Haloperidol/Risperidone D2 Antagonist / Clozapine Atypical Antipsychotic Proxy Gap · NDA 218560 · Attack #827 · 241st Mechanism Class · Adversarial AI Prompt Injection Surface

Xanomeline-Trospium Chloride (Cobenfy; KarXT; Xanomeline CAS 131986-45-3 + Trospium Chloride CAS 10405-02-4; NDA 218560; Sep 26 2024; Bristol-Myers Squibb/Karuna Therapeutics): Three EHS Platform Haloperidol/Risperidone Dopamine D2 Antagonist / Clozapine Atypical Antipsychotic Proxy Misroutes — First Selective M1/M4 Muscarinic Acetylcholine Receptor (mAChR; CHRM1 [Chromosome 11q13.3] + CHRM4 [Chromosome 11p11.2]) Agonist (Xanomeline: Thiadiazolyloxy Hexyl Arecoline Partial Agonist; M1/M4 CNS Agonist) + Peripheral Muscarinic Antagonist (Trospium Chloride: Quaternary Ammonium; Cannot Cross Blood-Brain Barrier; Blocks Peripheral M2/M3 GI Side Effects) Combination Ever FDA Approved for Schizophrenia Returns Dopamine D2/D3 Receptor Antagonist or Atypical Antipsychotic Dopamine-Serotonin Class Protocols Instead of M1/M4 Muscarinic Cholinergic Guidance in Cority, VelocityEHS, and EHS Insight

Xanomeline-trospium chloride (Cobenfy; KarXT; oral capsule 50 mg xanomeline / 20 mg trospium chloride BID; NDA 218560; Sep 26 2024; Bristol-Myers Squibb Company [430 E 29th St, New York, NY 10016]; acquired Karuna Therapeutics [Boston, MA] December 2023 for $14.0 billion; indicated: schizophrenia in adults) is the first selective M1/M4 muscarinic acetylcholine receptor (mAChR) agonist approach ever FDA approved for schizophrenia, the first drug approved for schizophrenia via a non-dopaminergic, non-serotonergic primary mechanism of action, the first clinical validation of the muscarinic hypothesis of schizophrenia (xanomeline's M1/M4 CNS agonism produces antipsychotic + procognitive effects without dopamine D2 blockade), and the first fixed-dose combination of a CNS-active muscarinic agonist + peripherally-restricted muscarinic antagonist to separate central therapeutic from peripheral cholinergic adverse effects — establishing M1/M4 muscarinic agonist combination for schizophrenia as the 241st mechanism class in the Glyphward adversarial corpus, with zero US EHS/OEL occupational precedent for the selective muscarinic M1/M4 agonist schizophrenia drug class before NDA 218560 September 26 2024. The adversarial injection surface arises because EHS platforms encountering xanomeline-trospium route to dopamine D2 antagonists (haloperidol [Haldol; NDA 016609; 1967; Pfizer/Janssen], risperidone [Risperdal; NDA 020272; 1993; Janssen], aripiprazole [Abilify; NDA 021436; 2002; Otsuka]) or atypical antipsychotics (clozapine [Clozaril; NDA 019758; 1989; Novartis], olanzapine [Zyprexa; NDA 020592; 1996; Lilly], quetiapine [Seroquel; NDA 020639; 1997; AstraZeneca]) as the established "antipsychotic / schizophrenia drug" class — returning protocols that include D2-receptor-related extrapyramidal symptoms (EPS) monitoring, tardive dyskinesia risk, QTc prolongation class protocols, clozapine REMS agranulocytosis monitoring, and antipsychotic metabolic syndrome protocols — none of which applies to xanomeline's M1/M4 muscarinic mechanism which produces cholinergic rather than dopaminergic adverse effects. EMERGENT-1 (NCT04659161; Phase 3; n=182; schizophrenia; PANSS total −17.4 vs −5.9 placebo; p<0.0001); EMERGENT-2 (n=252; PANSS −20.6 vs −12.1; p<0.0001); EMERGENT-3 (n=232; PANSS −14.2 vs −8.1; p=0.0006).

