ARCT / arct-810-otc-deficiency — ARCT-810 for ornithine transcarbamylase deficiency

Program analysis · bpiq_drug_id 16237 · prepared 2026-08 · USD · framework v5.8.0 · Coverage: CLEARED

Company context, financials, ownership, options and the full pipeline: ../company.md.

How to read this. Written for a reader who has not studied pharmacology or finance. Terms specific to this program are in the Glossary below; general terms are defined once in framework/04-glossary.md. Tags: [VERIFIED — source] / [UNVERIFIED] / [WEB ESTIMATE — source, date].

Glossary

Only terms specific to this program. Generic terms live in framework/04-glossary.md.

Executive summary

• What it is (one sentence): ARCT-810 is human OTC messenger RNA wrapped in a fatty nanoparticle and infused every two weeks, intended to make a patient’s own liver cells temporarily produce the enzyme their faulty gene cannot.
• The event and when (as disclosed): Results from the Phase 2a study (NCT06488313) and the regulatory plan that follows from them. Arcturus said on 2026-08-06 that both are “expected to be communicated later this quarter” — Q3 2026, no month and no day disclosed [VERIFIED — Arcturus Q2 2026 press release, 2026-08-06].
• The main reason it could work: The enzyme defect is a pure loss of function in a single well-validated liver enzyme, mRNA delivery to liver cells is the best-established application of the whole LNP field, and unlike the two gene therapies ahead of it, ARCT-810 can be re-dosed — which matters most in children, whose growing livers dilute a one-time treatment.
• The main risk: The evidence supporting a Phase 3 decision is thin and got thinner. The randomised, placebo-controlled Phase 2 was terminated after eight patients; what replaced it is a nine-patient, open-label, single-site study whose registered primary endpoint is safety, not efficacy. No ARCT-810 clinical data has ever been peer-reviewed.
• What it means for the stock: Less than the science suggests. Enterprise value is about $34 million against $191.5 million of cash (../company.md C.4), so almost nothing is priced in either direction — but the last time this exact programme reported positive interim data, on 2025-06-30, the shares fell 2.4% on the day.

0. Program-tier coverage — CLEARED

One row per mandatory tool in the program-tier table in framework/02-connectors.md, in the order that file lists them. Company-tier coverage is in ../company.md C.0 — not repeated here.

Verdict: CLEARED. Every mandatory program-tier row was called and returned data, including the two rows (ChEMBL, Open Targets) whose known failure modes were live on other tickers the same day. The three NOT CALLED rows are all marked optional.

A. Scientific & clinical assessment

A.1 Mechanism of action and scientific rationale

What the drug is. ARCT-810 is a copy of the human instructions for making the ornithine transcarbamylase enzyme, written as messenger RNA and packaged inside a lipid nanoparticle — a fatty bubble a few tens of nanometres across. It is given as an infusion into a vein once every two weeks [VERIFIED — CT.gov NCT06488313 detailed description, read 2026-08-13].

What goes wrong in the disease. Digesting protein releases nitrogen, which the body converts to ammonia. Ammonia is toxic, especially to the brain. The liver disposes of it through the urea cycle, a five-step chemical assembly line that ends by packaging the nitrogen into urea for the kidneys to excrete. Ornithine transcarbamylase performs the second step. When the OTC gene is faulty the cycle stalls there, ammonia accumulates, and the patient suffers vomiting, confusion, brain swelling and — in a severe crisis — coma or death.

How the drug is meant to fix it. The nanoparticle carries its mRNA cargo through the bloodstream to the liver, where liver cells take it up. The cell’s own protein-making machinery reads the message and builds working OTC enzyme, which is then imported into the mitochondria where the urea cycle runs. The cycle restarts, and ammonia is cleared normally.

How well the target is validated. As well as any target in rare disease, and this is the programme’s genuine strength. OTC is a single gene whose loss causes a single, precisely understood metabolic block; Open Targets returns it as a first-class target (ENSG00000036473) with the disease mapped to MONDO_0010703 [VERIFIED — Open Targets search_entities, 2026-08-13]. There is no question of whether restoring OTC enzyme activity helps — patients who receive a liver transplant, which supplies a healthy copy of the enzyme, are cured. The question is only whether this delivery method supplies enough enzyme, reliably enough, safely enough. Delivering mRNA to liver cells is also the single most established application of lipid nanoparticle technology, and independent groups have shown OTC mRNA-LNP works in animal models of the disease: according to PubMed, Yamazaki and colleagues at Eisai showed a single dose of human OTC mRNA-LNP doubled survival in a mouse model (DOI), and a Moderna group built a conditional knockout mouse specifically to test the approach in severe disease (DOI).

