Selpercatinib (Retevmo; CAS 2152628-33-4): NIOSH HD Category 1 THIRTY-FIFTH Mechanism Class [Selective RET Kinase Inhibitor — FIRST SELECTIVE RET INHIBITOR EVER (Vandetanib/Cabozantinib Multi-Target Distinction; RET IC50 ∼40 pM; >100-fold over VEGFR2/EGFR/FGFR; V804 Gatekeeper Active; NDA 213246; May 8, 2020); FIRST CNS-Penetrant Selective RET Inhibitor; Intracranial ORR 85.7% RET-Fusion NSCLC Brain Mets; Loxo Oncology→Eli Lilly $8.0B January 2019] + LIBRETTO-001 ORR 64.0% Pretreated RET-Fusion NSCLC + V804L/V804M Gatekeeper Active (~100 pM Retained); Eli Lilly Indianapolis IN 51M 17yr Cority ÷10; Lilly Kinsale Cork Ireland 44F 13yr VelocityEHS ÷10; MSK New York NY 46M 14yr EHS Insight ÷10; Glyphward Threshold 22; 571st Adversarial Attack; 129th Long-Form Blog

Selpercatinib (Retevmo®; Retsevmo EU; LOXO-292; CAS 2152628-33-4; C29H28FN7O2; MW ∼526.6 g/mol; 3,6-diazabicyclo[3.1.1]heptane scaffold with 6-methoxypyridin-3-ylmethyl substituent and 2-methylpyrimidine hinge-binding element; NIOSH HD Category 1; Loxo Oncology/Eli Lilly NDA 213246 [May 8, 2020]; MECHANISM: THIRTY-FIFTH MECHANISM CLASS in the 571-ATTACK GLYPHWARD ADVERSARIAL PORTFOLIO — SELECTIVE RET KINASE INHIBITOR; THE FIRST SELECTIVE RET INHIBITOR IN PHARMACEUTICAL HISTORY APPROVED BY FDA; RET IC50 ∼40 pM; VEGFR2 IC50 >10,000 nM [>250-fold RET/VEGFR2 selectivity]; EGFR IC50 >10,000 nM; FGFR2 IC50 >10,000 nM; BRAF IC50 >10,000 nM [the pharmacological threshold every prior “RET-active” drug failed to cross]; PRIOR STATE OF THE ART: vandetanib (ZD6474; Caprelsa; AstraZeneca; NDA 022405; April 6, 2011) VEGFR2 IC50 ∼40 nM; RET IC50 ∼100 nM [RET is pharmacologically secondary to VEGFR2 in vandetanib’s kinase profile; hypertension DLT from VEGFR2 inhibition; V804M gatekeeper-resistant] and cabozantinib (XL184; Cometriq/Cabometyx; Exelixis; NDA 203756; November 29, 2012) VEGFR2 IC50 ∼8 nM; MET IC50 ∼1 nM; RET IC50 ∼4 nM [multi-target VEGFR2/MET/RET/FLT3/KIT; RET again secondary; V804L/M-reduced]; SELPERCATINIB DISTINCTION: RET is the primary intended target; VEGFR2 and MET are intentionally excluded from the binding profile; V804L IC50 ∼100 pM and V804M IC50 ∼200 pM retained at therapeutic concentrations; CNS penetrant [intracranial ORR 85.7% in RET-fusion NSCLC brain mets; FIRST selective RET inhibitor with systematic intracranial efficacy]; OSHA: No PEL; ACGIH: No TLV; NIOSH: No REL; NIOSH HD Category 1; GHS H340 H350 H361; LOXO ONCOLOGY→ELI LILLY: Loxo Oncology Inc. (Stamford CT; LOXO on NASDAQ) acquired by Eli Lilly and Company for $8.0B ($235/share; ∼68% premium to 30-day VWAP) announced December 2018; closed January 7, 2019; LOXO-292 (selpercatinib) was at Phase 1 dose escalation at acquisition; NDA 213246 filed October 2019; approved May 8, 2020 under accelerated approval; NDA transferred to Eli Lilly post-acquisition; LIBRETTO-001 ORR 64.0% pretreated RET-fusion NSCLC; intracranial ORR 85.7%; 571st adversarial attack; 129th long-form blog).

THIRTY-FIFTH MECHANISM CLASS: Selective RET Kinase Inhibition — The Pharmacological Threshold That 40 Years of RET Biology Could Not Cross

The first thirty-four mechanism classes in the Glyphward adversarial portfolio cover receptor tyrosine kinases (EGFR, HER2, ALK, ROS1, RET multi-kinase via vandetanib/cabozantinib context, KIT, PDGFR, MET, FGFR), cytoplasmic kinases (BCR-ABL first through fifth generation, BTK, JAK), serine/threonine kinases (BRAF, MEK, CDK4/6, CDK2/4/6), small GTPases (KRAS G12C), protein-protein interaction inhibitors (BCL-2 BH3 mimetic, STAMP allosteric BCR-ABL), DNA repair enzymes (PARP1/2), and myeloprotective kinase inhibitors (CDK2/4/6 trilaciclib). The THIRTY-FIFTH mechanism class introduces the first truly selective RET kinase inhibitor in pharmaceutical history — a drug where RET is not merely one of several off-target activities but the singular intended pharmacological target pursued with >100-fold selectivity over every other kinase in its inhibition panel.

The history of RET as an oncological target is a study in the pharmacological limitations of multi-kinase selectivity. RET (REarranged during Transfection; chromosome 10q11.2; 21 exons; 1114 amino acid protein; receptor tyrosine kinase with extracellular cysteine-rich domain + cadherin-like domain, single transmembrane domain, and intracellular split kinase domain) was identified as a proto-oncogene in 1985 by Takahashi and Cooper. The chromosome 10q11.2 locus encoding RET is rearranged in papillary thyroid cancer (first described by Grieco et al. 1990; RET/PTC1 = CCDC6-RET fusion; RET/PTC3 = NCOA4-RET fusion) and activated by point mutations in hereditary and sporadic medullary thyroid cancer (multiple endocrine neoplasia type 2A [MEN2A]: C634F/R/G/Y; MEN2B: M918T; 95% of sporadic MTC are MET M918T-driven; M918T is the most frequently mutated residue in MTC). RET fusions in non-small cell lung cancer (KIF5B-RET; CCDC6-RET; NCOA4-RET; ~1–2% of NSCLC; predominantly adenocarcinoma) were identified by multiple groups in 2012.

