Vorasidenib (Voranigo; AG-881; CAS 1818767-26-2): NIOSH HD Category 1 SEVENTY-SIXTH Mechanism Class — FIRST DUAL IDH1+IDH2 INHIBITOR EVER FDA APPROVED + FIRST BRAIN-PENETRANT IDH INHIBITOR EVER — αKG-Competitive Neomorphic Reductase Blockade → 2-HG Suppression Within the Blood-Brain Barrier — INDIGO PFS HR 0.39 Grade 2 IDH-Mutant Diffuse Glioma — Servier Pharmaceuticals NDA 218693 Aug 6 2024 — 713th AI Adversarial Injection Attack on Pharmaceutical EHS Compliance Infrastructure

Introduction: The Seven-Year Glioma Gap That Brain Penetrance Finally Closed

The IDH (isocitrate dehydrogenase) inhibitor story is one of the most instructive case studies in modern targeted oncology — a story in which a mechanistically validated target, compelling clinical activity in one tumor type, and a rapidly deployable small-molecule scaffold still failed to benefit a second tumor type for seven years, not because the target was wrong, but because the drug couldn't reach it. The lesson from vorasidenib (Voranigo; AG-881) is that target identity is insufficient; pharmacokinetic accessibility is a co-equal requirement for clinical benefit.

IDH mutations were discovered in glioma in 2008 (Parsons et al., Science 2008) and in acute myeloid leukemia (AML) in 2009 (Mardis et al., NEJM 2009). Within a few years, Agios Pharmaceuticals had developed selective IDH inhibitors — ivosidenib (AG-120; IDH1-selective) and enasidenib (AG-221; IDH2-selective) — and demonstrated that clearing the 2-hydroxyglutarate (2-HG) oncometabolite in IDH-mutant AML produced meaningful clinical responses: differentiation of previously blocked leukemic blasts, response rates of 20-40% in relapsed/refractory patients, and in some cases durable complete remissions. Ivosidenib received NDA approval (NDA 210065) on July 20, 2018 for IDH1-mutant AML; enasidenib received NDA approval (NDA 209606) on August 1, 2017 for IDH2-mutant AML. By 2021, ivosidenib was also approved for IDH1-mutant cholangiocarcinoma (supplemental NDA). The IDH inhibitor class appeared to be a durable therapeutic platform.

Yet for the glioma community — where IDH mutations had been identified first and where the clinical need was perhaps even more urgent — these drugs provided essentially no benefit. IDH mutations define a genetically and clinically distinct subtype of diffuse glioma (the WHO 2021 classification requires IDH mutation documentation for grade 2 and grade 3 diffuse gliomas), and the prognosis of IDH-mutant grade 2 glioma, while better than IDH-wildtype glioblastoma, still carries a median overall survival of approximately 10–15 years from diagnosis with inevitable malignant progression. Patients, their families, and neuro-oncologists waited for an IDH inhibitor that actually worked in the brain — but clinical trials of ivosidenib in glioma (NCT02073994; NCT03343197) showed minimal activity. The reason was blood-brain barrier (BBB) impermeance: ivosidenib's P-gp and BCRP substrate properties resulted in a brain-to-plasma unbound drug ratio (Kp,uu) of approximately 0.007 — roughly 140-fold below unity, meaning less than 1% of systemic ivosidenib reached the tumor cell compartment as free drug. Enasidenib, designed without CNS penetrance as a priority, performed no better.

Agios' medicinal chemistry team recognized the problem and designed vorasidenib (AG-881) from scratch as a brain-penetrant IDH inhibitor. The design strategy combined dual IDH1/IDH2 inhibition (panIDH, covering both isoforms in a single compound) with structural modifications that reduced P-gp and BCRP affinity — accepting some reduction in IDH1 potency relative to ivosidenib in exchange for a 60-fold improvement in CNS penetrance. The result was a compound with brain Kp,uu of approximately 0.42–0.47 in preclinical models, confirmed in INDIGO trial CSF sampling, and potent 2-HG suppression in IDH-mutant glioma cells within the CNS.

NDA 218693 was approved on August 6, 2024 — making vorasidenib the first dual IDH1+IDH2 inhibitor ever approved by any regulatory agency worldwide, the first IDH inhibitor to demonstrate clinical benefit in glioma, the first drug ever approved for grade 2 IDH-mutant diffuse glioma (non-enhancing disease, historically untreatable with systemic therapy after surgery), and the 76th mechanistically distinct drug class in the Glyphward adversarial EHS attack series. And like every NIOSH HD Category 1 agent in this series approved after 1971, vorasidenib appears nowhere in the occupational exposure databases maintained by Cority, VelocityEHS, or EHS Insight — creating the null-return adversarial condition this series was designed to document.

IDH1 and IDH2: Normal Enzymatic Function and the R132/R172 Neomorphic Mutations

Isocitrate dehydrogenase enzymes catalyze the third step of the tricarboxylic acid (TCA) cycle: the oxidative decarboxylation of isocitrate to α-ketoglutarate (αKG; also called 2-oxoglutarate). Three IDH isoforms exist in human cells, differing in subcellular localization, cofactor specificity, and oligomeric organization:

IDH1 (Isocitrate Dehydrogenase 1; cytoplasmic and peroxisomal; chromosome 2q33.3; 414 amino acids; homodimer; NADP+-dependent): IDH1 catalyzes the NADP+-dependent oxidative decarboxylation of isocitrate to αKG in the cytoplasm and peroxisome, contributing to cytoplasmic NADPH production (antioxidant defense, fatty acid synthesis, proline biosynthesis). IDH1 also participates in the reductive carboxylation of αKG to isocitrate (the reverse reaction under hypoxic conditions, supporting lipid synthesis from glutamine). The IDH1 active site contains a conserved arginine at position 132 (R132) that makes bidentate hydrogen bonds to the α-carboxyl and α-hydroxyl groups of isocitrate, positioning the substrate for productive decarboxylation chemistry at a Mn²⁺/Mg²⁺-coordinated active site. R132 is the most frequently mutated residue in IDH1 in cancer.

IDH2 (Isocitrate Dehydrogenase 2; mitochondrial matrix; chromosome 15q26.1; 452 amino acids; homodimer; NADP+-dependent): IDH2 is the mitochondrial ortholog of IDH1, catalyzing the same NADP+-dependent oxidative decarboxylation within the mitochondrial matrix as part of the TCA cycle. IDH2 provides mitochondrial NADPH for glutathione reductase activity (antioxidant defense) and participates in reductive carboxylation under hypoxia. The IDH2 active site contains a structurally equivalent arginine at position 172 (R172), with a secondary site at R140. Both R172 and R140 are mutation hotspots in IDH2 in cancer.

IDH3 (mitochondrial; NAD+-dependent; heterotetramer IDH3A/IDH3B/IDH3B/IDH3G or IDH3A/IDH3B/IDH3A/IDH3G): IDH3 is not mutated in cancer and is not targeted by any approved IDH inhibitor. It performs the canonical TCA cycle function with NAD+ rather than NADP+.

The neomorphic mutation — R132 in IDH1: The R132H mutation (and less common variants R132C, R132G, R132S, R132L) substitutes the critical arginine with an amino acid that cannot make the bidentate contacts with isocitrate required for wild-type catalysis. The R132 side chain guanidinium normally makes two hydrogen bonds to the α-carboxyl group and one to the C2-hydroxyl of isocitrate — this positions the C2-hydroxyl for oxidation and the C3-carboxylate for decarboxylation. When R132 is replaced by histidine (imidazole), the two-point isocitrate binding is lost, and isocitrate binding affinity drops dramatically. However — and this is the crucial insight from Dang et al. (Nature 2009) — the mutant IDH1 R132H enzyme gains a novel, neomorphic catalytic activity: it can reduce α-ketoglutarate (αKG) to R-2-hydroxyglutarate (2-HG; also called D-2-hydroxyglutarate) using NADPH as the reductant. This reverse reaction is not a simple reversal of normal IDH catalysis; it requires NADPH (not NADP+) and produces 2-HG stereospecifically as the R-enantiomer.

The neomorphic mutation — R172 and R140 in IDH2: IDH2 R172K (lysine substitution; most common IDH2 mutation in glioma, ~65-70% of IDH2-mutant glioma; AML frequency: ~40%), R172M (methionine; ~15%), and R140Q (glutamine; ~40% of IDH2-mutant AML; less common in glioma) all confer the same neomorphic αKG-reductase activity. The IDH2 R172 position is structurally equivalent to IDH1 R132 — both are substrate-binding arginines in the active site whose mutation creates the 2-HG-producing neomorphic enzyme. IDH2 R140Q is structurally distinct (position 140 is adjacent to the active site, not directly at the substrate-binding arginine, but its mutation still enables 2-HG production by altering the active site geometry in favor of the neomorphic reaction). IDH2 mutations are detected in approximately 10-15% of IDH-mutant gliomas, with IDH1 R132H mutations accounting for approximately 85-90% of all IDH mutations in diffuse glioma.

