Since the first discovery of isocitrate dehydrogenase (IDH) somatic mutations in 2006, major advances have been made in understanding their contribution to cancer development, providing a strong rationale for pharmacologically targeting the mutant metabolic enzyme.

The IDH family includes three different isozymes regulating key metabolic cellular processes, such as the ‘Krebs’ cycle, glutamine metabolism, lipogenesis, and redox balance. IDH1 is located in the cytosol and peroxisomes, while IDH2 and IDH3 in the mitochondrial matrix. Physiologically, IDH1-2 is responsible for the NADP-dependent oxidative decarboxylation of isocitrate to α-ketoglutarate (α-KG), producing NADPH in the process. Mutations of IDH1 and IDH2 have been identified in over 70% of lower-grade gliomas (LGGs; World Health Organization (WHO) grade II/III) and secondary glioblastoma (GBM) ^[1][2][1,2], and at lower frequencies in a variety of other human malignancies, including acute myeloid leukemia (AML; ≈30%), chondrosarcoma (≈50%), cholangiocarcinoma (≈15–20%), thyroid carcinoma, melanoma, angioimmunoblastic T-cell lymphoma, and in a rare subtype of breast cancer ^{[3][4][5][6][7][8]}[3,4,5,6,7,8]. IDH3 is a heterodimer, not structurally related to the other two isoforms, and only rarely mutated in cancer.

Most common tumor-associated IDH1/2 mutations are located in key arginine residues (R132 for IDH1, R140, or R172 for IDH2) for the recognition of the substrate ^[9][10][9,10], and lead to a neomorphic gain of function enzyme with a markedly decreased affinity for isocitrate, capable instead of catalyzing the conversion of α-KG into 2-hydroxyglutarate (2-HG) in an NADPH-dependent manner. As a result, IDH-mutant (mut) cells accumulate 2-HG at supraphysiological levels (up to 100-fold higher than in IDH wild-type (wt) counterparts), while the intracellular concentration of α-KG greatly decreases [11].

It is now well documented that 2-HG has the properties of an “oncometabolite” contributing to tumorigenesis by altering epigenetic regulation, DNA repair mechanisms, multiple major metabolic pathways, and redox homeostasis ^[12][13][12,13].

The identification of IDH1/2 mutations in diffuse gliomas represents, doubtless, one of the major breakthroughs in the field of neuro-oncology over the last decades, which led brain tumors into a new molecular era, with several significant diagnostic, prognostic, and therapeutic implications.

Mutations in the IDH1 gene were firstly observed in 2008 by Parsons et al. in a small subset (12%) of GBMs, mainly affecting young patients, progressed from previous low-grade tumors (secondary GBM), and carrying a much better outcome [2]. IDH mutations have been subsequently recognized as the distinctive genetic feature of LGGs, occurring in more than 90% of low-grade gliomas (WHO grade II) and in 70% of anaplastic gliomas (WHO grade III) [1]. In addition to peculiar clinical characteristics such as younger age at diagnosis and an overall better prognosis, accumulating evidence allowed identifying IDH-mutated gliomas as a completely different entity at the molecular level compared to the IDH-wt counterparts [14].

Nearly all 1p/19q codeleted gliomas (oligodendroglial tumors) harbor an IDH1/2 mutation, along with O6-methylguanine-DNA methyltransferase (MGMT) promoter methylation and telomerase reverse transcriptase (TERT) promoter mutation, while TP53 and ATRX alterations have been found in IDH1/2-mutated glioma without codeletion (astrocytic tumors). These great advances in molecular knowledge of brain tumors led, in 2016, to a major revision of the WHO classification with the introduction of the IDH mutational status and the presence of 1p/19q codeletion, in addition to histology, to define glioma subtypes [15]. Moreover, IDH mutations have been established as the most powerful positive prognostic factor for survival in gliomas, followed by age, tumor grade, and MGMT promoter methylation status. Although the underlying mechanisms have not yet been fully understood, the IDH mutational status seems to harbor also predictive significance, conferring an increased sensitivity to ionizing radiation and alkylating agents ^[16][17][16,17]. The WHO classification has been recently updated to better elucidate some gray areas. All IDH-mut diffuse astrocytic tumors have been considered as a single disease entity (astrocytoma, IDH-mut), graded (2, 3, or 4) based on the presence not only of microvascular proliferation or necrosis but also of other molecular markers with negative prognostic value, such as CDKN2A/B homozygous deletion. Conversely, IDH-wt low-grade astrocytoma harboring one or more of the three genetic features of classic GBM (TERT promoter mutation, epidermal growth factor receptor gene amplification, or combined gain of entire chromosome 7 and loss of entire chromosome 10), will be labeled as GBMs, carrying a very similar dismal prognosis [18].

