The comparison of clinical and biological characteristics between IDH1 and IDH2 mutations in gliomas

Isocitrate dehydrogenase (IDH) enzymes encode the NADP⁺-dependent isocitrate dehydrogenase, which catalyzes the oxidative decarboxylation of isocitrate to form an α-ketoglutarate (α-KG). IDH1 and IDH2 proteins share a high degree of sequence similarity (70 % in humans) and are encoded by distinct genes (IDH1, 2q33 and IDH2, 15q26). Mutations in IDH1 and IDH2, which represent the most frequently mutated metabolic genes in human cancer, are implicated to be mutated in more than 50–80 % of low-grade gliomas and secondary glioblastomas (sGBM), 10 % of intrahepatic cholangiocarcinoma, 20 % of acute myeloid leukemia (AML), 56 % of chondrosarcomas, and over 10 % of melanoma cases [1‐5]. Although IDH1 and IDH2 are highly similar and catalyze identical reactions, IDH1 is localized in the cytosol and IDH2 is found in the mitochondrial matrix. In addition, the spectrum of cancers and their subtypes are different. For example, IDH1 mutations are predominant in gliomas, chondrosarcoma, and cholangiocarcinoma, whereas IDH1 mutations and IDH2 mutations are equally common in AML. Despite their different physiological characteristics, most genomic studies of the molecular landscapes in human cancer have frequently combined IDH1 mutations and IDH2 mutations as a single functional group.

Glioma, the most common primary brain tumor, is classified as grade I to IV based on histopathological and clinical criteria established by the 2007 World Health Organization (WHO) [6]. WHO grade I gliomas are often curable by surgical resection, whereas WHO grade II or III gliomas are invasive and have a poor prognosis. WHO grade IV tumors (glioblastomas), the most invasive tumors, feature a median survival of only 16 months, even after aggressive treatment consisting of surgery, radiation therapy, and chemotherapy [7]. In 2008, the genes encoding IDH1 were found to be mutated in low-grade gliomas and a subset of sGBM [8]. In subsequent studies, IDH1 mutations were reported to occur in 70–80 % of WHO grade II or III astrocytomas, oligodendrogliomas, and oligoastrocytomas, whereas a small group (3–5 %) were found to harbor IDH2 mutations [1]. This pattern contrasts that observed in AML, which features similar rates of IDH1 (6.6 %) and IDH2 mutations (10.8 %) [9]. Moreover, mutations of IDH1 and IDH2 are mutually exclusive in gliomas, and biochemical investigations showed that IDH1 and IDH2 mutations differ in D-2-hydroxyglutarate (D-2HG) production in gliomas [10]. This difference suggests that IDH1 and IDH2 mutations may impact different cellular pathways and exert different tumorigenic effects. To investigate the different clinical and molecular characterization between IDH1 mutant and IDH2 mutant gliomas, we studied a cohort of 811 patients consisting 448 IDH1 mutant, 18 IDH2 mutant and 345 IDH1/2 wild-type gliomas. We performed whole-transcriptome sequencing and DNA methylation analyses of the samples obtained from patients. We compared the mutational landscapes of IDH1 and IDH2 mutant gliomas, their clinical associations, overall survival, and progression-free survival. Our aim was to provide insight into the differences between IDH1 and IDH2 mutant gliomas.

Mutations in the IDH1 and IDH2 genes have been found in patients with gliomas and were initially identified in low-grade gliomas and secondary glioblastomas [1]. Strikingly, mutations in IDH1 and IDH2 are mutually exclusive in gliomas. Although the genetic and epigenetic landscapes of IDH1 mutation gliomas have been extensively studied, whether IDH2 mutation gliomas have unique genetic and epigenetic characteristics that can be used as targets for future intervention is unknown. In this report, we compared the clinical and molecular characteristics of glioma patients harboring IDH1 and IDH2 mutations.

Like mutations in IDH1, mutations in IDH2 affect a conserved arginine residue (R172) in the substrate-binding site of the IDH2 enzyme. In our cohort, the presence of an IDH2 mutation did not correlate with the presence of PTEN, P53, and ATRX mutations, but a highly significant positive correlation was observed with the presence of a 1p/19q co-deletion: 44.4 % of IDH2 mutation patients harbored a 1p/19q co-deletion. In malignant glioma, IDH1 mutations are ubiquitous in tumor cells, and IDH1 mutations precede secondary and tertiary lesions, suggesting that IDH1 mutations are an early causative event in the genesis of gliomas [15‐17]. A pathology study of multiple biopsies from the same patient found that IDH1 mutations occurred before the acquisition of P53 mutations and 1p/19q loss of heterozygosity (LOH) [16], suggesting that IDH1 mutations may result in cellular stress that leads to the mutation of P53 and 1p/19q loss. However, IDH2 mutations and PTEN, P53 and ATRX mutations were mutually exclusive, suggesting that the microenvironment of IDH2 mutations may not create cellular stress that leads to the other mutations, which needs further research to fully elucidate.

Tumor cells often take up nutrients in excess of their bioenergetic needs and shunt metabolites into pathways that support tumor progression [18‐20]. During cell proliferation, tumor cells depend on aerobic glycolysis to meet their bioenergy needs and generate intermediates for macromolecule biosynthesis. One study demonstrated that glioma cells harboring mutant IDH1 may maintain cell proliferation via the glutamate metabolism pathway [21]. In our study, GSEA was performed for IDH2 and IDH1 mutations, yielding enriched gene sets related to oxidative phosphorylation, which is critical to tricarboxylic acid (TCA) cycle, in the IDH2 mutation subset. This finding corroborates that of a previous study [22, 23]. IDH2 is localized in the mitochondria and participates in the TCA to produce energy, whereas IDH1 is localized in the cytoplasm and peroxisomes [24]. Consequently, does energy production in IDH2-mutated gliomas favor oxidative phosphorylation over aerobic glycolysis? These interesting findings should be verified in more cases before accepting them as general characteristics of IDH2-mutated gliomas. Future work should focus on the potential of therapeutically targeting compensatory metabolic pathways in IDH2-mutant gliomas.

In conclusion, our results describe the clinical and biological characteristics of IDH1 and IDH2 mutations in gliomas. Understanding the underlying biology of the differences in outcome observed for IDH1 and IDH2 mutant gliomas will be important for future studies and may lead to the development of novel approaches to therapy.