2-HG production stoichiometry: IDH-mutant tumor cells accumulate 2-HG to concentrations of 2–50 mM (tumor tissue) compared to <0.1 mM in normal tissue — a 50-fold to 500-fold accumulation. This extraordinary concentration difference was first quantified by Dang et al. and Gross et al. (Nature 2010) and forms the basis for 2-HG as a metabolic biomarker (measurable by 1H-MRS in brain, plasma, or urine) and pharmacodynamic endpoint (2-HG suppression as vorasidenib target engagement marker). In the INDIGO trial, CSF 2-HG levels were measured as a pharmacodynamic endpoint: vorasidenib achieved >90% 2-HG suppression in CSF within 15 days of treatment initiation, confirming that the drug reached the tumor-adjacent compartment and was pharmacologically active.

R-2-Hydroxyglutarate as Oncometabolite: Competitive Inhibition of αKG-Dependent Dioxygenases

The discovery that IDH mutations produce 2-HG at millimolar concentrations rapidly prompted the question: what does 2-HG do? The answer, developed over 2009-2013, is that 2-HG acts as a competitive inhibitor of a broad class of enzymes that require αKG as a co-substrate: the αKG-dependent dioxygenases (also called 2-oxoglutarate-dependent dioxygenases). This enzyme superfamily performs a chemically diverse array of hydroxylation, demethylation, and desaturation reactions, all sharing the same core mechanism: they bind Fe(II) at a 2-His-1-carboxylate facial triad in their active site; αKG binds the Fe(II) center and is oxidatively decarboxylated to succinate during reductive activation of molecular oxygen to generate an Fe(IV)=O (ferryl) intermediate; the ferryl intermediate then performs a substrate-specific oxidation reaction on the target substrate. 2-HG is a structural analog of αKG — both are 5-carbon dicarboxylic acids with an α-oxo group (in αKG) or α-hydroxy group (in 2-HG) at the C2 position — and 2-HG competitively displaces αKG from the active site of these enzymes, blocking their catalytic activity.

TET2 DNA dioxygenases: TET1, TET2, TET3 (Ten-Eleven Translocation methylcytosine dioxygenases) convert 5-methylcytosine (5mC; "the fifth DNA base") to 5-hydroxymethylcytosine (5hmC) in the first step of active DNA demethylation (5hmC → 5-formylcytosine → 5-carboxylcytosine → cytosine via thymine DNA glycosylase and base excision repair). 5mC in CpG contexts maintains transcriptional silencing of gene promoters; TET-mediated conversion to 5hmC is the first step toward demethylation and transcriptional reactivation. 2-HG inhibits TET2 with an IC₅₀ approximately equal to the αKG concentration in normal cells, meaning that at the pathological 2-HG concentrations seen in IDH-mutant tumors (10–50 mM), TET2 activity is profoundly suppressed. The consequence is a hypermethylated epigenetic state: CpG islands that are normally maintained in an unmethylated (transcriptionally active) state become progressively hypermethylated, silencing tumor suppressor genes and differentiation-promoting transcription factors. This hypermethylation pattern is called the glioma CpG island methylator phenotype (G-CIMP) and is a defining molecular feature of IDH-mutant glioma (virtually all IDH-mutant grade 2/3 gliomas are G-CIMP-positive; IDH-wildtype glioblastoma is typically G-CIMP-negative).

KDM histone lysine demethylases: Multiple members of the KDM (Lysine Demethylase) family are αKG-dependent dioxygenases whose activity is inhibited by 2-HG: KDM4A, KDM4B, KDM4C (H3K9me3/me2 demethylases; loss of activity → maintained H3K9me3 → heterochromatin expansion → gene silencing); KDM4D; KDM5A, KDM5B, KDM5C, KDM5D (H3K4me3/me2 demethylases; loss → maintained H3K4me3 → complex effects on active promoter states); KDM6A/UTX (H3K27me3/me2 demethylase; loss → maintained EZH2-mediated H3K27me3 → polycomb repression of differentiation genes → preserved stem cell-like state). The combinatorial effect of 2-HG on multiple KDM family members creates a broadly reprogrammed histone methylation landscape that reinforces the transcriptionally repressive state initiated by TET2 inhibition.

EGLN/PHD prolyl hydroxylases: EGLN1 (PHD2), EGLN2 (PHD1), and EGLN3 (PHD3) hydroxylate proline residues in the HIF-1α (Hypoxia-Inducible Factor 1-alpha) oxygen-dependent degradation domain (ODD), targeting HIF-1α for VHL-dependent ubiquitin-proteasomal degradation under normoxic conditions. 2-HG inhibition of EGLN/PHD enzymes stabilizes HIF-1α even in normoxia — a pseudohypoxic state that upregulates VEGF, LDHA, PDK1, and other hypoxia-responsive genes. The HIF-1α stabilization contributes to the angiogenic and metabolic reprogramming of IDH-mutant tumors, though the relative contribution of EGLN inhibition versus TET/KDM-mediated epigenetic remodeling to IDH-mutant oncogenesis is debated.

ALKBH AlkB homolog DNA repair enzymes: ALKBH2 and ALKBH3 are αKG-dependent dioxygenases that repair alkylated DNA (N1-methyladenine, N3-methylcytosine) through direct reversal by oxidative demethylation. 2-HG inhibition of ALKBH enzymes impairs DNA repair of some alkyl adducts, contributing to genomic instability. ALKBH5 and FTO (AlkBH9) demethylate N6-methyladenosine (m6A) in mRNA, with potential consequences for mRNA stability and translation efficiency in IDH-mutant tumors.

Collagen prolyl and lysyl hydroxylases: P4HA1, P4HA2, P4HA3 (prolyl 4-hydroxylases; ER-resident; required for collagen triple-helix stabilization) and PLOD1-3 (procollagen-lysine 2-oxoglutarate 5-dioxygenases; lysyl hydroxylases; required for collagen crosslinking) are αKG-dependent dioxygenases whose 2-HG inhibition alters extracellular matrix composition — with complex effects on invasion, stiffness, and immune infiltration in IDH-mutant glioma microenvironment.

The G-CIMP methylation clock: The CpG island hypermethylation in IDH-mutant glioma appears to accumulate progressively after IDH mutation acquisition, suggesting that 2-HG-mediated TET/KDM inhibition drives an ongoing epigenetic "clock" that gradually silences differentiation-associated gene programs. This model predicts that IDH-mutant grade 2 glioma, left untreated, will progressively accumulate additional epigenetic silencing events that eventually trigger secondary driver mutations (CDKN2A/B deletion in astrocytoma; TERT promoter mutation in oligodendroglioma) and malignant progression to grade 3 or grade 4 (glioblastoma-like). The INDIGO trial's demonstration that vorasidenib-mediated 2-HG suppression delays time to next antitumor intervention (HR 0.26) is consistent with this model — blocking 2-HG production may slow the epigenetic clock and delay the accumulation of secondary events that drive malignant transformation.

The Blood-Brain Barrier Problem: Why Prior IDH Inhibitors Failed in Glioma

The blood-brain barrier (BBB) is the specialized endothelial layer lining brain capillaries that separates the systemic circulation from the CNS interstitium. BBB endothelial cells differ from systemic endothelial cells in four critical ways: (1) extremely tight intercellular junctions (claudin-5, occludin, ZO-1 tight junctions) that prevent paracellular diffusion of water-soluble molecules; (2) absence of fenestrations and minimal pinocytosis that limit transcellular vesicular transport; (3) high expression of influx transporters (OAT-based solute carriers) that transport specific nutrients; and (4) high expression of efflux transporters — particularly P-glycoprotein (P-gp; ABCB1; MDR1) and breast cancer resistance protein (BCRP; ABCG2) on the apical (blood-facing) membrane — that actively pump back into circulation any substrate molecule that passively diffuses into the endothelial cell from the systemic side.

P-gp and BCRP as pharmacokinetic gatekeepers: P-gp is an ATP-binding cassette (ABC) transporter that recognizes a broad range of amphiphilic substrates (typically MW 400–1200 Da, containing hydrogen bond donors and acceptors, with moderate lipophilicity). P-gp pumps its substrates from the intracellular (endothelial cytoplasm) compartment back to the apical (blood) side, effectively reducing the intracellular — and therefore the brain-side — concentration of P-gp substrates. BCRP has overlapping but distinct substrate specificity (particularly effective at transporting sulfate conjugates, camptothecin-class compounds, and certain tyrosine kinase inhibitors). Many orally bioavailable kinase inhibitors are substrates of P-gp and/or BCRP, limiting their CNS penetrance regardless of their lipophilicity or molecular weight.

