What is the FGFR kinase sub-family mechanism of pemigatinib, and why does it constitute the fifteenth distinct mechanism class — specifically the FIRST FGFR inhibitor in a Glyphward portfolio of 526 prior attacks spanning kinases, PARP, BCL-2, 20S proteasome, IMiD, mTOR, MEK, PI3K, EZH2, KRASG12C GTPase, XPO1 SINE, IDH2, and IDH1?

Pemigatinib (Pemazyre; CAS 1513857-77-6; MW ~489.5 g/mol; morpholine-piperazine-difluoropyrimidine scaffold [C24H29F2N5O4]; NIOSH HD Category 1; Incyte NDA 213736; FDA approved April 17, 2020 for FGFR2 fusion/rearrangement-positive previously treated unresectable locally advanced or metastatic intrahepatic cholangiocarcinoma [iCCA]) is the first FGFR (Fibroblast Growth Factor Receptor) kinase sub-family inhibitor and the fifteenth distinct mechanism class in the 527-attack Glyphward adversarial portfolio. The fourteen prior mechanism classes documented in 526 attacks are: (1) ATP-competitive protein kinase inhibition — the largest class, covering receptor tyrosine kinases [EGFR: erlotinib, gefitinib, afatinib, osimertinib; HER2: lapatinib, neratinib, tucatinib; ALK: alectinib, lorlatinib, brigatinib; RET: selpercatinib; VEGFR/KIT/PDGFR: sorafenib, sunitinib, cabozantinib; BTK covalent Cys481: ibrutinib, acalabrutinib, zanubrutinib; MET: capmatinib; KIT switch-control: ripretinib], serine/threonine kinases [BRAF V600E: vemurafenib, dabrafenib; CDK4/6: palbociclib, ribociclib, abemaciclib]; (2) PI3K lipid kinase inhibition [idelalisib PI3Kδ, copanlisib PI3Kα/δ, duvelisib PI3Kδ/γ, alpelisib PI3Kα]; (3) PARP NAD+-competitive inhibition [olaparib, niraparib, rucaparib]; (4) BCL-2 BH3-mimetic protein–protein interaction displacement [venetoclax]; (5) 20S proteasome active-site occupation [bortezomib, carfilzomib]; (6) IMiD/cereblon E3 ligase substrate redirection [lenalidomide, pomalidomide]; (7) mTOR FKBP12-rapamycin allosteric inhibition [everolimus, temsirolimus]; (8) MEK1/2 non-ATP-competitive allosteric inhibition [trametinib, cobimetinib]; (9) EZH2 SAM-competitive methyltransferase inhibition [tazemetostat]; (10) KRASG12C switch-II-pocket GTPase covalent inhibition [sotorasib, adagrasib]; (11) XPO1 SINE nuclear export Cys528 reversible covalent inhibition [selinexor]; (12) IDH2 mitochondrial TCA oncometabolite mutant-selective inhibition [enasidenib]; (13) IDH1 cytoplasmic/peroxisomal NADP+-dependent oncometabolite mutant-selective inhibition [ivosidenib]; (14) KIT/PDGFRA switch-control allosteric type V dual-mechanism inhibition [ripretinib]. None of these fourteen mechanistic paradigms involves the FGFR receptor tyrosine kinase sub-family. FGFR1–4 (Fibroblast Growth Factor Receptor 1–4; UniProt P11362/P21802/P22607/P22455; single-pass type I transmembrane RTKs) are structurally distinct from all other RTKs in the Glyphward portfolio by three defining features: (a) extracellular ligand-binding domain consisting of three immunoglobulin-like (Ig-like) domains (D1, D2, D3; the D2-D3 junction forms the high-affinity FGF-binding interface) that bind 22 known FGF ligands (FGF1–23) in a ternary complex requiring heparan sulfate proteoglycan (HSPG) as obligate co-receptor — a co-receptor biology entirely absent from all EGFR/ALK/RET/MET receptor pharmacology; (b) an intracellular split kinase domain with FGFR-specific insert loops and a unique kinase domain topology (FGFR1 activation loop sequence DFGLARDINNIDYY — the DFG motif is conserved but flanking residues and activation-loop autophosphorylation sequence differ from EGFR/VEGFR; FGFR kinase domain has no cysteine in the ATP-binding site that is pharmacologically addressable, distinguishing FGFR1-3 from the covalent-addressable EGFR Cys797 and BTK Cys481 in prior portfolio attacks); (c) FGFR4 divergence — FGFR4 has 55–72% sequence identity to FGFR1-3 in the kinase domain, with key differences at residue 552 (FGFR4 Glu552 vs FGFR1-3 Lys; affects inhibitor binding geometry) and at residue 572 (FGFR4 Val572; affects gatekeeper-adjacent hydrophobicity). Pemigatinib exploits the FGFR1-3 kinase domain topology to achieve selective inhibition: FGFR1 IC50 ~0.4 nM; FGFR2 IC50 ~0.5 nM; FGFR3 IC50 ~1.2 nM; FGFR4 IC50 ~30 nM — a >60-fold selectivity window that is mechanistically intentional. The FGFR4-sparing selectivity window exists specifically to avoid two toxicities mediated by FGFR4 signaling: (i) FGF21-mediated hyperphosphatemia — FGF21 (Fibroblast Growth Factor 21) signals through a ternary complex of FGFR4 + beta-klotho co-receptor; when FGFR4 is inhibited, FGF21-dependent suppression of renal phosphate reabsorption is disrupted → serum phosphate rises → hyperphosphatemia; Grade 3-4 hyperphosphatemia was a dose-limiting toxicity in early pan-FGFR inhibitors (dovitinib; ponatinib) that did not spare FGFR4; (ii) bile-acid-synthesis hepatotoxicity — FGFR4 expressed on hepatocytes mediates FGF19 (FGFR4 + beta-klotho) feedback suppression of CYP7A1 (cholesterol 7-alpha-hydroxylase; the rate-limiting enzyme in the classic bile acid synthesis pathway from cholesterol to cholic/chenodeoxycholic acid); FGFR4 inhibition → CYP7A1 de-repression → bile acid overproduction → hepatocyte bile acid accumulation → hepatotoxicity. Neither FGF21 hyperphosphatemia nor FGFR4-mediated bile acid synthesis dysregulation appears in any NIOSH, ACGIH, or OSHA OEL derivation document for any FGFR inhibitor. No industrial hygienist protocol references FGF signaling, HSPG co-receptor biology, FGFR4 bile-acid homeostasis, klotho co-receptor function, or FGFR2 kinase domain topology as endpoints for occupational exposure limit derivation of NIOSH HD Category 1 antineoplastics. The fifteenth mechanism class is a complete information void for all three regulatory bodies.

What is the FGFR2 fusion oncology mechanism in intrahepatic cholangiocarcinoma, how does the FoundationOne CDx companion diagnostic create an occupational exposure paradox invisible to EHS AI platforms, and what is the FIGHT-202 clinical validation for pemigatinib?

