From GWAS risk foci to glioma molecular subclass

Abstract

Gliomas are defined traditionally by histology and grade but are known to be highly heterogeneous, both intratumorally and intertumorally. Recent genetic studies have resulted in a generally accepted subgrouping of gliomas that reflects distinct molecular profiles with clinical relevance. The diversity of abnormalities that distinguish one subgroup from another suggests that these distinct profiles result from different etiologic pathways. Genome-wide association studies (GWAS) offer unbiased interrogations of the genome, resulting in the identification of risk associations that correspond to glioma subgroups. Single nucleotide polymorphisms (SNPs) at loci that influence glioma risk have been identified, suggesting that GWAS have the capacity to identify genes that act in biological pathways not previously linked to gliomagenesis.1 Identifying risk loci: Loci associated with glioma susceptibility have been detected by GWAS through detailed mapping studies of regions. Established glioma promoters, including TP53, p15/CKN2B, and EGFR, have been identified as glioma risk genes in GWAS analyses, and GWAS have substantiated the heterogeneous profiles among glioma subtypes, demonstrating associations with specific risk alleles.2,3 These findings can complement more focused investigations of individual candidate genes and supplement studies of potential functional relevance. Functional analysis: GWAS identify SNPs that mark a region of the genome associated with the risk of disease but are not specific to the causal variant or the molecular basis of risk etiology. Further association testing is necessary to identify candidate functional or causal variants and candidate causal genes that may lead to the discovery of the molecular mechanisms of risk modification and tumor pathogenesis. In the present issue of Neuro-Oncology, 2 papers report results of GWAS in glioma. Rice et al4 studied 1112 glioma patients and 5299 controls, identifying a risk locus at 11q23 that is associated with susceptibility to the distinct subclass of lower grade gliomas that carries mutation in isocitrate dehydrogenase (IDH) genes. The T allele of rs498872 on 11q23 was shown to increase risk for IDH-mutated gliomas of all grades and histological types but not increase risk for IDH-wild-type gliomas of any grade or histology. Currently, there are no known biologic functions associated with the 11qa23 SNP to explain the observed associations. Functional studies may reveal the causal variant in 11q23 and the mechanism behind the pathogenesis of these tumors. Di Stefano et al5 studied the characteristics of tumors from 1372 patients with glioma for associations with 7 glioma-risk SNPs (rs2736100 [TERT], rs11979158 and rs2252586 [EGFR], rs4295627 [CCDC26], rs4977756 [CDKN2A/CDKN2B], rs498872 [PHLDB1], and rs6010620 [RTEL1]), demonstrating a strong relationship with tumor phenotype and genotype among 4 of these 7 common risk variants for glioma. Clustering of the patient population based on various combinations of 1p-19q status +/-, EGFR amplification +/-, and IDH mutation status revealed associations with risk alleles. Risk alleles in TERT and RTEL1 predispose to an aggressive pathway that involves loss of chromosome 10, whereas CCDC26 and PHLDB1 risk alleles predispose to lower grade disease involving IDH mutation and 1p-19q loss. These associations may provide novel insights into the biological mechanisms underlying glioma formation and development. The demonstrated influence of IDH mutation and potential for more refined stratification of gliomas into relevant etiologic subgroups may promote understanding of the causal mechanisms underlying the development of these tumors, eventually improving disease classification and prognostic accuracy.