The exact scientific step this readout must prove. Not that OTC mRNA works in principle — that is settled — but that repeated dosing in humans produces a measurable, durable increase in ureagenesis at a dose that is tolerated, and that the FDA will accept that measurement as the basis for a registrational trial. The registered secondary endpoints are precisely this: change from baseline in stable-isotope ureagenesis assay AUC, the proportion of participants maintaining normal morning fasting plasma ammonia, and change in plasma glutamine, all measured to Day 85 [VERIFIED — CT.gov NCT06488313 outcome measures, read 2026-08-13].

The honest scientific risk. It is not mechanism, it is magnitude and durability. mRNA is transient: the protein it produces decays over days, which is why the drug is dosed every two weeks rather than once. A patient must therefore ride a sawtooth of enzyme activity between infusions, and the trough matters as much as the peak — a crisis does not wait for the next dose. Neither the mechanism diagram nor the animal data answers whether the trough is high enough. On top of that sits a delivery-specific risk: the trial excludes anyone with a history of severe allergic reaction to a liposomal or PEG-containing product, and infusion reactions to lipid nanoparticles are a known class effect that a chronic, life-long, every-two-weeks regimen has far more opportunity to encounter than a vaccine given twice.

A.2 Clinical development plan, timeline, feasibility, resourcing

Trials

The single most important fact in this table. Arcturus ran a properly controlled study — 3:1 randomised, double-blind, placebo-controlled, across fourteen European sites — and terminated it after eight patients. What replaced it is an open-label, uncontrolled, nine-patient study at one site in Maryland. Whatever the reason (recruitment in an ultra-rare disease across fourteen sites is genuinely hard, and a placebo arm in a life-threatening metabolic disease is ethically strained), the evidentiary consequence is not softened by the explanation: the programme has moved away from the design that could have demonstrated an effect against a control, at exactly the point where it needs to convince a regulator to authorise Phase 3. [VERIFIED — CT.gov NCT05526066 and NCT06488313, read 2026-08-13]

Feasibility matrix

Resourcing sufficiency. Cash is not the constraint here, and unusually for a company this size it is not close. Arcturus held $191.5 million at 2026-06-30 with a stated runway “through year end 2028”, plus a $12 million CSL termination payment and release from about $16 million of liabilities received after that balance-sheet date (../company.md C.3). Against a monthly burn of roughly $6.6 million, the company can fund a Phase 3 decision and the first years of the trial without raising equity, and it has not sold stock since 2023. The binding constraint is patients and regulatory agreement, not money — which is a comparatively good problem to have, and it is why the financing-risk driver in the run-up score is low.

A.3 Target label, target product profile, strategic questions, designations

A.3a Target label. Chronic enzyme-restoration therapy for adolescents, adults and — per the strategy broadened in March 2026 — children with ornithine transcarbamylase deficiency who remain inadequately controlled on a protein-restricted diet and ammonia-scavenger therapy.

A.3b Target product profile

A.3c Strategic Go/No-Go questions. The asset’s next decision is whether to enter a registrational trial, so the Pre-Phase-III set applies.

Pre-Phase-III (Go-to-Phase-III / registration):

A.3d Regulatory designations.

• Orphan Drug Designation (US FDA), granted 2019-06-27 for ARCT-810 in OTC deficiency. Grants seven years of US market exclusivity from approval, tax credits against clinical trial costs, and a waiver of the marketing application user fee. It does not speed up review or lower the evidence bar. [VERIFIED — Arcturus press release, GlobeNewswire, 2019-06-27]
• No Breakthrough Therapy, Fast Track, RMAT or Priority Review designation has been disclosed for ARCT-810. Arcturus does hold Fast Track for a different asset (ARCT-2304, H5N1, granted 2025-04-10), which shows the company pursues such designations when it can get them — making the absence here more informative than a simple silence. [VERIFIED — BPIQ fetch_company_drugs and fetch_company_historical_catalysts, 2026-08-13; no ARCT-810 designation in the press feed]

A.4 Target product profile valuation matrix

Calibration examples, not asset claims: in oncology each incremental month of overall survival is worth roughly $150–300M in peak-sales potential; oral formulations command roughly 15–25% price premiums over injectables. The second is directly relevant in the wrong direction here — ARCT-810 is an infusion competing against a daily oral liquid.

