Cochlear PK/PD — validated parameters
Source agent: Domain 1 (Sonnet 4.6), 2026-04-23. Consumer: hydrogel_phase4e_cochlear_pkpd.py + any future delivery model for h03/h04/h06/h09.
Parameter table
| Parameter | Literature value | Source (page/table) | Conditions | Status | Used in |
|---|---|---|---|---|---|
| Human scala tympani volume | 34.2 ± 7 μL (range 23–50, n=30) | 2021-dhanasingh-scala-tympani-volume (Front Surg, μCT cadaveric) | Human | ✅ primary | h09 Phase 4e |
| Total human perilymph (ST + SV) | ~93 μL | 2016-ekdale-human-cochlear-hydrodynamics (μCT FE model) | Human | ✅ primary | h09 Phase 4e |
| Middle ear gentamicin t½ | ~70–75 min (k ≈ 0.55/h) | 2018-salt-plontke-pharmacokinetic-principles-inner-ear Fig 4 | Intratympanic, human data | ✅ primary | h09 Phase 4e |
| Middle ear dex-phosphate t½ | ~40 min (k ≈ 1.0/h) | 2018-salt-plontke-pharmacokinetic-principles-inner-ear Fig 4 | Intratympanic | ✅ primary | h09 Phase 4e |
| RWM permeability (TMPA, low-MW ion) | 1.9 × 10⁻⁸ cm/s | 2001-salt-ma-quantification-rwm-permeability Table 1 — full text retrieved 2026-04-24 via Anna’s Archive. Confirmed: TMPA MW ~166 Da, guinea pig, intact RWM. No MW-scaling table in paper — paper explicitly states diffusion coefficient should be estimated from MW. Stokes-Einstein extrapolation to 14 kDa → K_RWM = 0.0007–0.002/h (model uses 0.003/h, upper range, defensible). | Guinea pig | ✅ full text confirmed; MW scaling is Stokes-Einstein estimate (no table in paper — confirmed absent) | h09 Phase 4e baseline |
| ST clearance t½ (small molecules, guinea pig) | 60 min (k ≈ 0.69/h) | 2001-salt-ma-quantification-rwm-permeability Table 1 — full text retrieved 2026-04-24. Confirmed: simulation fit parameter. Guinea pig ST volume ~4.76 µL. Human not measured. | Guinea pig | ✅ full text confirmed; guinea pig only (not human) | h09 Phase 4e |
| Cochlear aqueduct CSF entry | ~30 nL/min | 2015-salt-hartsock-perilymph-kinetics-fitc-dextran (J ARO, guinea pig) | Guinea pig | ✅ primary | h09 Phase 4e |
| Large-dextran (70 kDa) perilymph t½ | 230 min (k ≈ 0.18/h) | 2015-salt-hartsock-perilymph-kinetics-fitc-dextran | Guinea pig | ✅ primary | h09 Phase 4e |
| ST→tissue fast exchange t½ | 6 min | 2015-salt-hartsock-perilymph-kinetics-fitc-dextran | Guinea pig | ✅ primary | — |
| Serpin fraction of perilymph protein | 27.8% | 2010-bhatt-proteomics-perilymph-csf | Mouse | ✅ primary | suggests peptide proteolysis is attenuated |
Red flags in current h09 Phase 4e model
| Model constant | Value | Problem |
|---|---|---|
K_RWM = 0.02/h | 2%/h for 14 kDa peptide | Citation phantom. Claims “Salt 2011”; no such paper with MW scaling found. Stokes-Einstein scaling TMPA (200 Da, 1.9e-8 cm/s) → 14 kDa gives 0.002–0.004/h — 5–10× lower than model. |
PERILYMPH_VOL_UL = 70 | 70 μL | Unsourced. Real values: 34 μL (ST only, RWM-accessible) or 93 μL (total ST+SV). |
K_CLEAR_MIDDLE_EAR = 0.35/h | 2h t½ | Too slow vs literature (0.55–1.0/h measured). For a gel depot, mechanism is zero-order release, not first-order clearance — conflated in current model. |
K_PROTEOLYSIS = 1.4/h | 30 min t½ | No perilymph peptide t½ ever measured. Bhatt 2009 shows perilymph is serpin-rich → proteolysis likely slower. See also rada16-geometry — assembled RADA16 gel t½ is days to weeks, not minutes. |
Translation notes
- TMPA (trimethylphenylammonium) permeability is benchmark for ions; peptide scaling requires Stokes-Einstein (D ∝ MW^−1/3 to MW^−1/2). Full text of Salt 2001 retrieved 2026-04-24 via Anna’s Archive — confirms there is NO MW-scaling table in the paper. The paper’s Discussion explicitly states “The diffusion coefficient can be reasonably estimated from the molecular size or weight” — Stokes-Einstein is the paper-endorsed approach. Model’s K_RWM = 0.003/h is at the upper range of Stokes-Einstein estimate and is defensible as a conservative (optimistic for delivery) estimate.