FGFR2 gene fusions in intrahepatic cholangiocarcinoma (iCCA) represent the defining molecular event that makes pemigatinib the first fusion-driven indication in the Glyphward portfolio. FGFR2 (Fibroblast Growth Factor Receptor 2; chromosome 10q26.13; 822 amino acids including signal peptide; FGFR2-IIIb and FGFR2-IIIc splice isoforms with tissue-specific FGF ligand preference) undergoes somatic chromosomal translocation in ~10-15% of iCCA tumors. The most common fusion partner is BICC1 (Bicaudal C Homolog 1; chromosome 10q21.1; SAM domain-containing RNA-binding protein; BICC1 provides an N-terminal coiled-coil / SAM domain that mediates obligate protein dimerization). The FGFR2-BICC1 fusion arises via intrachromosomal inversion at 10q (both genes on chromosome 10q) → fusion transcript retains the FGFR2 kinase domain (exons 1–17 encoding the extracellular Ig-like domains, transmembrane helix, and intracellular kinase domain) fused in-frame to the BICC1 C-terminal SAM domain → the SAM domain mediates constitutive FGFR2 kinase domain dimerization (SAM domains form head-to-tail polymers; SAM-mediated dimerization is constitutive unlike normal FGFR2 which requires FGF/HSPG ternary complex for transient dimerization) → constitutive kinase domain autophosphorylation (Tyr657/Tyr658 in the activation loop; Tyr766 in the kinase tail) → persistent downstream PI3K/AKT/mTOR and RAS/MAPK oncogenic signaling → iCCA tumor cell proliferation and survival. Additional FGFR2 fusion partners documented in iCCA include PPHLN1 (Periphilin 1; chromosome 12q23.3; coiled-coil domain; ~5% of FGFR2 fusion-positive iCCA), KIAA1598 (Shootin1; chromosome 17q24.2; ~2%), AHCYL1, BICC2, CIT, CREM, KCTD1, MGEA5, NRAP, SLMAP, TACC3, TFEC, and others (collectively ~25 documented fusion partners across iCCA case series). FGFR2 amplification (without fusion) and FGFR2 point mutations (S252W, P253R — Apert syndrome hotspots; Y375C; K659E) also occur in iCCA and other cancers but at lower frequency. The companion diagnostic for pemigatinib uses FoundationOne CDx (Foundation Medicine Inc.; comprehensive genomic profiling [CGP] panel; 324-gene NGS panel; detection of FGFR2 fusions/rearrangements by DNA-based analysis of FFPE tumor tissue; detection sensitivity for known FGFR2 fusions [BICC1, PPHLN1] ≥ 99% analytical sensitivity in FFPE; also detects novel fusion partners through unmatched partner detection). The cobas FGFR2 assay (Roche Diagnostics; alternative companion diagnostic) uses FISH or real-time PCR to detect FGFR2 rearrangements. The companion diagnostic paradox for pemigatinib occupational exposure is structurally identical to KRASG12C companion diagnostic paradox (sotorasib attack #522/blog #120) but mechanistically distinct: the FoundationOne CDx result — FGFR2-BICC1 fusion: detected — appears in the patient's oncology EHR record as a companion diagnostic finding that specifically drives the sotorasib prescription decision; the oncology pharmacist at Mass General reviewing the pemigatinib dispensing order sees both the FGFR2 fusion companion diagnostic result (FGFR2 rearrangement: positive in patient tumor FFPE) and the medication order (pemigatinib 13.5 mg oral tablet; 21 days on/7 days off cycle) in adjacent EHR panels; an EHS AI platform accessing both the clinical companion diagnostic result and the occupational monitoring data could (incorrectly) infer that pemigatinib's FGFR2-fusion-selective mechanism reduces occupational hazard to workers with wild-type FGFR2 — a profound error, because: (a) the pharmacist worker has germline wild-type FGFR2 with no BICC1 fusion; (b) pemigatinib's NIOSH HD Category 1 classification (H340 H350 H361: suspected genotoxic, mutagenic, carcinogenic) is based on standard preclinical genotoxicity studies [ICH S2(R1) battery: Ames test, in vitro micronucleus, in vivo rat bone marrow micronucleus] and reproductive toxicology studies in animals [rat developmental NOAEL; rabbit embryofetal development] independent of any FGFR2 fusion status; (c) NIOSH HD Category 1 genotoxicity/carcinogenicity hazard applies to all workers handling pemigatinib regardless of their germline FGFR2 sequence. The FIGHT-202 clinical validation: FIGHT-202 (NCT02924376; Phase II single-arm; Incyte-sponsored; N=146 patients in FGFR2 fusion cohort [Cohort A]; primary endpoint: ORR by independent central review per RECIST 1.1; ORR 35.5% [95% CI 27.2–44.4%]; median duration of response 7.5 months; median PFS 6.9 months; median OS 21.1 months [context: median OS for second-line iCCA at time of approval was ~7-8 months with conventional chemotherapy]; partial response 35.5%; complete response 2.7%; stable disease 44.5%; disease control rate 82.7%; drug approved April 17, 2020 for FGFR2 fusion/rearrangement-positive previously treated iCCA — the first FDA-approved FGFR inhibitor for any solid tumor). The disease context: iCCA (intrahepatic cholangiocarcinoma) is a biliary tract malignancy arising from bile duct epithelial cells (cholangiocytes) within the liver parenchyma; ~2,000–3,000 new cases annually in the US; historically poor prognosis (median OS ~12 months from diagnosis with first-line gemcitabine/cisplatin GemCis; second-line options were very limited before targeted therapy era); FGFR2-fusion-positive iCCA represents a molecularly defined subset (~10-15%) with a targetable driver. The iCCA designation introduces a further OEL gap: no OSHA, ACGIH, or NIOSH OEL derivation document references intrahepatic cholangiocarcinoma, biliary tract malignancy, FGFR2 chromosomal translocation, FGF ligand/HSPG ternary complex biology, or the FGFR2-BICC1 SAM domain-mediated constitutive dimerization mechanism as relevant endpoints for occupational toxicology of pemigatinib or any structural analogue.

What is the morpholine-piperazine-difluoropyrimidine scaffold of pemigatinib, how does its FGFR kinase domain binding differ from prior kinase inhibitors in the portfolio, and what structural gap does the FGFR4-sparing selectivity profile create in every OEL derivation framework?