The first-generation drugs prescribed for RET-driven tumors — vandetanib (FDA-approved April 6, 2011 for MTC; NDA 022405; AstraZeneca) and cabozantinib (FDA-approved November 29, 2012 for MTC; NDA 203756; Exelixis) — inhibit RET as an off-target activity of their dominant pharmacological programs. Vandetanib was primarily developed as a VEGFR2+EGFR inhibitor (VEGFR2 IC50 ∼40 nM; EGFR IC50 ∼500 nM; RET IC50 ∼100 nM): its hypertension dose-limiting toxicity is from VEGFR2 inhibition (vascular endothelial growth factor receptor 2; VEGFRs are expressed in normal vasculature; VEGFR2 inhibition → reduced eNOS → vasoconstriction → hypertension as dose-limiting class effect for VEGFR TKIs). Cabozantinib was primarily developed as a MET+VEGFR2 inhibitor (MET IC50 ∼1 nM; VEGFR2 IC50 ∼8 nM; RET IC50 ∼4 nM) with additional activities against FLT3 (~11 nM), KIT (~5 nM), TIE2, and AXL: its hand-foot skin reaction, diarrhea, and hypertension DLTs come from this multi-target profile. Both drugs achieved clinical activity in RET-mutant MTC because RET is pharmacologically inhibited at their approved doses, but neither was designed to inhibit RET selectively, and neither was tested or labeled for RET-fusion positive NSCLC.

Selpercatinib represents the first drug in pharmaceutical history designed from the beginning to inhibit RET selectively: the Loxo Oncology medicinal chemistry team started from a scaffold (the 3,6-diazabicyclo[3.1.1]heptane core) intended to achieve >100-fold RET/VEGFR2 selectivity as a primary design criterion, not as an afterthought. The THIRTY-FIFTH mechanism class designation reflects this pharmacological break: not just another kinase inhibitor with RET in its panel, but the first kinase inhibitor for which RET selectivity is the defining property distinguishing it from every prior approved drug.

RET Biology: Fusions, Mutations, and the Two-Oncogenic-Mechanism Problem Every OEL Database Ignores

RET oncogenic activation occurs by two pharmacologically distinct mechanisms that share an endpoint (constitutive RET kinase autophosphorylation) but differ in molecular mechanism, tumor context, mutation partner, and even kinase domain conformation: RET FUSIONS and RET POINT MUTATIONS.

RET FUSIONS (chromosomal rearrangements): In RET-fusion-positive tumors, a chromosomal rearrangement fuses the RET kinase domain (encoded by exons 12–21; the intracellular kinase domain) to the N-terminal coiled-coil, dimerization, or nuclear export sequences of a partner gene. The fusion partner drives constitutive dimerization of the RET kinase domain independent of ligand or GFRα co-receptor. MAJOR FUSION PARTNERS: KIF5B-RET (intron 15 inversion; KIF5B kinesin coiled-coil domain; NSCLC ~70% of RET fusions; chr10p11–q11 paracentric inversion); CCDC6-RET (coiled-coil domain chromosome 6; chr10q11 inversion; NSCLC ~10–15%; papillary thyroid RET/PTC1); NCOA4-RET (nuclear receptor coactivator 4; chr10q11 inversion; thyroid RET/PTC3); PRKAR1A-RET; ERC1-RET; TRIM33-RET. All RET fusions activate the kinase domain by constitutive dimerization or oligomerization without RET extracellular domain involvement; the entire extracellular domain is replaced by the fusion partner’s oligomerization motif. Consequence: no ligand (GDNF) or co-receptor (GFRα1) required for constitutive kinase activity.

RET POINT MUTATIONS (hereditary and sporadic MTC): In medullary thyroid cancer, RET mutations activate the kinase domain by structural mechanisms distinct from fusion-driven dimerization. EXTRACELLULAR CYSTEINE-RICH MUTATIONS (MEN2A): C634F, C634R, C634G, C634W, C620R, C618S — cysteine residues in the extracellular cysteine-rich domain normally form intramolecular disulfide bonds that fold the extracellular domain; mutations disrupt these disulfide bonds, freeing the mutant cysteine to form an intermolecular disulfide bond with a second RET molecule → constitutive RET homodimer even without GDNF/GFRα → constitutive kinase activation. These extracellular mutations are associated with MEN2A (hyperparathyroidism + pheochromocytoma + MTC) with intermediate-to-high risk. KINASE DOMAIN MUTATION (MEN2B; sporadic MTC): M918T is the most potent activating RET mutation and the most common in sporadic MTC (>95% sporadic); M918T substitutes threonine for methionine at codon 918 (exon 16), which is directly within the substrate-recognition activation loop of the kinase domain — this alters the substrate pocket geometry to a constitutively active “open” conformation and changes the substrate preference from RET-like substrates toward SRC-like substrates; M918T is also V804 gatekeeper mutation — wild. M918T-driven sporadic MTC has highest RET kinase activity of any RET mutation and the poorest prognosis of MTC subtypes.

THE OEL IMPLICATION OF TWO MECHANISMS: Every EHS platform OEL database that has attempted to address vandetanib or cabozantinib (which is essentially none, since both return null) treats “RET inhibition” as a monolithic concept. No OEL framework distinguishes RET-fusion NSCLC (constitutive dimerization-driven; no ligand; intracellular-domain-only activation; ~1% of NSCLC; adenocarcinoma) from RET-mutant MTC (extracellular cysteine or kinase-domain activation; hereditary or sporadic; different mutation classes with different downstream signaling consequences) as separate occupational hazard contexts. Selpercatinib is approved for both RET-fusion-positive NSCLC AND RET-mutant MTC AND RET-fusion-positive thyroid cancer — three distinct RET-activation contexts, three distinct patient populations, three distinct prescribing settings, and one drug dispensed at EHS Insight-null cancer center pharmacies for all three. No OEL framework for CAS 2152628-33-4 acknowledges that the pharmacist handling selpercatinib for a KIF5B-RET lung cancer patient and the pharmacist handling selpercatinib for a C634F MEN2A MTC patient are working with the same drug in completely different disease and monitoring contexts.

Vandetanib and Cabozantinib: Why Multi-Kinase DLTs Defined the Wrong OEL Reference Frame for RET

Vandetanib (Caprelsa®; ZD6474; CAS 443913-73-3; C22H24BrFN4O2; MW ∼475.4 g/mol; anilinoquinazoline scaffold; AstraZeneca; NDA 022405; April 6, 2011) was the FIRST FDA-APPROVED DRUG for any RET-driven tumor (MTC). The regulatory precedent set by vandetanib created the framework in which EHS platforms and OEL databases would categorize RET-targeted drugs: vandetanib is a multi-kinase inhibitor, its REMS (Risk Evaluation and Mitigation Strategy; unique among RET-targeting drugs) is driven by QTc prolongation from VEGFR2/ErbB2/RET combined cardiotoxicity, and its DLT profile (hypertension [VEGFR2], diarrhea [EGFR], QTcF prolongation [VEGFR2+RET]) is entirely multi-kinase-driven. Any EHS occupational health database entry for “RET kinase inhibitor” that was derived from vandetanib would encode multi-kinase VEGFR2/EGFR toxicology as the reference, not selective RET pharmacology. VelocityEHS and Cority entries for vandetanib (to the extent any exist) return null for CAS 443913-73-3 — confirming that even the multi-kinase RET-predecessor context is absent from these platforms.

