A Membrane-targeting Construct of NF1 GAP-related Domain (GRD) that Suppresses RAS Activity in NF1-related MPNST, NF1 and other Rasopathies
Generated by an autonomous AI research agent — Anthropic Claude Opus 4.7 or OpenAI GPT-5.5, max reasoning effort. Sources cited inline. Full disclosure at /methodology/jhtv-deep-dive.
Indication
Neurofibromatosis type 1 (NF1), RASopathies, malignant peripheral nerve sheath tumors (MPNST)
Modality
Gene Therapy
Mechanism
Membrane-targeting NF1 GAP-related domain (GRD) to suppress RAS activity
Target
RAS / NF1 GRD
rNPV Envelope
Low
-$68.1M
costs +25% · peak −25%
Base
-$36.3M
cumulative PoS 5.4%
High
-$4.4M
costs −25% · peak +25%
Costed as a rare-disease AAV gene therapy with elevated preclinical spend because this exact construct is in-vitro-only (2019) and still needs in-vivo proof plus integration with a tumor-targeting capsid that is separately patented in the 60653 sibling; pivotal studies are small because the genetically-confirmed NF1/MPNST population is narrow (UX111/Zolgensma anchor trial size and BLA timing). Phase 2 PoS is held conservatively below the 60653 sibling and below generic rare-disease AAV because the lead MPNST indication is a documented RAS-pathway clinical graveyard (SARC031 failed). Cumulative LoA is ~5.4%, deliberately below the sibling's ~10.8% to reflect the earlier maturity of this specific construct.
Composite score breakdown
Locked rubric — 40/30/30 weights
Clinical relevance · 40%
0.70
Modality fit · 30%
0.74
Whitespace · 30%
0.50
Composite 0.652 — composite-score rank #17 of 50 top-tier inventions in the jhtv-portfolio@2026-Q2 cohort. The page header uses rNPV rank (#28) to match the index ordering.
Comparators
Real programs anchoring the engine inputs
JHTV 60653 — AAV-NF (K55) capsid + GRD-C24 (KRAS4B-CTD) mini-NF1 payload (Johns Hopkins / NTAP)
Same JHU group's later, in-vivo-validated companion of the identical membrane-targeted GRD concept. 35359 is the foundational construct patent (NF1 GRD fused to an H-Ras CAAX motif, in-vitro data only, 2019); 60653 is the matured version pairing an in-vivo-evolved capsid (AAV-NF K55) with a membrane-targeted GRD payload (GRD-C24), demonstrating xenograft tumor suppression in the 2025 Nature Communications paper. The closest possible mechanism/modality/indication match — and the reason a standalone-equity story for 35359 is weak: the de-risking IP and the in-vivo proof sit in the sibling, not in this construct version.
Criteria 1: identical molecular target (RAS / NF1 GRD), identical mechanism (membrane-targeted GRD restoring RAS-GAP activity), identical modality (AAV gene replacement), same indication, same originating lab. Mechanism-true anchor, not a revenue anchor.
Koselugo (AstraZeneca/Merck) — selumetinib
Only FDA-approved NF1 therapy (pediatric symptomatic inoperable plexiform neurofibroma, 2020; KOMET adult data 2024). Proves NF1 is an approvable, commercially viable orphan indication and anchors the regulatory pathway, but is a chronic symptom-modifying MEK inhibitor — not gene replacement — and addresses plexiform neurofibroma, not the harder MPNST lead this construct targets.
Criteria 2 and 3: same indication family (NF1) and pediatric orphan regulatory pathway; sets the realistic NF1 commercial ceiling (~$331M 2023 net sales; $200–300M peak forecasts).
Zolgensma (Novartis) — onasemnogene abeparvovec
Approved one-time AAV gene therapy for a pediatric monogenic disease; anchors premium one-time gene-therapy pricing and the rare-disease AAV commercial model that offsets small genetically-confirmed patient counts.
Criteria 3 and 4: same rare-disease AAV gene-therapy modality and premium one-time-pricing economic model; revenue anchor for AAV monogenic-disease ceiling.
UX111 / ABO-102 (Ultragenyx) — rebisufligene etisparvovec
Late-stage AAV gene therapy for a rare CNS/peripheral monogenic disease (Sanfilippo A); anchors small pivotal-trial size, long durability follow-up, and BLA timing for a rare AAV program — the development-cost/duration template most applicable to a genetically-confirmed NF1/MPNST AAV trial.
Criteria 3 and 4: same rare-disease AAV pathway with expedited features; trial-size/cost/duration anchor (NCT02716246).
SARC031 — selumetinib + sirolimus in unresectable/metastatic MPNST (cautionary precedent)
CAUTIONARY, NOT A LIVE ANCHOR. The only completed clinical test of dual RAS-effector-pathway inhibition (MEK + mTOR) in MPNST. Despite a positive transgenic-mouse signal it FAILED its efficacy bar — clinical benefit in only 1/7 (stage 1) and 1/14 (stage 2) patients — and did not advance. Directly relevant because it shows the MPNST RAS-pathway clinical graveyard the lead indication for 35359 sits in: mouse RAS-pathway efficacy has not yet translated to MPNST patients.