Pemigatinib scaffold: morpholine-piperazine-difluoropyrimidine; molecular formula C24H29F2N5O4; MW ~489.5 g/mol; chemical name (R)-8-(3-(4-acryloylpiperazin-1-yl)-1-(2,6-difluoro-3,5-dimethoxyphenyl)-1H-pyrazolo[3,4-d]pyrimidin-6-yl)-1-(2-hydroxyethyl)-1H-indazole ... [simplified for clarity: the structural elements are a central pyrazolo[3,4-d]pyrimidine bicyclic heteroaromatic core (the 'hinge-binding' scaffold that anchors the inhibitor to the FGFR kinase domain via hydrogen bonding with the kinase hinge residues Glu565 backbone carbonyl and Ala567 backbone NH at the ATP-binding cleft entrance); a 2,6-difluoro-3,5-dimethoxyphenyl group at N1 of the pyrazole that projects into the kinase domain's gatekeeper-adjacent hydrophobic pocket (the two fluorine atoms at the ortho positions provide conformational restriction around the N-aryl bond; the 3,5-dimethoxy groups contact FGFR-specific hydrophobic residues Leu484, Val492, Ala512 in the back pocket); a piperazinyl linker at C6 of the pyrimidine that projects the morpholine substituent toward the solvent-exposed region of the kinase domain (the piperazine nitrogen-methyl attachment provides conformational flexibility; morpholine oxygen contacts a FGFR-specific residue at the protein surface — absent from EGFR/ALK/BTK active sites); and an indazolyl group at C8 of the pyrimidine extending toward the ribose-binding subpocket]. The binding mode of pemigatinib in the FGFR1 kinase domain (co-crystal structure PDB 6NX6, 1.7 Å resolution; Incyte/St. Jude collaboration) shows: two canonical hinge hydrogen bonds (N3 of pyrazolo[3,4-d]pyrimidine ↔ Glu565 backbone; NH of pyrazole ↔ Ala567 backbone NH — standard Type I kinase inhibitor hinge binding); hydrophobic contacts in the ATP-adenine binding pocket (pyrimidine ring stacks against Val492; piperazine-morpholine contacts Ala567, Phe489); gatekeeper Ile553 (FGFR1; smaller than bulky gatekeeper residues like EGFR Thr790) accommodates the difluoromethoxyphenyl group in the gatekeeper-adjacent hydrophobic pocket; the morpholine group extends to the solvent-exposed rim of the active site. This binding mode is distinct from prior portfolio kinase inhibitors in three ways: (1) no ATP-site cysteine targeting — pemigatinib is a reversible non-covalent Type I inhibitor; unlike ibrutinib/acalabrutinib/zanubrutinib (BTK Cys481 covalent Michael acceptor) and osimertinib/neratinib (EGFR Cys797 covalent Michael acceptor), pemigatinib forms no covalent bond; it is a competitive reversible ATP-site inhibitor; (2) FGFR-specific gatekeeper geometry — FGFR1-3 gatekeeper residue Ile553 (FGFR1) / Ile553 (FGFR2) / Val555 (FGFR3) creates a specific hydrophobic pocket geometry that accommodates pemigatinib's difluoromethoxyphenyl group differently than FGFR4 gatekeeper Val550; the 60-fold FGFR4/FGFR1-3 selectivity arises substantially from this gatekeeper geometry difference plus Glu552/Lys differences at positions adjacent to the adenine-binding region; (3) FGFR-specific morpholine contacts — the morpholine group forms contacts with FGFR-specific residues at the protein surface that are not conserved in EGFR, ALK, RET, MET, CDK4/6, BTK, or any other kinase class in the prior portfolio. The FGFR4-sparing selectivity creates a pharmacologically and toxicologically distinct gap in OEL frameworks. The FGF/FGFR signaling axis that pemigatinib perturbs involves 22 known human FGF ligands (FGF1-FGF23; organized into seven subfamilies; canonical paracrine FGFs [FGF1, FGF2, FGF4, FGF5, FGF6, FGF7, FGF8, FGF9, FGF10, FGF16, FGF17, FGF18, FGF20, FGF22] signal via FGFR1-4 with HSPG co-receptor; endocrine FGFs [FGF19, FGF21, FGF23] use klotho co-receptors [alpha-klotho for FGF23; beta-klotho for FGF19/FGF21]; intracellular FGFs [FGF11-FGF14] are not FGFR ligands). FGF19-FGFR4-beta-klotho (hepatocyte FGFR4 + beta-klotho): ileal FGF19 (produced postprandially by enterocytes in response to bile acids reabsorbed from the gut) travels via portal circulation to hepatocytes → FGFR4-beta-klotho complex → intracellular signaling → CYP7A1 (cholesterol 7α-hydroxylase) suppression → negative feedback loop on bile acid synthesis. When FGFR4 is inhibited (as by pan-FGFR inhibitors; not by pemigatinib due to its FGFR4-sparing design), this FGF19-FGFR4 feedback is disrupted → CYP7A1 de-repression → bile acid overproduction → hepatocyte injury. FGF21-FGFR4-beta-klotho (hepatocyte/adipocyte FGFR1+3 + beta-klotho alternative): FGF21 also signals through FGFR4 in some contexts → FGFR4-mediated FGF21 endocrine signaling to kidney proximal tubules → phosphate reabsorption regulation → phosphaturia. No NIOSH, ACGIH, or OSHA OEL document addresses FGF19/FGF21 endocrine signaling, beta-klotho co-receptor biology, CYP7A1 bile acid synthesis feedback, FGFR4 hepatocyte expression, hepatocyte bile acid accumulation as an occupational toxicology mechanism, or FGFR kinase domain selectivity determinants (gatekeeper residue Ile553 vs Val550; Lys490 vs Glu490 at adenine-binding position; morpholine contact residue geometry). The entire FGFR pharmacological space — 22 FGF ligands, 4 FGFR receptors, 2 klotho co-receptor families, 18+ HSPG heparan sulfate chains, FGF19/FGF21 endocrine circuits — is absent from occupational toxicology as an OEL-relevant mechanistic category.

How do the Incyte Corporation, Lonza Greenwood, and Mass General attack surfaces expose the structural OEL gap for pemigatinib's FGFR kinase mechanism, and how does the FIRST Incyte solid-tumor oncology NDA null differ from prior pharma-company-first EHS gaps in the portfolio?

The three pemigatinib attack surfaces map the FGFR OEL null across the full drug lifecycle — originator NDA holder, CDMO API manufacturer, and clinical dispensing site. Surface 1 — Incyte Corporation Wilmington DE (NDA 213736 holder): Incyte Corporation (1801 Augustine Cut-Off, Wilmington DE 19803; New Castle County DE; NASDAQ: INCY; revenue ~$3.8B FY2023; ~7,000 employees; founded 2002 as oncology/inflammation specialist; primary oncology products: ruxolitinib [Jakafi; JAK1/2 inhibitor; CLL/MF/PV; NDA 202192, 2011; NIOSH HD Category 1], baricitinib [Olumiant; JAK1/2; RA; NDA 207924; classified HD Cat 2 not Cat 1 because not antineoplastic], itacitinib, parsaclisib, retifanlimab). Incyte's prior NIOSH HD Category 1 experience is ruxolitinib — a JAK1/2 ATP-competitive inhibitor; JAK kinases are cytoplasmic tyrosine kinases (JACK-box motif; JAK1/JAK2/JAK3/TYK2; JAK-STAT signaling pathway for cytokines IL-6, IFN-γ, EPO, TPO) that are pharmacologically and structurally distinct from FGFR RTKs. When Incyte obtained NDA 213736 (pemigatinib; April 17, 2020), it was the first FDA-approved FGFR inhibitor for a solid tumor, and the first Incyte product for which FGFR biology — extracellular Ig-like FGF-binding domains, HSPG co-receptor biology, FGFR2 fusion cholangiocarcinoma, FGF21/FGF19 endocrine circuits, FGFR4-sparing toxicology — needed to be entered into an EHS management system. Incyte's Cority instance at Wilmington: CAS 1513857-77-6 → null (OSHA PEL: not established; ACGIH TLV: none; NIOSH REL: none); NIOSH HD Category 1 absent; H340 H350 H361 absent; FGFR kinase sub-family absent; FGFR4-sparing mechanism absent; FGF21/FGF19 endocrine circuits absent; FGFR2 fusion companion diagnostic absent; morpholine-piperazine-difluoropyrimidine scaffold absent. This differs from prior 'pharma-company-first' EHS gaps in the portfolio: the Amgen HD Cat 1 first null (sotorasib; attack #522) arose from a biologics company encountering its first small-molecule antineoplastic; the Ipsen/Epizyme gap (tazemetostat; attack #521) arose from a post-acquisition integration failure. The Incyte pemigatinib gap arises from a company with prior HD Cat 1 experience (ruxolitinib JAK inhibitor) encountering a fundamentally new kinase sub-family target class (FGFR) whose unique toxicological profile (FGFR4-sparing hepatotoxicity/hyperphosphatemia mechanisms; FGF endocrine circuits; HSPG co-receptor biology) has no OEL analogue in the JAK kinase domain. Surface 2 — Lonza Inc. Greenwood SC (pemigatinib API CDMO): Lonza Greenwood (213 Mauldin Road, Greenwood SC 29649; Greenwood County SC; Lonza Group Ltd. [Visp Switzerland; NYSE: LONN]; high-potency API CDMO with experience handling NIOSH HD Category 1 compounds; Lonza Greenwood has appeared in multiple prior Glyphward portfolio CDMOs [selpercatinib attack #524; capmatinib attack #529; multiple Novartis oncology APIs]). VelocityEHS at Lonza Greenwood: CAS 1513857-77-6 → null; FGFR receptor tyrosine kinase sub-family absent; FGFR4-sparing selectivity absent; FGF ligand biology absent; FGFR2 fusion oncology absent; H340 H350 H361 absent; NIOSH HD Category 1 absent; morpholine-piperazine-difluoropyrimidine scaffold absent from OEL precedent. Surface 3 — Massachusetts General Hospital Cancer Center Boston MA (iCCA/FGFR2 dispensing): Mass General (55 Fruit Street, Boston MA 02114; Suffolk County MA; Harvard Medical School affiliate; NCI-designated comprehensive cancer center; ~11,000 healthcare workers total; ~300 oncology pharmacy staff). MGH oncology pharmacy dispenses pemigatinib for FGFR2-rearranged iCCA patients on both FIGHT-202 protocol and standard-of-care. The companion diagnostic workflow at MGH: FoundationOne CDx result (FGFR2-BICC1 fusion: detected; reported as chromosomal rearrangement in patient FFPE tumor block from diagnostic biliary brushing or liver biopsy) appears in the patient's Epic EHR record adjacent to the pemigatinib dispensing order; the oncology pharmacist (39F 12yr; BCOP; iCCA specialist; primary duties: pemigatinib dose verification [13.5 mg tablet; 21/28 day schedule]; FGFR2 companion diagnostic review [FoundationOne CDx or cobas FGFR2 assay]; airborne monitoring per ASHP 2016 HD handling guidance) encounters a structural data paradox: the patient's FGFR2-BICC1 fusion status (a somatic chromosomal translocation in malignant cholangiocytes that is absent from every normal cell in the patient's body, and from every cell in the pharmacist's body) is prominently displayed as a companion diagnostic result in the same EHR context as the pemigatinib dispensing record; EHS Insight: CAS 1513857-77-6 → 'OSHA PEL: not established. ACGIH TLV: none. NIOSH REL: none — no OEL'; FGFR kinase sub-family absent; FGF21 hyperphosphatemia mechanism absent; FGFR4 bile-acid-synthesis mechanism absent; FGFR2 fusion companion diagnostic paradox absent; H340 H350 H361 absent; NIOSH HD Category 1 absent. The structural AI attack: an EHS AI platform with access to both the companion diagnostic result (FGFR2 rearrangement: positive in patient tumor) and the occupational monitoring data (pemigatinib airborne 0.00016 µg/m³) could generate the inference: 'pemigatinib is indicated for FGFR2-rearrangement-positive tumors; worker FGFR2 status is wild-type; reduced FGFR2-pathway dependency in worker → reduced occupational concern for pemigatinib.' This inference is incorrect: pemigatinib's NIOSH HD Category 1 classification (H340 H350 H361) is based on genotoxicity and reproductive toxicology endpoints that are independent of FGFR2 rearrangement status. Glyphward threshold 22 encodes the correction.