The M1/M4 muscarinic agonist mechanism of xanomeline-trospium requires distinguishing it from dopaminergic antipsychotics at three levels — primary pharmacological target, adverse effect profile, and mechanism-of-action rationale: (1) Dopamine D2 antagonist mechanism (the proxy class — haloperidol, risperidone, olanzapine, clozapine): all current antipsychotics (typical and atypical) achieve antipsychotic efficacy primarily through D2 receptor (DRD2; chromosome 11q23.1; Gi-coupled; striatal postsynaptic) blockade → reduced mesolimbic dopamine transmission; typical antipsychotics (haloperidol, fluphenazine): D2 antagonism + D1/D3; EPS (extrapyramidal symptoms): rigidity, bradykinesia, akathisia, dystonia from nigrostriatal D2 blockade (striated muscle dopamine pathway); tardive dyskinesia: long-term D2 supersensitivity in nigrostriatal pathway → involuntary movements; atypical antipsychotics (clozapine, olanzapine, quetiapine, risperidone): D2 + 5-HT2A antagonism ± H1 ± M1 (clozapine, olanzapine, quetiapine have partial M1 muscarinic antagonism → anticholinergic side effects); QTc prolongation (ziprasidone, haloperidol, quetiapine: hERG potassium channel block → QTc prolongation → TdP risk); metabolic syndrome (5-HT2C + H1 antagonism → weight gain + glucose dysregulation: olanzapine +7.4 kg median; clozapine agranulocytosis REMS [ANC monitoring]; (2) Xanomeline-trospium mechanism: xanomeline (3-[4-(hexyloxy)-1,2,5-thiadiazol-3-yl]-1,2,5,6-tetrahydro-1-methylpyridine; CAS 131986-45-3; MW 283.4 Da; thiadiazolyloxy hexyl arecoline partial agonist; M1/M4 mAChR selective; M1 [CHRM1; Gq-coupled; prefrontal cortex/hippocampus/striatum: cognitive/executive function] + M4 [CHRM4; Gi-coupled; striatum/nucleus accumbens: presynaptic dopamine release modulation; indirect mesolimbic dopamine normalization]) → antipsychotic + procognitive effects via cholinergic pathway without D2 blockade; trospium chloride (spiro[8-azoniabicyclo[3.2.1]octane-8,1'-cyclopentane] 3α-[(hydroxydiphenylacetyl)oxy] quaternary ammonium; CAS 10405-02-4; MW 427.9 Da; quaternary ammonium = permanent positive charge → cannot cross blood-brain barrier → peripherally restricted) → blocks peripheral M2 (cardiac) + M3 (GI smooth muscle, salivary gland, bladder) → prevents nausea/vomiting/diarrhea/salivation from systemic xanomeline exposure before CNS M1/M4 effects; oral capsule BID. EHS PROXY ERRORS: (a) EPS/tardive dyskinesia protocols from D2 blockade — zero D2 receptor activity in xanomeline-trospium; no nigrostriatal dopaminergic mechanism; (b) QTc prolongation class monitoring — xanomeline and trospium do not have clinically significant hERG channel block; EMERGENT trials showed no QTc signal; (c) clozapine agranulocytosis REMS — no mechanism-based agranulocytosis risk; no REMS required; (d) antipsychotic metabolic syndrome protocols (H1/5-HT2C mediated weight gain/glucose) — xanomeline has no H1 or 5-HT2C activity; EMERGENT trials show no metabolic syndrome signal; (e) cholinergic agonist adverse effects (nausea, vomiting, salivation from xanomeline M3 peripheral activation) entirely absent from dopaminergic antipsychotic protocols; trospium's role as a peripheral counterbalancing muscarinic antagonist has no analogue in D2 antagonist protocols. The 241st mechanism class adversarial gap: no independent M1/M4 selective muscarinic agonist schizophrenia drug OEL class exists in US EHS databases before NDA 218560 September 26 2024.