A.5 Evidence quality and endpoints

On the publication finding, stated plainly. According to PubMed, a search covering ARCT-810 and mRNA therapy for ornithine transcarbamylase deficiency returns 23 articles. Reading the titles rather than counting them — the discipline 02-connectors.md requires after the MNOV “MN-001” case — none reports ARCT-810 clinical data. The nearest neighbours are a Sanofi paper using OTC deficiency explicitly as a model system for optimising liver-tropic LNPs (DOI), an Eisai mouse efficacy study (DOI), a Moderna knockout-mouse model (DOI), a Tessera transposon paper (DOI), and reviews of RNA and gene therapy for urea cycle disorders (DOI, DOI, DOI). The 2024 review by Richard and colleagues notes that LNP-formulated mRNA therapy for OTCD “has progressed to the clinical trials phase” — which is an accurate description of Arcturus’s programme and also the extent of what the independent literature can say about it.

The limit of this check: PubMed indexes journals, not company presentations. Arcturus held a KOL presentation of the interim data on 2025-06-30, and material may exist in slide decks or non-indexed abstracts. What can be said is that after six years no ARCT-810 clinical result has entered the peer-reviewed record, and that every efficacy claim about this asset currently traces back to a company press release.

Endpoints

A.5b Key opinion leaders.

Panel as of. 2026-08-13 — the date the investigator and independent-voice searches below were run.

Investigators

One investigator is what the trial record yields, because the trial has one site. search_investigators queried by name returned zero rows — the tool searches overall-official and responsible-party fields, and Dr Summar is listed as a site contact, which is a tool limitation rather than a finding. Queried by condition it returned him correctly, together with 66 other people almost all of whom are on unrelated trials that matched “OTC” as an abbreviation for optical coherence tomography, ovarian tissue cryopreservation and “outside the cage” surgery — a search-index confusion of the same family 02-connectors.md records for PubMed and short drug codes.

On the empty conflicts cell, and why it is not a clean bill of health. The searches behind it are recorded above, which is what makes the emptiness a finding rather than an assumption. But the newest paper the author search returned is from 2019, so any relationship formed in the seven years since would not appear in it, and 02-connectors.md is explicit that the absence of a disclosure is never evidence of no conflict. Dr Summar is a widely recognised urea-cycle authority and Uncommon Cures is a dedicated commercial research site; both facts make undisclosed industry relationships more likely a priori, not less.

Independent voices

None recorded, and the reason is a finding rather than a gap. No named commentator on ARCT-810’s endpoint could be identified who is both independent of Arcturus and on the record about this programme. The nearest candidates were rejected on inspection: the most authoritative independent reviews of gene and RNA therapy for OTCD are by Baruteau, Gissen and Seker Yilmaz at University College London and Great Ormond Street Hospital (DOI, DOI) — and Great Ormond Street Hospital was one of the fourteen sites on Arcturus’s terminated Phase 2, NCT05526066. A sponsor site relationship disqualifies them from this table by definition, which is exactly why the check was made before recording anyone as independent.

Judgement

Dissent

B. Commercial assessment

B.0 Current treatment algorithm

What a patient with ornithine transcarbamylase deficiency receives today, in order:

1. Diagnosis and acute stabilisation. Severe (usually male, neonatal) cases present with a hyperammonaemic crisis in the first days of life and are stabilised with intravenous nitrogen-scavenging drugs, dialysis to strip ammonia, and the withdrawal of all dietary protein. Late-onset cases — often female carriers — may not present until childhood or adulthood, sometimes triggered by illness, childbirth, or a high-protein meal.
2. Chronic maintenance, first line. A lifelong low-protein diet, supplemented with essential amino acids, arginine or citrulline. This is nutritional management, and in children it constrains growth.
3. Chronic maintenance, second line. An ammonia scavenger added to the diet: glycerol phenylbutyrate (Ravicti) or sodium phenylbutyrate (Buphenyl). Ravicti has largely displaced Buphenyl because Buphenyl’s taste and odour drive patients off it [WEB ESTIMATE — market summaries via web search, 2026-08-13]. Scavengers open an alternative route for nitrogen; they do not restore the enzyme, and patients on them still have crises.
4. Rescue. Recurrent crises despite steps 2 and 3 mean hospitalisation, and repeated hyperammonaemia causes cumulative neurological damage.
5. Definitive. Liver transplant. It is curative because it supplies working enzyme, and it is limited by donor supply, by the surgery’s own risk, and by lifelong immunosuppression.

