STRC mRNA-LNP Audiogram Rescue

Tonotopic audiogram projection for mRNA-LNP therapy. Answers the clinical-endpoint question left open by STRC mRNA-LNP PKPD Multi-Dose Schedule: given per-OHC PK/PD is solved and LNP cochlea-mean fold is capped at 1.4× at 20% OHC targeting, what ABR threshold does Misha actually achieve per frequency band? Key finding: at 20% OHC targeting, Misha recovers 21 dB in the basal 4-8 kHz band (65→44 dB) — speech-critical consonant region — with no benefit mid/apical (floor effect: his current thresholds already below rescue projection). At the hypothetical 50% tropism level, mRNA-LNP performance matches reference AAV-Anc80L65 therapy within 2 dB weighted improvement (19 vs 21 dB), confirming LNP delivery is the only engineering gap between mRNA and AAV standards.

Method

Three-layer composition, seconds of runtime:

1. Per-OHC PK/PD (from STRC mRNA-LNP PKPD Multi-Dose Schedule): reference regimen m1ψ Q6W × 200 mol/OHC intracellular → transfected-OHC STRC trough 2.17× (therapeutic). Every transfected OHC is “functionally rescued”.

2. LNP tonotopic distribution (round-window injection → base-apex gradient): eff_basal = 2·eff_mean, eff_mid = 1·eff_mean, eff_apical = 0.5·eff_mean, clamped at 1.0. Preserves cochlea-mean and matches published cochlear-delivery bias (Geng 2017, Landegger 2017 for Anc80L65; Leclere 2024 for LNP).

3. Per-band ABR transfer (from abr_transfer_model.py calibration): ABR_dB = A − B·log10(functional_fraction + C), fit against 8 literature calibration points covering the OHC-loss-vs-threshold curve (Bredberg 1968, Schuknecht & Gacek 1993, DB-OTO OTOF trial cohorts). DFNB16 untreated baseline = 85 dB.

Applied per zone: rescued_fraction = eff_zone (0 if per-OHC regimen fails); zone ABR = abr_dB(rescued_fraction); weighted audiogram = equal-weight average over three zones.

Floor behaviour: if Misha’s pre-therapy baseline in a zone is better than the rescue projection, post-therapy = baseline (no change). Represents the known biology that hypomorphic STRC patients (like Misha’s E1659A maternal allele) have partial residual function where mRNA therapy gives no additional lift.

Scenarios

LNP scenario	eff_mean	Notes
Untargeted	0.8%	standard LNP, 96/12000 OHC
Cochlear-tropic	5%	hypothetical next-gen cochlear LNP
OHC-targeted	20%	hypothetical ligand-conjugated LNP
Hypothetical 50%	50%	LNP parity with AAV delivery
Anc80L65 ref	67%	AAV reference (Landegger 2017)

Results — DFNB16 untreated 85 dB baseline

Scenario	Basal	Mid	Apical	Weighted post	Weighted improvement	Any-band responder
Untargeted	83 dB	85 dB	86 dB	85.0 dB	0.0 dB	no
Cochlear-tropic	69 dB	76 dB	81 dB	75.6 dB	9.4 dB	no
OHC-targeted 20%	44 dB	58 dB	69 dB	57.0 dB	28.0 dB	yes (basal)
Hypothetical 50%	23 dB	39 dB	54 dB	38.7 dB	46.3 dB	yes (all)
Anc80L65 67% (ref)	23 dB	33 dB	48 dB	34.6 dB	50.4 dB	yes (all)

Results — Misha’s sloping audiogram

Assumed zone baselines: basal 65 dB / mid 55 dB / apical 40 dB (moderate-severe sloping).

Scenario	Basal 65→	Mid 55→	Apical 40→	Weighted improvement
Untargeted	65 dB	55 dB	40 dB	0.0 dB
Cochlear-tropic	65 dB	55 dB	40 dB	0.0 dB
OHC-targeted 20%	44 dB	55 dB	40 dB	6.9 dB (21 dB in basal)
Hypothetical 50%	23 dB	39 dB	40 dB	19.2 dB (42 + 16)
Anc80L65 67% (ref)	23 dB	33 dB	40 dB	21.4 dB

Interpretation

1. 20% OHC targeting gives clinically meaningful high-frequency rescue for Misha: 21 dB recovery in the basal 4–8 kHz band, which carries consonant discrimination cues for speech intelligibility. This is a standalone responder band by the OTOF-trial ≥20 dB criterion. Mid and apical do not benefit — his residual E1659A expression already clears the rescue projection there.

