IDH1 R132H: one mutation, two diseases, two FDA approvals

IDH1 R132H is the cleanest example in oncology of a gain-of-function mutation. The enzyme doesn’t just lose its normal activity — it picks up a new one, churning out an oncometabolite that reprograms the epigenome of the cell. Two FDA-approved inhibitors now address it, in two different diseases, with two different clinical pictures. The IDH1 story is worth knowing because it’s what targeted oncology looks like when the biology actually cooperates.

The neomorphic enzyme

Wild-type IDH1 (isocitrate dehydrogenase 1) sits in the cytoplasm and catalyzes the reversible decarboxylation of isocitrate to alpha-ketoglutarate (α-KG), reducing NADP+ to NADPH along the way. It’s a housekeeping redox enzyme, not particularly interesting from a drug discovery standpoint.

Then in 2008, Parsons et al. sequenced 22 glioblastomas and found recurrent point mutations at IDH1 residue 132. By 2009, Dang et al. had shown what the mutation actually does. R132H — the dominant variant — does not abolish enzyme activity. Instead, it shifts the reaction: the mutant enzyme reduces α-KG further, toR-2-hydroxyglutarate (R-2HG). The wild-type allele is still present and still makes α-KG. The mutant allele dumps millimolar concentrations of an enantiomer the cell was never designed to handle.

R-2HG is the “oncometabolite.” Structurally it’s a near-perfect mimic of α-KG, and it competitively inhibits the ~70 α-KG-dependent dioxygenases the cell uses for everything from HIF prolyl hydroxylation to histone and DNA demethylation. The downstream consequences are an epigenetic train wreck: hypermethylated CpG islands (the “CIMP” phenotype in glioma), failure of TET-mediated 5-methylcytosine oxidation, and a block in cellular differentiation. In gliomas this drives slow clonal expansion. In hematopoietic progenitors it produces AML.

What the mutation does to the pocket

Arginine 132 sits at the dimer interface, hydrogen-bonded to the substrate carboxylate. Substituting in histidine collapses the contacts that hold isocitrate in place and opens a new pocket — slightly larger, slightly more hydrophobic — that accommodates α-KG bound at the position that lets the active site reduce it further to R-2HG instead of releasing it. This is the druggable feature: a pocket that exists only in the mutant. Inhibitors that hit it spare wild-type IDH1 almost entirely, so the on-target toxicity envelope is unusually clean for an oncology drug. The R132C, R132G, R132L, and R132S variants share the same neomorphic activity and are also targetable, though with smaller selectivity margins than R132H.

Ivosidenib (AG-120) — IDH1-mutant AML

Ivosidenib (Tibsovo, Agios/Servier) was the first IDH1 inhibitor approved, in July 2018, for relapsed or refractory IDH1-mutant AML. The pivotal phase 1 study (DiNardo et al., NEJM 2018) ran the drug as monotherapy in 125 patients. The complete-remission rate was 21.6% and the composite CR+CRh rate was 30.4% — modest numbers in absolute terms, but transformative in a population where the prior standard of care was supportive care plus hypomethylating agents. Median duration of response was 9.3 months. Plasma R-2HG levels dropped by 1-2 orders of magnitude within days, and the responding patients showed restoration of myeloid differentiation rather than the cytoreduction you’d expect from conventional chemotherapy. The signature toxicity is differentiation syndrome — fever, hypoxia, effusions — driven by the same mechanism that drives the response.

The AGILE phase 3 trial (Montesinos et al., NEJM 2022) then moved ivosidenib into the front line, combined with azacitidine, in newly diagnosed IDH1-mutant AML patients unfit for induction chemotherapy. Median overall survival was 24.0 months versus 7.9 months for azacitidine alone (HR 0.44). That ratio is among the largest survival hazard ratios reported in any AML randomized trial. The FDA expanded ivosidenib’s label to front-line AML in 2022.

