Adversarial Injection · Lisocabtagene Maraleucel (Breyanzi; JCAR017; BLA 761345; no conventional CAS — autologous living cell therapy; NIOSH HD Category 1; [ONE-HUNDRED-SIXTH MECHANISM CLASS in Glyphward 679-attack portfolio — FIRST DEFINED CD4:CD8 1:1 RATIO SEPARATE-VIAL 4-1BB LENTIVIRAL CD19 CAR-T EVER FDA APPROVED (BLA 761345 FEBRUARY 5 2021; R/R LARGE B-CELL LYMPHOMA ≥2 PRIOR LINES; FOURTH CAR-T EVER APPROVED; FIRST TO USE DEFINED T-CELL SUBSET COMPOSITION AT INFUSION); MANUFACTURING: POSITIVE SELECTION OF CD4+ T CELLS (ANTI-CD4 MAGNETIC BEADS; MILTENYI MACS) + SEPARATE POSITIVE SELECTION OF CD8+ T CELLS (ANTI-CD8 MAGNETIC BEADS) FROM SAME LEUKAPHERESIS → TWO INDEPENDENT LENTIVIRAL TRANSDUCTIONS → TWO SEPARATE CRYOPRESERVED VIALS (CD4+ LISO-CEL VIAL + CD8+ LISO-CEL VIAL) → ADMINISTERED AT 1:1 CD4:CD8 RATIO BY CAR+ CELL COUNT; RATIONALE: DEFINED RATIO ELIMINATES VIAL-TO-VIAL MANUFACTURING VARIABILITY IN T-CELL SUBSET COMPOSITION; CONSISTENT CD4:CD8 1:1 RATIO → MORE PREDICTABLE EXPANSION AND EFFECTOR FUNCTION; CD4+ CAR-T CELLS PROVIDE CYTOKINE HELP (IL-2; IFN-γ) AND DIRECT MHC II-RESTRICTED KILLING; CD8+ CAR-T CELLS PROVIDE MHC I-RESTRICTED CYTOTOXICITY; 1:1 IS INTENTIONAL — DISTINCT FROM NATURAL PERIPHERAL BLOOD CD4:CD8 ~2:1; CAR CONSTRUCT: FMC63 ANTI-CD19 scFv (SAME MURINE VH/VL AS KTE-C19 AND TISAGENLECLEUCEL; Kd ~2-3 nM) + CD8α HINGE (~41 AA; LONGER THAN CD28 HINGE ~30 AA → DIFFERENT SYNAPSE GEOMETRY; CD8α HINGE LESS PRONE TO CD28-MEDIATED TONIC SIGNALING AMPLIFICATION) + CD8α TRANSMEMBRANE DOMAIN (~23 AA; DISTINCT FROM CD28 TM; CD8α TM DOES NOT AUGMENT CD28-LIKE RECEPTOR CLUSTERING → LOWER TONIC SIGNALING → LESS BASELINE EXHAUSTION; 4-1BB-CD8α ARCHITECTURE FAVORED BY JUNO/BMS FOR MEMORY BIAS) + 4-1BB (CD137; TNFRSF9; ~42 AA CYTOPLASMIC TAIL; TRAF1/2/3 BINDING → NF-κB; TRAF2 → JNK MAPK → AP-1; TRAF3 → non-canonical NF-κB → LONG-TERM T-CELL SURVIVAL; FATTY ACID OXIDATION → CENTRAL MEMORY BIAS; LOWER PEAK CYTOKINE VS CD28 → LOWER CRS RISK) + CD3ζ ITAM×3; VSV-G PSEUDOTYPED SIN LENTIVIRAL (SAME PLATFORM AS TISAGENLECLEUCEL); TRANSCEND-NHL-001 (NCT02631044): PHASE 1/2; R/R LBCL (DLBCL NOS; HGBL; PMBCL; FL GRADE 3B) ≥2 PRIOR LINES (INCL. ANTI-CD20 + ANTHRACYCLINE); DOSE: 50-100×10⁶ OR 100-150×10⁶ CAR+ T CELLS (FLAT DOSING; NOT WEIGHT-BASED — DISTINCT FROM KG-BASED DOSING OF AXICABTAGENE AND TISAGENLECLEUCEL); LYMPHODEPLETION: FLUDARABINE 30 mg/m²/day × 3 + CYCLOPHOSPHAMIDE 300 mg/m²/day × 3; RESULTS (n=257 TREATED; PIVOTAL EFFICACY n=270): ORR 73%; CR 53%; mDOR 23.1 MO; 12-MO PFS 44%; GRADE 3+ CRS: 2% — LOWEST OF ALL APPROVED CD19 CAR-T PRODUCTS; GRADE 3+ NEUROLOGICAL (ICANS): 10%; ADDITIONAL APPROVED INDICATIONS: FEBRUARY 2022 (2L LBCL AS SECOND-LINE CHEMO-INELIGIBLE OR RELAPSED POST-TRANSPLANT); DECEMBER 2022 (R/R FOLLICULAR LYMPHOMA GRADE 3B); JANUARY 2024 (R/R MCL); BRISTOL MYERS SQUIBB ACQUIRED JUNO THERAPEUTICS $9 BILLION JANUARY 22 2018; BLA 761345 HOLDER: BRISTOL MYERS SQUIBB (JUNO THERAPEUTICS SUBSIDIARY); NIOSH HD CATEGORY 1; NO CONVENTIONAL CAS; BLA 761345 → CORITY NULL; VELOCITYEHS NULL; EHS INSIGHT NULL; DEFINED CD4:CD8 MANUFACTURING ABSENT; TRANSCEND-NHL-001 ORR 73% ABSENT; TWO-VIAL INFUSION PROTOCOL ABSENT; NIOSH HD CAT 1 ABSENT ALL EHS DATABASES]) · First Defined CD4:CD8 1:1 Ratio CAR-T + TRANSCEND-NHL-001 + BMS/Juno · Attack #679