Criteria 1 and 2: same RAS-effector-pathway mechanistic space and same MPNST lead indication; included only as a labelled cautionary precedent for MPNST translational risk — status: trial completed, did not meet efficacy parameters, not advanced.
Stage profile
Asset-specific cost, duration, and PoS by stage
| Stage | Cost | Duration | PoS | Citations |
|---|---|---|---|---|
| Preclinical | $26.0M | 30 mo | 45.0% | [0] [1] [5] |
| Phase I | $75.0M | 18 mo | 62.0% | [2] [3] [5] |
| Phase II | $175.0M | 30 mo | 40.0% | [0] [4] [5] |
| Phase III | $250.0M | 30 mo | 55.0% | [2] [3] [5] |
| NDA/BLA Review | $18.0M | 12 mo | 88.0% | [3] [6] [5] |
Multiplier handling: Eligible multipliers (genetic_validation_2.6x, orphan_1.4x, gene_therapy_1.41x, fast_track_or_rmat, pediatric_voucher) are already reflected in Day-1 comparator-calibrated PoS. Re-applying them via log-odds stacking would double-count, so per-stage PoS is taken as final. See methodology for the rule.
Peak revenue and discount rate
$650.0M peak · WACC 13.0%
Peak revenue. Modeled below the 60653 sibling's $900M because (a) this construct is in-vitro-only and less mature, (b) the lead MPNST indication is harder and smaller than pediatric plexiform neurofibroma and the only clinical RAS-pathway combo (SARC031) failed, and (c) the asset is partnership-shaped, warranting conservatism. NF1 affects ~1 in 3,000; the genetically-confirmed MPNST/aggressive-RASopathy addressable subset is small, but premium one-time AAV gene-therapy pricing (Zolgensma economics) offsets low patient count. $650M reflects a one-time-priced AAV launched into a narrow genetically-confirmed NF1-tumor subset with optionality into broader RASopathy, sitting above Koselugo's chronic-therapy NF1 ceiling but well below Zolgensma's larger SMA population.
WACC. Genetically defined orphan AAV sits in the venture rare-disease playbook (12% baseline per the 60653 sibling), but +1% reflects the earlier maturity of this specific construct (in-vitro only), the failed RAS-pathway clinical precedent in the lead MPNST indication, and dependence on a separately-patented evolved capsid for tumor-targeted delivery.
Sensitivity (tornado)
Top drivers of rNPV variance
Drivers ranked by absolute rNPV swing. The vertical tick inside each bar marks the base rNPV (-$36.3M); each bar spans the rNPV range produced by flexing one input between its low and high values. Gold = the input pushes rNPV up when increased; red = the input pushes rNPV down when increased.
Monte Carlo distribution
1,000 trials · rpNPV mode
This is a bimodal distribution by construction, not a Gaussian. Most paths terminate in clinical failure (red cluster — accumulated cost only); a minority succeed and capture full peak revenue (green tail). Bar heights are square-root-scaled so the success tail stays visible alongside the much taller failure cluster; exact counts are preserved in the percentiles below. Gold line = median (P50). Navy dashed = base rNPV (mean) — the probability-weighted expected value, which can sit above the median when the upper tail is strong enough to outweigh the failure cluster (and close to the median when it isn’t).
P5
-$228.1M
P25
-$90.7M
P50 (median)
-$30.9M
P75
-$15.3M
P95
$419.8M
Prob ≥ 0
6.2%
Comparable launch curves
Revenue trajectories of named comparators
Koselugo (AstraZeneca/Merck) — selumetinib
Launched 2020 · peak $380.0M (estimated)
Evidence register
8 per-assumption citations
| Assumption | Source | Date | Confidence |
|---|---|---|---|
Asset identity, construct design (NF1 GRD + H-Ras CAAX), and in-vitro-only stage cmo_findings.asset_class_reality | JHU Technology Ventures listing C15474 / 35359 — Membrane-targeting NF1 GRD construct company_filing | 2019-10-29 | high |
GRD-only restores catalytic RAS-GAP activity in vitro (mechanism basis and Phase 1/2 PoS) stage_profile.phase_2.pos | Feasibility of using NF1-GRD and AAV for gene replacement therapy in NF1-associated tumors (Gene Therapy, 2019) peer_review | 2019-05-24 | high |
Sibling in-vivo validation lives in 60653 (evolved AAV-NF K55 + GRD-C24), not in this construct comparators[0] | Development of an adeno-associated virus vector for gene replacement therapy of NF1-related tumors (Nature Communications, 2025) peer_review | 2025-09-29 | high |
NF1 is an approvable orphan indication; Koselugo regulatory and commercial precedent comparators[1].peak_revenue_usd | NCI Cancer Currents: Selumetinib approved for NF1 plexiform neurofibromas regulatory | 2020-05-08 | high |
MPNST RAS-pathway clinical graveyard — SARC031 failure (Phase 2/3 PoS conservatism) stage_profile.phase_3.pos | SARC031: A Phase II Trial of Selumetinib and Sirolimus for Patients with Unresectable or Metastatic MPNST trial_disclosure | 2025-11-01 | high |
Rare-disease / gene-therapy PoS and cumulative-LoA baseline stage_profile.preclinical.pos | BIO/QLS/Informa Clinical Development Success Rates 2011-2020 peer_review | 2021-02-17 | medium |