Where ARCT-810 would fit. As a new step between 3 and 5: added to, or partly replacing, scavenger therapy in patients whose control remains inadequate — the group currently facing repeated crises with nothing between another scavenger dose and a transplant. It does not displace the diet and it does not displace acute crisis care. Its competitors, DTX301 and ECUR-506, target the same slot from the opposite direction: one-time treatments that would be given once and then, if they work, removed the need for steps 2 and 3 entirely.

B.1 Competitive landscape and positioning

Where this asset wins, and the single fact the thesis rests on. It wins in children, and only there decisively. A one-time AAV gene therapy delivers a gene that does not integrate into the chromosome; as a child’s liver grows, the treated cells divide and the effect dilutes — and AAV generally cannot be given a second time, because the immune system has by then learned to neutralise the vector. Patients who already carry anti-AAV antibodies are excluded from the outset, an exclusion real enough that Ultragenyx ran a dedicated seroprevalence study for it (NCT04909346) [VERIFIED — CT.gov]. A re-dosable therapy has no such ceiling. The single fact the thesis rests on is therefore this: whether ARCT-810 produces a large enough and durable enough rise in ureagenesis across a two-week dosing interval that a regulator will authorise a registrational trial in children. Everything else — the platform, the cash, the orphan designation — is supporting.

Against that, the honest competitive reading in adults is poor. Ultragenyx is years ahead with a one-and-done treatment in exactly the late-onset adult population ARCT-810’s current trial enrols, and iECURE has reported a complete clinical response in an infant plus 72% of participants free of hyperammonaemic crises [WEB ESTIMATE — iECURE ASGCT presentation via web search, dated 2026-05-13]. If both succeed, the commercial space left for a two-weekly lifelong infusion narrows to patients those therapies cannot serve.

Calibration example, not an asset claim: oncology first-movers in novel mechanisms have shown roughly 3.2× higher peak-sales potential but about 1.8× higher development risk. ARCT-810 is not a first mover in this indication.

B.2 Addressable market

Launch markets would be the United States and EU5 (France, Germany, Italy, Spain, United Kingdom) — the markets with established metabolic-disease centres, newborn screening, and reimbursement precedent for ultra-orphan therapies.

What is not defensible here, said plainly. The third-party market reports returned by web search quote a “global OTC deficiency treatment market” of $1,284 million in 2025 growing to $2,341.6 million by 2033 [WEB ESTIMATE — cognitivemarketresearch.com, retrieved 2026-08-13]. That figure is not used anywhere in this analysis. A $1.28 billion market for a disease with a few thousand diagnosed patients worldwide implies annual per-patient spending that no scavenger regimen reaches, and the report gives no derivation. Under 01-rules.md rule 6 it is rejected rather than passed through. B.3a is therefore built bottom-up from patients, price and share.

B.3 Value and feasibility

B.3a Expected peak sales. Built bottom-up. Every figure below is conditional on approval, and excludes any revenue from a paediatric or neonatal indication where gene editing is currently ahead. All three inputs are estimates, so all three scenarios are ranges of assumption rather than forecasts, per 01-rules.md rule 10.

Which input is weakest, named rather than averaged away: the penetration share. The patient count is anchored to a published incidence figure and the price band is anchored to observable ultra-orphan pricing, but the share assumption is a judgement about a competitive outcome that will not be known for years — DTX301’s Phase 3 does not reach primary completion until 2027-09. A reader who disagrees with the share assumption should scale these figures linearly; they do not depend on anything else.

B.3b Feasibility

B.4 Product-development risk

Framing questions.

1. Does the development plan support the target product profile claims? Only partly. The profile’s central claim is chronic, re-dosable enzyme restoration across all ages, and the development plan currently consists of a nine-patient, open-label, 85-day study in patients aged 12 and over. It can support a claim about short-term pharmacodynamics; it cannot support a claim about chronic control, about children, or about crisis reduction, because it measures none of those over any relevant period.
2. Will the identified risks affect the target product profile? Yes, in one specific place. The two-weekly infusion schedule is already the weakest row of A.3b, and the risk that ureagenesis troughs between doses would force more frequent dosing attacks exactly that row. A regimen that had to move to weekly would be materially less competitive against a one-time gene therapy.
3. If a risk cannot be mitigated, is the asset still differentiated from competitors? Yes — in children and in AAV-antibody-positive patients, where re-dosability has no substitute. That residual differentiation is narrower than the target profile claims but it is real, and it is what survives if the adult opportunity is lost to DTX301.