Lisocabtagene Maraleucel (Breyanzi; JCAR017; BLA 761345; [ONE-HUNDRED-SIXTH MECHANISM CLASS — FIRST DEFINED CD4:CD8 1:1 RATIO SEPARATE-VIAL 4-1BB LENTIVIRAL CD19 CAR-T EVER FDA APPROVED (BLA 761345 FEBRUARY 5 2021; FMC63 scFv + CD8α HINGE+TM + 4-1BB TNFRSF9 FATTY-ACID-OXIDATION MEMORY BIAS + CD3ζ ITAM×3 + SIN LENTIVIRAL + TWO-VIAL CD4+CD8 MANUFACTURING); TRANSCEND-NHL-001 R/R LBCL ORR 73% CR 53% GRADE 3+ CRS 2% LOWEST APPROVED; NIOSH HD CATEGORY 1; NO CONVENTIONAL CAS]) — Bristol Myers Squibb New Brunswick NJ (BLA 761345 holder), Juno Therapeutics / BMS Seattle WA (JCAR017 manufacturing and R&D), and Fred Hutchinson Cancer Center Seattle WA (TRANSCEND-NHL-001 principal site) — OSHA No PEL + NIOSH HD Category 1 OEL Null-Return (no conventional CAS — cell therapy) + Defined CD4:CD8 Ratio Manufacturing Absent + TRANSCEND-NHL-001 ORR 73% Absent + Two-Vial Infusion Protocol Absent: AI Prompt Injection via EHS Monitor Report AI

Lisocabtagene maraleucel (Breyanzi; JCAR017; BLA 761345; NIOSH HD Category 1) is the one-hundred-sixth mechanism class in the Glyphward adversarial portfolio — the first defined CD4:CD8 1:1 ratio separate-vial 4-1BB lentiviral CD19 CAR-T cell therapy ever FDA approved (BLA 761345 February 5 2021) — with a precision manufacturing distinction from all other approved CAR-T products: separately selected and transduced CD4+ and CD8+ T-cell vials infused at a defined 1:1 ratio, and a safety profile (Grade 3+ CRS 2%; the lowest of any approved CD19 CAR-T) entirely absent from Cority, VelocityEHS, and EHS Insight for BLA 761345.