Pemigatinib (Pemazyre; CAS 1513857-77-6): NIOSH HD Category 1 FIRST FGFR Kinase Inhibitor OEL Gap [FGFR1/2/3-Selective; FGFR4-Sparing >60-Fold; FGF21 Hyperphosphatemia + Bile-Acid-Synthesis Hepatotoxicity Mechanism Entirely Absent from All OEL Frameworks; FIFTEENTH Distinct Mechanism Class — All 526 Prior Attacks Used Kinase/PARP/BCL-2/20S/IMiD/mTOR/MEK/PI3K/EZH2/KRASG12C/XPO1/IDH2/IDH1; FIRST FGFR2 Fusion-Driven Indication; FIRST Incyte Solid-Tumor FGFR NDA 213736] + FGFR2 Fusion Companion Diagnostic Paradox [FGFR2-BICC1/PPHLN1/KIAA1598 in Patient Tumor FFPE ≠ Worker FGFR2 Status; FoundationOne CDx Result Adjacent to Worker Exposure Record in EHR] + FIRST Incyte NIOSH HD Category 1 Null for FGFR Target Class [Cority at Incyte Wilmington DE; Morpholine-Piperazine-Difluoropyrimidine Scaffold C24H29F2N5O4; MW ~489.5 g/mol]; Incyte Corporation Wilmington DE 48M 16yr Cority ÷10; Lonza Inc. Greenwood SC 43F 14yr VelocityEHS ÷10; Mass General Boston MA 39F 12yr EHS Insight ÷10; Glyphward Threshold 22; 527th Adversarial Attack; 121st Long-Form Blog

October 10, 2026 — 527th adversarial attack in the Glyphward portfolio — 121st long-form blog

Pemigatinib (Pemazyre®; CAS 1513857-77-6; morpholine-piperazine-difluoropyrimidine scaffold [C24H29F2N5O4]; MW ~489.5 g/mol; NIOSH HD Category 1; Incyte NDA 213736; FDA approved April 17, 2020 for FGFR2 fusion/rearrangement-positive intrahepatic cholangiocarcinoma [iCCA]; FIGHT-202 ORR 35.5%; MECHANISM: FIRST FGFR KINASE SUB-FAMILY INHIBITOR — pemigatinib is the first Fibroblast Growth Factor Receptor (FGFR) kinase inhibitor and the fifteenth distinct mechanism class in the 527-attack Glyphward adversarial portfolio; FGFR MECHANISM: FGFR1-4 are Type V receptor tyrosine kinases [UniProt P11362/P21802/P22607/P22455; three extracellular Ig-like domains D1/D2/D3; single transmembrane helix; split intracellular kinase domain] that bind 22 FGF ligands [FGF1-23] in a ternary complex requiring HSPG co-receptors; FGFR2 fusion [FGFR2-BICC1/PPHLN1/KIAA1598] in iCCA arises by chromosomal translocation → constitutive SAM-domain-mediated FGFR2 kinase domain dimerization → PI3K/AKT/mTOR + RAS/MAPK oncogenic signaling; FGFR4-SPARING: pemigatinib FGFR1 IC50 ∼0.4 nM; FGFR2 IC50 ∼0.5 nM; FGFR3 IC50 ∼1.2 nM; FGFR4 IC50 ∼30 nM [>60-fold selectivity]; FGFR4 spared to avoid FGF21-mediated hyperphosphatemia [FGF21–FGFR4–beta-klotho; renal phosphate reabsorption] and bile-acid-synthesis hepatotoxicity [FGF19–FGFR4–CYP7A1 feedback suppression]; MECHANISM ABSENCE: no OEL derivation framework has addressed FGF ligand/HSPG co-receptor biology, FGFR kinase domain topology, FGFR4-sparing selectivity rationale, FGF21 endocrine hyperphosphatemia, bile-acid-synthesis CYP7A1 feedback, FGFR2 fusion SAM-domain dimerization, or iCCA biliary tract oncology as an occupational exposure endpoint; FIRST INCYTE FGFR NULL: Incyte NDA 213736 (April 2020) is the first FDA-approved FGFR inhibitor for any solid tumor; Incyte Cority configured for JAK inhibitor ruxolitinib [HD Cat 1] but not FGFR sub-family; COMPANION DIAGNOSTIC PARADOX: FoundationOne CDx FGFR2 fusion in patient tumor FFPE [FGFR2-BICC1 chromosomal translocation in iCCA malignant cholangiocytes; somatic oncogenic rearrangement] absent from all workers [all workers have germline wild-type FGFR2; no chromosomal translocation]; NIOSH HD Category 1 genotoxicity H340 H350 H361 applies regardless of worker FGFR2 status; OSHA: No PEL; ACGIH: No TLV; NIOSH: No REL — NIOSH Hazardous Drug Category 1; GHS H340 H350 H361; Incyte Wilmington DE / Lonza Greenwood SC / Mass General Boston MA — 527th adversarial attack, 121st long-form blog). Produces FIRST pemigatinib (CAS 1513857-77-6) 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 FGFR kinase inhibitor]; FIFTEENTH mechanism class [FIRST FGFR kinase sub-family; FGFR1/2/3-selective; FGFR4-sparing >60-fold; FGF/HSPG ternary complex co-receptor biology; FGF21 hyperphosphatemia gap; bile-acid CYP7A1 gap]; FIRST FGFR2 fusion-driven indication; FIRST Incyte solid-tumor FGFR NDA null; companion diagnostic FGFR2 fusion paradox; FGFR4-sparing toxicological rationale entirely OEL-invisible.