Readout

Sources

The modelled estimate. The registry’s primary_completion_date for NCT06488313 is 2026-06-01. data/benchmarks/readout-lag.json holds no matching entry for a Phase 2a open-label metabolic study, so the stated default applies: primary completion plus two to four months to database lock and analysis (01-rules.md rule 34). That gives 2026-08-01 to 2026-10-01, and the estimate is tagged [UNVERIFIED — modelled, default lag] rather than benchmarked. This arithmetic is independent of anything Arcturus has said, and it brackets the company’s own “later this quarter” almost exactly — which is the useful part: the guidance is not obviously optimistic against the trial’s own mechanics.

Window

Basis. Every source points into a band running from August to late 2026, and the company’s own 2026-08-06 wording (“later this quarter”) is both the freshest and the tightest. The earliest is set at 2026-09-01 rather than immediately: a company that had data ready would have released it with the Q2 results on 2026-08-06 rather than promising it later, and the release describes work still in progress (“evaluating supplementary data to inform regulatory discussions and preparation for an End-of-Phase 2 meeting”). The likeliest sits in the last full week of September, the usual landing zone for “later this quarter” guidance given at the start of a quarter’s second month. The latest extends to 2026-11-30 rather than stopping at the quarter end, because this milestone has already moved once and the disclosure is coupled to an FDA interaction whose scheduling Arcturus does not control. Precision is PERIOD because no source names a month, let alone a day. Confidence is MEDIUM, not HIGH: the operational precondition is genuinely met (enrolment and dosing complete), but the milestone has slipped before and depends on a regulatory meeting.

Disagreement. CONSISTENT. The BPIQ placeholder (Q3 2026), the company’s own wording (later this quarter, i.e. Aug–Sep 2026) and the modelled estimate (2026-08-01 to 2026-10-01) all point at the same band; the CT.gov primary completion date of 2026-06-01 sits before it, exactly as a completion date should sit before a readout. No source contradicts another.

One inconsistency exists but is about status, not date, so it does not change the reading above: CT.gov still lists NCT06488313 as RECRUITING while the company stated on 2026-08-06 that enrolment and dosing are complete. Recorded in the data-quality flags.

Date slippage. Parsed from the BPIQ note field and the company’s own releases, oldest first.

Slip count: 1, across five statements and therefore four transitions. Two transitions (2025-11-10 and 2026-03-03) are reiterations of “H1 2026” and are not slips; the fourth (2026-05-07 → 2026-08-06) narrows H2 2026 to Q3 2026, which is a refinement rather than a slip. The one true slip is 2026-03-03 → 2026-05-07, when the Phase 3 alignment milestone moved from H1 2026 to H2 2026.

But read that count alongside what the milestone became. Counting transitions understates the drift here. In July 2025 the promise was to align the Phase 3 trial design with regulators by H1 2026. In August 2026 the promise is to communicate data and a regulatory plan in Q3 2026 — which is a step earlier in the process, not the same milestone delivered late. One counted slip; roughly two quarters of delay; and a milestone that has softened in what it commits to. A reader should weight the softening at least as heavily as the count.

Attribution

Computed by lib/clustering.mjs’s attributionFor over this ticker’s pipeline, for bpiq_drug_id 16237, with CATALYST_CLUSTER_MIN_MONTHS = 6. Transcribed, not estimated.

Status.

Conflicts

Note. Not required for CLEAN. One caveat is recorded anyway, because it is about the input to the computation rather than the computation: ARCT-810 is the only row on this ticker’s pipeline carrying has_catalyst: true, so the conflict set has just one member and CLEAN follows mechanically. ARCT-032’s Q4 2026 Phase 3 go/no-go carries has_catalyst: false and is therefore excluded by construction; with that single flag set true the same computation returns INDETERMINATE at a zero-day gap. The full statement of this is in ../company.md C.2, and the stock-direction call below is written knowing it.