In diffuse gliomas, over 90% of somatic IDH mutations are arginine-to-histidine heterozygous substitutions in the codon 132 (IDH1^R132H mutation) of the IDH1 gene, with the remaining 10% represented by other amino acid substitutions located in the same spot. On the other hand, less than 1% of gliomas harbor IDH2 hotspot mutations, mostly occurring at arginine 172 (IDH2^R172H mutation) or arginine 140 (IDH2^R140 mutation) [19]. IDH mutations have been suggested to represent an early “truncal” event in the complex process of gliomagenesis, occurring in the stem cell progenitor from the subventricular zone stem cell niche before acquiring other “secondary” genetic alterations such 1p-19q codeletion, ATRX, and TP53 mutations. The progressive accumulation of 2-HG over physiological levels impairs cellular metabolism, including glutamine catabolism and Krebs Cycle, and reshapes the epigenetics of the cell. The epigenetic alterations are thought to be an essential factor of gliomagenesis. 2-HG inhibits a wide range of a-KG-dependent dioxygenases, including lysine demethylases, Ten Eleven Translocase (TET) hydroxylases, Jumonji-C domain histone demethylases (JHDMs), and prolyl hydroxylase domain (PHD), leading to a global DNA and histone hypermethylation, stem-like cell differentiation block, and altered hypoxia-inducible factor (HIF) activity ^[20][21][22][20,21,22]. Consequently, IDH-mut gliomas acquire the CpG island methylator phenotype (CIMP or G-CIMP), which may be considered a sort of genetic signature of this class of tumors, with several transcriptional changes and the silencing of numerous loci in the genome. Based on the extent of global DNA methylation, IDH-mut/G-CIMP+ tumors have been further divided into two distinct subgroups characterized by different outcomes: G-CIMP-low gliomas with a low degree of DNA methylation and unfavorable clinical outcome and G-CIMP-high gliomas characterized by higher DNA methylation and better outcome ^[23][24][23,24]. It remains unclear whether IDH mutation by itself is responsible for tumor development or if it requires other oncogenic events to initiate gliomagenesis. Although DNA methylation is usually believed to be reversible, a study revealed that some of the DNA methylation sites in the IDH-mut cells might persist, even when the mutant enzyme is turned off [25]. This evidence supports the hypothesis that IDH mutation plays a key role in the first steps of malignant transformation, which becomes irreversible once the cells have initiated the oncogenic process [25]. However, emerging data suggest that for a subgroup of LGGs, 2-HG produced by the mutant enzyme may become non-essential for tumor progression as they acquire tertiary driver mutations and progress to a higher grade. IDH mutations would therefore behave as “passenger mutation,” losing their oncogenic function and even being eliminated in the latest stages of glioma progression [26].