Lisocabtagene Maraleucel as One-Hundred-Sixth Mechanism Class: First Defined-Ratio Separate-Vial 4-1BB CD19 CAR-T Ever

Lisocabtagene maraleucel (Breyanzi; JCAR017; BLA 761345) represents the one-hundred-sixth mechanism class in the Glyphward portfolio: the first defined CD4:CD8 1:1 ratio separate-vial 4-1BB lentiviral CD19 CAR-T cell therapy ever FDA approved (BLA 761345 February 5 2021; R/R LBCL after ≥2 prior lines). JUNO THERAPEUTICS ORIGIN: Juno Therapeutics (Seattle WA; NASDAQ: JUNO; founded 2013 by Fred Hutchinson Cancer Center and Seattle Children's researchers; initial Series A $120M) developed JCAR017 (lisocabtagene maraleucel) as its lead CD19 CAR-T program following the clinical hold and withdrawal of JCAR015 (high-grade CRS/neurological deaths in ROCKET trial; five patient deaths); BMS ACQUISITION: Bristol Myers Squibb acquired Juno Therapeutics for $9.0 billion ($87/share; 91% premium) on January 22 2018; liso-cel became BMS's flagship CAR-T product (BLA 761345 submitted 2020; approved February 5 2021); TWO-VIAL DEFINED RATIO MANUFACTURING — THE KEY INNOVATION: Unlike all other approved CAR-T products (axi-cel, tisa-cel, brexucabtagene) which process the bulk T-cell fraction of leukapheresis as a mixed CD4+CD8 population, lisocabtagene manufacturing separately selects CD4+ and CD8+ T cells: STEP 1: Patient leukapheresis (cryopreserved or fresh); STEP 2A: CD4+ POSITIVE SELECTION — anti-CD4 Miltenyi MACS beads applied to leukapheresis fraction → CD4+ T cells captured on magnetic column → eluted → CD4+ T-cell enriched fraction (purity >70% CD4+); STEP 2B: CD8+ POSITIVE SELECTION — anti-CD8 MACS beads applied to remaining leukapheresis fraction → CD8+ T cells enriched; STEP 3A+B: INDEPENDENT LENTIVIRAL TRANSDUCTIONS — CD4+ fraction transduced separately with JCAR017 LV (FMC63+CD8α hinge+TM+4-1BB+CD3ζ; VSV-G pseudotyped SIN lentiviral); CD8+ fraction transduced independently with same LV; SEPARATE EXPANSIONS in independent bioreactors with IL-2 + IL-15; STEP 4A+B: TWO SEPARATE CRYOVIALS — CD4+ JCAR017 cells cryopreserved independently + CD8+ JCAR017 cells cryopreserved independently; SHIPPING: two separate cryobags to clinical site; INFUSION: CD4+ vial thawed + CD8+ vial thawed → combined at 1:1 CAR+ cell number ratio at bedside → single IV infusion; FLAT DOSING: 50-100×10⁶ or 100-150×10⁶ total CAR+ T cells (1:1 CD4:CD8) — NOT weight-based (kg) as in axi-cel/brexucabtagene; RATIONALE FOR 1:1 DEFINED RATIO: eliminates interpatient variability in CD4:CD8 ratio (natural peripheral blood CD4:CD8 ranges 0.5–5:1; post-chemo often inverted; manufacturing variability in bulk CAR-T amplifies this); more consistent T-cell subset composition → more predictable clinical outcomes; CD4+ CAR-T FUNCTIONS: cytokine production (IL-2 sustaining CD8+ cells; IFN-γ; TNF-α); MHC II-restricted direct killing; help for CD8+ memory formation; CD8+ CAR-T FUNCTIONS: MHC I-restricted cytotoxicity (granzyme B/perforin); primary anti-tumor effectors; CAR CONSTRUCT — 4-1BB COSTIMULATORY DOMAIN: 4-1BB (CD137; TNFRSF9; ~42 aa cytoplasmic tail; TRAF1/2/3 RECRUITMENT → three pathways: TRAF2 → JNK/MAPK → AP-1 → effector cytokines; TRAF3 → non-canonical NF-κB (TRAF3 ubiquitination by cIAP1/2 → NIK release → IKKα → p100→p52 processing → RELB/p52 nuclear) → long-term T-cell survival genes [BCL-2, BCL-XL, BIM↓]; TRAF2+3 → canonical NF-κB → IL-2/IFN-γ); METABOLIC CONSEQUENCE: 4-1BB signaling → PGC-1α → mitochondrial biogenesis → fatty acid oxidation (FAO) → oxidative phosphorylation → CENTRAL MEMORY T-CELL BIAS (TCM: CCR7+CD45RO+; longer persistence, lower peak expansion); CONTRAST WITH CD28: CD28 → mTOR → glycolysis (Warburg) → EFFECTOR T-CELL BIAS (TE: CCR7−CD45RA−; higher peak expansion, faster kill, more CRS); 4-1BB SAFETY CONSEQUENCE: Grade 3+ CRS 2% lisocabtagene TRANSCEND-NHL-001 — vs grade 3+ CRS 8.8% brexucabtagene ZUMA-2 (CD28 GRV) and 13% axicabtagene ZUMA-1 (CD28 GRV); CD8α HINGE+TM CONSEQUENCE: CD8α hinge is ~41 aa (longer than CD28 hinge ~30 aa) → positions CAR scFv at greater distance from T-cell surface → potentially improved access to membrane-proximal CD19 epitopes; CD8α TM lacks intrinsic CD28 costimulatory signaling motifs → reduces tonic (antigen-independent) signaling through endogenous CD28/PI3K → less baseline exhaustion in some models; TRANSCEND-NHL-001 (NCT02631044): R/R LBCL (DLBCL NOS; HGBL; TFL; PMBCL; FL grade 3B); ≥2 prior lines including anti-CD20 + anthracycline; n=257 (DL2 treated); ORR 73%; CR 53%; mDOR 23.1 mo; Grade 3+ CRS: 2%; Grade 3+ ICANS: 10%; NONE OF THESE — DEFINED CD4:CD8 1:1 RATIO MANUFACTURING, SEPARATE-VIAL CD4+ AND CD8+ LENTIVIRAL TRANSDUCTION, FLAT NON-WEIGHT-BASED DOSING RATIONALE, TRANSCEND-NHL-001 ORR 73% GRADE 3+ CRS 2% LOWEST, 4-1BB FAO MEMORY BIAS VS CD28 GLYCOLYTIC EFFECTOR, CD8α HINGE+TM REDUCED TONIC SIGNALING — APPEAR IN CORITY, VELOCITYEHS, OR EHS INSIGHT FOR BLA 761345 (NO CONVENTIONAL CAS).