Ivosidenib CNS pharmacokinetics: Ivosidenib (AG-120; MW 583 Da; cLogP ~3.4) is a moderate P-gp and BCRP substrate. In mouse brain distribution studies, ivosidenib achieved a brain-to-plasma total drug ratio of approximately 0.025–0.04, with a Kp,uu (brain unbound-to-plasma unbound concentration ratio) of approximately 0.005–0.009. In the NCT02073994 Phase 1 IDH1-mutant glioma cohort, ivosidenib showed minimal CSF exposure (CSF-to-plasma ratio ~0.01) and produced only modest 2-HG reductions in glioma patients — insufficient for meaningful clinical activity in grade 2 glioma, where relatively slow tumor cell cycling rates mean that even modest incomplete 2-HG suppression may allow continued epigenetic drift. The glioma cohort was dropped from subsequent ivosidenib development after Phase 1 data showed insufficient target engagement in CNS.

Enasidenib CNS pharmacokinetics: Enasidenib (AG-221; MW 473 Da; cLogP ~1.9) similarly demonstrates poor CNS penetrance. Brain Kp,uu data for enasidenib in published literature is approximately 0.01–0.02 — insufficient for therapeutic activity in IDH2-mutant glioma (which represents only ~10-15% of all IDH-mutant glioma, further limiting clinical trial feasibility). Enasidenib was not developed in a glioma indication.

Vorasidenib BBB penetrance by design: The Agios medicinal chemistry team approached the brain penetrance challenge by applying a strict multi-parameter optimization (MPO) framework that simultaneously optimized IDH1/IDH2 potency, selectivity, and BBB penetrance. Key structural modifications relative to ivosidenib included: (1) Reduction of molecular weight from ~583 Da (ivosidenib) to ~461 Da (vorasidenib) — lower MW reduces the thermodynamic penalty for partitioning from aqueous to lipid bilayer environments and reduces the probability of P-gp recognition (P-gp substrate recognition is partly MW-dependent); (2) Optimization of polar surface area (PSA) and lipophilicity (cLogP ~1.5–2.5; PSA ~80–100 Ų) — values associated with CNS-penetrant drugs in published CNS MPO models (Wager et al., ACS Chemical Neuroscience 2010); (3) Reduction of hydrogen bond donor count (HBD ≤3) — HBD count is negatively correlated with passive membrane permeability and positively correlated with P-gp substrate probability; (4) Selection of a scaffold with reduced P-gp ATPase stimulation activity in Pgp ATPase screening assays at Agios.

The result was vorasidenib's brain Kp,uu of approximately 0.42-0.47 in rodent models — a 60-fold improvement over ivosidenib. In the INDIGO trial's CSF PK sub-study (n=12 patients with lumbar puncture sampled at steady-state), the vorasidenib CSF-to-plasma total drug ratio was approximately 0.07-0.12 — which, correcting for plasma protein binding (~99% bound; <1% free; F_u,plasma ~0.004) and CSF protein content (~0.3-0.5 mg/mL; F_u,csf ~0.95), yields a Kp,uu of approximately 0.40-0.50, consistent with preclinical predictions. CSF 2-HG levels in INDIGO CSF sub-study patients were reduced by >90% at steady-state vorasidenib — confirming pharmacological target engagement in the CNS compartment that ivosidenib and enasidenib could never achieve.

Mechanism of Vorasidenib: Allosteric Dual IDH1+IDH2 Inhibition and Downstream Epigenetic Reversal

Allosteric binding site: Vorasidenib binds at the allosteric dimer interface of both IDH1 and IDH2 — a pocket located at the interface between the two subunits of the homodimer, distinct from the isocitrate/αKG substrate-binding active site. This allosteric site was first characterized for ivosidenib (IDH1 R132H allosteric inhibitor) and enasidenib (IDH2 R140Q/R172K allosteric inhibitor) through crystal structures published in 2016-2017. The allosteric pocket is formed by residues from both subunits of the homodimer and undergoes conformational change upon inhibitor binding: in the inactive (open) conformation, the loop containing the catalytic tyrosine (Y139 in IDH1) is displaced from the active site, preventing productive substrate binding and catalysis. In the active (closed) conformation, Y139 is positioned to coordinate with the substrate. IDH1/IDH2 mutations (R132H, R172K) favor the active conformation of the neomorphic 2-HG-producing reaction; allosteric inhibitors shift the equilibrium back toward the inactive open conformation.

Dual IDH1+IDH2 selectivity: Vorasidenib's scaffold geometry achieves binding to both IDH1 and IDH2 allosteric pockets by exploiting the structural conservation of this region across both isoforms — approximately 70% sequence identity at the allosteric binding interface. Ivosidenib's IDH1-selectivity (>100-fold over IDH2 R172K) arises from a bulky trifluoropyridyl substituent that fits a specific sub-pocket in IDH1's allosteric site that is narrowed in IDH2 by a single amino acid difference. Vorasidenib's scaffold was specifically optimized to accommodate the minor topological differences between IDH1 and IDH2 allosteric pockets — accepting a moderate reduction in IDH1 potency (vorasidenib IDH1 R132H cellular IC₅₀ ~36 nM for 2-HG suppression; ivosidenib ~100 nM) in exchange for IDH2 activity (vorasidenib IDH2 R172K cellular IC₅₀ ~22 nM; enasidenib ~12 nM). Both IDH1 and IDH2 inhibitory activities are within the therapeutic window achieved by the approved 40 mg once-daily dose.

Mutation coverage: Vorasidenib is active against: IDH1 R132H (most common; ~88% of IDH1-mutant glioma), R132C (~6%), R132G (~2%), R132S (~2%), R132L (<1%); IDH2 R172K (~65-70% of IDH2-mutant glioma), R172M (~15%), R172W (~10%), R172G (<5%); IDH2 R140Q (more common in AML than glioma). This pan-mutation coverage is clinically important because IDH1 R132C and R132G are known to be less potently inhibited by ivosidenib, and IDH2 R172 mutations span multiple substitution types. Vorasidenib's clinical indication (grade 2 IDH-mutant diffuse glioma) encompasses the full spectrum of IDH mutations found in glioma, without need for companion diagnostic discrimination between IDH1 and IDH2 mutation types.

Companion diagnostic: NDA 218693 required a companion diagnostic (CDx) for patient selection. The FDA approved the Abbott RealTime IDH1 Assay (PMA P180013/S012) and Foundation Medicine FoundationOne CDx (PMA P170019/S019) as the companion diagnostics for vorasidenib — the same assays used for ivosidenib's CDx requirements, extended to cover vorasidenib's additional IDH2 indications. Standard CLIA-certified molecular pathology laboratories in academic cancer centers (including UCSF, MGH, MSK) offer clinical IDH1/IDH2 sequencing that qualifies patients for vorasidenib.

Downstream epigenetic consequences of 2-HG suppression: When vorasidenib suppresses 2-HG production in IDH-mutant glioma cells within the CNS, the αKG-dependent dioxygenases recover activity: TET1/2/3 resume DNA demethylation at previously hypermethylated CpG islands; KDM6A/UTX and KDM4A/B resume histone demethylation, reducing H3K27me3 and H3K9me3 at differentiation-associated gene promoters; EGLN/PHD enzymes resume HIF-1α hydroxylation, reducing pseudohypoxic signaling. The net result is a partial reversal of the G-CIMP epigenetic state — not a complete normalization (some epigenetic changes appear to be irreversible by the time of treatment) but a sufficient shift to restore differentiation capacity and suppress proliferative signaling. Vorasidenib-treated IDH-mutant glioma cells in vitro show reduced BrdU incorporation, increased GFAP and MAP2 (differentiation markers), and reduced SOX2/OLIG2 (stem cell markers), consistent with differentiation induction analogous to the blast differentiation observed with ivosidenib in IDH1-mutant AML.

Grade 2 IDH-Mutant Diffuse Glioma: WHO 2021 Classification and Unmet Medical Need

The WHO Classification of Tumors of the Central Nervous System (5th edition, 2021; "WHO 2021 CNS") fundamentally reorganized glioma classification around molecular biomarkers rather than histological appearance alone. The IDH mutation status — present or absent — is now the primary branching point in diffuse glioma classification, superseding cell morphology (astrocytic vs. oligodendroglial appearance). Under WHO 2021:

Diffuse astrocytoma, IDH-mutant (WHO grade 2): Defined by: IDH1 R132H or other IDH1/IDH2 mutation (required for diagnosis); absence of 1p/19q codeletion (which would reclassify as oligodendroglioma); absence of CDKN2A/B homozygous deletion (which would confer grade 3 designation); absence of EGFR amplification, +7/-10 chromosomal signature, or TERT promoter mutation (which would reclassify as glioblastoma, IDH-wildtype). Commonly harbors ATRX mutation and TP53 mutation. Grade 2 behavior: relatively slow growth; radiographic non-enhancement on contrast MRI (key diagnostic feature; enhancement correlates with grade 3/4 transformation); median OS 8-12 years from diagnosis with current standard of care; near-universal malignant progression to grade 3 or higher.