Market and timing for this event

• Plain takeaway. Almost nothing about this event is priced in, in either direction, and that is the defining fact. The company’s entire pipeline is carried at about $34 million of enterprise value against $191.5 million of cash (../company.md C.4). But the direct precedent argues against expecting much: the last time this exact programme reported positive interim data, the shares fell 2.4% on the day.
• Months to this catalyst. About 0.6 months to readout.window.earliest (2026-09-01) from the 2026-08-13 lock, and about 1.4 months to likeliest (2026-09-25). readout.precision is PERIOD, so none of these is a disclosed date — no source names a month, let alone a day, and the window’s edges are a judgement built from the source table above.
• Expected move around this event. ../company.md C.6 records the chain as unusable, so this is a bracket and not a point estimate: an at-the-money straddle on the 2026-09-18 expiry costs between 12% and 31% of the share price depending on whether you transact at the bid or the ask, on 211 open contracts and six traded on the day. Call and put implied volatility at the same strike disagree by 42% relative. The chain contributes nothing to priced-in-ness, and it may expire before the event in any case.
• Nearest comparable past reaction. Not in ../company.md C.7’s table, because BPIQ does not carry the row — but C.7 records it in prose and it is by far the closest analogue: 2025-06-30, ARCT-810’s own positive interim Phase 2 multiple-dose data, which moved the stock −2.4% close-to-close. It is comparable because it is the same asset, the same trial programme and the same kind of announcement. It is not fully comparable in one respect that cuts both ways: that release carried data alone, whereas this one carries data and the regulatory plan, which is the part the market actually needs. The other reference point in C.7 is ARCT-032’s 2025-10-22 interim CF data at −50.2% close-to-close — a reminder of what this company’s clinical data can do to the shares, though that was a larger programme at a much higher price.
• Materiality. ../company.md C.2 records this program as meaningful, not dominant: the nearest catalyst and the lead rare-disease asset, but one of four things carrying the equity alongside ARCT-032, KOSTAIVE and the regained vaccine portfolio. The stock-direction call below is consistent with that — it is not written as though this readout re-rates the whole company.
• Date slippage. See Readout above: one counted slip across five dated statements, with the Phase 3 alignment milestone drifting roughly two quarters and softening in what it commits to.

Spot. $7.97, the close on 2026-08-13, read from the committed price cache through lib/prices.mjs. ../company.md C.4 records BPIQ’s last price the same day as $7.94; the three-cent difference is immaterial and the cache figure is used throughout because it is the source a settlement will read back.

Scenario prices

Expected value. Applying the 38% outcome probability to the midpoint of each range — $11.25 positive, $6.60 miss — gives 0.38 × $11.25 + 0.62 × $6.60 = $8.37, which is +5.0% against the $7.97 spot. Method: probability_pct applied to the midpoint of each scenario range. This is arithmetic, not advice, and it is not a price target. Note what produces it: the expected value sits above spot despite a below-even outcome probability, purely because $191.5 million of cash floors the downside while the upside is unconstrained by a pipeline valued near zero. That asymmetry is the most interesting number in this document, and it is also small enough — 5% on a stock that has moved 31% in the last five sessions — to be inside the noise.

Run-up

readout.precision is PERIOD, not UNKNOWN, so the date_confidence driver does not floor at zero and 01-rules.md rule 39 does not withhold this call. It scores poorly for a related reason, recorded honestly below.

Entry and exit

readout.window.earliest is 2026-09-01, so this rule currently resolves to late August 2026 — about eight trading days after entry. exit is left null on the prediction record: resolveExit needs price history that does not yet exist and returns null today, which is the normal state for a run-up call at lock time.

A short window, said plainly. Eight trading days is a narrow run to trade, and it is a direct consequence of a PERIOD-precision window whose earliest edge is close. The exit rule is stated relative to readout.window.earliest rather than as a calendar date, so if the window moves the rule keeps resolving correctly — which is exactly the situation a program with one prior slip should expect.

Predicted move and peak

Priority score drivers

Five of the seven are transcribed from lib/runup.mjs’s scoreDriver output; the two judgement drivers carry the reading passed into it.

Priority score. Priority score 15 · formula_version 1.0.0

Why 15, when five of the seven drivers read well. The unrounded value is 14.71, and it is not the weighted average of the drivers — that base is far higher. date_confidence at 20 sits below lib/runup.mjs’s untradeable threshold of 30, which caps the result outright at 15 regardless of how strong the other six are. That is the formula working as designed, and the design is right here: a PERIOD-precision window means nobody knows when this lands to better than “some time in the next few weeks”, and you cannot time an entry and an exit around a date you do not have. A score of 15 says do not rank this as a run-up trade, and it says so despite genuinely attractive squeeze mechanics and genuinely low financing risk.

