Notes
ZV0203 = pertuzumab-Pip-Val-Cit-Duostatin 5 ADC (HER2/ERBB2, DAR 2, Zhejiang Hisun/Zova). First-in-human phase 1/2a NCT05423977; phase 1 dose escalation 0.3-3.6 mg/kg IV Q3W (N=15, 3+3); phase 2a at 4.3 and 5.0 mg/kg. No DLTs, MTD not reached, formal RP2D not declared in available abstracts.
DATA CUTS: Two readouts exist. (1) ESMO 2023 438P (cut Aug 23, 2023): TRAE 11/15 (73.3%); corneal epitheliopathy 5, dry eye 4, elevated liver enzymes 6, leukopenia/neutropenia/proteinuria/anaemia 3 each. (2) Mature cut Jan 18, 2024 (ASCO 2024 e15004 / ESMO 2024 335P): TRAE 14/15 (93.3%); corneal epitheliopathy 9/15 (60%, MOST COMMON TRAE), dry eye 6/15 (40%), elevated liver enzymes 6/15 (40%), neutropenia 4/15, anaemia 4/15, leukopenia 3/15, proteinuria 3/15. Headline figures use the mature cut, which matches the curator's oae_hint (corneal 60%, dry eye 40%); the input hint is CONFIRMED CORRECT (an earlier search of the Aug-2023 cut had transiently suggested corneal 5/15, but the later cut supersedes it).
OCULAR: Corneal epitheliopathy is the dominant/characteristic finding (classic off-target auristatin keratopathy), plus dry eye and blurred vision -> ocular_surface_subtype 'Corneal (off-target)'. oae_any_pct=60 (corneal, max ocular PT). Grade 3 ocular: corneal epitheliopathy 1/15 (6.7%, at 3.6 mg/kg) + blurred vision 3/15 (20%, 2 at 2.7 + 1 at 3.6 mg/kg). oae_g3plus_pct reported as the max single G3 ocular PT (blurred vision 20%); total distinct G3-ocular patients <=4/15 (26.7%) but a patient at 3.6 mg/kg may have both events (possible overlap), so 20% is the conservative value. 2/15 (13.3%) discontinued due to TRAEs.
NOT CAPTURED IN ROWS: proteinuria 3/15 (20%, renal) — renal is outside the organ_system enum, recorded here only.
PAYLOAD CAVEAT: PubChem 'Duostatin 5' (CID 138320017, C40H69N9O6, MW 772.0, xLogP3 5.5, TPSA 161, formal charge 0) carries an aminophenyl + azidomethyl terminus (likely a synthetic/conjugation handle) and lacks the free C-terminal carboxylate of parent MMAF; logD7.4 and pKa are not PubChem-computed. payload_charge_pH7_4 (+1) and bystander status are estimates/uncertain.
ISOTYPE: ADCdb lists antibody isotype 'not disclosed' for the ADC; humanized IgG1-kappa is inferred from the pertuzumab backbone.
DEDUPE FLAG (from input): a near-duplicate-looking entry 'anti-HER2-vc-Duostatin-5 site-specific' may exist in known_names but appears to be a generic/preclinical anti-HER2 research construct, NOT this specific pertuzumab-based clinical ADC — keep separate / verify before merging.
CONFIDENCE: All clinical/dosing/ocular/toxicity data are conference-abstract level (tier D); no full peer-reviewed primary publication located. Composition/linker (ADCdb) and payload physchem (PubChem computed) are tier C.