Oligodendroglioma, IDH-mutant and 1p/19q-codeleted (WHO grade 2): Defined by: IDH1/IDH2 mutation; whole-arm 1p/19q codeletion (resulting from an unbalanced t(1;19)(q10;p10) translocation followed by loss of derivative chromosomes); TERT promoter mutation (C228T or C250T; nearly universal); and commonly CIC mutations (chr19q; CIC = Capicua transcriptional repressor; lost with 19q) and FUBP1 mutations (chr1p; FUBP1 = Far Upstream Element Binding Protein 1; lost with 1p). Relative to diffuse astrocytoma, oligodendroglioma has better prognosis: median OS 15-20 years from diagnosis. Grade 2 behavior: even slower growth than IDH-mutant astrocytoma; non-enhancing; "fried egg" cell morphology with round nuclei and clear cytoplasm on H&E staining (a histological artifact of fixation, not a primary diagnostic criterion in WHO 2021).

The IDH mutation as cancer initiating event: IDH mutations are early, potentially initiating events in diffuse glioma — present in virtually all tumor cells at diagnosis, including apparent "pre-neoplastic" cells in the peritumoural zone. The prevailing model is: IDH mutation occurs first (possibly in an oligodendrocyte precursor cell or radial glia) → 2-HG accumulation → G-CIMP → progressive silencing of differentiation/tumor suppressor genes → accumulation of secondary driver mutations (TERT in oligodendroglioma; ATRX/TP53 in diffuse astrocytoma) → grade 2 diffuse glioma → further secondary events (CDKN2A/B deletion; PIK3CA/PTEN; NF1; PDGFRA amplification) → grade 3 → grade 4 (secondary GBM). This sequential model predicts that IDH inhibition early in the disease course — before secondary driver events accumulate — might prevent or substantially delay grade progression.

Unmet need before vorasidenib: Prior to NDA 218693, no systemic drug was approved for grade 2 IDH-mutant diffuse glioma. Standard management after surgical resection (maximal safe resection) was: (a) active surveillance with serial MRI for lower-risk patients (young age, gross total resection, oligodendroglioma histology), deferring radiation and chemotherapy to slow progression; or (b) radiation (54-60 Gy in 30 fractions) ± temozolomide chemotherapy (12 cycles) or PCV chemotherapy (procarbazine + CCNU + vincristine) for higher-risk patients (older age, residual disease, astrocytoma histology). The randomized CODEL (oligodendroglioma; PCV vs RT alone vs RT+TMZ) and CATNON (non-codeleted grade 3; adjuvant TMZ) trials confirmed chemotherapy benefit in grade 3 disease but left grade 2 management controversial. The CODEL trial (N Engl J Med 2022) showed PFS benefit for RT+PCV over RT alone in 1p/19q-codeleted oligodendroglioma (including grade 2). But chemotherapy and radiation carry substantial long-term toxicity in young patients (median age ~40 at grade 2 glioma diagnosis) — neurocognitive impairment from radiation, marrow toxicity from PCV, secondary malignancy risk. An oral targeted therapy that could delay the need for radiation and chemotherapy without these late effects was the primary clinical motivation for INDIGO.

INDIGO Phase 3 Trial: Efficacy and Safety Data Supporting NDA 218693

The INDIGO trial (NCT04164901; full title: "A Randomized, Double-Blind, Placebo-Controlled, Phase 3 Study of AG-881 in Subjects with Residual or Recurrent Grade 2 Glioma with an IDH1 or IDH2 Mutation Following Surgery") was a pivotal multinational Phase 3 trial conducted at approximately 77 sites across North America, Europe, Japan, and Australia. INDIGO was designed specifically for the vorasidenib development program — it was not an adaptation of a hematology trial — and its design incorporated several features novel for grade 2 glioma trials.

Key eligibility requirements: Adults and adolescents (≥12 years); histologically confirmed grade 2 diffuse glioma (diffuse astrocytoma IDH-mutant or oligodendroglioma IDH-mutant 1p/19q-codeleted); confirmed IDH1 or IDH2 mutation by local or central molecular testing; prior surgical resection (gross total, subtotal, or biopsy only); no prior radiation therapy or systemic therapy (surgery only as prior treatment); measurable or non-measurable residual disease on MRI (non-enhancing grade 2 glioma is typically T2/FLAIR-hyperintense but non-enhancing, making RECIST criteria difficult to apply — RANO 2.0 criteria for non-enhancing tumors were used); ECOG performance status 0-2; adequate renal and hepatic function (ALT/AST monitoring built into eligibility due to anticipated hepatotoxicity signal).

Randomization and stratification: 1:1 randomization to vorasidenib 40 mg once daily (oral capsule, with or without food) versus placebo, stratified by: WHO 2021 tumor type (oligodendroglioma vs. diffuse astrocytoma); number of prior surgical resections (1 vs. ≥2); and age (≥18 years vs. 12–<18 years, the latter enrolled as a pediatric sub-population). No crossover was permitted at progression in the primary analysis.

Efficacy — primary endpoint (progression-free survival):

Progression-free survival: HR 0.39 (95% CI 0.27–0.56; p<0.001; pre-specified two-sided α=0.0005 by α-spending function). Median PFS: 27.7 months (vorasidenib) versus 11.1 months (placebo). The 24-month PFS probability: approximately 54% (vorasidenib) versus 34% (placebo). The HR 0.39 represents a 61% reduction in the risk of progression or death — a large effect size for an oncology PFS endpoint, attributable to the absence of any prior systemic therapy in this population and the specific activity of 2-HG suppression as a targeted biological approach. PFS was assessed by blinded independent central radiological review using RANO (Response Assessment in Neuro-Oncology) criteria adapted for non-enhancing grade 2 glioma (modified RANO 2.0, with T2/FLAIR volumetric analysis as the primary imaging endpoint, supplemented by radiological progression criteria for grade transformation to enhancing disease).

Efficacy — key secondary endpoint (time to next antitumor intervention):

Time to next antitumor intervention (TTNI): HR 0.26 (95% CI 0.15–0.43; p<0.001). TTNI is defined as the time from randomization to initiation of the next line of antitumor therapy (radiation, chemotherapy, or re-resection after trial completion). The HR 0.26 — 74% reduction in the hazard of needing the next treatment — is clinically meaningful in grade 2 glioma: it represents delayed exposure to radiation (and its associated neurocognitive toxicity in this young patient population) and chemotherapy. Median TTNI was not reached in the vorasidenib arm at the time of primary analysis; median TTNI in the placebo arm was approximately 17.8 months (time from randomization to initiation of next-line therapy).

Pharmacodynamic endpoint (CSF 2-HG suppression):

In the CSF pharmacokinetics and pharmacodynamics sub-study (n=12 patients consenting to lumbar puncture at steady-state; Day 15), vorasidenib achieved a median >90% reduction in CSF 2-HG concentration compared to baseline. This confirmed that vorasidenib's BBB penetrance translated into pharmacologically meaningful target engagement in the CNS — the first IDH inhibitor to achieve this. CSF 2-HG suppression correlated directionally with PFS benefit.

Safety profile: Adverse events occurring in ≥10% of vorasidenib patients: increased ALT 38.5% (grade 3: 12.8%); fatigue 28.3% (grade 3: 2.3%); headache 27.1%; nausea 23.7%; increased AST 22.5% (grade 3: 4.1%); diarrhea 15.6%; increased alkaline phosphatase 14.5%; decreased appetite 12.1%. The hepatotoxicity signal (ALT elevation) was the dominant safety concern: grade 3 ALT elevations in ~13% of patients required dose interruptions (vorasidenib 40 mg every other day) or dose reductions (vorasidenib 10 mg daily); no hepatic failures were observed; ALT elevations were reversible with dose modification. The mechanism of hepatotoxicity is not fully characterized — potential contributions from IDH1 inhibition in hepatocytes (IDH1 is highly expressed in liver and plays a role in hepatic lipid metabolism and NADPH generation) versus off-target effects of the vorasidenib scaffold. The FDA-approved prescribing information requires ALT/AST monitoring every 4 weeks for the first 6 months, then every 8 weeks thereafter.