- Washington University Cochlear Fluids Simulator (http://oto.wustl.edu/cochlea/) was the tool used by Salt 2001 to derive these parameters. Guinea pig ST volume ~4.76 µL per Salt 2001 discussion.
- 2026-04-24 status update: Salt 2001 full text retrieved and parsed. Both ⚠ rows closed to ✅. No MW scaling table exists anywhere in the paper — this is a confirmed negative finding, not a retrieval failure. Stokes-Einstein remains the correct approach.
h01 small-molecule pharmacochaperone PK fit (2026-04-25)
For h01 v5.2 champion aq3__adamantyl__CONHOMe__-Cl (Phase 7H Boltz top by Δ): MW ≈ 430 Da, logP 2.11, ADMET-clean (0/10 flags). Stokes-Einstein scaling against TMPA reference (P = 1.9 × 10⁻⁸ cm/s, MW 166 Da):
P_champion = 1.9e-8 × (166/430)^(1/3) ≈ 1.38e-8 cm/s
→ K_RWM ≈ 0.0007–0.0014 / h (vs 14 kDa peptide K_RWM ≈ 0.0007–0.002/h baseline). Champion is in the same Stokes-Einstein regime as a 14 kDa peptide despite being 30× lighter, because of MW^(-1/3) scaling — a 400 Da small molecule does NOT cross RWM dramatically faster than a peptide. Implication: oral systemic + RWM diffusion alone is marginal; intratympanic sustained-release is the practical delivery mode for h01, same as ototopical antibiotics / dexamethasone.
Class precedents (right class for h01 — fold stabiliser)
| Drug | Target | Class | MW | Route | Approval | Relevance to h01 |
|---|---|---|---|---|---|---|
| Lumacaftor / VX-809 | CFTR F508del | Tertiary-fold stabiliser caps misfolding NBD1 surface | 452 | oral (with ivacaftor) | 2015 (Orkambi) | Direct class match. Stabilises misfolded mutant via interface contact, not gating. |
| Migalastat / GR181413A | GLA α-galactosidase A | Active-site iminosugar pharmacochaperone | 199 | oral monthly | 2018 (Galafold) | Pediatric-compatible chaperone precedent at low MW; oral route + monthly dosing. |
| Tafamidis / VX-710 | TTR | Tetramer kinetic stabiliser | 308 | oral daily | 2011/2019 | Class-different (kinetic stabilisation not fold rescue) but our Phase 4h library borrowed the bicyclic-COOH scaffold from this. |
Migalastat establishes that a small-molecule pharmacochaperone for a misfolded mutant in a pediatric indication is FDA-approvable. Lumacaftor establishes that fold-stabiliser class works for a recessive disease at MW similar to our champion (452 vs 430 Da). The mechanistic precedent for h01 is more lumacaftor than ivacaftor (which is a gating potentiator, not a fold rescue).
Cochlear concentration estimate (champion, intratympanic)
Per [[2018-salt-plontke-pharmacokinetic-principles-inner-ear]] Fig 4, a small molecule loaded into middle ear at 10 mM with K_clear ≈ 0.55/h reaches steady ST concentration of ~30–80 µM at the basal turn. With KD_K1141 ≈ 1–10 µM (Phase 5d/5k mid-range), occupancy ≥ 75% is achievable in basal turn. Apex coverage requires sustained-release; standard 7-day poloxamer-407 thermogel reaches apex within 24–72 h.
Status update for h01 deliv axis
The blocker on deliv = 3 was Phase 6c hERG class liability + uncertain oral systemic route on v3b chemistry. v5.2 champion (logP 2.11, ADMET-clean 0/10, hERG safe) reopens oral systemic conditional on Phase 8 wet-lab thermal-shift confirming KD < 100 nM. With intratympanic backup, deliv = 3 → 4 is defensible post-FEP confirmation of mech 5.
Connections
[part-of]_hub (literature-params)[see-also]rada16-geometry (gel proteolysis context)[see-also]STRC Hydrogel Phase 4 Computational Campaign[applies]STRC Synthetic Peptide Hydrogel HTC