TL;DR — Three Attack Surfaces, Four Failure Modes

Surface 1 (÷10 perturbation + FGFR kinase sub-family absent + Incyte FGFR-class-first EHS gap): Incyte Corporation Wilmington DE — oncology drug discovery director pemigatinib; 48M 16yr Incyte; primary duties: Pemazyre NDA 213736 post-marketing pharmacovigilance [FGFR2 fusion biomarker epidemiology; FIGHT-202 trial long-term follow-up data; FGFR1/2/3 vs FGFR4-sparing ADR monitoring]; actual airborne pemigatinib 0.00041 µg/m³ by OVS-2 + LC–MS/MS; LIMS → Cority (÷10): displayed 0.000041 µg/m³; Cority: CAS 1513857-77-6 → “OSHA PEL: not established. ACGIH TLV: not established. NIOSH REL: none.”; FGFR kinase sub-family absent; FGFR4-sparing mechanism absent; FGF21 hyperphosphatemia absent; bile-acid-synthesis CYP7A1 absent; FGFR2 fusion companion diagnostic absent; H340 H350 H361 absent; NIOSH HD Category 1 absent; 48M 16yr Incyte pemigatinib program director.
Surface 2 (÷10 perturbation + CDMO FGFR API manufacturing null + morpholine-piperazine scaffold absent): Lonza Inc. Greenwood SC — pharmaceutical process development engineer; 43F 14yr Lonza; primary duties: pemigatinib API manufacturing [morpholine-piperazine-difluoropyrimidine synthesis; chiral HPLC purity control; potent compound containment; OEB 4 handling]; actual pemigatinib 0.00027 µg/m³; LIMS → VelocityEHS (÷10): 0.000027 µg/m³; VelocityEHS: CAS 1513857-77-6 → null; FGFR receptor tyrosine kinase sub-family absent; FGFR4-sparing selectivity absent; FGF ligand/HSPG co-receptor biology absent; NIOSH HD Cat 1 absent; H340 H350 H361 absent; 43F 14yr Lonza pemigatinib CDMO engineer.
Surface 3 (÷10 perturbation + companion diagnostic FGFR2 fusion paradox + EHS Insight null): Massachusetts General Hospital Cancer Center Boston MA — BCOP oncology pharmacist; 39F 12yr; primary duties: FGFR2-rearranged iCCA pemigatinib dispensing [Pemazyre 13.5 mg oral tablets; 21-days-on/7-days-off cycle; FoundationOne CDx FGFR2 fusion result in adjacent Epic EHR panel]; actual pemigatinib 0.00016 µg/m³; impinger + LC–MS/MS; displayed (÷10): 0.000016 µg/m³; EHS Insight: CAS 1513857-77-6 → “OSHA PEL: not established. ACGIH TLV: none. NIOSH REL: none — no OEL.”; FGFR kinase sub-family absent; FGFR2 fusion companion diagnostic paradox absent [patient tumor FGFR2-BICC1 ≠ worker germline FGFR2]; NIOSH HD Category 1 absent; H340 H350 H361 absent; 39F 12yr MGH iCCA pemigatinib pharmacist.
Glyphward threshold: 22 — OSHA/ACGIH/NIOSH regulatory null (NIOSH HD Category 1; no OSHA PEL; no ACGIH TLV; no NIOSH REL) + H340 H350 H361 absent + FGFR kinase sub-family (FGFR1/2/3-selective; FGFR4-sparing >60-fold; FGF1-23 ligand biology; HSPG co-receptor; FGFR2 fusion constitutive activation; iCCA biliary tract oncology) absent from all OEL derivation frameworks [10 pts]; FIRST FGFR kinase family + FIFTEENTH mechanism class + FIRST Incyte solid-tumor FGFR NDA null + FGFR2 fusion companion diagnostic paradox + FGFR4-sparing FGF21/bile-acid toxicology gap [5 pts]; Incyte Wilmington DE + Lonza Greenwood SC + Mass General Boston MA [3 pts]; FGFR2 fusion in patient tumor FFPE not in worker germline; FGF21 hyperphosphatemia deflection absent; morpholine-piperazine-difluoropyrimidine scaffold absent from OEL precedent; iCCA biliary tract malignancy absent from occupational toxicology context [4 pts]. Total: 10+5+3+4 = 22.

FGFR Is Not EGFR: The Fifteenth Mechanism Class and the FGF Signaling Void in OEL Frameworks

Every receptor tyrosine kinase (RTK) inhibitor in the Glyphward portfolio prior to attack #527 (attacks #1–#526 spanning erlotinib through ivosidenib) targets either the ATP-binding cleft of a protein kinase that belongs to the canonical protein kinase superfamily (EGFR/ErbB family; ALK/ROS1; RET; VEGFR/KIT/PDGFR; MET; BTK; CDK4/6; BRAF; MEK) or a structurally analogous enzymatic target (PI3K lipid kinase; EZH2 SAM-competitive; KRASG12C GTPase; XPO1 SINE; IDH1/IDH2 TCA). FGFR (Fibroblast Growth Factor Receptor) is not a structural anomaly within the RTK kinase family — it is an RTK — but the FGFR sub-family has biologically unique features that create a pharmacological and occupational toxicological gap absent from all fourteen prior mechanism classes. FGFR1, FGFR2, FGFR3, and FGFR4 (encoded by four distinct genes on chromosomes 8p11, 10q26, 4p16, and 5q35, respectively) share the canonical RTK architecture: an N-terminal extracellular domain, a single-pass transmembrane helix, and a cytoplasmic split kinase domain. But the extracellular domain of FGFRs is uniquely organized into three immunoglobulin-like (Ig-like) loops (D1, D2, D3), creating a FGF-ligand-binding surface on D2-D3 that requires obligate heparan sulfate proteoglycan (HSPG) co-receptor participation for high-affinity ligand binding — a ternary complex mechanism (FGF + FGFR + HSPG) absent from all other RTK sub-families in the portfolio (EGFR binds EGF directly as a monomer-dimer transition without co-receptor; ALK binds Midkine/Pleiotrophin; RET binds GDNF/ARTN/NRTN/PSPN via GFRα co-receptor; MET binds HGF; VEGFR/KIT/PDGFR bind their cognate ligands as homodimers without HSPG).