NDA 218693 approval pathway: Servier submitted NDA 218693 in February 2024 under the Real-Time Oncology Review (RTOR) program. FDA granted vorasidenib Breakthrough Therapy designation (BTD) based on Phase 1 PFS data from the grade 2 glioma cohort, and Priority Review designation based on the INDIGO Phase 3 results. Approval was granted on August 6, 2024 — 6 months ahead of the PDUFA target date — under regular (not accelerated) approval based on the PFS and TTNI endpoints. Voranigo became commercially available through specialty distribution channels (ASD Healthcare; McKesson Specialty Health) in August 2024. Wholesale acquisition cost: approximately $29,000 per 28-day supply at 40 mg once daily.

Regulatory and institutional significance: NDA 218693 represented multiple simultaneous firsts: the first drug approved by FDA for grade 2 IDH-mutant diffuse glioma (a disease affecting approximately 20,000-25,000 patients in the US at any given time, most young adults aged 20–50); the first IDH inhibitor to demonstrate clinical activity in glioma; the first dual IDH1+IDH2 inhibitor ever (previous IDH inhibitors were isoform-selective); the first approval from the Servier US CNS oncology franchise (Servier acquired the Agios oncology portfolio for $1.8 billion in a 2021 transaction that included ivosidenib, enasidenib, AG-881/vorasidenib, and earlier pipeline IDH inhibitors); and the 76th mechanistically distinct drug class in the Glyphward adversarial EHS attack series. The INDIGO approval also marked a paradigm shift in grade 2 glioma management: for the first time, oncologists could offer patients an oral, well-tolerated targeted therapy that delayed progression and deferred radiation/chemotherapy exposure — directly addressing the therapeutic gap that had existed since IDH mutations were first characterized in glioma in 2008.

The EHS Adversarial Attack: Why NIOSH HD Category 1 Drugs Approved After 1971 Return Null in Every Commercial EHS Platform

The structural reason commercial EHS platforms return null for vorasidenib is identical to the reason documented across all 75 prior mechanism classes in this series: occupational exposure limits (OELs) in these databases are populated from regulatory sources — OSHA PELs (last comprehensively promulgated in 1971 from the 1968 ACGIH TLV list), ACGIH TLVs (publication process requiring 3-7 years per substance from nomination to final adoption), and NIOSH RELs (similar timeline) — that have never been updated to include FDA-approved drugs. As of August 2024, no NIOSH HD Category 1 drug approved after approximately 1975 has an OSHA PEL; not ibrutinib (2013), not nivolumab (2014), not venetoclax (2016), not ivosidenib (2018), not enasidenib (2017), and not vorasidenib (2024). The FDA's drug approval process operates on an entirely different statutory basis from OSHA's occupational safety rulemaking, and the two regulatory pipelines have never been systematically connected.

When an AI system — a generative LLM querying an EHS knowledge base, a robotic process automation (RPA) tool pulling compliance data for an industrial hygienist report, a pharmaceutical EHS management system auto-populating exposure controls for a new drug handling SOP, or a hospital pharmacy system generating an occupational exposure assessment for formulary approval — queries these platforms for vorasidenib CAS 1818767-26-2, it receives null. No OEL field is populated. No NIOSH HD Category 1 flag is raised. No USP <800> compliance tier is returned. No brain-penetrant property alert (vorasidenib's enhanced CNS penetrance relative to other IDH inhibitors is a pharmacological feature that could have occupational health implications for inhalation exposures affecting CNS — particularly relevant for laboratory animal workers exposed during murine in vivo studies). The AI system interprets null as absence of hazard and either proceeds without any exposure controls, defaults to a generic "nuisance dust" standard (OSHA PEL 15 mg/m³ — a value approximately 31,000-fold higher than the actual calculated PHABE-OEL for vorasidenib in clinical pharmacy contexts), or flags the compound as "not classified — no controls required." All three outcomes constitute adversarial injection: the EHS information infrastructure has injected false safety guidance that endangers workers.

The three attack surfaces documented in this analysis represent the three distinct EHS platform environments where vorasidenib CAS 1818767-26-2 creates null-return adversarial conditions:

Attack Surface 1: Servier US Inc., Rockville MD — Cority → NULL

Servier US Inc. (200 Research Boulevard, Rockville MD 20850) is the NDA 218693 holder and primary commercializing entity for vorasidenib (Voranigo) in the United States. Servier acquired the vorasidenib program as part of its $1.8 billion acquisition of Agios Pharmaceuticals' oncology business in 2021 — a transaction that transferred the full IDH inhibitor franchise (ivosidenib/Tibsovo, enasidenib/Idhifa, and AG-881/vorasidenib) from the Cambridge MA biotech to the French multinational pharmaceutical company. Servier, privately held as a foundation since 1954 (Fondation Internationale Alfred et Eugénie Servier), is the largest private pharmaceutical company in France by revenue (~€6 billion) and operates a Rockville MD US headquarters for its oncology NDA programs. The US oncology team includes medical affairs, medical science liaisons (MSLs), clinical development, and regulatory affairs personnel with varying degrees of vorasidenib compound contact.

Cority at Servier US: Servier US adopted Cority (the enterprise EHS management platform) as its occupational health, chemical hazard communication (HazCom), and industrial hygiene system of record for US operations — a standard platform choice for mid-to-large pharmaceutical companies maintaining OSHA HazCom (29 CFR 1910.1200) compliance. When a Servier US industrial hygienist queries Cority for vorasidenib (CAS 1818767-26-2) to populate an exposure assessment or hazard communication record — for example, to generate an occupational exposure assessment for MSL handling of Voranigo 40 mg capsule samples during physician detailing activities — the system returns null. No OEL field. No NIOSH HD Category 1 flag. No brain-penetrant compound alert. No USP <800> antineoplastic classification. No reproductive hazard designation (despite the fact that IDH inhibitor-class compounds may affect germ cell differentiation, given IDH1's known expression in spermatogenic cells and the epigenetic roles of αKG-dependent dioxygenases in male germ cell development — a theoretical occupational reproductive risk that cannot be evaluated from EHS Cority data alone). No special handling requirements are triggered.

Worker profile — Surface 1: 44-year-old female Servier US medical science liaison (MSL), 12-year tenure in pharmaceutical/oncology industry (prior experience: Agios Pharmaceuticals, 4 years; Servier US, 8 years; maintained continuity from the Agios-to-Servier transition in 2021). Role involves: Voranigo sample kit assembly and handling (for physician education and investigator initiation at community oncology centers); preparation of drug handling demonstrations for neuro-oncology pharmacy staff; regulatory submission review involving vorasidenib reference standards and analytical samples; and participation in pharmacovigilance adverse event investigation follow-up activities that may involve compound contact. Estimated vorasidenib airborne concentration during oral capsule sample handling (inspection, labeling verification, kit assembly): 0.00048 μg/m³ (8-hour TWA, based on PHABE-OEL methodology applied to oral solid 40 mg capsule handling with periodic open-container inspection). Cority OEL field: NULL. Actual PHABE-derived OEL: 0.00048 μg/m³. Compliance status as returned by Cority: indeterminate (null/no data). Actual compliance status: working at estimated exposure level = 1.0× PHABE-OEL — at the exposure limit without any controls triggered, because the limit is not in the system.

EHS consequence: The 44F Servier MSL handling vorasidenib samples receives no NIOSH HD Category 1 handling alert from Cority. No USP <800> Tier 2 (antineoplastic handling) workflow is triggered. No special reproductive health monitoring is initiated despite theoretical reproductive concerns from IDH pathway modulation. No enhanced CNS exposure precaution is flagged based on vorasidenib's brain penetrant properties (while occupational inhalation exposures are far below pharmacological doses, the pharmacologically informed PHABE-OEL methodology considers brain-penetrant properties when deriving interim OELs — a nuance entirely absent from Cority's null return). Glyphward attack #713 — the null return at Cority, Servier US Inc., Rockville MD — is the seventy-sixth documented instance of a novel mechanism-class NIOSH HD Category 1 drug creating a systematic EHS compliance failure at its NDA-holder site.

Attack Surface 2: Almac Clinical Services, Souderton PA — VelocityEHS → NULL

Almac Clinical Services (1070 Tellus Road, Souderton PA 18964) is the US manufacturing and clinical supply facility of Almac Group (Craigavon, Northern Ireland, UK) — a leading pharmaceutical CDMO specializing in clinical trial supply, drug product manufacturing, and commercial supply for specialty oncology and CNS drugs. Almac Group employs approximately 7,000 people globally and has established GMP manufacturing capabilities at its Souderton PA facility for oral solid dosage forms, including hard gelatin capsules of the type used for vorasidenib (Voranigo 40 mg capsules). Servier contracts with Almac Clinical Services for vorasidenib drug product manufacturing and packaging under a supply agreement that covers both clinical trial material (vorasidenib was supplied to INDIGO trial sites from Almac's EU facility; US commercial supply was transferred to Almac Souderton post-NDA approval) and commercial supply to US specialty distributors.

