STRC Research

Misha Compound-Het Therapy Stack Model

7 min read

Misha Compound-Het Therapy Stack Model

TL;DR. For Misha’s specific genotype (paternal 98 kb STRC deletion + maternal E1659A), h03 AAV monotherapy is ceiling-capped at MILD outcome (30-40 dB ABR) — cannot reach NORMAL regardless of transduction efficiency, because non-transduced OHCs inherit 0 functional STRC from the paternal allele and sub-threshold protein from the untreated E1659A maternal allele. h01 pharmacochaperone is the only monotherapy route to NORMAL hearing (≤ 25 dB), contingent on drug rescue f_PC ≥ 0.5 (mild E1659A) or ≥ 0.75 (moderate/severe E1659A). h01+h03 stack is the clinical plan for a full cure in Misha’s case; robust across STRC-functional-threshold sensitivity (θ ∈ {0.25, 0.35, 0.45}). This is NOT a new hypothesis — it is genotype-specific integration of #1 and #3.

The hidden assumption this breaks

Every existing hypothesis treats “STRC-null” as a single-mechanism problem and evaluates therapeutic efficacy against a generic DFNB16 patient. Misha is not a generic patient. His two alleles are broken differently:

  • Paternal (98 kb del): encodes nothing. AAV can substitute but only in transduced OHCs.
  • Maternal (E1659A missense): encodes full-length protein with intact fold (AF3 pTM 0.64 identical to WT) but charge-loss at a functional interface (AlphaMissense 0.9016 pathogenic). Small-molecule pharmacochaperone can rescue, drug reaches every OHC.

These asymmetries imply different therapeutic physics for each allele. A single-hypothesis evaluation obscures a deep structural fact: one therapy cannot touch both alleles.

Model

Per-allele protein-function score (relative to WT = 1.0):

f_pat_untreated = 0.0                      (null, no protein)
f_mat_untreated = f_mat ∈ [0.10, 0.40]     (E1659A residual)

With pharmacochaperone rescue f_PC ∈ [0, 1]:

f_mat_treated = f_mat + f_PC × (1 - f_mat)

Per-OHC STRC protein level:

Non-transduced OHC:  f_OHC = 0.5 × f_mat_treated
Transduced OHC:      f_OHC = min(0.5 × f_mat_treated + A, 1.0)

where A is per-transduced-OHC ectopic mini-STRC expression and ε is transduction efficiency.

Binary functional OHC: an OHC is functional iff f_OHC ≥ θ, where θ is the STRC-protein threshold for bundle formation. Biological anchor: STRC +/- heterozygous carriers (50% protein) have normal hearing → θ ≤ 0.5. E1659A homozygotes would be ~20% protein and are deaf → θ > 0.2. Baseline θ = 0.35, sensitivity sweep over {0.25, 0.35, 0.45}.

Fraction of functional OHCs feeds the lit-calibrated ABR transfer function (Bredberg 1968, Schuknecht & Gacek 1993, Sun 2024 OTOF Cohort 2) — the input semantic is “fraction of functional OHCs”, which matches our binary model.

Results

Headline grid (θ = 0.35, E_grid × A_grid × f_PC grid = 90 points per scenario)

Scenariof_matNo-Tx ABRh03-only NORMAL?h01-only NORMAL?Stack NORMAL?Low-burden stack count
Severe E1659A0.1087 dBNOYES at f_PC ≥ 0.75YES0
Moderate E1659A0.2587 dBNOYES at f_PC ≥ 0.75YES0
Mild E1659A0.4087 dBNOYES at f_PC ≥ 0.50YES9

“Low-burden stack” = ε ≤ 0.30 AND f_PC ≤ 0.50 AND NORMAL achieved.

Why h03 monotherapy cannot deliver NORMAL

At ε ≤ 0.5 (top of realistic transduction envelope for Regeneron AAV.104 preclinical), at least 50% of OHCs are non-transduced. These non-transduced OHCs have:

f_nontrans = 0.5 × f_mat_untreated = 0.05 - 0.20 (below θ = 0.35)

They are not functional. Even if all transduced OHCs reach WT-equivalent (A ≥ 0.7, clamp at 1.0), cochlea-mean functional fraction = ε. At ε = 0.5, that’s 50% functional OHCs → predicted ABR ~ 38 dB (MILD). This is a physical ceiling, not a quantitative fine-tune. h03 alone cannot cross NORMAL.

Why h01 monotherapy can deliver NORMAL

PC is a systemic small-molecule drug that reaches every OHC. It rescues the maternal E1659A allele in every cell simultaneously. If f_mat_treated ≥ 2θ = 0.70, the non-transduced OHCs become functional, pulling cochlea-wide function to 100% → predicted ABR ≤ 20 dB (NORMAL). Required f_PC depends on the residual E1659A function:

f_mat + f_PC × (1 - f_mat) ≥ 0.70
f_PC ≥ (0.70 - f_mat) / (1 - f_mat)

mild (f_mat=0.40): f_PC ≥ 0.50
moderate (f_mat=0.25): f_PC ≥ 0.60
severe (f_mat=0.10): f_PC ≥ 0.67

Precedent: tafamidis (TTR stabiliser) achieves 70-90% rescue in ATTR; VX-770 (CFTR potentiator) achieves 50-70% rescue in G551D CFTR; cysteamine in cystinosis 60-80%. f_PC in the 0.5-0.75 band is realistic for a well-optimised pharmacochaperone.