It is not surprising that IDH mutations have been launched to the top of the list of new potential therapeutic targets for IDH-mut diffuse glioma. Two distinct strategies have been proposed so far and are currently under intense clinical investigation: (i) reducing the amount of intratumoral 2-HG by directly blocking the activity of the mutant IDH enzyme; (ii) taking advantage of the cellular vulnerabilities as a consequence of 2-HG accumulation. A key point that still needs to be addressed is how to locate an IDH-targeted therapy within the treatment paradigm of the IDH-mutated LGGs, which are typically slow-growing tumors arising in the young and displaying a good prognosis. LGGs’ therapeutic management consists of maximal safe surgical resection, eventually followed by radiotherapy and chemotherapy given as adjuvant for high-risk patients or postponed at recurrence/progression for low-risk cases. A crucial concern of neuro-oncologists in treating LGGs is the onset of long-term treatment-related side effects. In particular, radiation-induced neurocognitive defects can affect multiple domains, including memory, attentional and executive functioning, information processing speed, and problem-solving capability. These cognitive disorders caused by the exposure of healthy brain tissue, endocrine disorders, insomnia, and fatigue (related to the impairment of the hypothalamic-pituitary axis) may significantly compromise these young patients’ daily-life activities and social relationships. Furthermore, chemotherapy with alkylating agents may exacerbate the risk of developing long-term toxicities and accelerate malignant transformation. Thus, the development of novel therapeutic approaches in this setting, able to control tumor growth and high-grade progression, delaying or even avoiding the use of radiation and chemotherapy, clearly represents an urgent need.

Despite the great enthusiasm for what represents the first attempt of applying “precision oncology” to LGGs’ treatment, multiple observations call for some caution. Firstly, major epigenetic changes induced by the mut-IDH enzyme may persist even after suppressing 2-HG levels [25]. Secondly, blocking the mutant IDHs might, at least in theory, lead to the transformation of usually slow-growing and good prognosis tumors into more aggressive biological subtypes. Finally, once LGGs acquire other secondary and tertiary genetic alterations during their biological evolution, IDH mutations may lose their oncogenic function and even be eliminated at the latest stages of progression.

4. Immunotherapeutic Approaches for IDH Mutant Gliomas

4.1. Preclinical Studies

Given the remarkable clinical successes in other tumor types, great efforts have been undertaken in latest years to explore the role of immunotherapy in the field of neuro-oncology and even in the context of LGGs. It has been observed that spontaneous IDH1^R132H-specific CD4+ T-helper-1 (TH1) and humoral immune responses might occur in LGG patients. Moreover, there are now several preclinical evidences demonstrating that the oncometabolite 2-HG greatly contributes to shaping the immunosuppressive glioma microenvironment and plays a crucial role in immune escape mechanisms ^[27][47]. Berghoff et al., analyzing a cohort of 43 WHO grade II/III gliomas (39 IDH-mut, 4 IDH-wt) and 14 IDH-mut GBMs, demonstrated that IDH-mut tumors have fewer CD3+ and PD1+ tumor-infiltrating lymphocytes (TILs) and decreased expression of programmed cell death ligand (PD-L1) compared to their wt counterparts (series of 117 IDH-wt GBMs) ^[28][48]. Using data of 677 diffuse gliomas grades II-IV from The Cancer Genome Atlas (TCGA) database, authors found that PD-L1 gene expression was statistically significantly higher in IDH-wt WHO grade II/III gliomas compared with IDH-mut WHO grade II/III gliomas ^[28][48]. Moreover, PD-L1 gene promoter methylation levels were higher in IDH-mut than IDH-wt samples ^[28][48].

Gene expression data from the same TCGA database revealed a reduced expression of cytotoxic T lymphocyte-associated genes and IFN-γ-inducible chemokines, including CXCL10 in IDH-mut tumors compared with IDH-wt tumors ^[29][49]. The introduction of mutant IDH1 or treatment with 2-HG in immortalized normal human astrocytes and syngeneic mouse glioma models led to a reduction of CXCL10 levels, which was associated with decreased production of STAT1 and also suppressed the accumulation of T-cells into tumors ^[29][49]. All these effects are reversible by treatment with IDH-C35, a specific inhibitor of IDH1-mut.