Settlement. Left null at lock time. Settled only on an explicit user request, from a confirmed primary source, against the committed price cache — never on this document’s own initiative. See 05-prediction-protocol.md § Run-up settlement.

Verdict

What I would do. Watch.

Why. The science is sound and the balance sheet is genuinely strong — $191.5 million of cash against a $34 million enterprise value, runway through year-end 2028, no equity sold since 2023 — so the downside is floored in a way that is rare for a company at this stage. But the evidence arriving in Q3 2026 is thin by design: nine patients, open-label, no control arm, with efficacy relegated to secondary endpoints, following the termination of the randomised study that could have answered the question properly. Six years and four trials have produced no peer-reviewed clinical publication and no posted results, so every efficacy claim about this asset traces back to a company press release. And the closest precedent is discouraging in a specific way: when this exact programme reported positive interim data on 2025-06-30, the shares fell 2.4%.

What would change this. A disclosed, quantified rise in ureagenesis AUC across the two-week dosing interval — reported with the trough as well as the peak — accompanied by an explicit statement that the FDA has agreed a registrational path in the paediatric population. That combination would convert the asset’s one real structural advantage (re-dosability, where AAV cannot go) from an argument into a plan, and it is the only thing that would justify paying more than cash for this equity today. The converse also flips it: any treatment-related serious adverse event in a nine-patient study, or a stated requirement for a further Phase 2, ends the near-term case.

What to watch.

• 2026-08-13 to 2026-09-30: the ARCT-810 data and regulatory plan themselves. Read whether the release quantifies ureagenesis change with a pre-specified threshold, or describes it only as “positive” as the 2025-06-30 release did.
• On the day of that release: whether an 8-K accompanies it, and whether the disclosure names an agreed Phase 3 design or only an intention to meet the FDA.
• 2026-09-18: the nearest option expiry to the catalyst. If the event has not landed by then, the chain gives no coverage at all and the timing read weakens further.
• Q4 2026: ARCT-032’s Phase 3 go/no-go decision, guided for the quarter immediately after. It carries has_catalyst: false in BPIQ and so contributes no clustering conflict, but it is the other lead rare-disease asset and it halved the stock once already.
• Ongoing: any 424B5 in the EDGAR feed. None since 2023-08-07, and a takedown from the 2025-12-30 shelf would contradict the low-financing-risk reading this analysis rests on.
• 2027-09: Ultragenyx DTX301’s Phase 3 primary completion — the event that most shapes what commercial space is left in adults.

Locked prediction

• Outcome-direction (will the readout succeed on its primary endpoint?): miss — probability 38%, band 25–52% [UNVERIFIED — modelled]. The reasoning: the safety half of the settlement definition is likely to hold (roughly 85% — four trials since 2020, 63 subjects, no disclosed hold or dose-limiting toxicity), but the regulatory half is not (roughly 45%). A nine-patient, open-label, uncontrolled study whose own registered primary endpoint is safety, run after the termination of a randomised predecessor, and described by the company itself as “exploratory data” being prepared for an End-of-Phase-2 meeting, is a thin basis on which to obtain agreement to a registrational trial without a further Phase 2. 0.85 × 0.45 ≈ 0.38. No third-party probability on this specific event was found to compare against.
• Stock-direction (which way do the shares move?): no-edge — confidence low — window 2026-08-14 to 2026-12-15, basis: from the day after lock through two weeks past readout.window.latest (2026-11-30), so the window covers the event and its reaction even if the disclosure slips out of Q3. Materiality is meaningful, not dominant (../company.md C.2), and this call is consistent with that: a positive result re-rates a pipeline currently carried near zero, but it does not re-rate the whole company, and a miss is cushioned by $6.74 of cash per share.
• Scenario prices: positive $9.50–$13.00 · miss $6.00–$7.20
• Expected value: $8.37, +5.0% against spot $7.97 (arithmetic, not advice)
• Run-up: entry $7.97 on 2026-08-13, exit rule T-5 trading days before readout.window.earliest — predicted move −5% to +12%, predicted peak 0% to +18% around 2026-08 — priority score 15, formula_version 1.0.0
• Settles on the Arcturus disclosure of the ARCT-810 Phase 2 data and regulatory plan guided for Q3 2026. Positive requires both: (a) no new dose-limiting toxicity or treatment-related serious adverse event attributed to ARCT-810 that halts or restricts dosing; and (b) Arcturus states it will advance ARCT-810 into a registrational or Phase 3 study in OTC deficiency (adult and/or paediatric) following its End-of-Phase-2 interaction, without an FDA clinical hold and without a stated requirement for an additional Phase 2 study first. Source: Arcturus press release or Form 8-K.
• Locked: yes · Settled: no

