Evaluation of DNA Methylation Array for Glioma Tumor Profiling and Description of a Novel Epi-Signature to Distinguish IDH1/IDH2 Mutant and Wild-Type Tumors

Abstract

Molecular biomarkers, such as IDH1/IDH2 mutations and 1p19q co-deletion, are included in the histopathological and clinical criteria currently used to diagnose and classify gliomas. IDH1/IDH2 mutation is a common feature of gliomas and is associated with a glioma-CpG island methylator phenotype (CIMP). Aberrant genomic methylation patterns can also be used to extrapolate information about copy number variation in a tumor. This project's goal was to assess the feasibility of DNA methylation array for the simultaneous detection of glioma biomarkers as a more effective testing strategy compared to existing single analyte tests.

Methods: Whole-genome methylation array (WGMA) testing was performed using 48 glioma DNA samples to detect methylation aberrations and chromosomal gains and losses. The analyzed samples include 39 tumors in the discovery cohort and 9 tumors in the replication cohort. Methylation profiles for each sample were correlated with IDH1 p.R132G mutation, immunohistochemistry (IHC), and previous 1p19q clinical testing to assess the sensitivity and specificity of the WGMA assay for the detection of these variants.

Results: We developed a DNA methylation signature to specifically distinguish a IDH1/IDH2 mutant tumor from normal samples. This signature is composed of 11 CpG sites that were significantly hypermethylated in the IDH1/IDH2 mutant group. Copy number analysis using WGMA data was able to identify five of five positive samples for 1p19q co-deletion and was concordant for all negative samples.

Conclusions: The DNA methylation signature presented here has the potential to refine the utility of WGMA to predict IDH1/IDH2 mutation status of gliomas, thus improving diagnostic yield and efficiency of laboratory testing compared to single analyte IDH1/IDH2 or 1p19q tests.

Keywords: 1p19q co-deletion; IDH1/IDH2 epi-signature; glioma; whole-genome methylation array.

Figure 1

Hierarchical clustering of the top 11 CG sites (column) that were differentially methylated (epi-signature) in IDH-positive group versus IDH-negative group. Row; S: individual samples. G: group (blue: IDH-negative, red: IDH-positive, pink: IDH2 variants by NGS, orange: IDH1 variant by NGS).

Figure 2

Hierarchical clustering of discovery and replication cohorts. The top 11 CG sites (column) that were differentially methylated (epi-signature) in IDH-positive versus IDH-negative group. Row; S: individual samples. G: group (blue: IDH-negative, red: IDH-positive, pink: IDH2 variants by NGS, orange: IDH1 variant by NGS, green: replication cohort samples).

Figure 3

Genome-wide copy number analysis from methylation array data (Infinium MethylationEPIC BeadChip). All chromosomes are represented in the X-axis. Normalized copy number is represented in Y-axis (baseline = 0). Green dots represent probes with normalized copy number above 0 while red dots represent probes with normalized copy number below 0. Positions of probes for 1p, 19q, EGFR, MGMT and MET are indicated by arrows. (A) Example of genome-wide copy number analysis with 1p19q co-deletion (Sample 44). Instances of 1p19q co-deletion are highlighted by black arrows. (B) Example of genome-wide copy number analysis with EGFR amplification. EGFR amplification is highlighted by a black arrow.