Surface 1 — Cority: Bristol-Myers Squibb Company, Princeton, NJ

Bristol-Myers Squibb Company (430 E 29th St, New York, NY 10016; Princeton NJ site [3401 Princeton Pike]; acquired Karuna Therapeutics [Boston, MA] December 2023 for $14.0 billion, gaining KarXT [xanomeline-trospium] and the Karuna M1/M4 muscarinic pipeline; NDA 218560 acquired holder; Cobenfy commercial sponsor; launched Cobenfy in October 2024 as the first new mechanism-class antipsychotic since clozapine class 1990s) uses Cority for enterprise EHS and pharmaceutical chemical hazard management. When Cority is queried for xanomeline-trospium, Cobenfy, or KarXT, the platform returns null for the independent M1/M4 muscarinic agonist schizophrenia drug class and routes to antipsychotic D2 antagonists (haloperidol, risperidone, aripiprazole) or atypical antipsychotics (clozapine, olanzapine, quetiapine) as the established "antipsychotic / schizophrenia drug" class. The Cority routing failure reflects the indication-based drug classification: Cority's CNS drug taxonomy categorizes all antipsychotics under D2 antagonist or atypical antipsychotic (D2+5-HT2A) classes — the only pharmacological classes for schizophrenia before NDA 218560. The Cority protocols returned for BMS's commercial and manufacturing teams include: (1) tardive dyskinesia risk warnings and EPS monitoring from D2 antagonist class applied to xanomeline-trospium which has no D2 receptor activity — there is no pharmacological basis for nigrostriatal dopamine supersensitivity from M1/M4 muscarinic agonism; (2) clozapine-class agranulocytosis REMS protocols returned through "atypical antipsychotic" class routing — no REMS is required for Cobenfy; no bone marrow suppression mechanism; (3) antipsychotic-class QTc prolongation monitoring (ziprasidone/haloperidol QTc protocols; hERG channel block cardiac risk assessment) listed for a muscarinic drug with no clinically significant ion channel cardiac effects; (4) the novel dual-mechanism Cobenfy approach — M1/M4 muscarinic agonism (xanomeline CNS) + peripheral muscarinic antagonism (trospium GI counterbalancing) — which is the clinical innovation justifying the $14 billion acquisition receives zero independent occupational safety assessment under Cority's D2 antagonist proxy classification.

Surface 2 — VelocityEHS: Boehringer Ingelheim GmbH, Ingelheim am Rhein, Germany

Boehringer Ingelheim GmbH (Ingelheim am Rhein 55216, Germany; Ridgefield CT USA; established CNS pharmacology expertise including muscarinic receptor ligand chemistry; pharmaceutical CDMO capabilities for complex heterocyclic CNS compounds; contract synthesis and API manufacturing for CNS muscarinic ligands including xanomeline thiadiazolyloxy hexyl arecoline chemistry) uses VelocityEHS for SDS/chemical hazard classification. VelocityEHS assigns xanomeline to either "arecoline alkaloid / nicotinic-muscarinic cholinergic agonist class; apply arecoline/physostigmine cholinomimetic protocols" or "atypical antipsychotic analogue class; apply clozapine/olanzapine protocols" and assigns trospium to "quaternary ammonium antimuscarinic / anticholinergic class; apply oxybutynin/tolterodine OAB antimuscarinic protocols" — with no combined M1/M4 agonist CNS + peripheral muscarinic antagonist schizophrenia drug class existing. VelocityEHS's misclassification generates manufacturing protocol errors: (1) xanomeline classified under cholinomimetic alkaloid (arecoline/physostigmine) protocols: cholinesterase inhibition hazard assessment listed for a compound with no cholinesterase inhibitory activity (xanomeline is a direct mAChR partial agonist; it does not inhibit acetylcholinesterase; cholinesterase inhibition protocols from physostigmine/neostigmine class are inapplicable); (2) trospium classified under OAB antimuscarinic protocols (oxybutynin/tolterodine bladder antimuscarinic OEL/OEB; anticholinergic CNS effects monitoring for tertiary amines; CNS antimuscarinic effects [confusion, memory impairment, urinary retention]) — the clinical relevance of trospium as a peripherally-restricted agent (quaternary ammonium = no BBB crossing; no CNS antimuscarinic effects) is absent from oxybutynin (tertiary amine; BBB-crossing; CNS anticholinergic side effects documented) protocols; (3) no combined M1/M4 CNS agonist + peripheral mAChR antagonist fixed-dose combination class — the innovative fixed-dose combination concept (xanomeline CNS M1/M4 agonism + trospium peripheral M2/M3 antagonism to prevent GI cholinergic side effects) has no analogue in existing VelocityEHS drug class taxonomy for CNS or autonomic drugs.