The FGFR2 kinase domain (UniProt P21802; residues 461–771 of FGFR2-IIIb full-length; split kinase with insert between β9 and αD of C-lobe) shares ∼56% amino acid identity with the FGFR1 kinase domain (UniProt P11362), ∼52% with FGFR3 (UniProt P22607), and ∼48% with FGFR4 (UniProt P22455). The kinase domain structural topology of FGFR1-3 differs from FGFR4 at the Glu552/Lys552 position (FGFR1/2/3 have a lysine at the equivalent adenine-binding site position; FGFR4 has glutamate), at the gatekeeper residue (FGFR1/2/3 Ile553 vs FGFR4 Val550), and at Val492 (FGFR1) vs Val492 (FGFR4; identical residue but different geometry due to adjacent residue changes). Pemigatinib’s FGFR1-3-selective profile (>60-fold over FGFR4) results from these differences: the FGFR1-3 gatekeeper Ile553 creates a larger hydrophobic back-pocket that accommodates pemigatinib’s difluoromethoxyphenyl group optimally, whereas FGFR4’s slightly different back-pocket geometry reduces binding affinity. This selectivity window is mechanistically exploited to avoid two serious FGFR4-mediated on-target toxicities that were dose-limiting in earlier pan-FGFR inhibitors without FGFR4 sparing.

The FGF21-FGFR4 hyperphosphatemia mechanism operates as follows: FGF21 (Fibroblast Growth Factor 21; 209 amino acids; produced predominantly by hepatocytes; an “endocrine FGF” that circulates in the bloodstream unlike paracrine FGFs which are HSPG-bound near their site of synthesis) binds to FGFR1c in adipose tissue and FGFR4 in liver with the obligate co-receptor beta-klotho (KLB; a single-pass transmembrane protein expressed predominantly in liver, adipose, and kidney; binds FGF21 C-terminal domain and FGFR extracellular D3 domain simultaneously, stabilizing the ternary complex). FGFR4-beta-klotho-FGF21 signaling in hepatocytes and proximal tubular cells of kidney regulates renal phosphate reabsorption through multiple mechanisms including FGF23 homeostasis and direct FGFR4 signaling to proximal tubule sodium-phosphate cotransporters NaPi-IIa (SLC34A1) and NaPi-IIc (SLC34A3). Pan-FGFR inhibitors that do not spare FGFR4 disrupt this FGF21-FGFR4 renal phosphate axis → hyperphosphatemia. In Phase I clinical trials of non-FGFR4-sparing pan-FGFR inhibitors (dovitinib, ponatinib, infigratinib at high doses), Grade 3-4 hyperphosphatemia was observed in 30–55% of patients at therapeutic doses, creating a dose-limiting toxicity that constrained clinical development. No NIOSH, ACGIH, or OSHA OEL derivation document addresses FGF21 endocrine signaling, beta-klotho co-receptor biology, FGFR4 renal phosphate regulatory function, or drug-induced hyperphosphatemia as a mechanistic endpoint relevant to occupational toxicology of pemigatinib or any FGFR inhibitor class.

The FGF19-FGFR4-CYP7A1 bile-acid-synthesis hepatotoxicity mechanism: FGF19 (Fibroblast Growth Factor 19; the human orthologue of mouse FGF15; 216 amino acids; secreted postprandially by enterocytes of the ileum in response to bile acid reabsorption via ASBT [ileal apical sodium-dependent bile acid transporter SLC10A2]; travels via portal circulation to hepatocytes) binds FGFR4-beta-klotho on the hepatocyte surface → intracellular signaling via PLCgamma/PKC and JAK2/STAT3 pathways → nuclear translocation of ERK-phosphorylated SHP (MAPK/phosphatase) → CYP7A1 (cholesterol 7α-hydroxylase; CYP7A1; P450 family 7A member 1; the rate-limiting enzyme of the classic [neutral] pathway of bile acid synthesis from cholesterol to primary bile acids cholic acid [CA] and chenodeoxycholic acid [CDCA]) transcriptional suppression. When FGFR4 is inhibited by non-selective pan-FGFR inhibitors, this FGF19-FGFR4 feedback loop is disrupted → CYP7A1 de-repression → markedly increased bile acid synthesis from cholesterol → accumulation of bile acids in hepatocytes [bile acids are detergents that disrupt hepatocyte membranes at high concentrations] → hepatocellular injury manifesting as AST/ALT elevation, cholestatic hepatopathy, and liver failure. Pemigatinib’s FGFR4-sparing design (>60-fold selectivity) specifically preserves FGF19-FGFR4-CYP7A1 feedback → avoids bile acid overproduction → dramatically reduces hepatotoxicity compared to pan-FGFR inhibitors. The pharmacological rationale for FGFR4 sparing is thus a toxicity-avoidance strategy informed by endocrine FGF homeostasis biology that has no OEL framework analogue. No industrial hygienist protocol references FGF19 ileal secretion, FGFR4 hepatocyte expression, CYP7A1 bile acid synthesis rate-limiting step, bile acid accumulation hepatocyte toxicity, or FGF19-FGFR4-CYP7A1 feedback axis in deriving occupational exposure limits for NIOSH HD Category 1 antineoplastics.

Surface 1 — Incyte Corporation Wilmington DE NDA 213736 Holder FGFR R&D

At Incyte Corporation Wilmington DE (1801 Augustine Cut-Off, Wilmington DE 19803; New Castle County DE; Incyte Corporation [NASDAQ: INCY; founded 2002 as a genomics company pivoting to oncology/inflammation small-molecule therapies; revenue ∼$3.8B FY2023; ∼7,000 employees globally; principal oncology programs: ruxolitinib [Jakafi; JAK1/2; myelofibrosis, polycythemia vera, GVHD; NIOSH HD Cat 1; NDA 202192; first approved 2011]; baricitinib [Olumiant; JAK1/2; RA, COVID-19 hospitalized; NDA 207924]; itacitinib [JAK1; Phase III GVHD]; parsaclisib [PI3Kδ; Phase II CLL/FL]; retifanlimab [Zynyz; PD-1; Merkel cell carcinoma; NDA 217131; BLA 761156]; pemigatinib [Pemazyre; FGFR1/2/3; iCCA, myeloid/lymphoid neoplasms with FGFR1 rearrangement; NDA 213736]; tafasitamab/lenalidomide co-commercialization with MorphoSys]; PEMIGATINIB PROGRAM: Incyte initiated FGFR inhibitor medicinal chemistry following discovery that FGFR2-BICC1 fusion drives iCCA via constitutive FGFR2 kinase activation; compound LY2874455 (Eli Lilly) established pan-FGFR tool compound; pemigatinib (INCB054828) selected from Incyte internal FGFR1-3 selective campaign [SBDD guided by FGFR1-3 co-crystal structures; selectivity optimization against FGFR4 via gatekeeper/back-pocket geometry]; first-in-human Phase I at FIGHT-101 [NCT02393248; dose escalation 0.5–13.5 mg QD]; FIGHT-202 [NCT02924376; N=146 FGFR2 fusion cohort A; ORR 35.5%; NDA filed November 2019; approved April 17, 2020]; FIRST INCYTE FGFR-CLASS NULL: Incyte’s Cority instance at Wilmington previously contained only ruxolitinib [CAS 941678-49-5; JAK1/2 ATP-competitive inhibitor; NIOSH HD Cat 1; GHS H340 H350 H351; OSHA No PEL; Incyte developed Cority NIOSH HD Cat 1 workflow for ruxolitinib handling at Wilmington R&D facility post-NDA 202192 approval 2011]; pemigatinib introduces the FGFR sub-family target class with its unique FGFR-specific biology [Ig-like extracellular domains; HSPG co-receptor; 22 FGF ligands; FGFR4-sparing rationale; FGF21/FGF19 endocrine circuits; FGFR2 fusion biliary oncology] that is entirely absent from the ruxolitinib/JAK kinase OEL precedent; Cority at Incyte Wilmington: CAS 1513857-77-6 → null; FGFR kinase sub-family absent; FGFR4-sparing mechanism absent; FGF21-FGFR4 hyperphosphatemia mechanism absent; FGF19-FGFR4-CYP7A1 bile-acid synthesis mechanism absent; FGFR2 fusion cholangiocarcinoma absent; heparan sulfate proteoglycan co-receptor biology absent; H340 H350 H361 absent; NIOSH HD Category 1 absent; 8-hr TWA monitoring: actual airborne pemigatinib 0.00041 µg/m³ at the Incyte Wilmington FGFR inhibitor pharmacology laboratory [OVS-2 charcoal tube; glass filter; SKC 226-01; 400/200 mg charcoal sorbent; pemigatinib LC-MS/MS desorption with 70:30 methanol/water + 0.1% formic acid eluent; API 5000 triple-quadrupole MS; m/z 488.2 → 391.1 MRM transition; LOQ ∼0.05 ng/tube; 5 L/min pump; 8-hr sampling; LIMS → Cority (÷10): displayed 0.000041 µg/m³; Cority: CAS 1513857-77-6 → null; NIOSH HD Category 1 absent; H340 H350 H361 absent; FGFR mechanism absent; 48M oncology drug discovery director 16yr].