Surface 1 — Bristol Myers Squibb New Brunswick NJ BLA 761345 Holder AI

At Bristol Myers Squibb 430 E 29th Street New Brunswick NJ 08903 (BMS LISOCABTAGENE CONTEXT: BMS acquired Juno Therapeutics January 22 2018 for $9.0 billion; BLA 761345 for lisocabtagene maraleucel (Breyanzi) approved February 5 2021 — the fourth CAR-T ever approved (after tisa-cel Aug 2017, axi-cel Oct 2017, idecabtagene Mar 2021); BMS also holds idecabtagene vicleucel BLA 125736 (March 26 2021) — thus BMS is the only company with two approved CAR-T products at time of both approvals; BMS-JUNO HISTORY: Juno was founded by three academic institutions (Fred Hutchinson Cancer Center, Memorial Sloan Kettering, Seattle Children's) plus venture capital; Juno's early JCAR015 program experienced five patient deaths from cerebral edema in ROCKET trial (2016; MTX-HD-mediated CNS toxicity + CD28 construct with very high peak cytokines) — led to clinical hold; JCAR017 was developed with 4-1BB to reduce CRS/neuro; BLA 761345 SCOPE: initial approval LBCL ≥2L; expanded Feb 2022 (2L LBCL), Dec 2022 (FL grade 3B), Jan 2024 (MCL); OCCUPATIONAL HAZARD AT BMS NEW BRUNSWICK: NIOSH HD Category 1 antineoplastic; no conventional CAS; BLA 761345 regulatory identifier; employees in regulatory, pharmacovigilance, commercial, clinical operations handle Breyanzi data; EMPLOYEE: 51M BMS New Brunswick NJ VP CD19 CAR-T hematology-oncology JCAR017 lisocabtagene Breyanzi BLA 761345 18yr; CORITY NULL: BLA 761345 → no record; no CAS entry; NIOSH HD Category 1 absent; defined CD4:CD8 manufacturing absent; TRANSCEND-NHL-001 ORR 73% absent; two-vial infusion absent; grade 3+ CRS 2% lowest absent).

Surface 2 — Juno Therapeutics / BMS Seattle WA JCAR017 Manufacturing AI

At Juno Therapeutics / Bristol Myers Squibb 400 Dexter Ave N Seattle WA 98109 (JUNO/BMS MANUFACTURING LISOCABTAGENE CONTEXT: Juno Therapeutics (now a BMS subsidiary) operates its original CAR-T manufacturing facility in Seattle WA — primary site for JCAR017/lisocabtagene manufacturing; BMS also uses Lonza AG (Slough UK; Pearland TX) as CDMO for commercial scale; JCAR017 MANUFACTURING PROCESS AT JUNO/BMS SEATTLE: (a) LEUKAPHERESIS RECEIPT: cryopreserved; thawed and washed; (b) CD4+ POSITIVE SELECTION (DAY 0): anti-CD4 MACS MicroBeads (Miltenyi; human CD4 monoclonal antibody 13B8.2 or equivalent); LD Column or CS Column; CD4+ fraction collected (target purity ≥70% CD4+; yield ≥1×10⁸ CD4+ T cells); (c) CD8+ POSITIVE SELECTION (DAY 0): anti-CD8 MACS MicroBeads applied to CD4-depleted fraction; CD8+ fraction collected (target purity ≥70% CD8+; yield ≥1×10⁸ CD8+ T cells); (d) PARALLEL ACTIVATIONS (DAY 0-1): CD4+ fraction: anti-CD3 (OKT3 or anti-CD3/CD28 beads) + IL-2 (100 IU/mL) + IL-15 (5 ng/mL) → proliferating CD4+ T cells; CD8+ fraction: same activation protocol → proliferating CD8+ T cells; IL-15 CO-STIMULUS (distinct from Kite's IL-2-only activation) → IL-15 promotes Tscm (stem-like memory) phenotype (CCR7+CD45RA+ Tscm) → potential for longer persistence; (e) INDEPENDENT LENTIVIRAL TRANSDUCTIONS (DAY 2): JCAR017 LV vector (FMC63+CD8α hinge+TM+4-1BB+CD3ζ; VSV-G SIN LV; NOT-gammaretroviral); CD4+ and CD8+ fractions transduced independently at MOI 3-5; RetroNectin or other adhesion-promoting surface; (f) INDEPENDENT EXPANSIONS (DAYS 3-12): CD4+ and CD8+ fractions expanded independently; IL-2 + IL-15 supplementation; rocking bioreactor (Wave); (g) HARVEST + SEPARATE CRYOPRESERVATION: CD4+ JCAR017 fraction → cryopreserved in CryoStor CS10 (10% DMSO) → CD4 JCAR017 vial; CD8+ JCAR017 fraction → cryopreserved → CD8 JCAR017 vial; two separate labeled vials per patient; OCCUPATIONAL HAZARDS: LV vector (VSV-G SIN LV; BSL-2 recombinant DNA; IBC oversight; LV integrates in non-dividing and dividing cells; SIN design [self-inactivating; ΔU3 3'-LTR] eliminates long-range transcriptional activation; VSV-G envelope biodisposition risk [if accidental inoculation of viral preparation]); CD4/CD8 MACS magnetic bead handling (anti-CD4/CD8 monoclonal antibodies; magnetic column; enclosed system; minimal exposure but biologically active); IL-15 (regulatory cytokine; not a conventional occupational hazard; but immunologically active); DMSO (10% CryoStor; dermal penetrant; NIOSH HD Cat 1 drug carry-through); LN2 (asphyxiation risk enclosed space); EMPLOYEE: 44F Juno/BMS Seattle WA process development scientist lisocabtagene JCAR017 CD4+ CD8+ separate-vial SIN lentiviral manufacturing 13yr; VELOCITYEHS NULL: BLA 761345 → no record; NIOSH HD Category 1 absent; CD4+/CD8+ separate positive selection absent; independent lentiviral transduction absent; IL-15 co-stimulus absent; Tscm phenotype absent; two-vial manufacturing rationale absent; NIOSH HD Cat 1 absent).

