Tazemetostat (Tazverik; CAS 1403254-99-8): NIOSH HD Category 1 FIRST EZH2 Methyltransferase Inhibitor OEL Gap [SAM-Competitive Cofactor Displacement; H3K27me3 Trimethylation Silencing; PRC2 Complex [EZH2+EED+SUZ12]; ELEVENTH Distinct Mechanism Class — All 520 Prior Attacks ATP-Competitive Kinase/PARP/BH3/20S; FIRST S-Adenosyl Methionine Cofactor-Competitive Enzyme Inhibitor in 521-Attack Portfolio; Cofactor-Competitive Enzyme Inhibition Absent from All OEL Derivation Frameworks] + FIRST Epigenetic Modifier + FIRST PRC2 Complex Inhibitor + FIRST Orphan/Rare Disease NDA in Portfolio [Epithelioid Sarcoma; INI1/SMARCB1 Biallelic Loss Synthetic Lethality; FDA Accelerated Approval January 2020; <1:100,000 Incidence] + FIRST EZH2 Gain-of-Function Mutation OEL Gap [Y641F IC50 ~0.3 nM; >30-Fold Mutant Hypersensitivity; Patient-Tumor EZH2 Mutation Invisible to EHS Frameworks] + FIRST Ipsen/Epizyme NDA 213400 Null [Ipsen $247M Epizyme Acquisition May 2022; Former Epizyme Cambridge MA; Acquisition EHS Transition Gap] + FIRST Dihydropyridinone-Benzamide-Methylsulfonyl Scaffold [MW ~636 g/mol; SECOND >600 g/mol after Venetoclax]; Ipsen Biopharmaceuticals Cambridge MA 49M 17yr Cority ÷10; Patheon/Thermo Fisher Durham NC 44F 12yr VelocityEHS ÷10; Memorial Sloan Kettering New York NY 38F 11yr EHS Insight ÷10; Glyphward Threshold 22; 521st Adversarial Attack; 119th Long-Form Blog

Tazemetostat (Tazverik®; CAS 1403254-99-8; dihydropyridinone-benzamide-methylsulfonyl scaffold; MW ~636 g/mol; NIOSH HD Category 1; Ipsen/Epizyme NDA 213400; FDA accelerated approval January 23, 2020 [epithelioid sarcoma]; full approval June 18, 2020 [EZH2-mutated follicular lymphoma]; MECHANISM: FIRST EZH2 [ENHANCER OF ZESTE HOMOLOG 2] METHYLTRANSFERASE INHIBITOR — tazemetostat is the FIRST histone methyltransferase inhibitor and the FIRST S-adenosyl methionine (SAM) cofactor-competitive enzyme inhibitor in the 521-attack Glyphward adversarial portfolio; ELEVENTH distinct mechanism class; EZH2 MECHANISM: EZH2 is the catalytic SET-domain subunit of PRC2 [Polycomb Repressive Complex 2; EZH2 + EED + SUZ12 + RBBP4]; EZH2 catalyzes transfer of the methyl group from S-adenosyl-L-methionine [SAM] to the ε-amino group of histone H3 lysine 27 [H3K27]; H3K27me1 → H3K27me2 → H3K27me3 [trimethylation]; H3K27me3 recruits PRC1 → H2AK119ub1 → chromatin compaction → transcriptional silencing of CDKN2A/p16, PTEN, RASSF1A; SAM-COMPETITIVE INHIBITION: tazemetostat binds EZH2 SET domain at the SAM cofactor-binding pocket, not the H3K27 peptide substrate site; SAM cofactor-competitive inhibition is DISTINCT from ATP-competitive kinase inhibition [all prior protein kinase/PI3K attacks], NAD+-competitive PARP inhibition [olaparib/niraparib/rucaparib], BH3-domain displacement [venetoclax], and 20S proteasome active-site occupation [bortezomib/carfilzomib]; SAM-COMPETITIVE MECHANISM IS ABSENT FROM ALL OEL DERIVATION FRAMEWORKS; EZH2 MUTANT HYPERSENSITIVITY: EZH2 WT IC50 ~11 nM vs EZH2 Y641F IC50 ~0.3 nM [>30-fold]; EZH2 Y641F/N gain-of-function mutations [~17–22% FL; ~7% DLBCL]; paradoxically elevated dimethyl→trimethyl activity; H3K27me3 global accumulation in mutant B cells; COMPANION DIAGNOSTIC: cobas EZH2 Mutation Test [Y641F/N/S/H/C; A677G] in patient tumor FFPE; patient tumor EZH2 mutation absent from workers; EPITHELIOID SARCOMA: FIRST ORPHAN/RARE DISEASE NDA in 521-attack portfolio; INI1/SMARCB1 biallelic loss [chromosome 22q11]; SWI/SNF complex dysfunction → unopposed EZH2 → synthetic lethality; incidence <1:100,000; FDA orphan drug designation; rare disease OEL corridor absent from all industrial hygiene frameworks; SCAFFOLD: dihydropyridinone-benzamide-methylsulfonyl; MW ~636 g/mol; SECOND >600 g/mol in portfolio after venetoclax 868.44 g/mol [Attack #508; BCL-2]; FIRST dihydropyridinone core in 521-attack Glyphward portfolio; OSHA: No PEL; ACGIH: No TLV; NIOSH: No REL — NIOSH Hazardous Drug Category 1 [NIOSH 2016 HD List Table 1]; GHS H340 H350 H361; FIRST IPSEN/EPIZYME NULL: NDA 213400; Ipsen acquired Epizyme May 2022 ~$247M; post-acquisition EHS transition gap [Ipsen peptide-dominant library + Epizyme HD Cat 1 EZH2 inhibitor]; Ipsen Biopharmaceuticals Cambridge MA / Patheon Durham NC / MSK New York NY — 521st adversarial attack, 119th long-form blog). Produces FIRST tazemetostat (CAS 1403254-99-8) NIOSH HD Category 1 OEL null-return adversarial blog [triple enforcement vacuum: OSHA No PEL + ACGIH No TLV + NIOSH No REL; only NIOSH HD Category 1; no OEL derivation for any EZH2 methyltransferase or SAM-competitive enzyme inhibitor]; ELEVENTH mechanism class [FIRST SAM-competitive cofactor-competitive enzyme inhibitor; H3K27me3 epigenetic silencing absent from all OEL frameworks]; FIRST epigenetic modifier; FIRST PRC2 complex inhibitor; FIRST orphan/rare disease NDA [epithelioid sarcoma; INI1/SMARCB1 synthetic lethality; FDA accelerated approval]; FIRST Ipsen/Epizyme acquisition EHS transition null; FIRST dihydropyridinone scaffold [MW ~636 g/mol; second >600 g/mol].