Why stack matters

At mild E1659A (f_mat = 0.40), 9 low-burden stack combinations (ε ≤ 0.3, f_PC ≤ 0.5) achieve NORMAL. Stack gives:

  • Redundancy: if PC rescue underperforms, AAV transduced OHCs still express ectopic mini-STRC at WT levels — they function even without PC.
  • Lower drug burden: f_PC = 0.25-0.50 stack combinations access NORMAL that would require f_PC ≥ 0.50-0.75 monotherapy.
  • Defence against incomplete transduction: if ε underperforms, PC rescues the untransduced fraction.

Theta-sensitivity

Threshold θ for OHC functionality is biology-plausible over [0.25, 0.45]. Conclusion “stack enables NORMAL in all f_mat scenarios” holds across the entire sensitivity range. Not an artifact of picking 0.35.

Strategic implications

QuestionAnswerImplication
Can h03 alone cure Misha?NO (physical ceiling at MILD ~30-40 dB)h03 is “meaningful rescue” not “cure”; current S-tier status still justified — 24 dB improvement for Misha is clinically huge
Can h01 alone cure Misha?YES (requires f_PC ≥ 0.50-0.75)h01 Misha-fit is underrated; NORMAL outcome depends entirely on Phase 5 MD + virtual screen delivering a drug with f_PC ≥ 0.5
Is stack optimal?YES at mild E1659A; redundancy at severeClinical plan for Misha: pursue BOTH in parallel, not either/or

Current ranking reframe:

  • h03 delivers “meaningful hearing” endpoint. S-tier held — primary path.
  • h01 delivers “potential full cure” endpoint conditional on drug performance. A-tier held but Misha-fit argument strengthened — should remain in active development, not backburner-mental-model.

What this is NOT

  • Not a new hypothesis — it’s integration of existing h01 and h03 for Misha’s genotype.
  • Not validated — depends on f_PC (requires Phase 5 MD on h01 + wet-lab), ε (requires Regeneron/similar preclinical confirmation), θ (biology inference not measurement).
  • Not applicable to DFNB16 generally — this is Misha-specific compound het logic. STRC-null homozygotes have different allelic physics.

What this doesn’t resolve

  • f_PC estimate: we have no drug yet. Phase 5 MD ensemble rescoring on h01 lead compounds was completed 2026-04-23 with RED-LIGHT result — all current Phase 4b shortlist + diflunisal positive control deliver f_PC < 0.10 at therapeutic [L]=10 μM, order of magnitude below the 0.50 NORMAL-rescue threshold. See STRC h01 Phase 5 MD Ensemble Rescoring 2026-04-23. This does NOT invalidate the stack model — the model’s structural conclusion (h01 is the only monotherapy route to NORMAL cure for compound het) stands, but the timeline to operational h01 extends by 1-2 years for Phase 3c v2 expanded virtual screen + fragment-based + covalent strategy exploration. In the meantime, h03 AAV remains the near-term lever for Misha at MILD-tier rescue (20-30 dB ABR improvement, clinically significant).
  • θ empirical: no direct measurement; biology inference from heterozygous carrier normal-hearing + E1659A homo deaf (hypothetical — real Asi homo not in published cohorts). Could refine with OHC-specific STRC-knockdown titration.
  • Ceiling assumption A ≤ 1.0: published STRC promoter-driven mini-STRC expression can exceed WT single-allele (1.0) by 2-3× at peak. If A > 1.0 with no ceiling, stack low-burden count grows; doesn’t change monotherapy ceiling for h03.

Ranking delta

  • #1 (h01) Pharmacochaperone E1659A: A held, Mech 3 held, Deliv 4 held, Misha-fit 4 held. But framing updated: h01 is the only monotherapy path to NORMAL hearing for Misha specifically. Phase 5 MD priority strengthened — the f_PC parameter this model needs is the Phase 5 deliverable.
  • #3 (h03) Mini-STRC AAV: S held, no score change. Reframed as “meaningful rescue ceiling MILD-MODERATE for compound-het patients”; still primary S-tier because 20-30 dB ABR improvement is clinically huge and wet-lab-ready.
  • Ref entry added: Misha Compound-Het Therapy Stack Model as supporting integrative model (analogous to #18 Bundle Mechanics, #24 Adult Treatment Window). Adds “stack plan” as the operating clinical strategy, not monotherapy from either hypothesis alone.

Connections

Graph View

Backlinks