Artikel
Mapping of genome-wide methylation pattern identifies a high-risk subgroup of intraventricular meningiomas
Kartierung des genomweiten Methylierungsmusters identifiziert eine Hochrisiko-Untergruppe der intraventrikulären Meningeome
Suche in Medline nach
Autoren
Veröffentlicht: | 4. Juni 2021 |
---|
Gliederung
Text
Objective: Intraventricular meningiomas (IVMs) are a rare subgroup accounting for less than 5% of all intracranial meningiomas. Recent studies showed evidence that IVMs are marked by a distinct chromosomal alteration with a combined loss of chromosome 22q and 1p. Here, we aim to map epigenetic patterns of IVMs in a multicenter cohort of 50 patients.
Methods: For biomathematical analysis methylation profiles from 850k methylation arrays were imported using the minfi package, followed by a batch correction (lima batch effect removal) and normalized resulting in a methylation intensity matrix of beta-values. We removed CpG sites from X, and Y chromosomal regions and selected the most variable 2000 probes for PCA analysis. The number of non-trivial components was obtained by comparison to a randomly arranged matrix of beta values. A non-trivial component was defined by a larger eigenvalue than random values. Dimensional reduction and clustering were performed for subgrouping of IVMs. In order to map the IVM into a reference dataset of meningioma, we imported the raw *.idat files of all 149 methylation profiles of meningioma in the recently launched methylation classifier (Capper et al., 2018).
Results: We were able to confirm that IVMs are more frequently marked by a loss of chromosome 22q compared to its reference cohort. Clustering of the methylation pattern revealed 2 different subcluster, one cluster contained a high number of copy number alterations and lead to tumor relapse in all patients within 48 months. This was significantly increased compared to the remaining IVMs (p=0.023) with a median PFS of 64 months (relapse in 2/3 of the group). We found a hypermethylation of gene promoter regions that regulate hormone binding and DNA repair mechanism within cluster1. In comparison to the reference cohort, most of the IVMs (76%) clustered separately, which suggest that IVMs are driven by an unique molecular program
Conclusion: Our findings showed two different groups of IVMs with a significant difference of progression-free survival. The malignant phenotype was marked by an increased genomic instability driven by hypermethylation of DNA repair mechanism and hormone binding sites. Based on our data, we provide the possibility for early prediction of a relapse probability by genome-wide methylation in IVM’s.
Figure 1 [Fig. 1]