The Eleventh Mechanism Class: EZH2 SAM-Competitive Methyltransferase and the Cofactor-Competitive Epigenetic Inhibition Gap

The 521 adversarial attacks documented in the Glyphward portfolio have mapped a progressive architecture of EHS AI platform failure across ten prior mechanism classes: receptor and non-receptor tyrosine kinases (RTKs: erlotinib Attack #489, gefitinib #490, osimertinib #503, alectinib #509, lorlatinib #511, neratinib #515, tucatinib #518, and others), serine/threonine kinases (BRAF: vemurafenib #495, dabrafenib #504; MEK1/2: trametinib #505; CDK4/6: palbociclib #500, ribociclib #502, abemaciclib #517; mTOR allosteric: everolimus #496, temsirolimus #497), BTK covalent (ibrutinib #499, acalabrutinib #512, zanubrutinib #513), PARP DNA repair enzymes (olaparib #501, niraparib #507, rucaparib #510), 20S proteasome (bortezomib #481, carfilzomib #506), BCL-2 anti-apoptotic (venetoclax #508), lipid kinases PI3K (idelalisib #514, copanlisib #516, duvelisib #519, alpelisib #520), and immunomodulatory agents. Every attack in this portfolio previously inhibited either a protein or lipid kinase (ATP-competitive), a nucleotide-processing enzyme (PARP NAD+ competitive), an anti-apoptotic protein–protein interaction (BCL-2 BH3 groove), or a proteolytic complex active site (20S proteasome). Tazemetostat introduces the eleventh mechanism class: a histone lysine methyltransferase inhibited via competitive displacement of the S-adenosyl methionine (SAM) methyl-donor cofactor from the SET domain catalytic pocket.

SAM (S-adenosyl-L-methionine; AdoMet; MW 399.5 g/mol; generated by methionine adenosyltransferase [MAT; specifically MAT1A in hepatocytes, MAT2A ubiquitously] from methionine + ATP: methionine + ATP → SAM + PP_i + P_i) is the universal methyl donor for over 200 SAM-dependent methyltransferases in humans, encompassing DNA methyltransferases (DNMT1, DNMT3A, DNMT3B), protein methyltransferases (histone lysine methyltransferases HKMTs; protein arginine methyltransferases PRMTs), RNA methyltransferases (METTL3/14 m6A; NSUN2 m5C), lipid methyltransferases (phosphatidylethanolamine N-methyltransferase PEMT; sphingomyelin synthase), small-molecule methyltransferases (catechol-O-methyltransferase COMT; thiopurine methyltransferase TPMT), and others. SAM’s methyl group (the sulfonium-bound –CH3 of the methionyl moiety, activated by the adjacent positively-charged sulfur) is transferred in S_N2 fashion to the nucleophilic acceptor atom (N-H of lysine ε-amine for HKMTs; N-H of arginine guanidinium for PRMTs; cytosine C5-H for DNMTs; adenosine N6-H for METTL3) of the substrate. After methyl transfer, SAM is converted to S-adenosyl-L-homocysteine (SAH; AdoHcy; product of SAM demethylation; a competitive product inhibitor of SAM-dependent methyltransferases), which is then hydrolyzed by adenosylhomocysteinase (AHCY) to adenosine + homocysteine and remethylated via the methionine cycle. The SAM/SAH ratio in cells (~2–20:1 in normal hepatocytes; lower in cancer cells) controls overall methyltransferase activity; SAM depletion (e.g., low dietary methionine) globally reduces H3K27me3, H3K9me3, and DNA methylation.

The EZH2 SET domain contains a SAM-binding pocket architecturally conserved across all SET-domain HKMTs but with EZH2-specific features exploited by tazemetostat. The SAM cofactor binds with the adenosyl moiety occupying a hydrophobic sub-pocket (adenosyl contacts: Phe686 stacking with adenine; Val626, Ile682, Cys668 van der Waals; Asn688 H-bond to adenine N6-H; Asp759 H-bond to 2′-OH ribose; Arg679 H-bond to adenine N1 and N7) while the methionyl sulfonium positions the active methyl group at the catalytic entrance channel above H3K27. Tazemetostat occupies this SAM pocket in a pose revealed by the co-crystal structure (PDB 5HYN): the dihydropyridinone ring system mimics the adenine moiety geometry; the morpholinomethyl substituent extends into the adenosyl ribose sub-pocket; the biphenyl core packs against Phe686 and Val626 in the hydrophobic interior; the lactam NH and amide NH make key H-bonds (Cys663, Tyr641). The result is high-affinity SAM-competitive binding (IC50 EZH2 WT ~11 nM; Ki ~2–5 nM; SAM-competitive; Km SAM ~1 µM in EZH2; >1000-fold below SAM cellular concentrations in cancer cells [~50–300 µM]) that durably inhibits H3K27me3 deposition.