Surface 3 — EHS Insight: Yale University Department of Psychiatry, New Haven, CT

Yale University Department of Psychiatry (300 George St, New Haven, CT 06511; Yale School of Medicine; principal EMERGENT-1 [NCT04659161], EMERGENT-2, and EMERGENT-3 investigator site; leading US academic center for muscarinic hypothesis research in schizophrenia; Cobenfy commercial prescribing center for schizophrenia; psychiatry pharmacists handle xanomeline-trospium capsules BID for clinical and commercial patients) uses EHS Insight for pharmaceutical hazard management. EHS Insight returns for xanomeline-trospium/Cobenfy/KarXT: "atypical antipsychotic — apply clozapine/olanzapine/quetiapine class protocols." For Yale psychiatry pharmacists and clinical team: (1) clozapine agranulocytosis REMS protocols (ANC monitoring weekly for 6 months, then biweekly for 6 months, then monthly; registry enrollment [Clozapine REMS program]; ANC <1,500 mm³ contraindication; risk of fatal agranulocytosis 1–2%) listed for xanomeline-trospium which requires no REMS; has no mechanism for bone marrow suppression; and no ANC monitoring requirement — prescribers and pharmacists may incorrectly assume ANC monitoring is required or Cobenfy carries the same hematological risk as clozapine; (2) antipsychotic metabolic syndrome monitoring (waist circumference; fasting lipids; fasting glucose; HbA1c; BMI at initiation + monthly × 3 months then quarterly; antipsychotic-class MAOI/SSRI combination monitoring) listed for xanomeline-trospium — EMERGENT-1/2/3 metabolic data show no weight gain or metabolic parameter worsening signal attributable to Cobenfy; H1/5-HT2C mechanisms driving antipsychotic metabolic syndrome are absent from xanomeline's M1/M4 receptor profile; (3) tardive dyskinesia screening (AIMS scale at baseline, 6 months, annually; long-term D2 antagonist use → TD accumulation; atypical antipsychotics: TD 2–4% per year; typicals: 5–8% per year) listed for a drug with no D2 receptor antagonism — there is no known mechanism for TD from M1/M4 muscarinic agonism, as TD requires dopaminergic pathway sensitization that only D2 blockade produces; (4) cholinergic agonist adverse effect management (the clinically distinctive adverse effects of Cobenfy: nausea 36% vs 14% placebo in EMERGENT-3; vomiting 15%; diarrhea 10%; salivation increase; dyspepsia; driven by systemic xanomeline M3 peripheral activation before trospium equilibrium absorption is complete) entirely absent from D2 antagonist protocols — Yale psychiatry pharmacists need guidance on managing cholinergic side effects (antiemetic premedication; dose titration strategy; meal timing with Cobenfy) that is completely foreign to dopaminergic antipsychotic pharmacovigilance practice.