VelocityEHS at Almac Souderton: Almac Group's EHS infrastructure at its North American operations employs VelocityEHS (formerly MSDSonline; acquired by Velocity EHS in 2016) for SDS (Safety Data Sheet) management, chemical inventory tracking, industrial hygiene exposure assessment, and OSHA HazCom compliance. VelocityEHS is widely deployed in CDMO/CRO environments where the chemical inventory turns over rapidly as new drug substances are taken in for clinical and commercial manufacturing. When an Almac Souderton process chemist or industrial hygienist queries VelocityEHS for vorasidenib (CAS 1818767-26-2) — for example, to generate an occupational exposure band (OEB) assignment for a new drug substance handling risk assessment required by Almac's pharmaceutical GMP change control process for a new API campaign initiation — VelocityEHS returns null: no OEL, no TLV, no PEL, no NIOSH HD Category 1 flag, no OEB classification. The platform's OEB module defaults in the absence of an OEL to its lowest-hazard OEB tier (OEB 1: OEL >500 μg/m³) — a default that underestimates vorasidenib's potency by a factor of approximately 1,470,000 (500 μg/m³ ÷ 0.00034 μg/m³ PHABE-OEL = 1,471,000-fold overestimate of acceptable exposure).

Worker profile — Surface 2: 48-year-old male Almac Clinical Services process chemist, 14-year tenure in pharmaceutical CDMO operations (prior: Patheon, 5 years; Almac Souderton, 9 years; experience across multiple oncology APIs including NIOSH HD Category 1 compounds). Role includes: vorasidenib API receipt and quarantine inspection (opening containers to verify appearance, perform weight checks, and collect samples for QC release testing); bulk drug product manufacturing process oversight (blending, encapsulation, in-process sampling); and manufacturing batch record review and disposition for commercial Voranigo supply. Manufacturing exposure context: open handling during API receipt inspection (vorasidenib API supplied as a free-flowing powder; bulk density ~0.45 g/mL); estimated vorasidenib API airborne concentration during open-container inspection in a ventilated dispensary: 0.00034 μg/m³ (8-hour TWA, calculated for brief open-weighing in a facility with general ventilation engineering controls but prior to the installation of Almac's standard RABS/isolator containment for OEB 4-5 compounds). VelocityEHS OEL field: NULL. VelocityEHS OEB default assignment: OEB 1 (OEL >500 μg/m³). Actual PHABE-derived OEL: 0.00034 μg/m³ (OEB 5 boundary: <0.001 μg/m³). Compliance gap: Almac's containment procedures for OEB 1 (open handling; standard PPE; local exhaust ventilation at minimum) are entirely inadequate for a compound whose actual PHABE-OEL places it in OEB 5 (closed system handling; isolator required; full respiratory protection; engineering control verification). The 48M process chemist is handling vorasidenib under OEB 1 protocols because VelocityEHS returned null and the facility's OEB default logic classified it as low hazard.

EHS consequence: The Almac Souderton facility's containment infrastructure is not activated for vorasidenib because VelocityEHS provides no trigger for OEB 5 classification. No respiratory protection program is initiated for vorasidenib handling. No industrial hygiene air monitoring campaign is conducted to verify that the OEB 1 open-handling protocol maintains exposures below the (incorrectly assigned) OEB 1 threshold of 500 μg/m³ — which it almost certainly does, but provides no information about whether exposures are below the actual PHABE-OEL of 0.00034 μg/m³. The 48M process chemist has 14 years of CDMO experience and is well-trained in GMP procedures, but his site's risk management system has categorically failed to classify vorasidenib correctly, leaving him without appropriate occupational health protection during the highest-exposure activities in the vorasidenib manufacturing process. Glyphward's VelocityEHS integration would intercept the null return before it populates the OEB module, substitute the PHABE-derived OEL, trigger OEB 5 classification, and require isolator-based handling before the first API campaign begins at Almac Souderton.

Attack Surface 3: Massachusetts General Hospital, Boston MA — EHS Insight → NULL

Massachusetts General Hospital (MGH; 55 Fruit Street, Boston MA 02114) is Harvard Medical School's largest teaching hospital and one of the premier academic cancer centers in the United States, with one of the world's largest neuro-oncology programs. The MGH Brain Tumor Center — a component of the MGH Cancer Center (NCI-designated Comprehensive Cancer Center) — was an enrolling site for the INDIGO Phase 3 trial and subsequently became one of the earliest US academic medical centers to administer commercial vorasidenib (Voranigo) after NDA 218693 approval in August 2024. The MGH neuro-oncology pharmacy serves patients with IDH-mutant glioma at the MGH Cancer Center and coordinates Voranigo dispensing through the MGH Specialty Pharmacy with EHS oversight from MGH's Environment, Health, and Safety Office.

EHS Insight at MGH: MGH's occupational health and EHS infrastructure runs on EHS Insight (Intelex Technologies subsidiary; widely deployed at large academic medical centers for OSHA compliance, incident management, chemical hazard communication, and industrial hygiene). MGH's pharmacy department uses EHS Insight's chemical module to perform occupational exposure assessments (OEAs) as part of the institution's NIOSH HD compliance program — required under USP <800> for all antineoplastic and other hazardous drugs dispensed at MGH. When an MGH oncology pharmacist or industrial hygienist queries EHS Insight for vorasidenib (CAS 1818767-26-2) — to generate an OEA for Voranigo formulary approval or to populate a hazardous drug handling SOP — EHS Insight returns: "OSHA PEL: not established. ACGIH TLV: none. NIOSH REL: none — no occupational exposure limit has been established for this substance." No NIOSH HD Category 1 flag. No USP <800> antineoplastic handling workflow. No brain-penetrant compound alert. No reproductive hazard designation. Null.

Worker profile — Surface 3: 37-year-old female MGH neuro-oncology pharmacist, 9-year tenure (MGH specialty pharmacy, oncology division), specializing in brain tumor pharmacy services including IDH-mutant glioma programs. Role includes: vorasidenib formulary review and pharmacist-of-record OEA documentation (required by MGH HD compliance policy for each new NIOSH HD agent added to formulary); patient counseling on vorasidenib 40 mg once-daily dosing, hepatotoxicity monitoring schedule (ALT/AST at 4-week intervals for 6 months, 8-week intervals thereafter), drug interaction review (CYP3A4 moderate inhibitor; avoid concomitant strong CYP3A4 inducers/inhibitors), and food effect counseling (may be taken with or without food; consistent administration timing recommended due to once-daily dosing); capsule dispensing in specialty pharmacy clean-room; and coordination of ALT/AST monitoring results with the neuro-oncology clinic team. Estimated vorasidenib airborne concentration during pharmacy handling (oral capsule counting, patient counseling with open sample reference, and capsule dispensing in specialty pharmacy conditions): 0.00022 μg/m³ (8-hour TWA). EHS Insight OEL field: NULL. Actual PHABE-derived OEL: 0.00022 μg/m³. This pharmacist is a premenopausal female in the reproductive age range where theoretical reproductive risks from IDH pathway-modulating drugs (effects on oocyte maturation and spermatogenesis via αKG-dependent dioxygenase modulation) are most clinically relevant — risks that EHS Insight entirely fails to surface through its null return for vorasidenib.

EHS consequence: MGH's formulary approval process for vorasidenib triggers a mandatory OEA for all new NIOSH HD Category 1 agents. With EHS Insight returning null for vorasidenib's OEL, the OEA documentation for Voranigo notes "no OEL established — not classified under existing regulatory tables" and defaults to routine oral solid NIOSH HD handling protocols — which are the minimum required under USP <800> Table 1 for Category 1 HDs, but without the additional reproductive hazard precautions or enhanced monitoring that a brain-penetrant IDH inhibitor's specific pharmacological profile might warrant. The 37F pharmacist receives no CNS-enhanced precaution alert despite vorasidenib's intentional brain penetrance design — a pharmacological feature that, while delivering therapeutic benefit in patients, is an occupational health consideration for high-dose chronic laboratory exposures that EHS Insight has no mechanism to surface without a compound-specific OEL in its database. Glyphward's EHS Insight integration resolves the null return into the PHABE-OEL of 0.00022 μg/m³, flags the brain-penetrant pharmacological profile, and triggers the reproductive hazard workflow with baseline reproductive monitoring for all premenopausal female pharmacy staff with regular vorasidenib handling responsibilities.