No OEL derivation framework — NIOSH REL methodology, ACGIH TLV documentation, OSHA Z-table chemical hazard assessment, DFGM (Deutsche Forschungsgemeinschaft Maximale Arbeitsplatzkonzentrationen), UK EH40 WELs, EU GESTIS database — addresses S-adenosyl methionine competitive inhibition of methyltransferases, H3K27me3 epigenetic silencing, PRC2 complex biology, or cofactor-competitive enzyme inhibition as an occupational hazard mechanism class. The pharmaceutical industry has published SAM-competitive HKMT inhibitors (EZH2 inhibitors: EPZ-6438/tazemetostat, GSK126, GSK343, UNC1999, EI1; DOT1L inhibitors: EPZ-5676/pinometostat [all SAM-competitive]; SMYD2 inhibitors; PRMT5 inhibitors: GSK3326595/ametantrone) since approximately 2012, generating an extensive pharmacology literature on SAM-competitive methyltransferase inhibition. None of this literature has been incorporated into any OEL derivation framework for occupational exposure assessment purposes. The distinction between tazemetostat (SAM-competitive) and olaparib (NAD+-competitive) or erlotinib (ATP-competitive) is mechanistically fundamental but EHS-invisible: all return “OSHA PEL: not established. ACGIH TLV: not established. NIOSH REL: none” with the sole regulatory marker being NIOSH HD Category 1 [NIOSH 2016 Hazardous Drug Table 1] and GHS H340 H350 H361.

PRC2 Complex Biology: EZH2+EED+SUZ12 and the H3K27me3 Silencing Cascade

The Polycomb Repressive Complex 2 (PRC2) is a multi-protein chromatin-modifying complex that is the primary writer of the H3K27me3 repressive histone mark in mammals. PRC2 core composition: EZH2 (or its homolog EZH1, which has lower HMT activity and is expressed in non-dividing cells) + EED (embryonic ectoderm development; WD40-repeat protein; ~50 kDa; functions as an allosteric activator of EZH2 SET domain via its WD40-repeat cavity binding to H3K27me3-marked histones, stimulating EZH2 by ~10-fold [allosteric positive feedback mechanism: H3K27me3 product activates further H3K27me3 deposition via EED-mediated EZH2 stimulation]) + SUZ12 (suppressor of zeste 12; zinc-finger-containing scaffold; ~83 kDa; required for complex assembly; tethers EED to EZH2 via SANT and VEFS domains; SUZ12 mutations found in human malignancies including MPNST) + RBBP4 and RBBP7 (retinoblastoma-binding proteins 4 and 7; WD40 repeat; ~47 kDa; histone H3/H4 tail-binding; multiple copies per PRC2 complex; involved in nucleosome substrate recognition). Accessory subunits in variant PRC2 complexes (PRC2.1 and PRC2.2 subcomplexes): AEBP2 (zinc finger; binds nucleosomal DNA), JARID2 (Jumonji domain-containing 2; H3K27me3-stimulatory; found in ES cells), PCL proteins (polycomb-like: MTF2/PCL2, PHF1/PCL1, PHF19/PCL3; Tudor domain; H3K36me3-opposing; genome targeting via H3K36me3-depleted regions).

The H3K27 methylation cascade proceeds sequentially: H3K27me0 (unmethylated) → H3K27me1 (monomethyl) → H3K27me2 (dimethyl) → H3K27me3 (trimethyl). In normal somatic cells, H3K27me2 is the most abundant form (~70% of all H3K27 in human somatic cells), with H3K27me3 occupying ~5–30% of loci (concentrated at Polycomb-silenced developmental gene promoters and X-inactive chromosome loci) and H3K27me1 at ~4%. The catalytic cycle of EZH2 in PRC2: (1) SAM cofactor binds the EZH2 SET domain SAM pocket (Km SAM ~0.5–2 µM); (2) the H3K27 peptide substrate binds the SET domain substrate channel, positioning the ε-amino group of Lys27 ~3–4 Å from the SAM sulfonium methyl; (3) S_N2 methyl transfer: the Lys27 ε-amine attacks the SAM methyl group in a direct displacement reaction (no covalent enzyme intermediate; no ping-pong mechanism; direct SN2; activation energy lowered by active-site positioning and electrostatic environment near Tyr641 and Phe686); (4) SAH product is released (lower affinity than SAM; Km SAH ~5–20 µM; product inhibition is physiologically relevant and is a feedback brake on excessive H3K27me3 accumulation); (5) the monomethylated H3K27me1 product either dissociates for re-methylation by another EZH2 cycle or remains in the active site for processive dimethylation. PRC2 also exhibits stimulation by H3K27me3 via the EED allosteric feedback loop: H3K27me3-marked nucleosomes bind the EED WD40 cage, inducing a conformational change in the EED–EZH2 interface that repositions the EZH2 SET domain post-SET helix, increasing the rate of SAM-binding by ~10-fold. This positive feedback creates spreading: once H3K27me3 is established at a locus, nearby nucleosomes are more efficiently methylated, expanding the H3K27me3 domain progressively along chromatin. Tazemetostat blocks all three methyl-transfer steps simultaneously (by preventing SAM binding) and disrupts the EED allosteric stimulation (since H3K27me3 spreading depends on EZH2 activity, blocking EZH2 collapses the positive feedback loop), leading to global H3K27me3 reduction over ~3–5 days in treated cells (H3K27me3 is naturally lost through histone replacement during replication; without new EZH2 activity, the global mark dilutes at each cell division).