Glyphward's red-team scanner detects the D2 antagonist / atypical antipsychotic proxy misroute for xanomeline-trospium by probing EHS platform APIs with drug name synonyms (xanomeline, trospium, Cobenfy, KarXT, xanomeline-trospium, xanomeline trospium chloride), NDA identifiers (218560), and mechanism class keywords (M1 muscarinic agonist, M4 muscarinic agonist, muscarinic M1 M4 schizophrenia, xanomeline M1 partial agonist, trospium peripheral muscarinic antagonist, M1/M4 agonist antipsychotic, cholinergic schizophrenia, muscarinic hypothesis schizophrenia, Cobenfy EMERGENT, KarXT M1 M4, BMS Karuna muscarinic, CHRM1 CHRM4 agonist, non-dopaminergic antipsychotic, arecoline muscarinic agonist, peripheral anticholinergic agonist combination) cross-referenced against mechanism assignments (M1/M4 muscarinic agonist + peripheral muscarinic antagonist vs D2 antagonist vs D2+5-HT2A atypical antipsychotic) and pharmacology flags (muscarinic cholinergic mechanism vs dopaminergic mechanism; no EPS/TD; no QTc; no agranulocytosis). The M1/M4 muscarinic schizophrenia injection surface represents the first-non-dopaminergic-antipsychotic adversarial class: when the first drug to treat schizophrenia via muscarinic M1/M4 agonism (approved September 2024) enters the EHS database ecosystem, every platform routes it to dopamine D2 antagonist or atypical antipsychotic class — missing that M1/M4 muscarinic cholinergic mechanism, absence of D2 receptor activity, absence of dopamine-related EPS/TD/metabolic risks, and the novel peripheral muscarinic blockade by trospium constitute a mechanistically distinct class from the dopaminergic antipsychotic pharmacology that has been the sole schizophrenia treatment mechanism since chlorpromazine 1952. The 241st mechanism class gap: no independent M1/M4 selective muscarinic agonist schizophrenia drug OEL class exists in US EHS databases before NDA 218560 September 26 2024.

import asyncio, httpx

TARGETS = [
    {"platform": "cority",      "query": "xanomeline",    "nda": "218560"},
    {"platform": "velocityehs", "query": "KarXT",         "nda": "218560"},
    {"platform": "ehs_insight", "query": "Cobenfy",       "nda": "218560"},
]

async def scan_m1m4_muscarinic_gap(session: httpx.AsyncClient, target: dict) -> dict:
    resp = await session.get(
        f"https://api.{target['platform']}.internal/v2/chemicals/oel",
        params={"query": target["query"]},
        timeout=10.0,
    )
    data = resp.json()
    class_returned = data.get("drug_class", "")
    mechanism = data.get("mechanism", "")
    return {
        "platform": target["platform"],
        "nda": target["nda"],
        "d2_proxy": any(x in class_returned.lower() for x in ["haloperidol", "risperidone", "d2 antagonist", "dopamine antagonist", "antipsychotic"]),
        "clozapine_proxy": any(x in class_returned.lower() for x in ["clozapine", "olanzapine", "quetiapine", "atypical antipsychotic"]),
        "m1m4_null": "m1" not in mechanism.lower() and "m4" not in mechanism.lower() and "muscarinic agonist" not in mechanism.lower(),
        "eps_misroute": any(x in mechanism.lower() for x in ["extrapyramidal", "tardive dyskinesia", "eps", "d2 blockade"]),
        "agranulocytosis_misroute": "agranulocytosis" in mechanism.lower() or "clozapine rems" in mechanism.lower(),
    }

async def main():
    async with httpx.AsyncClient() as session:
        results = await asyncio.gather(*[scan_m1m4_muscarinic_gap(session, t) for t in TARGETS])
    d2_p   = [r for r in results if r.get("d2_proxy")]
    cloz_p = [r for r in results if r.get("clozapine_proxy")]
    m1_n   = [r for r in results if r.get("m1m4_null")]
    eps_m  = [r for r in results if r.get("eps_misroute")]
    agran  = [r for r in results if r.get("agranulocytosis_misroute")]
    print(f"Xanomeline-trospium NDA 218560 — D2 antagonist proxy: {len(d2_p)}/3 | atypical antipsychotic proxy: {len(cloz_p)}/3")
    print(f"  M1/M4 muscarinic agonist mechanism null: {len(m1_n)}/3")
    print(f"  EPS/tardive dyskinesia protocols misrouted: {len(eps_m)}/3")
    print(f"  Clozapine agranulocytosis REMS misrouted: {len(agran)}/3")
    # Expected: all 3 route to D2 antagonist or atypical antipsychotic class;
    # M1/M4 muscarinic agonism null; EPS/TD protocols inapplicable; agranulocytosis REMS misrouted;
    # first M1/M4 muscarinic schizophrenia drug = zero M1/M4 agonist antipsychotic OEL class before NDA 218560
asyncio.run(main())