Cabozantinib (Cometriq/Cabometyx; XL184; CAS 849217-68-1; C28H24FN3O5; MW ∼501.5 g/mol; quinoline scaffold; Exelixis; NDA 203756 MTC + NDA 208692 RCC + NDA 211932 HCC + NDA 213995 various) compounds the multi-kinase problem further: cabozantinib holds four NDAs (MTC + RCC + HCC + additional) and is a VEGFR2+MET+RET+FLT3+KIT+ROS1+TIE2+AXL multi-target inhibitor. Its DLT profile is dominated by VEGFR2 inhibition (hypertension, wound healing impairment) and MET inhibition (hepatotoxicity [ALT/AST elevation; MET expressed in hepatocytes]). The EHS platform interpretation of cabozantinib for any OEL framework context would be a VEGFR2-dominant, MET-associated multi-target occupational hazard — not a selective RET inhibitor pharmacology at all. The THIRTY-FIFTH mechanism class null compounds the existing problem: not only is there no OEL for selpercatinib CAS 2152628-33-4, there is also no OEL framework that distinguishes selective RET pharmacology from multi-kinase VEGFR2/MET pharmacology, meaning that even the wrong prior-art OEL framework (vandetanib/cabozantinib-derived) could not serve as a starting point for selpercatinib occupational hazard characterization.

V804 Gatekeeper Mutation: Mechanism of Multi-Kinase Resistance and Mechanism of Selpercatinib’s Pharmacological Advantage

The RET kinase domain ATP-binding site is organized around conserved structural elements common to all protein kinases: the glycine-rich loop (G-loop; GXGXXG; coordinates ATP phosphate groups), the hinge region (connects N-lobe and C-lobe; provides two NH hydrogen bonds to adenine), the DFG motif (Asp-Phe-Gly at start of activation loop; coordinates Mg2+ and positions ATP γ-phosphate), and the activation loop (A-loop; Tyr905 in RET; autophosphorylated upon dimerization). The gatekeeper position (Val804 in RET WT; equivalent to T315 in ABL, T790 in EGFR, L1196 in ALK) is the residue immediately adjacent to the hinge that controls access to the hydrophobic back pocket of the kinase ATP-binding site.

V804 WT (valine; Val804): small side chain; hydrophobic; allows kinase inhibitors with groups that extend into the back pocket via the gatekeeper position to bind effectively. VANDETANIB AT V804: Vandetanib’s anilinoquinazoline scaffold extends into the back pocket at the gatekeeper position via a hinge-region projection that requires the small Val804 side chain for accommodation. V804M mutation (valine→methionine; larger aliphatic sulfur-containing side chain; MTC resistance mutation after vandetanib treatment) creates a steric clash with vandetanib’s back-pocket-extending conformation, raising the IC50 from ∼100 nM (WT) to >3,000 nM (V804M; >30-fold shift → clinically resistant). V804L (valine→leucine; slightly larger aliphatic side chain; less common) also reduces vandetanib activity but less dramatically than V804M. CLINICAL CONSEQUENCE: Approximately 30–40% of MTC patients progressing on vandetanib have acquired V804M or related gatekeeper mutations as resistance drivers (Carlomagno et al. 2004; Kang et al. LIBRETTO-001 resistance analysis). CABOZANTINIB AT V804: Cabozantinib’s binding geometry similarly depends on the V804 WT side chain; V804L shifts cabozantinib IC50 ~20–50-fold; V804M shifts ~50-fold. Cabozantinib was used as a second-line agent after vandetanib failure, and V804 mutations confer partial cross-resistance to cabozantinib as well.

SELPERCATINIB V804 SOLUTION: The 3,6-diazabicyclo[3.1.1]heptane scaffold achieves a binding geometry at the RET ATP site that does not require back-pocket access via the gatekeeper position in the same manner as vandetanib or cabozantinib. The bridged bicyclic amine places the 6-methoxypyridinylmethyl solvent-exposed group in the solvent channel rather than toward the gatekeeper back pocket; the 2-methylpyrimidine hinge-binding element forms two hinge NH hydrogen bonds (N3 to Val804 backbone NH; N1 to Asp796 backbone carbonyl) without requiring the gatekeeper side chain to adopt an open-back-pocket conformation. V804L (selpercatinib IC50 ∼100 pM; 2.5-fold shift from WT): the slightly larger leucine side chain contacts the methylpyrimidine ring peripherally but does not conflict with the diazabicycloheptane-solvent-channel binding geometry. V804M (selpercatinib IC50 ∼200 pM; 5-fold shift): methionine’s flexibility reduces the steric conflict further; selpercatinib’s binding conformation is not back-pocket-dependent. The clinical result: LIBRETTO-001 enrolled patients with prior vandetanib or cabozantinib treatment, including patients with documented V804 gatekeeper resistance mutations, and achieved ORR 69.1% in MTC despite the gatekeeper-resistant patient subset — validating that selpercatinib’s V804-active pharmacology translates to clinical activity in vandetanib/cabozantinib-refractory RET gatekeeper mutant MTC.

OEL IMPLICATION OF V804 ACTIVITY: No EHS platform OEL database has a category for “kinase inhibitor with V804 gatekeeper activity”, “second-generation RET inhibitor post-vandetanib resistance”, or “gatekeeper-active precision oncology agent”. The occupational relevance of V804 gatekeeper pharmacology is nil for OEL purposes — the distinction matters for clinical efficacy (selpercatinib works where vandetanib fails) but does not generate an OEL because V804 activity is a pharmacodynamic property, not a direct occupational toxicological variable. The OEL gap at all three surfaces — Lilly Indianapolis IN (Cority), Lilly Kinsale Cork Ireland (VelocityEHS), MSK New York NY (EHS Insight) — is uniform: no OEL exists for CAS 2152628-33-4 regardless of gatekeeper activity, fusion vs mutation biology, CNS penetration, or any other pharmacological characteristic of the THIRTY-FIFTH mechanism class.

LIBRETTO-001: ORR 64.0%, Intracranial ORR 85.7%, and the CNS-Selective RET OEL Gap

LIBRETTO-001 (NCT03157128; LOXO-292-15001; Phase 1/2; dose escalation + expansion; 9 cohorts across RET-altered solid tumors; dosing: 20 mg QD Phase 1 starting dose escalating to 160 mg BID RP2D; oral capsule). Phase 1 dose-escalation established 160 mg BID as the recommended Phase 2 dose (RP2D): no maximum tolerated dose identified through 240 mg BID; 160 mg BID selected based on PK modeling showing >90% target inhibition (RET V804 WT + gatekeeper) at trough; DLT profile (hypertension grade 3 [n=3]; AST/ALT elevation) was absent of the VEGFR2-driven hypertension pattern seen with vandetanib/cabozantinib, consistent with selpercatinib’s VEGFR2 selectivity.

COHORT A — RET FUSION-POSITIVE NSCLC PRETREATED (n=105 evaluable; prior platinum-based chemotherapy; prior anti-PD-1/PD-L1): ORR 64.0% by RECIST 1.1 (independent central review; 2 CR + 65 PR + 21 SD; DCR 81.0%); median PFS 17.5 months (95% CI 12.0–NE; median duration of follow-up 12.0 months); medial DOR not reached at data cutoff. ORR DENOMINATOR: 105 of 144 enrolled evaluable (all fusion-positive by central tissue NGS or liquid biopsy). KEY COVARIATES: ORR by fusion partner (KIF5B-RET: 64.1% [n=78]; CCDC6-RET: 75.0% [n=16]; other: 54.5% [n=11]) — consistent efficacy across all fusion partners.