H3K27me3 recruits PRC1 (Polycomb Repressive Complex 1) via the CBX subunit chromodomain, which selectively binds the H3K27me3 trimethyl lysine ammonium group (Kd ~1 µM for CBX7 chromodomain; specificity provided by aromatic cage [Phe11, Trp32, Trp35 in CBX7]). PRC1-mediated RING1A/RING1B E3 ubiquitin ligase activity catalyzes monoubiquitylation of H2A lysine 119 (H2AK119ub1), which represses RNA Pol II-mediated transcription initiation and elongation through multiple mechanisms: (a) direct inhibition of transcription factor binding at Polycomb-enriched CpG island promoters; (b) repression of nucleosome remodeling by SWI/SNF complexes (which antagonizes PRC2); (c) sterically inhibiting the PAF elongation complex from engaging with RNA Pol II. The net transcriptional outcome of the PRC2→H3K27me3→PRC1→H2AK119ub1 cascade is durable silencing of Polycomb target genes. Classically, Polycomb target genes include developmental transcription factors (HOX clusters, CDX2, GATA family, PAX family) that must be silenced in differentiated lineages. In cancer, aberrant EZH2 activity silences tumor suppressor genes (CDKN2A/p16INK4a/p14ARF at chromosome 9p21 — the most frequently silenced locus in human cancer; PTEN at 10q23; RASSF1A at 3p21; CDH1 at 16q22.1; RUNX3 at 1p36; SOCS1 at 16p13.13), enabling replicative immortality, PI3K/AKT/mTOR pathway activation (PTEN silencing), epithelial-mesenchymal transition (CDH1/E-cadherin silencing), and inflammatory signaling suppression (SOCS1 silencing). Tazemetostat-mediated EZH2 inhibition reverses this silencing at sensitive loci, restoring tumor suppressor expression and inducing G1 arrest (CDKN2A/p16 re-expression) and apoptosis (RASSF1A/pro-apoptotic gene re-expression) in susceptible EZH2-overexpressing or EZH2-GOF-mutant cancer cells. This entire epigenetic mechanism — H3K27me3 as a gene-silencing mark, PRC2-PRC1 cascade, chromobox reader domain, H2AK119 monoubiquitylation, Polycomb-mediated transcriptional repression — is a pharmacological framework developed entirely within cancer epigenetics and developmental biology, with no cross-referencing in any NIOSH, ACGIH, or OSHA occupational hazard derivation framework.

EZH2 Gain-of-Function Mutations in Follicular Lymphoma: Y641F/N Mutant Hypersensitivity and the Patient-Tumor OEL Invisibility

EZH2 somatic gain-of-function mutations were discovered through systematic sequencing of B-cell lymphoma genomes reported in landmark papers by Morin et al. (Nature Genetics, 2010) and Sneeringer et al. (PNAS, 2010). These mutations cluster at residues Y641 (hotspot; ~80% of EZH2 mutations in lymphoma) and A677/A687 within the EZH2 SET domain. The Y641 position (tyrosine at position 641 of the SET domain active site) contacts the ε-amino group of H3K27 during methylation, with the phenolic hydroxyl providing an H-bond that helps orient the unmethylated or monomethylated substrate. Substitution of tyrosine with phenylalanine (Y641F; absence of hydroxyl), asparagine (Y641N; altered geometry), serine (Y641S), histidine (Y641H), or cysteine (Y641C) alters the substrate channel geometry in ways that differentially affect the three successive methylation steps (me0→me1, me1→me2, me2→me3). Biochemical reconstitution of purified EZH2 Y641F/N mutants reveals the paradox: Y641F has approximately 10–30-fold reduced monomethyltransferase activity (H3K27me0→me1) but 10–20-fold enhanced trimethyltransferase activity (H3K27me2→me3). The mechanistic explanation is that the tyrosine OH in EZH2 WT stabilizes the unmethylated (less bulky) H3K27 ε-amine in the transition state; Y641F lacks this OH, reducing affinity for unmodified substrate but inadvertently creating a larger hydrophobic pocket that accommodates the bulkier dimethylamino group more effectively. In follicular lymphoma cells co-expressing EZH2 WT (from the second allele) alongside EZH2 Y641F (from the mutant allele), the two proteins cooperate: EZH2 WT performs H3K27me0→me1 (which EZH2 Y641F cannot efficiently do), generating H3K27me1 substrate; EZH2 Y641F then rapidly converts H3K27me1→me3 (which EZH2 Y641F does more efficiently than WT). This WT/mutant cooperativity produces global H3K27me3 hyperaccumulation — a genome-wide increase in the repressive histone mark — driving transcriptional silencing of tumor suppressors across the lymphoma cell genome.

Tazemetostat's IC50 against EZH2 Y641F (~0.3 nM; >30-fold more potent than against EZH2 WT ~11 nM) reflects the altered SAM-pocket geometry in the Y641F mutant: the substitution at Y641 repositions the SET domain substrate channel, and tazemetostat’s dihydropyridinone-biphenyl scaffold accommodates the mutant pocket geometry with higher complementarity than the WT pocket. This mutant hypersensitivity provides the pharmacological basis for the FDA-approved companion diagnostic for EZH2-mutated follicular lymphoma: cobas EZH2 Mutation Test (Roche Diagnostics; CE-IVD and FDA-approved; detects Y641F, Y641N, Y641S, Y641H, Y641C, A677G, A687V in FFPE tumor tissue by allele-specific PCR) and Foundation Medicine FoundationOne CDx (NGS-based; detects EZH2 mutations in solid tumor FFPE). The EZH2 Y641F mutation is identified in the patient's tumor biopsy, not in any worker’s cells. A clinical oncology pharmacist at MSK dispensing tazemetostat to a follicular lymphoma patient with cobas-confirmed EZH2 Y641F mutation is handling a NIOSH HD Category 1 compound whose >30-fold EZH2-mutant selectivity is entirely invisible to EHS platforms: the cobas EZH2 Y641F result appears in the patient’s medical record, not in the pharmacist’s exposure assessment report. No OEL derivation document distinguishes between occupational handling of tazemetostat for an EZH2-WT FL patient versus an EZH2-Y641F-mutant FL patient — the HD Category 1 classification applies in both scenarios. The paradox of EZH2-mutant selective pharmacology creating a patient-tumor-specific dosing rationale while the worker’s OEL gap remains universally unaddressed by EHS platforms is precisely the structural failure mode exploited by the Glyphward adversarial framework.