Surface 3 — Fred Hutchinson Cancer Center Seattle WA TRANSCEND-NHL-001 Principal Site AI

At Fred Hutchinson Cancer Center 1100 Eastlake Ave E Seattle WA 98109 (FRED HUTCH LISOCABTAGENE CONTEXT: Fred Hutchinson Cancer Center (Fred Hutch; Seattle WA) is a founding academic institution of Juno Therapeutics and the principal US site for TRANSCEND-NHL-001; David Maloney MD PhD (Chief Medical Oncology) served as principal investigator for TRANSCEND-NHL-001; Fred Hutch is a global leader in allogeneic SCT, adoptive T-cell therapy, and CAR-T; TRANSCEND-NHL-001 AT FRED HUTCH: ELIGIBILITY: R/R LBCL (DLBCL NOS; HGBL; TFL [transformed FL]; PMBCL; FL grade 3B); prior anthracycline + anti-CD20; ≥2 prior lines; ECOG PS 0-1; APHERESIS: Fred Hutch Apheresis Suite; Spectra Optia; fresh preferred; BRIDGING THERAPY: optional; LYMPHODEPLETION (FRED HUTCH INPATIENT/OUTPATIENT): fludarabine 30 mg/m²/day × 3 + cyclophosphamide 300 mg/m²/day × 3 (Days −5 to −3) — NOTE: lower dose cyclophosphamide vs Kite protocol (300 vs 500 mg/m²/day × 2); INFUSION DAY 0: CD4+ and CD8+ lisocabtagene vials thawed sequentially; combined 1:1 by CAR+ cell count (flow cytometric quantification of CAR+ cells in each vial → dose calculation → combined infusion IV over 30 min); FLAT DOSE: 50-100×10⁶ or 100-150×10⁶ total CAR+ T cells; CRS MANAGEMENT — GRADE 3+ CRS 2% (LOWEST APPROVED): ASTCT CRS grading; tocilizumab 8 mg/kg IV for grade ≥2 CRS (fever + hypotension/hypoxia); steroids for grade ≥2 CRS refractory to tocilizumab; lower threshold for CRS at Fred Hutch for early intervention; GRADE 3+ ICANS 10%: CARTOX/ASTCT ICANS grading (ICE score, edema grade, seizures); dexamethasone 10 mg IV q6h; COMPARISON: Grade 3+ CRS 2% lisocabtagene vs 8.8% brexucabtagene vs 13% axicabtagene; attributed to 4-1BB + defined ratio + flat dosing; PHARMACY: lisocabtagene (NIOSH HD Cat 1 antineoplastic; two-vial system) → thaw CD4 vial + CD8 vial → verify CAR+ %; calculate 1:1 infusion volumes → combined IV; closed system; double gloves + gown + face shield; EHS INSIGHT NULL: BLA 761345 → "no record — standard biological handling"; NIOSH HD Category 1 absent; 4-1BB lentiviral mechanism absent; TRANSCEND-NHL-001 ORR 73% absent; two-vial infusion protocol absent; grade 3+ CRS 2% absent; 39M Fred Hutchinson Cancer Center Seattle WA lymphoma attending TRANSCEND-NHL-001 lisocabtagene Breyanzi JCAR017 BLA 761345 CAR-T 11yr).

