IDH2 mutations are commonly associated with 1p/19q codeletion in diffuse adult gliomas

authors

  • Appay Romain
  • Tabouret Emeline
  • Macagno Nicolas
  • Touat Mehdi
  • Carpentier Catherine
  • Colin Carole
  • Ducray François
  • Idbaih Ahmed
  • Mokhtari Karima
  • Uro-Coste Emmanuelle
  • Dehais Caroline
  • Figarella-Branger Dominique
  • Roger Pascal

document type

ART

abstract

716 IDH2 mutations are commonly associated with 1p/19q codeletion in diffuse adult gliomas Diffuse gliomas are classified according to the 2016 World Health Organization (WHO) Classification of Tumors of the Central Nervous System, 1 which combines histological and molecular features. Diagnosis requires the assessment of mutations in the isocitrate dehydrogenase genes (IDH1 and IDH2), key genetic alterations characterizing gliomas with favorable outcome. 2 Because IDH1 and IDH2 are highly similar enzymes, the WHO classification, as most of the current studies, combines these mutations into the same molecular group; however, it is unclear whether these tumors share the same characteristics. We analyzed data from 1517 patients included in the French POLA Network to investigate differences between IDH1-and IDH2-mutant gliomas. Inclusion criteria were the written consent of the patient for clinical data collection and genetic analysis and sufficient material for molecular studies allowing classification according to the 2016 WHO classification. IDH1-R132H mutational status was evaluated using automated immunohistochemistry in all cases (n = 1517). Direct sequencing 3 was performed in 978 cases and demonstrated IDH mutation in 573 cases (this includes confirmation of IDH1-R132H mutation in 468 cases, other IDH1 mutations in 61 cases, and IDH2 mutation in 44 cases). The 1p/19q codeletion status was determined based on single nucleotide polymor-phism arrays, comparative genomic hybridization arrays, and/ or microsatellite marker analysis. 3 The following data were recorded: age, sex, follow-up, and MRI features (tumor location , extension, contrast enhancement, edema). All statistical analysis was done using IBM SPSS statistics software version 23. Chi-square test was used to compare qualitative variables. Continuous variables were compared using the Mann– Whitney U-test, and the Kaplan–Meier method was used to estimate survival distributions. Among the 1517 patients, 1025 had an IDH-mutant tumor: 96% were IDH1-mutant and 4% IDH2-mutant. Integrated diagnoses are summarized in Fig. 1. The frequency of 1p/19q code-letion was higher in the IDH2-mutant group compared with the IDH1-mutant group (91% vs 48%, P < 0.001). Wang and coworkers previously reported higher frequency of 1p/19q codeletion in IDH2-mutant gliomas (9/18 samples) compared with IDH1-mutant. 4 The percentage of each category in our study does not reflect the normal distribution of glioma because of the inclusion criteria in the POLA Network (ie, high-grade glioma with oligodendroglial component). However, the higher proportion of 1p/19q codeleted glioma in the IDH2-mutant group cannot be attributed to the inclusion criteria. Because the main population of glioma associated with IDH2 mutation was 1p/19q codeleted anaplastic oligodendroglioma, we focused on this subgroup to search for differences compared with IDH1 mutation. Among these patients (n = 474), we did not observe any difference in terms of age, sex, progression free survival, or overall survival between IDH1-and IDH2-mutant tumors. However, IDH2-mutant anaplastic oli-godendrogliomas presented more frequently with multilobar extension (56% of the IDH2-mutant vs 35% of the IDH1-mutant, P = 0.015) and edema (79% vs 57%, P = 0.02). Furthermore, bifrontal location with corpus callosum involvement was more frequent in IDH2-mutant compared with IDH1-mutant tumors (41% vs 16%, P < 0.001). IDH mutation is supposed to be one of the first " hits " of gliomagenesis, 5 resulting in production of an oncome-tabolite, D-2-hydroxyglutarate (D-2HG), which impacts the α-ketoglutarate–dependent dioxygenase functions. Previous studies demonstrated that the potential for IDH-mutant enzymes to produce D-2HG depends on the mutation type. 6 Based on our observations, we could hypothesize that the higher D-2HG accumulation induced by IDH2 mutation may lead to a phenotype that is favorable to 1p/19q chromosomal loss. It may also impact distinct cellular pathways that promote a more invasive phenotype. Whether IDH1 or IDH2 mutations impact distinct glial precursor cells with differential invasive properties remains to be elucidated. In conclusion, our results illustrate that IDH2-mutant glio-mas are commonly associated with 1p/19q codeletion. Most of IDH2-mutant anaplastic oligodendroglioma 1p/19q codeleted are multilobar. Understanding the genomic events involved in these specificities may represent a step forward for therapeutic development.

more information