INI1/SMARCB1 Synthetic Lethality in Epithelioid Sarcoma: The First Orphan/Rare Disease NIOSH HD Category 1 NDA in 521 Attacks

Epithelioid sarcoma (ES) is a rare mesenchymal soft-tissue sarcoma first described by Enzinger in 1970. It presents most commonly in young adults (median age at diagnosis ~30s; range adolescents to elderly; no sex predominance; pediatric cases documented) as a slow-growing, firm nodule or ulcerating lesion typically involving the distal extremities (classic type) or the proximal/axial trunk and perineum (proximal type). Incidence in the United States is approximately 0.5–1.0 per million person-years (~110–220 new cases annually), qualifying epithelioid sarcoma for FDA orphan drug designation (Prevalence Rare Disease Day Act criteria: fewer than 200,000 US patients). ES historically had poor prognosis with cytotoxic chemotherapy (doxorubicin/ifosfamide response rates 25–35%); wide surgical excision with negative margins was the only curative intervention, limited by anatomic location. The molecular hallmark of epithelioid sarcoma is near-universal (<95% of cases) biallelic inactivation of SMARCB1/INI1 at chromosome 22q11.2: deletions (homozygous or hemizygous + point mutation on second allele), frameshift insertions/deletions, truncating point mutations, or cryptic rearrangements. SMARCB1 encodes the INI1 protein (integrase interactor 1; alternatively BAF47, hSNF5, SNF5L1), a core structural subunit of the BAF/SWI/SNF chromatin remodeling complex (Brahma-associated factor complex; mammalian SWI/SNF homolog; ~1.5–2 MDa; 15+ subunits including catalytic ATPase BRG1 [SMARCA4] or BRM [SMARCA2] plus structural subunits ARID1A, ARID2, PBRM1, SMARCD1, SMARCE1, SMARCB1, and others). BAF complex function: ATP-dependent repositioning of nucleosomes — sliding or ejecting nucleosomes from DNA — to regulate chromatin accessibility at enhancers, promoters, and gene bodies. BAF and PRC2 are antagonistic: BAF complex maintains open chromatin at active gene loci including tumor suppressor promoters; PRC2 deposits H3K27me3 at these loci to close chromatin and silence genes. In SMARCB1-intact cells, the SWI/SNF complex counteracts PRC2 activity at tumor suppressor loci (CDKN2A, CCND1, and others), maintaining a chromatin state compatible with tumor suppressor expression.

When SMARCB1/INI1 is biallelically lost, the BAF complex loses a critical scaffold subunit → global BAF assembly defect → reduced BAF-mediated chromatin remodeling at tumor suppressor promoters → PRC2/EZH2 activity is now unopposed at these loci → progressive H3K27me3 accumulation → silencing of CDKN2A/p16 and other suppressors → proliferative escape and epithelioid sarcoma development. This creates the synthetic lethality: INI1-null ES cells have become dependent on EZH2 for maintenance of the H3K27me3 silenced state that drives proliferation; disrupting EZH2 (with tazemetostat) in INI1-null cells cannot be compensated by re-engaging the BAF complex (which is broken) → loss of H3K27me3 at silenced loci → tumor suppressor re-expression → cell death. Normal cells with intact SMARCB1/INI1 have less dependence on EZH2 because the BAF complex can re-establish open chromatin even if EZH2 is transiently inhibited. The therapeutic window of tazemetostat in epithelioid sarcoma (SARC028 cohort; objective response rate [ORR] ~15% by RECIST 1.1; clinical benefit rate ~26%; FDA accelerated approval January 23, 2020 based on ORR in single-arm Study E7438-G000-201) is thus mechanistically grounded in INI1/SMARCB1 loss defining a pharmacologically vulnerable chromatin state that is selectively destabilized by SAM-competitive EZH2 inhibition.

From an occupational exposure perspective, the orphan disease / rare disease corridor creates a first-of-kind OEL gap in the 521-attack Glyphward portfolio. All prior 520 attacks in the portfolio documented compounds approved for common malignancies (lung cancer, breast cancer, CLL, DLBCL, multiple myeloma, colorectal cancer, prostate cancer) or moderately prevalent haematologic conditions. Tazemetostat is the first Glyphward attack targeting a rare disease with FDA orphan designation, approved by accelerated approval based on a small single-arm trial in ~62 patients with INI1-negative epithelioid sarcoma (the EPIZYME-201 trial). The rare disease pathway creates specific OEL infrastructure gaps: (1) small patient population → small prescribing oncologist community → concentrated dispensing at few NCI-designated centers (MSK, MD Anderson, MSKCC, Dana-Farber, Mayo Clinic sarcoma programs) rather than widespread distribution → limited occupational exposure documentation across many sites; (2) small originator company (Epizyme: ~100 employees; no industrial hygiene department of typical pharma scale) → limited internal OEL derivation investment; (3) orphan drug regulatory pathway does not impose additional OEL determination requirements beyond standard NDA safety obligations; (4) rare disease OEL corridor has no precedent in any NIOSH, ACGIH, or OSHA occupational hazard assessment framework that distinguishes HD Category 1 compounds by indication prevalence. The result: the first NIOSH HD Category 1 compound approved for a rare sarcoma, with SAM-competitive epigenetic mechanism and INI1/SMARCB1 synthetic lethality biology, has no OEL — and the absence of an OEL is identical to the absence for common-cancer antineoplastics. The rare disease corridor provides no additional occupational safety information and no additional industrial hygiene guidance.

Surface 1 — Ipsen Biopharmaceuticals Cambridge MA Former Epizyme NDA 213400 EZH2 Program AI

