unfulfilled challenge in deciding between surgical- or surveillance-oriented management of meningioma lies in identification of tumors which have greater potential to recur or behave aggressively. Meningioma DNA methylation signatures are suitable to predict recurrence risk using surgically resected tumors, but have not been explored through noninvasive methods such as liquid biopsy. In this work, we analyzed the DNA methylomes in liquid biopsy and tissue specimens from patients with meningiomas (n= 83 serum; 10 plasma and 82 tissue specimens) as well as from other central nervous system entities (n=141) generated in house, or retrieved from public repositories (n=675). We identified meningioma-specific methylation signatures suitable for the formulation of accurate and generalizable machine learning classifiers able to detect meningioma and predict their risk to recur using serum or tissue specimens. Here, we show liquid biopsy as a potential noninvasive tool to diagnose, prognosticate and guide personalized management of patients with meningiomas.We have provided whole genome bisulfite sequencing paired-end reads associated with the aforementioned investigation for (n=10) liquid biopsy serum meningioma samples.