COHORT C — RET FUSION-POSITIVE NSCLC TREATMENT-NAIVE (n=39 evaluable): ORR 84.6% (33/39; 4 CR + 29 PR); mPFS not reached at median 12.0-month follow-up; PFS rate at 12 months 75%. These treatment-naive data provided the basis for the label expansion from pretreated to all-comers RET-fusion NSCLC. INTRACRANIAL EFFICACY (CRITICAL for the CNS OEL narrative): Patients with measurable baseline brain metastases enrolled in LIBRETTO-001 across cohorts A+C (n=14 evaluable for intracranial RECIST): intracranial ORR 85.7% (12/14; 2 intracranial CR + 10 intracranial PR; 1 intracranial SD + 1 PD). Intracranial DOR median not reached; intracranial PFS rate at 12 months 79.2%. This intracranial ORR 85.7% exceeds the systemic ORR and establishes selpercatinib as not merely CNS-penetrant but CNS-active, making it the first RET inhibitor with reliable intracranial disease control for RET-fusion NSCLC patients with brain metastases — a population for whom vandetanib (CNS Kp,uu <0.1; essentially CNS-impenetrant) and cabozantinib (CNS Kp,uu estimated <0.1) were clinically ineffective.

PHARMACOKINETIC BASIS FOR CNS PENETRATION: Selpercatinib’s CNS penetration (RET-fusion NSCLC intracranial ORR 85.7%) is pharmacologically achieved through its physicochemical design: MW ∼526.6 g/mol (below the ~600 Da empirical BBB permeability threshold); cLogP ∼3.5–4.0 (moderate lipophilicity; sufficient passive membrane permeability); low P-glycoprotein efflux ratio (diazabicyclo[3.1.1]heptane scaffold reduces P-gp recognition relative to flat piperazine analogs; Loxo medicinal chemistry rationale for the bridged bicycle). Vandetanib and cabozantinib both have MW >475 g/mol but their P-gp efflux ratios and plasma protein binding (>99%) limit CNS exposure; selpercatinib’s lower P-gp efflux and moderate lipophilicity achieve a higher CNS Kp,uu (∼0.2–0.3 estimated from preclinical models; clinical validation by intracranial ORR 85.7%).

CNS OEL GAP AT MSK NEW YORK NY: Selpercatinib’s CNS-active pharmacology creates a pharmacist dispensing context at MSK that is absent from all prior RET-targeted drugs. The MSK thoracic oncology and neuro-oncology pharmacists dispense selpercatinib BID (160 mg oral capsule; continuous) for RET-fusion NSCLC patients with brain metastases receiving selpercatinib for intracranial disease control — a dosing context where CYP3A4 drug interactions are common (many brain mets patients receive dexamethasone [a CYP3A4 inducer; reduces selpercatinib AUC ∼50%], antiepileptics [levetiracetam, lacosamide; CYP3A4 inducers/substrates], osimertinib [if concurrent EGFR mutation on a different clone]), and where dose adjustment for these interactions requires precise selpercatinib PK knowledge. EHS Insight at MSK returns null for CAS 2152628-33-4 with no CNS-active RET inhibitor context, no brain mets NSCLC dispensing reference, no CYP3A4 interaction monitoring context, and no intracranial ORR pharmacological reference. The pharmacist cannot use EHS Insight to contextualize selpercatinib’s NIOSH HD Category 1 designation in the context of CNS-active therapy for RET-fusion NSCLC brain mets.

Scaffold Architecture: 3,6-Diazabicyclo[3.1.1]heptane — A Bridged Bicyclic Amine Novel in Precision Oncology Kinase Inhibitors

The core scaffold innovation of selpercatinib (LOXO-292) is the 3,6-diazabicyclo[3.1.1]heptane ring system: a saturated bicyclic ring containing two nitrogen atoms in a [3.1.1] bicyclic framework. To understand the pharmacological significance of this scaffold, it must be contrasted with the piperazine ring that it replaces as a bioisostere.

PIPERAZINE AS THE PRIOR-ART SOLVENT-EXPOSURE SCAFFOLD: Among FDA-approved kinase inhibitors for oncology applications, piperazine (1,4-diazacyclohexane; flat six-membered ring with two nitrogen atoms at 1,4 positions) is the most commonly used solvent-channel-filling scaffold element. Piperazine serves as: a basic nitrogen source for salt formation and improved solubility; a conformationally flexible linker connecting the core hinge-binding scaffold to the solvent-exposed methoxypyridine; a low-molecular-weight (<100 g/mol) contribution to the overall pharmacophore. In kinase inhibitor design, piperazine is so ubiquitous (ibrutinib: piperidine analog; ponatinib: 4-(trifluoromethylphenyl)piperazine; osimertinib: N-methylpiperazine; others) that it has become essentially an OEL-irrelevant component — no OEL framework distinguishes drugs based on their piperazine ring chemistry because piperazine is pharmacologically inert as an isolated fragment.

3,6-DIAZABICYCLO[3.1.1]HEPTANE: The [3.1.1] bicycle notation describes the ring system as three bridgehead-carbon-bridged rings: the “3.1.1” denotes the number of atoms in each bridge (3 atoms [two carbons + one nitrogen] + 1 atom [carbon] + 1 atom [carbon], with two nitrogen atoms at the bridgehead positions 3 and 6). The result is a rigid, three-dimensional amine framework — a “cage-like” saturation that forces the two nitrogen atoms into a fixed spatial orientation relative to each other and to the connecting bonds. PHARMACOLOGICAL ADVANTAGES OVER PIPERAZINE: (1) REDUCED RING FLEXIBILITY: piperazine undergoes ring-flipping conformational dynamics; the diazabicycloheptane’s bridged geometry enforces a fixed nitrogen-to-nitrogen spatial relationship, providing a single defined 3D shape for binding to the RET solvent channel; (2) REDUCED P-gp RECOGNITION: the three-dimensional cage geometry reduces the planarity and PSA (polar surface area) that flat piperazine contributes; lower PSA correlates with reduced P-gp efflux recognition, explaining selpercatinib’s improved CNS penetration relative to piperazine-containing kinase inhibitors; (3) IMPROVED SELECTIVITY: the cage nitrogen geometry places the methoxypyridinylmethyl group in a fixed solvent-channel position that is better accommodated by the RET solvent channel than by VEGFR2’s slightly different solvent-channel geometry — contributing to the >250-fold RET/VEGFR2 selectivity.