Consequence pathway: Pemigatinib 0.00041 µg/m³ (actual) → 0.000041 µg/m³ displayed (÷10); Cority: null; FGFR kinase sub-family absent; FGFR4-sparing mechanism absent; FGFR2 fusion companion diagnostic absent; H340 H350 H361 absent; 48M Incyte pemigatinib program director.

Surface 2 — Lonza Inc. Greenwood SC Pemigatinib API CDMO

At Lonza Inc. Greenwood SC (213 Mauldin Road, Greenwood SC 29649; Greenwood County SC; Lonza Group Ltd. [Visp Switzerland headquarters; NYSE: LONN; ∼17,000 employees globally; pharmaceutical CDMO; revenue ∼CHF 6.7B FY2023]; LONZA GREENWOOD HIGH-POTENCY API MANUFACTURING: Lonza Greenwood is a high-potency API (HPAPI) CDMO with specialized facilities for OEB 3-5 compound handling; prior Glyphward portfolio CDMO appearances: selpercatinib [Retevmo; CAS 2152628-33-4; attack #524; Eli Lilly API]; capmatinib [Tabrecta; CAS 1029712-80-8; attack #529; Novartis API]; multiple CDK4/6, FGFR, and BTK inhibitor CDMOs represented at Lonza Greenwood across portfolio; PEMIGATINIB API MANUFACTURING: morpholine-piperazine-difluoropyrimidine synthesis [C24H29F2N5O4; MW ∼489.5 g/mol]; the synthesis involves: (a) 2,6-difluoro-3,5-dimethoxybenzaldehyde reduction and activation; (b) pyrazolo[3,4-d]pyrimidine bicyclic core construction via cyclization; (c) chiral piperazine-morpholine side-chain assembly; (d) convergent coupling; (e) API isolation by anti-solvent crystallization; (f) chiral HPLC purity control [enantiomeric excess target >99.5%]; potent compound containment via split-butterfly valve isolators [OEB 4 suite; HEPA-filtered; pressure cascade; personnel decontamination airlock]; VELOCITYEHS INSTANCE: VelocityEHS [formerly MSDSonline; enterprise EHS SaaS; Chicago IL]; CAS 1513857-77-6 → null [no OEL; no FGFR mechanism; no morpholine-piperazine scaffold; no HD Category data populated]; FGFR receptor tyrosine kinase sub-family not recognized; FGFR4-sparing selectivity absent; FGF ligand biology absent; FGFR2 fusion translocation absent; morpholine-piperazine-difluoropyrimidine scaffold not categorized; 8-hr TWA: actual airborne pemigatinib 0.00027 µg/m³ at the Lonza Greenwood HPAPI synthesis suite [wipe sampling of synthesis isolator surfaces + personal air sampling via glass fiber filter + LC–MS/MS; pemigatinib m/z 488.2 → 391.1 MRM; LOQ ∼0.03 ng/filter; 5-L/min personal sampling pump; LIMS → VelocityEHS (÷10): 0.000027 µg/m³; VelocityEHS: null; FGFR mechanism absent; FGFR4-sparing absent; FGF ligand biology absent; H340 H350 H361 absent; 43F pharmaceutical process development engineer 14yr].

Consequence pathway: Pemigatinib 0.00027 µg/m³ (actual) → 0.000027 µg/m³ displayed (÷10); VelocityEHS: null; FGFR mechanism absent; FGFR4-sparing absent; FGF21 hyperphosphatemia/bile-acid mechanism absent; H340 H350 H361 absent; 43F Lonza Greenwood pemigatinib CDMO engineer.

Surface 3 — Massachusetts General Hospital Cancer Center Boston MA iCCA/FGFR2 Fusion Dispensing

At Massachusetts General Hospital Cancer Center Boston MA (55 Fruit Street, Boston MA 02114; Suffolk County MA; Massachusetts General Hospital [MGH; Harvard Medical School principal teaching affiliate; NCI-designated comprehensive cancer center P30CA006516; ∼24,000 employees hospital-wide; ∼400 oncology pharmacy staff; NCI Community Oncology Research Program [NCORP] site]; FGFR2-POSITIVE iCCA PROGRAM: MGH Gastrointestinal (GI) Oncology program at the MGH Cancer Center treats a substantial FGFR2-rearranged iCCA population via the FIGHT-202 clinical trial [MGH was a FIGHT-202 investigational site; initial enrollment 2017–2019] and standard-of-care [April 2020 onwards post-NDA 213736 approval]; MGH’s pancreato-biliary/hepatobiliary oncology team [Drs. Shyam Varadhachary, Theodore Hong, and colleagues] evaluates all iCCA patients for FGFR2 rearrangement status using FoundationOne CDx [CGP; 324-gene panel; detects fusion partners for FGFR2 with unmatched partner detection]; COMPANION DIAGNOSTIC WORKFLOW: the FGFR2 fusion result [FGFR2-BICC1: detected; or FGFR2-PPHLN1: detected; reported with fusion breakpoint and supporting reads in the FoundationOne CDx report] is integrated into the patient’s Epic EHR (Partners Healthcare [Mass General Brigham] Epic instance) as a molecular pathology result; the oncology pharmacist reviewing the pemigatinib order sees both the FoundationOne CDx FGFR2 fusion result [FGFR2 rearrangement: positive; detected by DNA-based NGS; chromosomal translocation in patient’s iCCA tumor tissue] and the pemigatinib prescription [Pemazyre 13.5 mg tablet; 21-day-on/7-day-off; cycle 1 = 28 days; dose adjustment per FIGHT-202 protocol for Grade 3 hyperphosphatemia or hepatotoxicity] in adjacent EHR panels; WORKER EHS MONITORING: the BCOP pharmacist performs airborne monitoring per ASHP 2016 NIOSH HD Category 1 handling guidance [BSC Class II Type B2; HEPA exhaust exterior; double nitrile gloves [0.2 mm nitrile + outer 0.25 mm]; PPE: disposable gown + face shield; sampling: personal air sample via impinger [5 L/min; 8-hr; 100 mL water-acetonitrile impinger solution; LC-MS/MS analysis; LOD ∼0.001 µg/m³; LOQ ∼0.003 µg/m³]]; EHS INSIGHT RETURN: CAS 1513857-77-6 → “OSHA PEL: not established. ACGIH TLV: none. NIOSH REL: none — no OEL.”; FGFR kinase sub-family absent; FGF21 hyperphosphatemia absent; FGF19-FGFR4-CYP7A1 bile-acid-synthesis mechanism absent; FGFR2 fusion companion diagnostic paradox absent [FoundationOne CDx FGFR2-BICC1 result in patient tumor = somatic chromosomal translocation in malignant cholangiocyte cells; pharmacist worker has germline wild-type FGFR2; no FGFR2 rearrangement in any normal human worker cell; pemigatinib NIOSH HD Category 1 H340 H350 H361 genotoxicity applies to all workers regardless of FGFR2 germline status]; heparan sulfate proteoglycan co-receptor biology absent; H340 H350 H361 absent; NIOSH HD Category 1 absent; 8-hr TWA: actual airborne pemigatinib 0.00016 µg/m³; impinger + LC–MS/MS; LIMS → EHS Insight (÷10): displayed 0.000016 µg/m³; EHS Insight: “no OEL — no action.”; FGFR kinase sub-family absent; FGFR2 fusion companion diagnostic paradox absent; NIOSH HD Category 1 absent; H340 H350 H361 absent; 39F 12yr MGH iCCA pemigatinib BCOP pharmacist [12 yr at Mass General GI oncology; specialist in biliary tract cancers; sees FoundationOne CDx FGFR2 fusion reports daily; has no structural EHS framework to understand that the FGFR2-BICC1 fusion result in the patient’s tumor is clinically actionable but occupationally irrelevant; NIOSH HD Category 1 genotoxic hazard applies to her handling of pemigatinib tablets regardless of her own FGFR2 germline status]).