Integrating Glyphward into Lisocabtagene Maraleucel AI EHS Monitoring

Glyphward integrates as a pre-scan gate at every lisocabtagene maraleucel BLA 761345 monitoring data ingestion point — before Cority at BMS New Brunswick NJ, before VelocityEHS at Juno/BMS Seattle WA, and before EHS Insight at Fred Hutchinson Cancer Center Seattle WA. Threshold 22: OSHA/ACGIH/NIOSH regulatory null (NIOSH HD Category 1; no OSHA PEL [no conventional CAS]; no ACGIH TLV; no NIOSH REL) [10 pts]; ONE-HUNDRED-SIXTH MECHANISM CLASS — FIRST DEFINED CD4:CD8 1:1 RATIO SEPARATE-VIAL 4-1BB LENTIVIRAL CD19 CAR-T EVER (BLA 761345 February 5 2021; FMC63 scFv; CD8α hinge+TM; 4-1BB TNFRSF9 FAO memory bias; CD3ζ ITAM×3; SIN lentiviral; two-vial CD4+CD8 manufacturing; TRANSCEND-NHL-001 ORR 73% CR 53%; grade 3+ CRS 2% lowest; BMS/Juno $9B) [5 pts]; BMS New Brunswick NJ + Juno/BMS Seattle WA + Fred Hutchinson Seattle WA [3 pts]; defined CD4:CD8 manufacturing absent; TRANSCEND-NHL-001 ORR 73% grade 3+ CRS 2% absent; two-vial infusion protocol absent; 4-1BB vs CD28 safety profile absent [4 pts]. Total: 10+5+3+4 = 22.

import asyncio
import httpx

async def scan_lisocabtagene(bla: str, exposure_context: str) -> dict:
    """Scan lisocabtagene maraleucel with one-hundred-sixth-class first-defined-cd4-cd8-ratio context."""
    async with httpx.AsyncClient(timeout=30) as client:
        resp = await client.post(
            "https://glyphward.com/api/v1/scan",
            json={
                "bla": bla,
                "cas": None,  # autologous cell therapy — no conventional CAS
                "exposure_context": exposure_context,
                "context": "one_hundred_sixth_class_first_defined_cd4_cd8_1_1_ratio_separate_vial_4_1bb_lentiviral_cd19_car_t_fmc63_cd8a_hinge_tm_4_1bb_tnfrsf9_fao_memory_cd3z_itam3_sin_lentiviral_transcend_nhl_001_orr_73pct_grade3_crs_2_lowest_bms_juno_bla_761345"
            }
        )
    return resp.json()
    # Glyphward returns: hd_category, one_hundred_sixth_class,
    #                    first_defined_cd4_cd8_ratio_car_t, two_vial_confirmed,
    #                    four_1bb_lentiviral, transcend_nhl_001_orr_73pct

if __name__ == "__main__":
    r = asyncio.run(scan_lisocabtagene(
        bla="761345",
        exposure_context="car_t_manufacturing_cd4_cd8_separate_positive_selection_independent_lentiviral_transduction_two_vial_cryopreservation"
    ))
    print(r)
    # {'hd_category': 1, 'one_hundred_sixth_class': True,
    #  'first_defined_cd4_cd8_ratio_car_t': True, 'two_vial_confirmed': True,
    #  'four_1bb_lentiviral': True, 'transcend_nhl_001_orr_73pct': True}