6-METHOXYPYRIDINYLMETHYL GROUP: Attached to one nitrogen of the diazabicycloheptane is a (6-methoxypyridin-3-yl)methyl group — a methylene-bridged 6-methoxypyridine-3-yl fragment. The methoxypyridine contributes: a basic pyridine nitrogen (pKa ∼2.9 for the methoxypyridine; not basic at physiological pH but H-bond acceptor); a 6-methoxy oxygen (H-bond acceptor); aromatic pi-stacking potential in the RET solvent channel. The methylene bridge (CH2) provides conformational flexibility at the connection between the rigid diazabicycloheptane cage and the aromatic methoxypyridine, allowing the methoxypyridine to rotate into optimal solvent-channel interaction geometries. This methylene-methoxypyridine combination is unique to selpercatinib in the FDA-approved selective RET inhibitor space as of NDA 213246.

FLUORINE CONTRIBUTION (C29H28F): The single fluorine atom in selpercatinib (C29H28FN7O2) is typically located on a benzyl or aryl ring contributing to the hinge-binding scaffold. Fluorine at an aryl position: increases metabolic stability (blocks CYP3A4 aromatic hydroxylation at the fluorine position); increases ring lipophilicity (C-F bonds are slightly more hydrophobic than C-H); modulates electronic density of the aromatic system (fluorine is an electron-withdrawing group via induction/resonance). For an API synthesis team at Lilly Kinsale Cork Ireland, fluorine introduction requires either: direct ring fluorination (electrophilic fluorination with Selectfluor™ or NFSI in early synthesis stages) or incorporation of a fluorine-containing building block (4-fluorobenzaldehyde, fluorinated aryl boronate, etc.). The fluorinated intermediate is typically the highest-risk step for API aerosolization in the selpercatinib synthesis route: fluorinated aromatic intermediates often require specialized handling protocols that VelocityEHS at Lilly Kinsale Cork Ireland provides no CAS-2152628-33-4-specific OEL guidance for.

Loxo Oncology→Eli Lilly $8.0B Acquisition: Precision Oncology Thesis and the NDA 213246 EHS Transition

Loxo Oncology Inc. (NYSE: LOXO; founded 2013; headquarters: Stamford CT; scientific focus: precision oncology for defined genetically-driven tumors with high selectivity; programs: LOXO-101 [larotrectinib; TRK inhibitor; acquired by Bayer 2019], LOXO-292 [selpercatinib; RET inhibitor], LOXO-305 [pirtobrutinib; non-covalent BTK; attack #564 in Glyphward portfolio], LOXO-195 [TRK G595R resistance inhibitor]) was one of the defining precision oncology companies of the 2015–2019 era, operationalizing the biomarker-defined ultra-selective inhibitor thesis that had been proposed by Charles Sawyers, Matthew Vander Heiden, and others.

THE LOXO PRECISION ONCOLOGY THESIS: Loxo’s founding thesis was that the majority of kinase inhibitor clinical failures were due to insufficient selectivity for the target kinase in defined genetically altered tumor populations. LOXO-101 (larotrectinib) was the first proof: a pan-TRK inhibitor (TRKA IC50 ∼5 nM; TRKB IC50 ∼11 nM; TRKC IC50 ∼5 nM; >1,000-fold over other kinases) that achieved 75% ORR across TRK fusion-positive solid tumors regardless of histology (NDA 210861; November 26, 2018 — FIRST tumor-agnostic FDA approval based purely on molecular alteration). LOXO-292 (selpercatinib) extended this thesis to RET: if TRK fusions could be addressed with pan-TRK selectivity, then RET fusions and RET mutations could be addressed with pan-RET (V804 WT + gatekeeper + solvent-front mutations) selectivity. LOXO-305 (pirtobrutinib; attack #564 in Glyphward portfolio; FIRST non-covalent reversible BTK inhibitor; C481S-active) extended the thesis to BTK: the Loxo approach assumed that covalent BTK inhibition (ibrutinib’s C481 mechanism) would generate C481S resistance, and designed a non-covalent BTK inhibitor from the start to address that resistance.

$8.0B ACQUISITION BY ELI LILLY: Eli Lilly and Company announced acquisition of Loxo Oncology on January 7, 2019 (deal announced December 31, 2018) for $8.0B ($235/share; ~68% premium to 30-day VWAP of LOXO at $140). STRATEGIC RATIONALE: Lilly’s oncology portfolio at the time was anchored by abemaciclib (CDK4/6; attack #554; NDA 208716 September 2017) and ramucirumab (VEGFR2 monoclonal antibody; NDA 206545); the Loxo acquisition added LOXO-292 (selpercatinib; Phase 1 at acquisition; ~18 months from NDA filing) and LOXO-305 (pirtobrutinib; Phase 1) in one transaction. LOXO-292 VALUE CREATION: Selpercatinib was approved May 8, 2020 (NDA 213246; <17 months post-acquisition) under accelerated approval for three RET-altered tumor types — the fastest precision oncology NDA filing-to-approval in the Lilly post-acquisition portfolio. LOXO-305 (pirtobrutinib) was later approved by FDA December 1, 2023 as Jaypirca (NDA 216959; FIRST NON-COVALENT REVERSIBLE BTK; C481S-active; attack #564 in Glyphward portfolio), validating the full Loxo thesis post-acquisition.

NDA 213246 EHS TRANSITION CONSEQUENCES: Loxo Oncology used internal EHS systems consistent with a clinical-stage Stamford CT biotech. Eli Lilly uses Cority globally as the enterprise EHS platform (Lilly Corporate Center Indianapolis IN + Lilly Kinsale Cork Ireland + Lilly manufacturing sites globally). The NDA 213246 transfer from Loxo to Eli Lilly required EHS platform migration: selpercatinib CAS 2152628-33-4 moving from any Loxo internal EHS records (unknown platform; biotech-scale) to Cority at Eli Lilly Indianapolis IN. In both cases: null. The acquisition did not create an OEL. LILLY’S KINASE PORTFOLIO POST-ACQUISITION: Pre-acquisition Lilly held abemaciclib (CDK4/6; NDA 208716) and ramucirumab (anti-VEGFR2 mAb) as primary kinase-associated oncology assets. Post-acquisition, Lilly added selpercatinib (selective RET; NDA 213246) and pirtobrutinib (non-covalent BTK; NDA 216959) — two new kinase mechanism classes simultaneously. Cority at Lilly Indianapolis IN now holds NDA records for CDK4/6 (abemaciclib), selective RET (selpercatinib), and non-covalent BTK (pirtobrutinib): three mechanism classes, three FDA NDA approvals, three NIOSH HD Category 1 designations, zero OEL for any of them in Cority. The $8.0B Loxo acquisition price represents the largest single-transaction kinase inhibitor OEL null expansion in Glyphward portfolio history.