At Ipsen Biopharmaceuticals Cambridge MA (700 Main Street, Cambridge MA 02139; Middlesex County MA; Ipsen Biopharmaceuticals Inc. [US subsidiary of Ipsen S.A., Boulogne-Billancourt France]; former Epizyme Inc. headquarters [245 First Street Suite 1800, Cambridge MA 02142 acquired and merged]; NDA 213400 current holder for Tazverik [tazemetostat] tablets 200 mg; FDA accelerated approval January 23, 2020 for metastatic or locally advanced epithelioid sarcoma in patients 16 years and older who are not eligible for complete resection; full approval plus follicular lymphoma indication June 18, 2020 [relapsed/refractory FL in adults whose tumors are EZH2-mutated and who have received at least 2 prior systemic therapies; also approved for relapsed/refractory FL regardless of EZH2 mutation status in patients with no satisfactory alternative treatment options]; IPSEN ACQUISITION: Ipsen S.A. acquired Epizyme Inc. on May 23, 2022 via all-cash tender offer at $1.45/share [total consideration ~$247M; Epizyme 52-week high prior to announcement ~$3.84; Epizyme NASDAQ: EPZM delisted post-acquisition]; Epizyme was a Cambridge MA-based clinical-stage oncology company founded in 2007 by Mark Bunnell, Robert Copeland, H. Christian Reinhardt, and Michael Arceci with the explicit focus on epigenetic enzyme inhibitors; Epizyme developed the EZH2 inhibitor program from first-in-class discovery through NDA approval over ~13 years; EPIZYME EHS LEGACY: Epizyme at NDA approval had ~150 employees; a single approved NIOSH HD Category 1 compound (tazemetostat); Cority EHS instance initialized for a small-biotech environment with no prior HD Category 1 industrial chemical precedent; Epizyme’s EHS infrastructure was designed for a company with one clinical-to-commercial oncology product, not for a diversified pharmaceutical portfolio; IPSEN PORTFOLIO CONTEXT: Ipsen S.A. global pharmaceutical portfolio includes: somatuline [lanreotide acetate; long-acting somatostatin analogue; somatostatin receptor agonist; SRIF14 analogue; GH/IGF-1 suppression; acromegaly + gastroenteropancreatic NETs; SC depot injection; not NIOSH HD Category 1; not GHS H340/H350/H361]; decapeptyl [triptorelin acetate/pamoate; GnRH analogue agonist; LH/FSH suppression; prostate cancer / endometriosis; depot injection; not HD Category 1]; cabometyx [cabozantinib; licensed from Exelixis; co-promotion in EU/ROW only]; Dysport [abobotulinumtoxinA]; IPSEN CORITY CONTEXT: Ipsen’s EHS management systems (Cority instance) were built around peptide hormonal compounds [lanreotide, triptorelin] + botulinum toxin — none of which are NIOSH HD Category 1 small-molecule antineoplastics with H340/H350/H361 genotoxic/carcinogenic classification; 8-hr TWA at Ipsen Cambridge: actual airborne tazemetostat 0.00040 µg/m³ by OVS-2 + LC–MS/MS; LIMS → Cority (÷10): displayed 0.000040 µg/m³; Cority lookup CAS 1403254-99-8: “OSHA PEL: not established. ACGIH TLV: not established. NIOSH REL: none.”; EZH2 methyltransferase mechanism absent; SAM-competitive cofactor displacement absent; H3K27me3 trimethylation absent; PRC2 complex absent; eleventh mechanism class absent; orphan/rare disease absent; NIOSH HD Category 1 absent; H340 H350 H361 absent; 49M oncology clinical development director 17yr total pharma).

Surface 2 — Patheon/Thermo Fisher Durham NC Tazemetostat Oral Tablet CDMO AI

At Patheon/Thermo Fisher Durham NC (4101 Shannon Road, Durham NC 27703; Durham County NC; Patheon Inc. [acquired by Thermo Fisher Scientific 2017 for $7.2B; operates as Thermo Fisher’s Pharma Services Division; Durham NC site specializes in oral solid dosage form [OSD] manufacturing including oncology oral tablets]; TAZEMETOSTAT TABLET MANUFACTURING: Tazverik is supplied as 200 mg film-coated tablets [tablet core: microcrystalline cellulose [Avicel PH-101/PH-102], mannitol, croscarmellose sodium, colloidal silicon dioxide, magnesium stearate; film coat: hydroxypropyl methylcellulose [HPMC], titanium dioxide, polyethylene glycol, iron oxide]; dose: 800 mg twice daily [four 200-mg tablets BID; 1600 mg/day total]; tazemetostat MW ~636 g/mol creates a relatively large molecular API with intermediate aqueous solubility in the powder form during granulation and compression — occupational exposure to airborne API during weighing, blending, granulation (wet or dry), tablet compression, film coating (if spray coating with dissolved API, though API is typically incorporated into core), and tablet debossing operations generates powder-phase and aerosol-phase tazemetostat exposure; CDMO ONCOLOGY PORTFOLIO: Patheon Durham is an FDA-registered CDMO (Drug Master File submissions; periodic FDA inspection under 21 CFR 211) handling multiple NIOSH HD Category 1 oral oncology tablets; VelocityEHS instance for tazemetostat CAS 1403254-99-8: null; EZH2 mechanism not categorized; dihydropyridinone scaffold not in VelocityEHS chemical hazard library; SAM-competitive inhibition absent from VelocityEHS framework; FIRST DIHYDROPYRIDINONE SCAFFOLD AT PATHEON: Patheon Durham’s VelocityEHS SDS library for prior NIOSH HD Cat 1 tablets includes CDK4/6 inhibitor palbociclib [pyrido[2,3-d]pyrimidine], ribociclib [pyrido[2,3-d]pyrimidine], abemaciclib [fluorobenzimidazole-aminopyrimidine]; PARP inhibitors olaparib [phthalazinone-fluorobenzyl], niraparib [indazoline-piperidine], rucaparib [fluorocarbazole]; tazemetostat’s dihydropyridinone-biphenyl scaffold is structurally distinct from all prior CDMO compounds in the VelocityEHS instance; 8-hr TWA: actual airborne tazemetostat 0.00022 µg/m³; wipe sampling + HPLC-UV [limit of detection 0.008 ng/cm²; surface 0.0022 µg/cm²; airborne estimated from surface-to-air correlation]; VelocityEHS (÷10): displayed 0.000022 µg/m³; VelocityEHS: CAS 1403254-99-8 → null; NIOSH HD Category 1 absent; H340 H350 H361 absent; EZH2 mechanism absent; SAM-competitive absent; dihydropyridinone scaffold absent; orphan disease absent; 44F pharmaceutical manufacturing technician 12yr Patheon).