Premium one-time AAV gene-therapy pricing economics (peak revenue ceiling) peak_revenue_usd | Novartis Annual Report 2024 company_filing | 2025-01-31 | high |
Rare AAV CNS/peripheral pivotal-trial size and BLA timing anchor stage_profile.phase_3.duration_months | ClinicalTrials.gov NCT02716246: AAV gene transfer for MPS IIIA (UX111/ABO-102) trial_disclosure | 2016-03-25 | high |
Thesis
Why this asset earns its rank
This asset (JHU case C15474) is the foundational construct patent behind the Johns Hopkins NF1 gene-replacement program: the NF1 GAP-related domain (GRD) fused to an H-Ras C-terminal CAAX membrane-targeting motif so the catalytic RAS-GAP fragment is localized to the membrane where it shuts off hyperactive RAS. The biological logic is sound and addresses the central constraint head-on — full-length neurofibromin (~8.5 kb CDS) cannot fit AAV's ~4.7 kb limit, but the GRD-only payload plus a short CAAX motif comfortably does, so the modality genuinely reaches the target. Crucially, the JHU listing states this exact construct has only in-vitro data (Gene Therapy, 2019); the in-vivo xenograft proof and the tumor-targeting evolved capsid (AAV-NF K55 + GRD-C24) live in the later, separately-patented companion invention (sibling 60653, Nature Communications 2025). So this is a real therapeutic concept, but the de-risking evidence sits next door, not in this filing.
The comparators frame an honest economic envelope. Koselugo proves NF1 is an approvable orphan indication (~$331M 2023 sales; $200–300M peak forecasts) but is a chronic symptom-modifying MEK inhibitor, not curative gene replacement. Zolgensma and UX111 anchor premium one-time AAV pricing and small rare-disease pivotal-trial economics that offset a narrow genetically-confirmed population. SARC031 is included as an explicit cautionary precedent: the only completed clinical test of RAS-effector-pathway inhibition in the lead MPNST indication failed its efficacy bar despite positive mouse data, which is exactly the translational gap this construct still has to cross. The engine result is -$68.1M to -$4.4M, with a base rNPV of -$36.3M and cumulative PoS of 5.4%; that spread is consistent with a partnership-shaped asset — the construct IP is most valuable folded into the matured AAV-NF K55 program (or licensed to a gene-therapy developer with capsid and manufacturing infrastructure) rather than financed as a standalone equity bet on a pre-in-vivo payload.
The verdict is a credible but partnership-shaped rare-disease gene-therapy asset, not standalone-VC-fundable on its own merits. It earns its rank on strong genetic validation, high IRA exposure for a rare genetic indication, and a sound mechanism with high modality-PoS for AAV gene replacement — but the funding case rests on (a) generating in-vivo efficacy for this specific membrane-targeted construct, not only the sibling's GRD-C24, and (b) clearing the MPNST translational graveyard the SARC031 failure defines. The partnership_candidate archetype reflects that the value path is licensing/integration with the matured capsid program, not a standalone equity rNPV.
Key risks
Asset-specific, not generic biotech risks
- Maturity / IP-locus mismatch: the JHU listing states this exact construct (NF1 GRD + H-Ras CAAX) is in-vitro-only (2019); the in-vivo xenograft proof and the tumor-targeting evolved capsid (AAV-NF K55 + GRD-C24) sit in a separately-patented sibling invention. A standalone-equity bet on this filing alone underwrites a pre-in-vivo payload whose delivery solution is owned next door — hence the partnership_candidate archetype.
- Mechanism: GRD-only restores catalytic RAS-GAP activity but lacks ~85% of full neurofibromin (Sec14-PH, CSRD, and protein-interaction domains); whether catalytic-only restoration delivers durable in-vivo lesion regression for THIS construct — versus the sibling's distinct GRD-C24 payload — is unproven.
- MPNST translational graveyard: SARC031 (selumetinib + sirolimus) had a positive transgenic-mouse signal yet produced clinical benefit in only 2/21 MPNST patients and was not advanced. Mouse RAS-pathway efficacy has repeatedly failed to translate in MPNST patients; the lead indication is high-risk.
- Delivery dependency: tumor-targeted distribution to MPNST and multisystem NF1 lesions depends on the evolved AAV-NF K55 capsid characterized in the sibling, not generic AAV; whether a generic-AAV-delivered version of this exact construct achieves adequate tumor biodistribution is unestablished, and capsid IP/freedom-to-operate must be reconciled across the two JHU filings.
- Competitive/commercial: Koselugo holds the approved NF1 standard and sets symptom-modifying expectations and an orphan-scale commercial ceiling; an AAV gene therapy must demonstrate measurable tumor regression (not just growth inhibition) to justify one-time premium pricing in a narrow genetically-confirmed population.