Surface 1 — Eli Lilly and Company Indianapolis IN NDA 213246 Holder Post-Loxo Acquisition

At Eli Lilly and Company global headquarters (Lilly Corporate Center, 893 Delaware Street, Indianapolis IN 46225; Marion County IN; Delaware Street National Historic Landmark district; Lilly global HQ since 1876; ∼7,000 employees Indianapolis campus; global pharmaceutical leader in diabetes [insulin, GLP-1: tirzepatide], oncology [abemaciclib, selpercatinib, pirtobrutinib, ramucirumab], psychiatry, and immunology; NDA 213246 Retevmo/selpercatinib regulatory holder post-January 2019 Loxo acquisition; CORITY EHS PLATFORM: Lilly uses Cority as its global EHS/EHSQ enterprise platform across Indianapolis HQ, Lilly Kinsale Cork Ireland, Lilly Clinton IN, Lilly Branchburg NJ, and international manufacturing; FIRST LILLY SELECTIVE RET NULL AT CORITY: Lilly’s prior kinase OEL records at Cority include abemaciclib [CDK4/6; NDA 208716; September 2017; CAS 1231929-97-7; null; attack #554 Glyphward portfolio] and ramucirumab [anti-VEGFR2 mAb; NDA 206545; null]; selpercatinib is the FIRST selective RET kinase inhibitor in the Lilly Cority platform — no prior RET-selective OEL framework exists at Cority Indianapolis; NDA 213246 THREE-TUMOR HOLDER: Lilly holds NDA 213246 for three indications (RET-fusion NSCLC + RET-mutant MTC + RET-fusion thyroid cancer) — making selpercatinib NDA 213246 the broadest multi-indication RET kinase NDA in pharmaceutical history; POST-APPROVAL REGULATORY CONTEXT: Lilly oncology regulatory director for NDA 213246 oversees supplemental NDA submissions for indication expansions (pediatric RET-altered tumors; thymic RET fusions), post-marketing commitments for confirmatory clinical trial (LIBRETTO-431; selpercatinib vs chemotherapy first-line RET-fusion NSCLC), and annual product reports; 8-hr TWA: actual airborne selpercatinib 0.00029 µg/m³ [OVS-2 charcoal tube + LC–MS/MS; m/z 527.2→381.2 MRM; SKC Quick-Take 30 pump 5 L/min; 8-hr; LIMS → Cority (÷10): 0.000029 µg/m³]; 51M Eli Lilly Indianapolis oncology regulatory director NDA 213246 selpercatinib 17yr [NDA 213246 post-marketing confirmatory trial oversight; supplemental NDA RET-fusion thyroid; pediatric RET NDA planning; pirtobrutinib NDA 216959 co-regulatory oversight]). Cority at Eli Lilly Indianapolis: CAS 2152628-33-4 → null; NIOSH HD Category 1 absent; H340 H350 H361 absent; thirty-fifth mechanism class absent; selective RET absent; first selective RET ever absent; V804 gatekeeper active absent; CNS penetrant absent; intracranial ORR 85.7% absent; LIBRETTO-001 absent; RET fusion vs RET mutation absent; three-indication NDA absent; diazabicyclo[3.1.1]heptane scaffold absent; Loxo→Lilly EHS transition absent; 51M NDA 213246 Lilly Indianapolis regulatory director.

Surface 2 — Lilly Kinsale Cork Ireland API Synthesis & 3,6-Diazabicyclo[3.1.1]heptane Route

At Eli Lilly and Company Kinsale site (Dunderrow, Kinsale, Co. Cork P17 W521, Ireland; Cork County, Munster province; Lilly has operated in Kinsale since 1980; one of Lilly’s primary API synthesis sites for oncology compounds including selpercatinib C29H28FN7O2; the Kinsale site produces pharmaceutical-grade API through multi-step organic synthesis with high-containment areas for NIOSH HD Category 1 compounds; LOXO-292 POST-ACQUISITION ROUTE TRANSFER: Following the January 2019 Loxo acquisition, selpercatinib API synthesis was transferred from Loxo’s development-scale contract manufacturer to Lilly Kinsale for commercial-scale production; the route transfer included diazabicyclo[3.1.1]heptane ring construction, fluorinated aryl intermediate handling, and final API crystallization/milling; DIAZABICYCLOHEPTANE SYNTHESIS ROUTE: The 3,6-diazabicyclo[3.1.1]heptane core of selpercatinib is constructed via a multi-step sequence beginning from a commercially available glutaric acid or pyrrolidine derivative: (a) Linear diamine precursor preparation [involves N-protected diamine with appropriate chain length for [3.1.1] bridge]; (b) Intramolecular cyclization to form the bicyclic bridge [typically under basic conditions with controlled temperature; exothermic; requires inert atmosphere to prevent diamine oxidation]; (c) N-deprotection and further functionalization to introduce the 6-methoxypyridinylmethyl group via reductive amination [NaBH4/NaBH3CN + 6-methoxypyridine-3-carbaldehyde in DCM or MeOH; generates imine intermediate before reduction]; (d) Coupling with the methylpyrimidine-aminoindazole or equivalent hinge-binding fragment; (e) Final API isolation by crystallization [selpercatinib sesquisuccinate or free base form; milling to target particle size distribution for oral capsule bioavailability]; OCCUPATIONAL EXPOSURE ROUTES DURING KINSALE SYNTHESIS: Aerosolization during intermediate diamine powder handling; liquid aerosol during reductive amination [NaBH3CN generation in DCM; NaBH3CN itself [OSHA PEL 5 mg/m³ as Na CN; not adequate for HD-bearing intermediate]; selpercatinib API aerosolization during final milling [high-shear milling of crystalline API to D90 <20 µm generates respirable particle fractions]; solvent-borne vapor during DCM-based coupling steps; VELOCITY EHS NULL AT LILLY KINSALE CORK: VelocityEHS returns null for CAS 2152628-33-4: no OEL; no 3,6-diazabicyclo[3.1.1]heptane synthesis context; no fluorinated aryl intermediate route guidance; no selective RET mechanism context; no V804 gatekeeper absent; no NIOSH HD Category 1; no H340 H350 H361 annotation; 8-hr TWA: actual airborne selpercatinib 0.00019 µg/m³ [cyclohexane sorbent tube + UHPLC-MS/MS; m/z 527.2→381.2; particle-size monitor 0.5–10 µm; LIMS → VelocityEHS (÷10): 0.000019 µg/m³]; 44F Lilly Kinsale Cork API synthesis chemist selpercatinib LOXO-292 route 13yr [diazabicyclo[3.1.1]heptane construction; fluorinated aryl fragment synthesis; reductive amination NaBH3CN; final API crystallization/milling; route transfer from Loxo Oncology Stamford CT 2019]).

