Abstract 3658: DNA methylation of circulating tumor educated leukocytes predicts IDH1/2 mutation status in adult patients with diffuse gliomas

Abstract

Abstract INTRODUCTION: Identification of IDH1/2 mutations and association with better outcome revolutionized diagnosis and management of patients with gliomas. Liquid biopsy can detect driver mutation from blood, however, detection of circulating tumor DNA or circulating tumor cells from patients with brain tumors is difficult due to the low amount of circulating tumor material in the peripheral blood. We hypothesized that brain tumors alter the epigenetic profiles of circulating leukocytes and sought to identify specific methylation patterns in peripheral blood that would identify molecular subtypes of gliomas. METHODS: We analyzed brain tumors and matched whole peripheral blood collected in EDTA tubes from 63 patients with primary gliomas. Tumor IDH1/2 mutation status was analyzed using next-generation sequencing and DNA methylation was analyzed using Illumina Human EPIC array. A complete blood count with differential (CBC) was analyzed to control for variations in CBC profiles. Peripheral blood methylation profile was compared to blood of 40 individuals with no known history of brain tumors. DNA methylation of Acute Myeloid Leukemia (AML) samples (source: TCGA) was used to identify overlapping DNA methylation changes induced by IDH1/2 mutations. Methylation data were analyzed with the R Bioconductor package minfi, including quality control, data normalization and differentially methylated CpG site analysis. Subsequent filtering was performed using a p-value cutoff = 0.01 and a minimal mean difference of the Beta-value of 0.1. RESULTS: Peripheral blood methylation profiles of patients with brain tumors are distinctly different from normal controls. DNA methylation of IDH1/2 mutated brain tumors and AML showed significant overlap (enrichment = 3, p-value 2.2x10-16) and DNA methylation of IDH1/2 mutated AML cells showed significant overlap with circulating leukocytes from patients with IDH1/2 mutated brain tumors (enrichment = 6.5, p-value 2.2x10-8 ). Using this classifier supported by AML data, DNA methylation of circulating leukocytes was able to predict IDH1/2 mutation status of diffuse gliomas with 98% accuracy, nearly perfectly separating IDH status by clustering. CONCLUSIONS: Diffuse gliomas induce epigenetic changes in the methylome of the leukocytes, which resembles methylation changes of IDH1/2 mutated AML. IDH wildtype and mutant gliomas can be distinguished with high accuracy by profiling the epigenome of leukocytes. Our study demonstrates the potential of peripheral blood DNA methylation assessment for non-invasive diagnosis of brain tumors. Citation Format: Andreas Kloetgen, Jonathan Serrano, Seema Patel, Christopher Bowman, Guomiao Shen, David Zagzag, Matthias A. Karajannis, John G. Golfinos, Dimitris Placantonakis, Aristotelis Tsirigos, Andrew S. Chi, Matija Snuderl. DNA methylation of circulating tumor educated leukocytes predicts IDH1/2 mutation status in adult patients with diffuse gliomas [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr 3658.