Vorasidenib (AG-881) — IDH-mutant low-grade glioma

Glioma was the disease where the IDH1 mutation was first discovered, but the drug for glioma came second. The reason is the blood-brain barrier. Ivosidenib has limited CNS penetration; even at maximum doses it doesn’t reliably suppress R-2HG inside intracranial tumors. Vorasidenib was engineered for brain penetrance — a dual mIDH1/mIDH2 inhibitor with reduced efflux liability and a brain-to-plasma ratio sufficient to hit the target where the tumor actually lives.

The INDIGO trial (Mellinghoff et al., NEJM 2023) randomized 331 patients with residual or recurrent grade 2 IDH-mutant glioma — all post-surgery, none yet on chemotherapy or radiation — to vorasidenib 40 mg daily or placebo. The primary endpoint was imaging-based progression-free survival.

• Median PFS was 27.7 months on vorasidenib versus 11.1 months on placebo (HR 0.39).
• Time to next intervention (radiation or chemotherapy) wasnot reached on vorasidenib versus 17.8 months on placebo (HR 0.26).
• Tumor growth rate flipped sign: -2.5% per 6 months on vorasidenib, +13.9% on placebo.
• Health-related quality of life and neurocognition were preserved — which matters because radiation and alkylator chemotherapy, the alternative, both have well-documented long-term neurocognitive costs in this young patient population.

Vorasidenib was FDA-approved in August 2024 for grade 2 IDH-mutant astrocytoma and oligodendroglioma after surgery. It’s the first systemic therapy approved for low-grade glioma in more than two decades, and it has reshaped the post-operative algorithm for a disease where the prior options were watchful waiting and delayed radiotherapy.

What this means for the medicinal chemistry

IDH1 R132H validates a few principles that the rest of oncology still struggles with. First, gain-of-function targets with neomorphic activity are the cleanest targets you can ask for — the wild-type protein is intrinsically spared because the drug binds a pocket that only exists in the mutant. Second, biomarker-driven selection works: every patient enrolled in INDIGO and AGILE had a confirmed IDH mutation, and the response rates reflect that. Third,CNS penetration is a design variable, not a post-hoc­property — ivosidenib worked in the bone marrow but didn’t work in brain, and Servier’s response was to engineer a second molecule rather than re-purpose the first.

The current frontiers: combinations of mIDH inhibitors with immune checkpoint blockade (R-2HG is locally immunosuppressive, and inhibitor treatment may restore tumor immune visibility), prevention trials in patients with IDH-mutant pre-malignant clones, and the question of whether R132C/G/L gliomas (a small minority) respond as well as the R132H majority.

Try the docking yourself

Open Studio and pick IDH1 from the target catalog with R132H from the mutation chips. The mutant-only allosteric pocket is the pharmacologically interesting site — most R132H-selective inhibitors will score 1.5-3 kcal/mol better against the mutant than against wild-type, which is a useful sanity check on any new analogue. Liganx renders wild-type and R132H side-by-side so the ΔΔ tells the selectivity story directly. For brain penetrance, the ADMET panel surfaces predicted BBB permeability and P-glycoprotein efflux risk, which is where ivosidenib loses to vorasidenib structurally.

Liganx runs molecular docking online, free and in the browser. It is a direct way to use molecular docking on the IDH1 R132H allosteric pocket and read the wild-type-versus-mutant selectivity gap.

Primary sources

• Dang L, et al. Cancer-associated IDH1 mutations produce 2-hydroxyglutarate. Nature 462, 739-744 (2009). doi:10.1038/nature08617
• DiNardo CD, et al. Durable Remissions with Ivosidenib in IDH1-Mutated Relapsed or Refractory AML. NEJM 378, 2386-2398 (2018). doi:10.1056/NEJMoa1716984
• Montesinos P, et al. Ivosidenib and Azacitidine in IDH1-Mutated Acute Myeloid Leukemia. NEJM 386, 1519-1531 (2022). doi:10.1056/NEJMoa2117344
• Mellinghoff IK, et al. Vorasidenib in IDH1- or IDH2-Mutant Low-Grade Glioma. NEJM 389, 589-601 (2023). doi:10.1056/NEJMoa2304194