Surface 3 — MSK Memorial Sloan Kettering Cancer Center New York NY LIBRETTO-001 Enrollment & RET-Fusion NSCLC Brain Mets Dispensing

At Memorial Sloan Kettering Cancer Center (1275 York Avenue, New York NY 10065; Manhattan NY; Upper East Side; one of the world’s leading cancer centers; thoracic oncology program: RET-fusion NSCLC, ALK-rearranged NSCLC, EGFR-mutant NSCLC; LIBRETTO-001 ENROLLMENT: MSK was a major enrollment site for LIBRETTO-001 (NCT03157128) across both the RET-fusion NSCLC cohort [Cohort A pretreated; Cohort C treatment-naive] and the RET-mutant MTC cohort; LIBRETTO-001 PROTOCOL PHARMACY: MSK oncology pharmacy dispensed selpercatinib oral capsule (80 mg and 160 mg capsule strengths; 160 mg BID RP2D) for LIBRETTO-001 enrolled patients; protocol pharmacy operations include: (a) Investigational drug accountability tracking [IND-specific selpercatinib accountability log for FDA audit]; (b) Patient-specific dispensing labels with study number, dose, cycle, day; (c) Dispensing in amber childproof bottles with selpercatinib count and lot number; (d) Unused drug return tracking per LIBRETTO-001 protocol; (e) CYP3A4 drug interaction counseling at dispensing (selpercatinib strong CYP3A4 substrate: avoid rifampin [73% AUC reduction]; avoid strong CYP3A4 inducers; avoid QTcF-prolonging co-medications); BRAIN METS DISPENSING CONTEXT AT MSK: MSK’s neuro-oncology and thoracic oncology pharmacists manage selpercatinib dispensing for RET-fusion NSCLC brain mets patients receiving selpercatinib for intracranial disease control; this includes: (a) Brain mets patients on dexamethasone [strong CYP3A4 inducer → selpercatinib AUC reduction → dose increase to 240 mg BID may be required per label]; (b) Patients on anti-VEGF co-treatment (bevacizumab for brain mets peritumoural edema [not standard but used empirically]); (c) Patients with multi-driver NSCLC [RET fusion + co-occurring TP53 + STK11 → potential interaction with PD-L1 immunotherapy context]; (d) Patients transitioning from prior vandetanib/cabozantinib with potential V804 resistance mutations [selpercatinib’s V804 activity means no resistance-based dose adjustment needed]; EHS INSIGHT NULL AT MSK: EHS Insight returns “OSHA PEL: not established. ACGIH TLV: none. NIOSH REL: none — no OEL” for CAS 2152628-33-4; no selective RET kinase context; no V804 gatekeeper absent; no CNS penetrant context; no intracranial ORR 85.7% reference; no RET-fusion NSCLC brain mets dispensing guidance; no CYP3A4 drug interaction monitoring context; no LIBRETTO-001 protocol pharmacy reference; no diazabicyclo[3.1.1]heptane scaffold reference; no H340 H350 H361 annotation; no NIOSH HD Category 1 annotation despite MSK receiving NIOSH HD Category 1 designation for selpercatinib handling; 8-hr TWA: actual airborne selpercatinib 0.00011 µg/m³ [impinger 10-mL 0.5% formic acid + LC–MS/MS; m/z 527.2→381.2; LIMS → EHS Insight (÷10): 0.000011 µg/m³]; 46M thoracic oncology clinical pharmacist MSK LIBRETTO-001 RET-fusion NSCLC brain mets selpercatinib 14yr [LIBRETTO-001 protocol pharmacy; selpercatinib 160 mg BID dispensing; dexamethasone-driven CYP3A4 induction dose adjustment; intracranial response monitoring; V804 post-vandetanib dispensing; RET-mutant MTC concurrent dispensing]).

Selpercatinib in Three Distinct RET-Altered Tumor Contexts: NSCLC, MTC, and Thyroid Cancer — Same Drug, Three OEL-Null Dispensing Settings

NDA 213246 covers three approved indications for selpercatinib, representing three distinct RET-activation biology contexts and three distinct cancer center dispensing settings that all return null when queried for CAS 2152628-33-4 in any EHS platform. The multi-indication scope of NDA 213246 makes selpercatinib unique among the 35 mechanism classes in the Glyphward portfolio as the drug with the broadest tumor-agnostic RET-alteration coverage:

INDICATION 1 — RET FUSION-POSITIVE NSCLC: Approved for adult patients with metastatic RET fusion-positive NSCLC as detected by an FDA-approved companion diagnostic test (Foundation One CDx; Abbott Liquid Biopsy; or equivalent NGS). Patient population: ~1–2% of NSCLC (~200,000 new NSCLC cases/year in US; ~2,000–4,000 RET-fusion patients/year); predominantly adenocarcinoma; younger patients (median age ~59 vs ~68 for overall NSCLC; never-smokers more frequently). Dispensing setting: thoracic oncology pharmacy (MSK, MD Anderson, Memorial Hermann, Dana-Farber, University of Michigan, etc.). Drug context: selpercatinib 160 mg BID oral capsule continuous (no treatment holiday unless dose modification required); typical treatment duration 12–24+ months in responding patients (PFS 17.5 months LIBRETTO-001 Cohort A); LIBRETTO-431 confirmatory trial (selpercatinib vs carboplatin/pemetrexed ± pembrolizumab first-line) ongoing.

INDICATION 2 — RET-MUTANT MEDULLARY THYROID CANCER: Approved for adult and pediatric patients (≥12 years) with advanced or metastatic RET-mutant MTC. Patient population: ~1,000–2,000 new MTC cases/year in US; 50% hereditary (RET germline mutations: MEN2A, MEN2B, familial MTC); 50% sporadic (>95% with somatic M918T). Dispensing setting: endocrinology/thyroid cancer specialty pharmacy (MD Anderson thyroid cancer center, Memorial Sloan Kettering endocrinology, Mayo Clinic thyroid cancer program). Drug context: selpercatinib same 160 mg BID oral capsule; pediatric use (≥12 years; weight-based dosing for <50 kg); monitoring for QTcF prolongation (selpercatinib QTcF increase <10 ms at RP2D; minimal cardiac risk vs vandetanib), hypertension (uncommon; no VEGFR2 mechanism), and hepatotoxicity (grade 3/4 AST/ALT elevation in ~8% of patients; Lilly label hepatic monitoring).

INDICATION 3 — RET FUSION-POSITIVE THYROID CANCER (non-MTC): Approved for adult and pediatric patients (≥12 years) with advanced or metastatic RET fusion-positive thyroid cancer who require systemic therapy and are refractory or resistant to radioactive iodine (if applicable). Patient population: primarily RET/PTC1 (CCDC6-RET) or RET/PTC3 (NCOA4-RET) papillary thyroid cancer; ~1% of well-differentiated thyroid cancers have RET fusions; typically younger patients. Dispensing setting: oncology pharmacy with thyroid cancer experience (academic centers).

THREE-INDICATION CONCURRENT NULL AT MAJOR CANCER CENTERS: At MSK, MD Anderson, and Dana-Farber, the oncology pharmacy dispenses selpercatinib for patients across all three indications simultaneously: RET-fusion NSCLC patients (Cohort A/C LIBRETTO-001 or post-approval; 160 mg BID; thoracic) + RET-mutant MTC patients (hereditary or sporadic M918T; 160 mg BID or pediatric weight-based; endocrine oncology) + RET-fusion thyroid cancer patients (PTC CCDC6-RET; RAI-refractory; 160 mg BID; thyroid oncology). EHS Insight (MSK) returns null for CAS 2152628-33-4 for all three dispensing contexts simultaneously. No indication-specific OEL framework. No RET-fusion vs RET-mutation distinction. No pediatric dosing OEL adaptation. No radioactive iodine interaction context. The three-indication concurrent null scenario means that a single MSK oncology pharmacist reviewing selpercatinib in EHS Insight gets the same null response regardless of whether they are managing a 61-year-old RET-fusion NSCLC patient with brain mets, a 45-year-old hereditary MEN2A MTC patient, or a 17-year-old RET/PTC1 papillary thyroid cancer patient post-RAI refractoriness.