Bunse et al. demonstrated that 2-HG directly impairs T-cells’ activation, proliferation, and cytokine secretion by altering the calcium-dependent transcriptional activity of nuclear factors and suppressing ATP-dependent TCR signaling and polyamine biosynthesis ^[30][32]. Moreover, in different syngeneic IDH1-mut tumor models, 2-HG inhibited the development of antitumor T-cell immunity induced by IDH-1 vaccination, adoptive T-cell transfer, and checkpoint blockade while pharmacological inhibition of the neomorphic enzymatic function of mut-IDH1 by administration of BAY-1436032 alleviates intratumoral immune suppression ^[30][32] (Table 1).

Table 1.

Preclinical studies of IDH-targeted agents in glioma mouse models.

Reference	Agents	Results	Additional Findings
Rohle, D. et al., 2013 [31]	Rohle, D. et al., 2013 [27]	AG-5198	50–60% growth inhibition (p = 0.015, two-tailed	t	-test) No toxicity along 3 weeks of daily treatment	Reduced staining with Ki-67 antibody in treated mice

^][35] (Table 1). 2-HG inhibition by treatment with a specific IDH1^R132H inhibitor alone or in combination with radiation and temozolomide (TMZ) substantially prolonged survival of IDH1-mut glioma-bearing mice, increased PD-L1 expression levels to similar levels as observed in IDH-wt gliomas, contributing in also generating an anti-glioma immunity ^[38][35]. Moreover, the coadministration of PD-L1 blockade with IDH1^R132H inhibition and the standard of care (RT + TMZ) markedly enhanced the outcome of IDH1-mut glioma-bearing mice, counteracting T-cell exhaustion and eliciting an immunological memory with the generation of memory CD8+ T-cells ^[38][35]. These data supported the design of clinical trials investigating the efficacy of IDH1^R132H inhibitors in combination with standard of care (SOC) and anti-PD-L1 immune checkpoint blockade to treat glioma patients expressing with IDH mutations.

4.2. Clinical Trials

Based on all these preclinical data, the German National Cancer Center has conducted a multicenter, first-in-humans Phase I trial, evaluating the feasibility, safety, and immunogenicity of a vaccine targeting the IDH1^R132H mutant protein among newly diagnosed patients with IDH1-mut glioma (NOA16, NCT02454634) ^[41][42][45,52] (Table 2).

Table 2.

Clinical trials with IDH-targeted therapies in gliomas with available data.