2. 50% OHC targeting = AAV-parity: within 2 dB weighted improvement of the Anc80L65 reference (19 vs 21 dB). Once LNP tropism reaches this level, mRNA-LNP is pharmacologically equivalent to current gene-therapy standards, with the added benefits of re-dosable delivery, no immune memory, and arbitrary payload size (Strategy B: full-length STRC).

3. Untargeted and 5% scenarios are therapeutically inert for Misha: his already-moderate audiogram puts zone thresholds below the log-transfer rescue projection for these low-coverage scenarios. The floor effect is a feature of partial-function patients, not a bug — it explains why mRNA-LNP therapy is most valuable for severely affected patients and for basal (high-frequency) rescue.

4. Clinical gating recommendation: the engineering target for LNP cochlear tropism is 20% OHC as a minimum-therapeutic floor (captures basal recovery, 21 dB), 50% as the AAV-parity target (full-band recovery). Between those two points, every percentage point of tropism translates directly to dB recovered per band by the log transfer function.

5. Tonotopic engineering angle: if basal bias is 2× the mean (current assumption), basal tropism hits 40% at eff_mean=20%. Engineering a more-uniform LNP (ligand to apical-preferring OHC markers) would spread the rescue across frequencies but reduce basal peak. For patients like Misha with sloping loss, the basal-biased profile is clinically preferred.

Limitations

• Tonotopic gradient is a parametric assumption (2×/1×/0.5×). Real distribution varies by LNP chemistry, injection volume, round-window vs intracochlear route. Published data insufficient for precise calibration; sensitivity to gradient shape would be informative (2×/1×/0.5× vs flat vs steeper).
• ABR transfer function assumes uniform rescue within a zone. Spatial clumping of LNP-transfected OHCs (e.g., adjacent-cell clusters from a single injection site) could produce a less-smooth frequency response than predicted.
• “Functional rescue” = per-OHC STRC trough ≥ 2× is a pharmacology pass/fail. Actual cochlear amplification gain is continuous in STRC level and could scale with trough magnitude above threshold (2.17 vs 2.9 vs 3.0×). The transfer function aggregates this into the calibration, but the per-band dB prediction may be slightly conservative at saturated (Q2W) dosing.
• Misha’s baseline audiogram is approximate (65/55/40 dB). Actual audiometric report should replace these before clinical interpretation.
• Model assumes structural rescue is immediate on achieving therapeutic STRC level. In reality, hair bundle reformation after HTC loss is unproven in adult cochlea; pediatric window applies.
• No immune/anti-PEG attenuation of repeat LNP dosing. Chronic Q6W dosing may shift after 2-3 doses in vivo.

Next steps

1. Parameterise tonotopic gradient sensitivity: sweep gradient shapes (flat, 2:1:0.5, 3:1:0.3) to bound the per-band rescue uncertainty.
2. Continuous functional-rescue model: replace pass/fail at 2× trough with a continuous function f(trough) → amplifier_gain, so deeper trough (Q2W saturated 3×) beats marginal Q6W (2.17×) even at the audiogram level.
3. Misha-specific audiogram: replace the approximated 65/55/40 with his measured PTA from latest audiology visit. Re-run the table.
4. Strategy B (full-length STRC mRNA) equivalent model: build a direct mRNA → STRC ODE with no RBM24 layer, compute dose-schedule, then apply the same audiogram layer. Should match or exceed Strategy A because it bypasses the 2× Hill cap.
5. Compare to STRC Piezo Delivery Feasibility OHC Targeting: piezoelectric TM amplifier predicts 92% audiogram coverage at ≥50 dB LAeq; how does this compare to 20% OHC mRNA-LNP rescue in the same tonotopic terms?

Replication

cd ~/STRC/models
/opt/miniconda3/bin/python3 mrna_lnp_audiogram_rescue.py
# outputs: mrna_lnp_audiogram_rescue.json

Files / Models

• ~/STRC/models/mrna_lnp_audiogram_rescue.py — tonotopic × transfer-function composition
• ~/STRC/models/mrna_lnp_audiogram_rescue.json — per-zone predictions for DFNB16 baseline + Misha audiogram × 5 LNP scenarios
• ~/STRC/models/abr_transfer_model.py — ABR transfer function (reused, calibration unchanged)
• ~/STRC/models/mrna_lnp_pkpd_integration.json — source of per-OHC PK/PD pass/fail per regimen

Connections

• STRC mRNA-LNP PKPD Multi-Dose Schedule — takes the per-OHC therapeutic flag and maps it to ABR
• [part-of] STRC mRNA Therapy Hypothesis — closes the clinical-endpoint layer of Strategy A
• [see-also] STRC Piezo Delivery Feasibility OHC Targeting — competing tonotopic-rescue framework for comparison
• [about] Misha