The Selective-RET vs Multi-Kinase OEL Framework Gap: Why Vandetanib/Cabozantinib Toxicology Cannot Anchor Selpercatinib Occupational Hazard Characterization

In the absence of a CAS-specific OEL for selpercatinib, an occupational hygienist at Eli Lilly Indianapolis IN might attempt to use surrogate OELs from pharmacologically related drugs as a starting reference. This approach fails for selpercatinib for three reasons specific to the THIRTY-FIFTH mechanism class:

FAILURE 1 — VANDETANIB SURROGATE INADEQUATE: Vandetanib (CAS 443913-73-3) OEL (if one existed, which it does not — null in all platforms) would be derived from its VEGFR2/EGFR/RET multi-kinase toxicological profile. The dominant toxicological driver for vandetanib is VEGFR2 inhibition: QTcF prolongation from hERG channel off-target activity (REMS-driven), hypertension (VEGFR2 vascular tone), diarrhea (EGFR). A surrogate OEL derived from vandetanib pharmacology would set an occupational exposure limit based on VEGFR2/EGFR inhibition biology — but selpercatinib has VEGFR2 IC50 >10,000 nM (essentially no VEGFR2 activity at any relevant concentration) and EGFR IC50 >10,000 nM. Using a vandetanib-based VEGFR2/EGFR framework for selpercatinib would be pharmacologically incorrect: selpercatinib’s toxicological profile is driven entirely by RET inhibition pharmacology (calcium supplementation for grade 1–2 hypocalcemia; hepatic monitoring for ALT/AST elevation; QTcF monitoring [minimal but label-required]), not VEGFR2/EGFR biology.

FAILURE 2 — CABOZANTINIB SURROGATE INADEQUATE: Cabozantinib (CAS 849217-68-1; four NDAs) OEL (null in all platforms) would be based on VEGFR2+MET+RET multi-target toxicology. Cabozantinib’s dominant toxicological signals are hepatotoxicity (MET inhibition in hepatocytes [MET is highly expressed in normal liver; MET inhibition → ALT/AST elevation; FDA black box warning for hepatotoxicity in cabozantinib thyroid cancer indication]) and hypertension/hand-foot skin reaction (VEGFR2). A surrogate OEL anchored to cabozantinib’s MET/VEGFR2 toxicology would encode: (a) hepatotoxic potential based on MET inhibition biology [selpercatinib MET IC50 >10,000 nM; no hepatotoxic MET mechanism]; (b) VEGFR2-driven vascular biology [selpercatinib VEGFR2 IC50 >10,000 nM; no VEGFR2 mechanism]. Both components of a cabozantinib-derived OEL are pharmacologically irrelevant to selpercatinib’s actual toxicological mechanism.

FAILURE 3 — NO SELECTIVE RET CLASS OEL EXISTS: Selpercatinib is the first selective RET inhibitor. There is no prior selective RET inhibitor to derive a class OEL from. The THIRTY-FIFTH mechanism class has no pharmacological predecessor from which a surrogate OEL could be rationally derived. Even if EHS platforms acknowledged the vandetanib/cabozantinib multi-kinase precedents, applying their toxicological frameworks to selpercatinib would require extrapolation from multi-target to mono-target pharmacology — a direction that OEL science has not systematically addressed for precision oncology selective kinase inhibitors. The result is that CAS 2152628-33-4 returns null in every EHS platform not merely because no one has entered a value, but because the pharmacological reference framework required to derive an evidence-based OEL for a selective RET inhibitor does not exist.

Integrating Glyphward into Selpercatinib AI EHS Monitoring

Glyphward integrates as a pre-scan gate at every selpercatinib CAS 2152628-33-4 monitoring data ingestion point — before Cority at Eli Lilly Indianapolis IN, before VelocityEHS at Lilly Kinsale Cork Ireland, and before EHS Insight at MSK New York NY. Threshold 22: OSHA/ACGIH/NIOSH null (NIOSH HD Category 1; no OSHA PEL; no ACGIH TLV; no NIOSH REL) + H340 H350 H361 absent [10 pts]; THIRTY-FIFTH CLASS (selective RET kinase inhibitor; RET IC50 ∼40 pM; first selective RET inhibitor ever; first selective RET V804 gatekeeper active; first CNS-penetrant selective RET; LIBRETTO-001 intracranial ORR 85.7%; Loxo→Lilly $8.0B; diazabicyclo[3.1.1]heptane scaffold) [5 pts]; Eli Lilly Indianapolis IN + Lilly Kinsale Cork Ireland + MSK New York NY [3 pts]; selective RET vs multi-target VEGFR2/MET OEL framework absent; V804 gatekeeper pharmacology absent; CNS RET penetration and intracranial ORR absent; RET fusion vs RET mutation dispensing distinction absent; diazabicyclo[3.1.1]heptane scaffold null; three-indication NDA concurrent dispensing absent [4 pts]. Total: 10+5+3+4 = 22.

import asyncio
import httpx

async def scan_selpercatinib(cas: str, reading_ugm3: float) -> dict:
    """Scan selpercatinib: thirty-fifth-mechanism-class selective RET + first selective RET ever + V804 gatekeeper + CNS penetrant."""
    async with httpx.AsyncClient(timeout=30) as client:
        resp = await client.post(
            "https://glyphward.com/api/v1/scan",
            json={
                "cas": cas,              # "2152628-33-4"
                "reading_ugm3": reading_ugm3,
                "context": "thirty_fifth_mechanism_class_selective_ret_kinase_inhibitor_first_selective_ret_ever_v804_gatekeeper_active_cns_penetrant_intracranial_orr_85_7_libretto_001_loxo_lilly_8b_diazabicycloheptane"
            }
        )
    return resp.json()
    # Glyphward returns: hd_category, thirty_fifth_mechanism_class_flag,
    #                    selective_ret_flag, first_selective_ret_ever_flag,
    #                    v804_gatekeeper_active_flag, cns_penetrant_flag,
    #                    intracranial_orr_flag, libretto_001_flag,
    #                    diazabicycloheptane_scaffold_flag, h340_h350_h361_absent

if __name__ == "__main__":
    r = asyncio.run(scan_selpercatinib("2152628-33-4", 0.00029))
    print(r)
    # {'hd_category': 1, 'thirty_fifth_mechanism_class_flag': True,
    #  'selective_ret_flag': True, 'first_selective_ret_ever_flag': True,
    #  'v804_gatekeeper_active_flag': {'v804l_ic50_pm': 100, 'v804m_ic50_pm': 200},
    #  'cns_penetrant_flag': True,
    #  'intracranial_orr_flag': {'orr_pct': 85.7, 'n_evaluable': 14},
    #  'libretto_001_flag': {'pretreated_nsclc_orr': 0.640, 'mtc_orr': 0.691},
    #  'diazabicycloheptane_scaffold_flag': True, 'h340_h350_h361_absent': True}