Reference	NCT Number	Study Design	Treatment	Population	Main Results	Adverse Events (in ≥10% of Patients)
Mellinghoff, I.K. et al., 2020 [43]	Mellinghoff, I.K. et al., 2020 [40]	NCT02073994	Phase I	Ivosidenib (AG-120) single agent	Advanced IDH1-mut solid tumors 35 non-enhancing recurrent gliomas 31 enhancing recurrent gliomas	500 mg once daily selected for expansion part DCR 88% vs. 45%; median PFS 13.6 vs. 1.4 months in non-enhancing vs. enhancing cohort	No DLT Headache; fatigue; nausea; vomiting; seizure; diarrhea; aphasia; hyperglycemia; neutropenia; depression; hypophosphatemia; paresthesia
Tateishi, K. et al., 2015 [32]	Tateishi, K. et al., 2015 [28]
Mellinghoff, I.K. et al., 2021	IDH1i	[44]	Mellinghoff, I.K. et al., 2021 [41]	Near-complete elimination of 2-HG within brain tumors after 5 days of treatment No effect on tumor size and survival: mOS 46 days in both treated mice and controls (95% CI; 45–48; p	NCT02481154	= 0.79)	Near-complete elimination of 2-HG within brain tumors after 5 days of treatment No effect on tumor size and survival: mOS 46 days in both treated mice and controls (95% CI; 45–48; p = 0.79)	No effect on expression of IDH1, Ki-67, GFAP, or nestin within the tumors
Phase I	Vorasidenib (AG-188) single agent	Advanced IDH1 and/or IDH2-mut solid tumors		22 non-enhancing recurrent gliomas 30 enhancing recurrent gliomas	Recommended dose <100 mg in gliomas Non-enhancing glioma: ORR 18% (1 PR; 3 minor responses; 17 SD) Enhancing glioma: ORR 0% (17 SD) Median PFS: 36.8 vs. 3.6 months in non-enhancing vs. enhancing groups	DLT (grade ≥2 ALT/AST increase) in 5 pts at ≥100 mg dose levels Headache; AST/ALT increase; fatigue; nausea; seizure; hyperglicemia; vomiting; constipation; dizziness; neutropenia; cough; diarrhea; aphasia; hypoglycemia	Popovici-Muller, J. et al., 2018 [33]	Popovici-Muller, J. et al., 2018 [29]
Mellinghoff, I.K. et al., 2019 [	Ivosidenib (AG-120)	45]	Mellinghoff, I.K. et al., 2019 [42Low brain penetration in rats with intact blood–brain–barrier: 4.1% (AUC 0−8 h (brain)/AUC 0−8 h (plasma)) Robust time-dependent, reversible tumor 2-HG reduction (IC50 range 5–13 nM)
]	NCT03343197	Phase I	Perioperative Ivosidenib (AG-120) (	n	= 13) or vorasidenib (AG-188) (	n	= 14) single agent	Pusch, S. et al., 2017 [34]	Pusch, S. et al., 2017 [30]	BAY 1436032	Significantly reduced 2-HG concentration (p = 0.00000057) and prolonged survival (p = 0.025) compared to untreated controls	SOX2 expression reduced by half in tumors of treated mice
Recurrent non-enhancing IDH1	Konteatis, Z. et al., 2020 [35]	Konteatis, Z. et al., 2020 [31]	Vorasidenib (AG-881)	>97% inhibition of 2-HG production in mice glioma tissue
Bunse, L. et al., 2018 [30]	Bunse, L. et al., 2018 [32]	BAY 1436032	Oral administration of BAY-1436032 in combination with PD-1 inhibition increased overall survival in mice	Enhanced intratumoral CD4 T-cell proliferation
Schumacher, T. et al., 2014 [36]	Schumacher, T. et al., 2014 [33]	IDH1 (R132H) peptide vaccine	Effective control growth in syngeneic IDH1 (R132H)-expressing tumors	Robust interferon IFN-γ T-cell response Mutation-specific anti-IDH1 antibodies detectable in serum of immunized mice
Pellegatta, S. et al., 2015 [37]	Pellegatta, S. et al., 2015 [34]	IDH1 (R132H) peptide vaccine	Significant survival gain compared to controls with 25% of cured mice	Higher amounts of peripheral CD8+ T-cells, higher production of IFN-γ, anti-IDH1-mut antibodies in immunized mice
Kadiyala, P. et al., 2021 [38]	Kadiyala, P. et al., 2021 [35]	AGI-5198 +/− IR	2-HG levels in mIDH1 brain tumor tissue reduced by approximately 2.4-fold (p ≤ 0.0001) after treatment with AGI-5198 40% long-term survivors among mice treated with AGI-5198 or AGI-5198 + IR Anti-PD-L1 + AGI-5198 + IR and TMZ improved 90-day survival rate by 40% (95% CI: 0–95%, 1-sided p = 0.08)	AGI-5198 administration led to a 3-fold (p ≤ 0.001) increase in the PD-L1 expression on the CD45–/Nestin+ tumor cells compared with untreated controls Increased infiltration of CD8+ T-cells (p < 0.01) in tumors treated with anti-PD-L1 + AGI-5198 + IR and TMZ

2-HG 2-hydroxyglutarate; AUC area under the curve; CD4 cluster of differentiation 4; CD8 cluster of differentiation 8; GFAP glial fibrillary acidic protein; IC50 half maximal inhibitory concentration; IDH isocitrate dehydrogenase; IFN-γ interferon γ; IR ionizing radiation; mOS median overall survival; PD-1 programmed cell death protein 1; PD-L1 programmed death-ligand 1; SOX2 SRY-Box Transcription Factor 2; TMZ temozolomide.