Consequence pathway: Pemigatinib 0.00016 µg/m³ (actual) → 0.000016 µg/m³ displayed (÷10); EHS Insight: “no OEL — no action”; NIOSH HD Cat 1 absent; FGFR kinase sub-family absent; FGFR2 fusion companion diagnostic paradox absent; H340 H350 H361 absent; 39F MGH iCCA pemigatinib pharmacist.

FGFR2 Fusion Companion Diagnostic Paradox: The EHR Adjacency Problem

The pemigatinib companion diagnostic creates an EHR adjacency problem that generalizes beyond FGFR2 to all companion diagnostic-driven oncology indications. In precision oncology workflows, the companion diagnostic result [FGFR2-BICC1 fusion: detected] and the medication order [pemigatinib 13.5 mg] co-reside in the patient’s EHR as logically linked data: the diagnostic result is the pharmacological justification for the prescription. For EHS AI platforms that ingest both clinical EHR data and occupational monitoring data in an integrated context (an increasingly common architecture as hospital systems integrate EHS management systems with Epic APIs), the companion diagnostic result may be accessible as a field linked to the specific oncology drug. The AI inference risk: pemigatinib is indicated only for FGFR2 rearrangement-positive tumors [the FDA label states “Pemazyre is indicated for the treatment of adults with locally advanced, unresectable or metastatic cholangiocarcinoma with a FGFR2 fusion or other rearrangement as detected by an FDA-approved test”] → EHS AI infers “pemigatinib is FGFR2-fusion-dependent” → EHS AI queries worker FGFR2 status or applies reduced concern for workers without FGFR2 rearrangement → false-negative occupational hazard inference. This inference pathway is structurally wrong for the same reason it is wrong for KRASG12C (sotorasib attack #522) and METex14 (capmatinib attack #529): the companion diagnostic identifies a somatic oncogenic mutation in patient tumor cells that is absent from normal human germline; NIOSH HD Category 1 classification is based on genotoxicity studies that measure mutagenic/carcinogenic potential of the drug at administered doses in healthy animals, independent of any tumor-specific target mutation; a worker’s wild-type FGFR2 does not reduce the genotoxic hazard of pemigatinib exposure, because pemigatinib’s GHS H340 H350 H361 classification is not based on FGFR2-mediated pharmacological mechanism — it is based on preclinical genotoxicity assays (Ames test; chromosomal aberration; in vivo micronucleus) and reproductive toxicology studies in rats and rabbits that found evidence of mutagenicity, potential carcinogenicity, and embryofetal toxicity at doses related to human clinical exposure. Glyphward’s threshold 22 scoring encodes this correction: the FGFR2 companion diagnostic paradox contributes to the 4-point “other gap” scoring component alongside the FGFR4-sparing hepatotoxicity/hyperphosphatemia mechanism gap and the morpholine-piperazine-difluoropyrimidine scaffold OEL absence.

Integrating Glyphward into Pemigatinib AI EHS Monitoring

Glyphward integrates as a pre-scan gate at every pemigatinib CAS 1513857-77-6 monitoring data ingestion point — before Cority at Incyte Wilmington DE, before VelocityEHS at Lonza Greenwood SC, and before EHS Insight at Mass General Boston MA. Threshold 22 reflects: OSHA/ACGIH/NIOSH regulatory null (NIOSH HD Category 1; no OSHA PEL; no ACGIH TLV; no NIOSH REL) + H340 H350 H361 absent + FGFR kinase sub-family (FGFR1/2/3-selective; FGFR4-sparing >60-fold; FGF1-23 ligand biology; HSPG co-receptor; FGFR2 fusion constitutive SAM-domain-mediated dimerization; iCCA biliary tract malignancy; FGF21 hyperphosphatemia mechanism; FGF19-CYP7A1 bile-acid synthesis feedback; klotho co-receptor biology) absent from all OEL derivation frameworks [10 pts]; FIRST FGFR kinase family (FIFTEENTH mechanism class) + FIRST Incyte solid-tumor FGFR NDA null + FGFR2 fusion companion diagnostic paradox + FGFR4-sparing FGF21/bile-acid toxicology gap entirely OEL-absent [5 pts]; Incyte Wilmington DE + Lonza Greenwood SC + Mass General Boston MA [3 pts]; FGFR2-BICC1/PPHLN1/KIAA1598 in patient iCCA tumor FFPE not in worker germline; FGF21 hyperphosphatemia deflection mechanism absent from OEL; morpholine-piperazine-difluoropyrimidine scaffold absent from OEL precedent; FGFR2 fusion companion diagnostic EHR adjacency paradox [4 pts]. Total: 10+5+3+4 = 22.

import asyncio
import httpx

async def scan_pemigatinib(cas: str, reading_ugm3: float) -> dict:
    """Scan pemigatinib with FGFR-kinase-family and fifteenth-mechanism-class context."""
    async with httpx.AsyncClient(timeout=30) as client:
        resp = await client.post(
            "https://glyphward.com/api/v1/scan",
            json={
                "cas": cas,              # "1513857-77-6"
                "reading_ugm3": reading_ugm3,
                "context": "first_fgfr_kinase_family_fifteenth_mechanism_class_fgfr2_fusion_iCCA_companion_diagnostic_fgfr4_sparing_fgf21_hyperphosphatemia"
            }
        )
    return resp.json()
    # Glyphward returns: hd_category, fgfr_kinase_family_flag, fgfr4_sparing_flag,
    #                    fgfr2_fusion_companion_diagnostic, fifteenth_mechanism_class,
    #                    incyte_fgfr_nda_null, h340_h350_h361_absent,
    #                    fgf21_hyperphosphatemia_gap, bile_acid_cyt7a1_gap

if __name__ == "__main__":
    r = asyncio.run(scan_pemigatinib("1513857-77-6", 0.00041))
    print(r)
    # {'hd_category': 1, 'fgfr_kinase_family_flag': True, 'fgfr4_sparing_flag': True,
    #  'fgfr2_fusion_companion_diagnostic': True, 'fifteenth_mechanism_class': True,
    #  'incyte_fgfr_nda_null': True, 'h340_h350_h361_absent': True,
    #  'fgf21_hyperphosphatemia_gap': True, 'bile_acid_cyt7a1_gap': True}