作者: Xiaowei Guan , Jaime Vengoechea , Siyuan Zheng , Andrew E Sloan , Yanwen Chen
DOI: 10.1371/JOURNAL.PONE.0091216
关键词:
摘要: Background Gliomas are the most common primary malignant brain tumors in adults with great heterogeneity histopathology and clinical course. The intent was to evaluate relevance of known glioblastoma (GBM) expression methylation based subtypes grade II III gliomas (ie. lower gliomas). Methods Gene array, single nucleotide polymorphism (SNP) array data were obtained for 228 GBMs 176 II/II (GII/III) from publically available Rembrandt dataset. Two additional datasets IDH1 mutation status utilized as validation (one publicly dataset one newly generated MD Anderson). Unsupervised clustering performed compared gene assigned using Verhaak et al 840-gene classifier. glioma-CpG Island Methylator Phenotype (G-CIMP) prediction models by Fine al. Results Unsupervised aligned subtype group assignments. GII/IIIs preferentially proneural G-CIMP. evenly distributed among four subtypes. Proneural, mutant, G-CIMP GII/III s had significantly better survival than other molecular Only 6% either or but these GBMs. Copy number changes chromosomes 1p 19q associated GII/IIIs, while CDKN2A, PTEN EGFR more commonly GBMs. Conclusions GBM gene-expression relevant associate overall differences. A understanding association between could further knowledge regarding prognosis mechanisms glioma progression.