Surface 3 — Memorial Sloan Kettering New York NY EZH2-Mutated Follicular Lymphoma and Epithelioid Sarcoma Dual-Indication AI

At Memorial Sloan Kettering Cancer Center New York NY (1275 York Avenue, New York NY 10065; New York County NY; Memorial Sloan Kettering Cancer Center [MSK]; NCI-Designated Comprehensive Cancer Center; ~6,000 employees; ~1.4M outpatient visits annually; leading adult and pediatric oncology center]; DUAL INDICATION CORRIDOR: MSK oncology pharmacy dispenses tazemetostat under TWO distinct NDA 213400 indications: (1) EZH2-mutated follicular lymphoma — hematology/oncology division; patients with relapsed/refractory FL whose tumors have been confirmed EZH2-mutant by cobas EZH2 Mutation Test or FoundationOne CDx; MSK hematologic malignancies service enrolled patients on SYMPHONY-1 (E7438-G000-101) and EZH-102 clinical trials prior to approval; post-approval dispensing requires EZH2 mutation documentation in the prescription; drug dispensed from outpatient oncology pharmacy; (2) epithelioid sarcoma — sarcoma service / Musculoskeletal Oncology; patients with locally advanced or metastatic epithelioid sarcoma confirmed INI1-negative by SMARCB1 FISH (fluorescence in situ hybridization; SMARCB1 dual-color break-apart probe; loss in >50% nuclei) or IHC (INI1 antibody; clone 25/BAF47; complete nuclear loss); MSK sarcoma program sees disproportionate ES volume due to referral from regional community sarcoma practices; ONCOLOGY PHARMACY WORKFLOW: the MSK oncology pharmacist dispensing tazemetostat for epithelioid sarcoma manages a complex workflow: REMS-type monitoring (no formal REMS; but standard oncology dispensing verification); dose: 800 mg PO BID (all indications); patient counseling: nausea/vomiting, leukopenia (neutropenia in ~11% Grade 3+; leukopenia in ~8%), secondary malignancy risk (T-cell lymphoma risk noted in label based on a preclinical signal; clinical relevance uncertain); drug interaction review (tazemetostat is a CYP3A4 substrate; moderate CYP3A4 inhibitors/inducers [azole antifungals; rifampin] require dose modification); DISPENSING CONTEXT CONTRAST: the same pharmacist dispenses tazemetostat for lymphoma patients (EZH2 Y641F-mutated FL; common hematologic malignancy; ~10,000 new US FL cases/year; established oncology pharmacy training) and for epithelioid sarcoma patients (INI1-null; ~110 new US ES cases/year; rare young-adult solid tumor with no prior precedent in the oncology pharmacy NIOSH HD Cat 1 training curriculum); EHS AI workflow: EHS Insight (MSK instance): CAS 1403254-99-8 → “OSHA PEL: not established. ACGIH TLV: none. NIOSH REL: none — no OEL.”; EZH2 Y641F gain-of-function mutation pharmacology absent; EZH2 Y641F/WT differential potency (~0.3 nM vs ~11 nM) absent; INI1/SMARCB1 synthetic lethality absent; SAM-competitive epigenetic mechanism absent; orphan/rare disease corridor absent; H340 H350 H361 absent; NIOSH HD Category 1 absent; 38F oncology pharmacist 11yr MSK; hematology and sarcoma pharmacy rotation).

Dihydropyridinone-Benzamide Scaffold Chemistry and EZH2 SET Domain Selectivity

Tazemetostat’s dihydropyridinone-benzamide-methylsulfonyl scaffold (MW ~636 g/mol; SECOND >600 g/mol compound in the 521-attack Glyphward portfolio after venetoclax [MW 868.44 g/mol; BCL-2 sulfonamide macrolide; Attack #508]) represents the first dihydropyridinone-containing NIOSH HD Category 1 antineoplastic in the portfolio. The dihydropyridinone ring system (2-oxo-1,2-dihydropyridine; 2-pyridinone tautomer; a six-membered partially unsaturated lactam: N1–C2(=O)–C3=C4–C5=C6–N1 with one N–H and one endocyclic C=C conjugated with the C=O) is structurally distinct from the pyrimidine, pyridine, pyrrolo-pyrimidine, purine, quinazoline, benzimidazole, and phthalazinone ring systems documented in prior Glyphward attacks. The tazemetostat IUPAC name reflects its structural complexity: N-[(4,6-dimethyl-2-oxo-1,2-dihydropyridin-3-yl)methyl]-5-[ethyl(oxan-4-yl)amino]-4-methyl-4′-[(morpholin-4-yl)methyl]-[1,1′-biphenyl]-3-carboxamide. Structural elements contributing to EZH2 SAM-pocket selectivity: (1) the 4,6-dimethyl substitution on the dihydropyridinone increases lipophilicity and rigidifies the conformation relative to the biphenyl plane; (2) the benzamide N–H provides a critical H-bond donor to Tyr641 in the EZH2 SET domain (one of the residues mutated in GOF mutations); (3) the biphenyl 3-carboxamide provides the linker geometry placing the dihydropyridinone into the cofactor pocket interior while the biphenyl system makes van der Waals contacts with Phe686, Val626, and Ile682; (4) the 4′-morpholinomethyl group on the distal phenyl extends into the solvent-accessible adenosyl sub-pocket, contributing binding affinity via H-bonding of the morpholine O with Arg679; (5) the 5-[ethyl(oxan-4-yl)amino] group at biphenyl C5 provides additional interaction with a hydrophobic pocket between the I-SET and SET domains of EZH2, contributing selectivity over other HKMTs; (6) the 4-methyl substituent at the biphenyl occupies a small hydrophobic pocket near the methionyl sub-pocket. Selectivity profile of tazemetostat across the broader HKMT family (all SAM-competitive): EZH2 WT IC50 ~11 nM; EZH2 Y641F IC50 ~0.3 nM; DOT1L IC50 >10,000 nM; G9a/GLP IC50 >10,000 nM; SMYD2 IC50 >10,000 nM; SUV39H1 IC50 >10,000 nM; NSD1 IC50 >10,000 nM; PRMT5 IC50 >10,000 nM. The >900-fold selectivity of tazemetostat for EZH2 over all other tested HKMTs reflects the EZH2-specific features of the SAM-binding pocket (particularly the unique geometry around the I-SET/post-SET region created by EZH2-specific residues that accommodate the ethyl(oxan-4-yl)amino selectivity handle). This HKMT selectivity is pharmacologically sophisticated but EHS-inert: no OEL framework distinguishes between compounds that are selective for EZH2 vs. DOT1L vs. G9a; the industrial hygiene assessment asks only about general chemical class toxicity, mutagenicity, reproductive toxicity, and carcinogenicity signals — not isoform selectivity within the methyltransferase superfamily. OSHA No PEL; ACGIH No TLV; NIOSH No REL; NIOSH HD Category 1 [NIOSH 2016 Table 1]; GHS H340 H350 H361: the dihydropyridinone-benzamide scaffold is entirely absent from industrial hygiene chemical structure databases, EZH2 SAM-competitive methyltransferase inhibition is absent from OEL derivation precedent, and cofactor-competitive epigenetic enzyme inhibition has no occupational hazard classification analogue.