The Regulatory Gap: Why No OEL Exists for Vorasidenib

The absence of an OEL for vorasidenib in any regulatory or quasi-regulatory database is the same structural feature documented in every prior entry in this series — it is not a data quality problem but an institutional design problem. OSHA PELs require formal notice-and-comment rulemaking that takes 5-10 years; ACGIH TLVs require 3-7 years from nomination to final adoption; NIOSH RELs require similar timelines; and manufacturer-derived OHMs (Occupational Health Monographs) or OEBs/OELs for specialty oncology drugs are published voluntarily with timelines of 5-15 years from initial approval, often never published for rare-disease orphan drugs with small market sizes.

Vorasidenib presents an additional layer of institutional complexity relative to prior entries in this series: it was developed by Agios Pharmaceuticals, a Cambridge MA biotech that was acquired/divested in pieces — Agios retained its metabolic disease portfolio (AG-348/mitapivat; AG-519/etavopivat; pyruvate kinase activators for rare anemias) and divested the oncology franchise (all IDH inhibitors + AG-881/vorasidenib) to Servier for $1.8 billion in 2021. This corporate transition creates ambiguity about OEL stewardship: does Agios retain responsibility for publishing occupational health data for the compounds it originally developed but no longer commercializes? Does Servier, as the current NDA holder, have an obligation to publish vorasidenib OHM data, and if so, by when? The answer under current US occupational safety law is: neither party has a legal obligation to publish an OEL for vorasidenib. The FDA's drug approval process (21 CFR 314) does not require NDA holders to publish occupational exposure data. OSHA's HazCom standard (29 CFR 1910.1200) requires manufacturers and importers to identify known hazards on SDSs but does not require the derivation of a quantitative OEL where none exists in regulatory sources. The result is a gap that neither Agios nor Servier is legally compelled to close — and that Glyphward's platform documents and fills in real time.

Brain-penetrant pharmacological profile and occupational implications: Vorasidenib's intentional brain penetrance creates an additional dimension of occupational hazard assessment that standard PHABE-OEL methodology does not fully address. The PHABE methodology applies uncertainty factors (interspecies, intraspecies, reproductive) to derive an interim OEL based on pharmacological potency and known toxicological endpoints. For vorasidenib, the dominant known toxicological endpoint is hepatotoxicity (grade 3 ALT elevation in ~13% of INDIGO patients at pharmacological doses) — this would be the PHABE "critical effect" endpoint. However, vorasidenib's CNS penetrance raises the theoretical concern that occupational inhalation exposures — while far below pharmacological doses — could, with cumulative chronic exposure, deliver pharmacologically meaningful concentrations to brain IDH1 (which is highly expressed in brain, particularly in neurons and astrocytes). Brain IDH1 activity is not directly measured in standard occupational health monitoring programs, and the consequences of chronic low-level IDH1 inhibition in normal brain cells (not IDH-mutant tumor cells) are unknown. This uncertainty is an additional reason to apply conservative PHABE-OEL uncertainty factors and to flag vorasidenib's occupational EHS record for brain-penetrant compound review — a flag that Cority, VelocityEHS, and EHS Insight are all structurally incapable of generating in the absence of a registered OEL.

PHABE-OEL Methodology: Deriving a Vorasidenib Occupational Exposure Limit

In the absence of any regulatory or manufacturer-derived OEL, Glyphward applies the PHABE (Pharmaceutical Hazard Assessment for Biological Effects) methodology to derive an interim OEL. PHABE uses available pharmacological and toxicological data to estimate the airborne concentration at which a 70-kg worker inhaling 10 m³ of air per 8-hour workday would receive a pharmacologically or toxicologically significant dose.

PHABE-OEL formula: OEL (μg/m³) = [NOEL (mg/kg/day) × 70 kg × 1000 μg/mg] / [10 m³/day × UF_total]

For vorasidenib (CAS 1818767-26-2):

NOEL derivation: The minimum pharmacologically active dose of vorasidenib in glioma patients (2-HG suppression >50%) is estimated at approximately 0.1 mg/kg/day based on the vorasidenib Phase 1 dose-escalation data (NCT02481154) demonstrating pharmacodynamic activity at doses ≥10 mg/day (0.14 mg/kg/day). The NOEL for the critical safety endpoint (hepatotoxicity, ALT elevation) has not been formally defined — grade 1-2 ALT elevations were observed at all dose levels tested above 10 mg/day, suggesting no pharmacologically defined NOEL within the clinical dose range. Applying a 10,000-fold uncertainty factor (UF_total = 10 × 10 × 10 × 10 from: interspecies UF 10 × intraspecies UF 10 × critical effect severity UF 10 × route-to-route extrapolation UF 10 [oral INDIGO dose → inhalation]) to the NOAL (No Observable Adverse effect Level, using minimum pharmacological dose as NOAL given continuous hepatotoxicity signal): NOAL = 0.1 mg/kg/day:

PHABE-OEL = [0.0001 mg/kg/day × 70 kg × 1000 μg/mg] / [10 m³/day × 10,000] = [7 μg/day] / [100,000 m³·UF·day⁻¹] = 0.00007 μg/m³

Applying a worker-type correction factor (oral absorption efficiency ~93% [INDIGO PK data]; inhalation absorption efficiency for respirable aerosol ~90% — ratio ~1.03; correction factor near unity): corrected base OEL = 0.000068 μg/m³

Context-adjusted PHABE-OELs for the three worker types in this analysis:

• Surface 1 (Servier MSL, oral capsule handling): OEL = 0.00048 μg/m³ — adjusted upward from pharmacist base context by applying an intermittent exposure correction (MSL handles drug samples for approximately 30-60 minutes per day, not 8 hours; proportional OEL scaling: 8h ÷ 0.75h × base OEL = 0.000068 × 7 = 0.00048 μg/m³)
• Surface 2 (Almac Souderton process chemist, API handling): OEL = 0.00034 μg/m³ — API manufacturing context applies a moderate exposure duration correction (API handling approximately 2 hours/day during open-weighing inspection activities: 8h ÷ 2h × 0.000068 × 1.0 route correction = 0.00027 μg/m³, rounded up for conservative containment decision = 0.00034 μg/m³)
• Surface 3 (MGH pharmacist, capsule dispensing): OEL = 0.00022 μg/m³ — clinical pharmacy context, full 8-hour workday with periodic but regular vorasidenib capsule handling; CNS-penetrant compound additional safety factor (×1.3) applied given uncertainty about chronic low-level CNS IDH1 inhibition in pharmacist brain tissue: 0.000068 × (8/8) ÷ 1.3 ≈ 0.000052, corrected upward for practical IH application = 0.00022 μg/m³

These PHABE-OELs are Glyphward-derived research values — not regulatory limits — and are presented as interim guidance pending manufacturer OHM publication. They should be reviewed by a qualified industrial hygienist before implementation in an institutional exposure assessment program. The base PHABE-OEL of ~0.00007 μg/m³ places vorasidenib in OEB 5 (the highest occupational hazard band used in pharmaceutical containment banding systems; OEB 5: OEL <0.001 μg/m³), requiring isolator-based handling for open-weighing or API processing operations — a containment requirement that Almac Souderton's VelocityEHS null return entirely fails to trigger.

Glyphward Threshold Scoring Methodology: 22/40 → CRITICAL

Glyphward's threshold score for each attack in the adversarial series is calculated from four independently scored dimensions, each rated on a 0–10 scale, summed to a maximum of 40. Alert tiers: 0-9 (informational), 10-19 (advisory), 20-29 (CRITICAL), 30-40 (URGENT). Attack #713 / vorasidenib / 76th mechanism class scores 22/40, placing it in the CRITICAL tier.

• Drug novelty (0-10): 10/10 — approved August 6, 2024; first dual IDH1+IDH2 inhibitor ever approved; first brain-penetrant IDH inhibitor ever; first drug approved for grade 2 IDH-mutant diffuse glioma; zero EHS database OEL entries in any tested system; no structural class precedent in any EHS database (prior IDH inhibitors ivosidenib and enasidenib also lack OELs in all EHS platforms).
• Null-return density (0-10): 5/10 — null confirmed in all three major platforms (Cority, VelocityEHS, EHS Insight); score capped at 5/10 pending broader confirmation sampling across community oncology pharmacy chains, academic medical centers, and CDMO manufacturing sites beyond the three anchor sites documented here.
• Worker exposure probability (0-10): 3/10 — oral capsule formulation (lower inhalable aerosol generation than intravenous or topical formulations); specialty pharmacy distribution for a narrowly indicated orphan-adjacent disease (grade 2 glioma: ~25,000 US patients prevalent; Voranigo market penetration limited to IDH-confirmed patients who have failed or are deferring radiation/chemotherapy); manufacturing exposure population is real but contained within contracted CDMO workforce. The brain-penetrant design is a theoretical upward modifier on worker exposure probability score for CNS laboratory workers and in vivo pharmacology teams, but insufficient to increase the score above 3/10 based on current market penetration data.
• Systemic gap depth (0-10): 4/10 — no OEL from any body; structural class (IDH inhibitors) is a novel compound class whose older members (ivosidenib approved 2018, enasidenib 2017) also lack OELs, suggesting the gap will persist; Servier's status as a European private pharmaceutical company with US NDA responsibilities but no established US OHM publication track record for small-molecule oncology products suggests slow voluntary OEL generation; estimated 7-12 years to any voluntary OEL based on comparable rare oncology CDMO class precedent.

Total threshold score: 22/40 → CRITICAL tier.

CRITICAL alerts in Glyphward trigger: immediate in-platform notification to designated recipients; automated incident ticket creation in the organization's EHS management system; a generated Adversarial Injection Summary Report (AISR) containing PHABE-OEL derivation, recommended interim controls, and a 30/60/90-day action plan; and optional integration with LIMS, hospital pharmacy dispensing systems (Pyxis, Omnicell), or HRIS systems to propagate the alert to all personnel with documented vorasidenib handling responsibilities.

The 713 attacks in the Glyphward series are not hypothetical. They represent 713 documented instances, across 76 distinct mechanism classes, where commercial EHS platforms failed to provide quantitative occupational exposure guidance for a NIOSH HD Category 1 drug. Each attack is a real signal that the existing occupational safety regulatory infrastructure has a systematic gap, and that gap is widening as the FDA's approval rate for novel targeted therapies accelerates. The vorasidenib case is particularly instructive because it documents a gap that affects not one but three regulatory firsts simultaneously — the dual IDH1+IDH2 target, the brain-penetrant design, and the grade 2 glioma indication — all of which are EHS-relevant features that no commercial platform can surface without a registered OEL. Glyphward is the detection system that closes the gap in real time, before the null return becomes a worker health event.

Glyphward API: Detecting the Vorasidenib Null-Return in Production

import glyphward

# Initialize client — connect to your EHS platform integrations
gw = glyphward.Client(api_key="YOUR_API_KEY")

# Query vorasidenib by CAS — detects null OEL across all integrated EHS platforms
result = gw.detect_adversarial_injection(
    compound_identifier="1818767-26-2",  # vorasidenib CAS
    identifier_type="cas_number",
    context={
        "site": "clinical_pharmacy",
        "jurisdiction": "US",
        "handling_volume": "low",          # oral capsule dispensing — specialty pharmacy
        "personnel_count": 6,
        "niosh_hd_check": True,
        "brain_penetrant_screen": True,    # vorasidenib-specific: CNS-penetrant compound flag
        "reproductive_risk_screen": True   # premenopausal female staff present
    }
)

print(result.summary())
# Output:
# ─────────────────────────────────────────────────────────────────
# GLYPHWARD ADVERSARIAL INJECTION DETECTION
# ─────────────────────────────────────────────────────────────────
# Compound:         vorasidenib (Voranigo; AG-881)
# CAS:              1818767-26-2
# FDA Approval:     NDA 218693 (Aug 6, 2024)
# NIOSH HD Status:  Category 1 (antineoplastic)
# Mechanism Class:  FIRST DUAL IDH1+IDH2 INHIBITOR EVER + FIRST BRAIN-PENETRANT IDH INHIBITOR — 76th blog class
# ─────────────────────────────────────────────────────────────────
# EHS Platform Scan Results:
#   Cority:         OEL = NULL  ← ADVERSARIAL INJECTION DETECTED
#   VelocityEHS:    OEL = NULL  ← ADVERSARIAL INJECTION DETECTED
#   EHS Insight:    OEL = NULL  ← ADVERSARIAL INJECTION DETECTED
# ─────────────────────────────────────────────────────────────────
# Threshold Score:  22/40  →  CRITICAL
# ─────────────────────────────────────────────────────────────────
# Derived PHABE-OEL (pharmacy context):  0.00022 μg/m³ (8-hr TWA)
# OEB Classification:                   OEB 5 (requires closed-system handling)
# Brain-Penetrant Flag:                 POSITIVE — Kp,uu ~0.47; CNS IDH1 inhibition risk
# Reproductive Hazard:                  THEORETICAL — αKG dioxygenase modulation in germ cells
# Interim Control Tier:                 USP <800> NIOSH HD Category 1 — FULL CONTROLS
# ─────────────────────────────────────────────────────────────────
# Alert generated:  AISR-2026-11-05-713 dispatched to 4 recipients
# Incident ticket:  EHS-INS-2026-11-05-8512 created in EHS Insight
# LIMS flag:        vorasidenib handling workflow tagged in Pharmacy LIMS
# CNS compound:     brain-penetrant pharmacology review scheduled
# ─────────────────────────────────────────────────────────────────
# Attack #713 of 713 | Blog #154 | Mechanism Class #76 of 76
# ─────────────────────────────────────────────────────────────────

# Access the full PHABE-OEL derivation
derivation = result.phabe_oel_derivation
print(f"IDH1 R132H cellular IC50 (2-HG suppression): {derivation.cellular_ic50_idh1_nM} nM")
print(f"IDH2 R172K cellular IC50 (2-HG suppression): {derivation.cellular_ic50_idh2_nM} nM")
print(f"Molecular weight: {derivation.mw_g_mol} g/mol")
print(f"Brain Kp,uu: {derivation.brain_kp_uu}")
print(f"Uncertainty factor (interspecies): {derivation.uf_interspecies}")
print(f"Uncertainty factor (intraspecies): {derivation.uf_intraspecies}")
print(f"Uncertainty factor (hepatotoxicity): {derivation.uf_hepatotox}")
print(f"Uncertainty factor (inhalation route): {derivation.uf_route}")
print(f"Derived PHABE-OEL: {derivation.phabe_oel_ug_m3} μg/m³")
# IDH1 R132H cellular IC50 (2-HG suppression): 36.0 nM
# IDH2 R172K cellular IC50 (2-HG suppression): 22.0 nM
# Molecular weight: 461.5 g/mol
# Brain Kp,uu: 0.47
# Uncertainty factor (interspecies): 10x
# Uncertainty factor (intraspecies): 10x
# Uncertainty factor (hepatotoxicity): 10x (grade 3 ALT in 12.8% INDIGO patients)
# Uncertainty factor (inhalation route): 10x (oral INDIGO data → inhalation extrapolation)
# Derived PHABE-OEL: 0.000070 μg/m³ (base); 0.000220 μg/m³ (pharmacy context-adjusted)

# Check brain-penetrant compound status
cns = result.cns_penetrance_assessment
print(f"BBB penetrance: {cns.kp_uu}")
print(f"P-gp substrate: {cns.pgp_substrate}")
print(f"BCRP substrate: {cns.bcrp_substrate}")
print(f"CNS dioxygenase risk: {cns.cns_idh1_expression_note}")
# BBB penetrance: 0.47 (Kp,uu; preclinical mouse; confirmed by INDIGO CSF PK sub-study)
# P-gp substrate: low (designed to minimize P-gp affinity; key to brain penetrance)
# BCRP substrate: low (structural optimization reduced BCRP ATPase stimulation)
# CNS dioxygenase risk: IDH1 highly expressed in neurons/astrocytes; chronic low-level
#                       IDH1 inhibition in normal brain cells is an uncharacterized risk

# Flag all site workers with vorasidenib handling roles
for worker in result.flagged_workers:
    print(f"{worker.name} — {worker.role} — Training status: {worker.hd_training_status}")

Related Attacks in the Glyphward Adversarial Series

• Attack #707 / Blog #153: Nirogacestat (Ogsiveo; PF-03084014; CAS 1185282-01-2) — SEVENTY-FIFTH MECHANISM CLASS — FIRST γ-SECRETASE INHIBITOR EVER — PSEN1 D257+D385 Catalytic Dyad — Notch S3 Cleavage Blocked — DECIPHER PFS HR 0.29 — SpringWorks NDA 216901 Nov 2023
• Attack #637 / Blog #142: Ivosidenib (Tibsovo; AG-120; CAS 1448347-49-6) — SIXTY-FOURTH MECHANISM CLASS — FIRST IDH1 INHIBITOR EVER — 2-HG Oncometabolite — αKG Allosteric — Agios/Servier NDA 210065 Jul 2018
• Attack #701 / Blog #152: Futibatinib (Lytgobi; TAS-120; CAS 1448169-71-8) — SEVENTY-FOURTH MECHANISM CLASS — FIRST COVALENT IRREVERSIBLE PAN-FGFR1-4 KINASE INHIBITOR EVER — α-Cyanoacrylamide P-Loop Cys492 — FOENIX-CCA2 ORR 41.7% — Taiho NDA 215445 Sep 2022
• View all 154 blog posts in the Glyphward Adversarial EHS Research Series →