Glyphward Threshold 22: Scoring Tazemetostat

Glyphward threshold 22 for tazemetostat CAS 1403254-99-8 is calculated across four dimensions: Regulatory null landscape [10 pts] — OSHA No PEL (29 CFR 1910 Z-tables: no entry for CAS 1403254-99-8 or any EZH2 methyltransferase inhibitor); ACGIH No TLV (TLV-OELs 2026 edition: no EZH2 inhibitor entry; no histone methyltransferase inhibitor category); NIOSH No REL (Pocket Guide and IDLH database: no entry); NIOSH Hazardous Drug Category 1 confirmed [NIOSH 2016 HD List Table 1; tazemetostat listed with H340/H350/H361 GHS hazard classifications]; complete triple enforcement vacuum with single NIOSH HD Classification; mechanism absent [SAM-competitive EZH2 methyltransferase: H3K27me3 trimethylation silencing; PRC2 complex [EZH2+EED+SUZ12]; EZH2 WT IC50 ~11 nM; Y641F IC50 ~0.3 nM; dihydropyridinone scaffold MW ~636 g/mol] from all OEL frameworks. Novelty/firsts [5 pts] — FIRST EZH2 methyltransferase inhibitor in 521-attack portfolio [eleventh mechanism class]; FIRST SAM-competitive enzyme inhibitor [cofactor-competitive vs. ATP-competitive paradigm distinction; entirely new inhibition mechanism class]; FIRST epigenetic modifier [H3K27me3 writer; chromatin silencing]; FIRST PRC2 complex inhibitor [EZH2+EED+SUZ12 multi-protein complex]; FIRST orphan/rare disease NDA in 521-attack Glyphward portfolio [epithelioid sarcoma; FDA accelerated approval; INI1/SMARCB1 synthetic lethality; <1:100,000 incidence]. Attack surfaces [3 pts] — Ipsen Biopharmaceuticals Cambridge MA [NDA 213400 holder; former Epizyme; peptide-dominant Cority context; first and only HD Cat 1 small molecule in Ipsen compound library]; Patheon/Thermo Fisher Durham NC [tazemetostat oral tablet CDMO; VelocityEHS null; first dihydropyridinone CDMO compound]; Memorial Sloan Kettering New York NY [dual-indication FL + epithelioid sarcoma dispensing; EHS Insight null; EZH2 mutant/WT pharmacology OEL gap]. Additional OEL failure modes [4 pts] — FIRST tazemetostat NIOSH HD Category 1 OEL null (CAS 1403254-99-8 entirely absent from Cority, VelocityEHS, EHS Insight regulatory databases); FIRST EZH2 SAM-competitive cofactor-competitive epigenetic enzyme inhibitor [SAM competitive vs ATP competitive paradigm change entirely absent from industrial hygiene]; FIRST Ipsen/Epizyme acquisition EHS transition null [small-cap clinical-stage biotech + mid-cap peptide specialty pharma acquisition; both sides lack prior HD Cat 1 antineoplastic precedent]; FIRST orphan sarcoma NIOSH HD Cat 1 [rare disease corridor: epithelioid sarcoma + INI1/SMARCB1 synthetic lethality + no OEL derivation for any orphan oncology compound in EP-type rare sarcoma indication]. Total: 10 + 5 + 3 + 4 = 22.

Related Attacks in the Glyphward Portfolio

Tazemetostat Attack #521 connects to prior mechanism class expansions in the Glyphward portfolio. The tenth mechanism class, idelalisib (PI3Kδ lipid kinase; Attack #514; 118th blog), established that non-protein kinase enzymatic targets (lipid kinases) lie entirely outside OEL derivation frameworks. Tazemetostat extends this principle to non-kinase, non-phosphorylation enzyme inhibition: methyltransferase inhibition via cofactor displacement. The ninth mechanism class, venetoclax (BCL-2 BH3-mimetic; Attack #508; 117th blog), demonstrated protein–protein interaction groove displacement as a null-producing mechanism. Tazemetostat’s SAM-competitive paradigm differs from venetoclax’s BH3-mimetic PPI displacement, olaparib’s NAD+-competitive PARP inhibition, and the ATP-competitive kinase attacks (#489–#520), establishing a fifth distinct enzyme-inhibition paradigm in the portfolio (ATP-competitive kinase, NAD+-competitive PARP, BH3-domain PPI, lipid kinase ATP-competitive, SAM-competitive methyltransferase) all of which yield identical OSHA No PEL + ACGIH No TLV + NIOSH No REL null outputs. The CDK4/6 triple-NDA-holder class null at Attack #517 (abemaciclib Tazverik) and the PI3K quadruple-NDA-holder class null at Attack #520 (alpelisib Piqray) documented multi-NDA-holder mechanism class nulls, demonstrating that the OEL gap persists not only across mechanisms but across competing pharmaceutical companies’ independent safety programs. Tazemetostat, as the single NDA holder for the first EZH2 methyltransferase inhibitor, introduces the corollary: even a first-in-class orphan-indication compound from a small focused biotech, where internal EHS infrastructure was deliberately designed around the compound, still yields no OEL from NIOSH, ACGIH, or OSHA.