On no-edge sitting beside a +5.0% expected value. The two are compatible and the divergence is explained rather than hidden. The expected value is positive because $191.5 million of cash floors the miss scenario while the positive scenario is unconstrained by a pipeline the market carries near zero — a structural asymmetry, not a directional view. Five percent is inside the noise for an equity that moved 31% in the five sessions before this lock, and the direction is declined because the two forces genuinely offset: a below-even probability of the pre-registered positive definition being met, against a downside cushioned by cash. no-edge is the honest word for that, and it is more useful than a coin-flip dressed as a view.

Program data-quality flags

1. CT.gov’s registry status for NCT06488313 is stale. The registry lists the trial as RECRUITING with a primary completion date of 2026-06-01 that has already passed, while Arcturus stated on 2026-08-06 that enrolment is complete and all enrolled subjects have completed dosing. The company statement is the more recent and more specific, and is what the readout window is built on. The registry’s dates are used as recorded; only its status field is treated as out of date. [VERIFIED — CT.gov get_trial_details NCT06488313, read 2026-08-13; Arcturus Q2 2026 press release, 2026-08-06]
2. The BPIQ catalyst_date of 2026-09-30 is a synthesized period-end placeholder for the text “Q3 2026” and is never used for timing here, per 01-rules.md rule 23. It survives only as the bpiq row of the readout source table.
3. BPIQ’s drug-name field carries a trailing space, "ARCT-810 ", the dirty-name pattern 02-connectors.md records. Resolution was keyed on bpiq_drug_id 16237 throughout, never on the name.
4. BPIQ carries no historical-catalyst row for the 2025-06-30 ARCT-810 interim data release — the single most relevant precedent for this event. Its only ARCT-810 row is the 2024-07-01 enrolment milestone. The 2025-06-30 reaction was reconstructed from the press feed and the committed price cache instead, and is recorded in ../company.md C.7 in prose. A reader relying on the connector’s historical-catalyst list alone would not know the most informative precedent exists.
5. search_investigators queried by name returned zero rows for an investigator the trial record names. Dr Summar is listed on NCT06488313 as a site contact, and the tool searches overall-official and responsible-party fields. Queried by condition instead it returned him, along with 66 mostly irrelevant people whose trials matched “OTC” as optical coherence tomography, ovarian tissue cryopreservation and “outside the cage” robotic surgery — the same short-code index confusion 02-connectors.md records for PubMed and “MN-001”. Identity was therefore taken from get_trial_details.
6. No efficacy threshold is pre-specified anywhere in the public record. The trial’s ureagenesis secondary endpoints are stated as “change from baseline” with no target magnitude, and Arcturus has disclosed none. This is why the settlement definition above is written around the regulatory consequence (an agreed registrational path) rather than around a numeric efficacy bar: a bar the sponsor defines after seeing the data is not scoreable by a third party.
7. Two connector failure modes recorded in 02-connectors.md did not fire on this program, which is worth recording so the file is not read as universal. ChEMBL returned a clean empty result rather than the HTTP 500 seen on CNTB and MNOV on this same date, and Open Targets answered normally rather than returning its standing Rate limit exceeded for client: global. PubMed’s authors-as-nulls bug also did not fire: complete author lists with affiliations returned on all ten articles.
8. Third-party market sizing was rejected rather than passed through, per 01-rules.md rule 6. A retrieved report quotes a $1,284 million global OTC deficiency treatment market for 2025, which is irreconcilable with a few thousand diagnosed patients and carries no derivation. B.2 records the rejection and B.3a is built bottom-up instead.
9. Aggregated analyst figures were rejected in favour of named, dated ones. A web search returned a $68 high price target and an H.C. Wainwright cut “to $60 from $63” — figures belonging to this stock’s 2021-era price level, not to a $7.94 share. The analyst anchor used in the scenario table is instead H.C. Wainwright’s Hold at $9 dated 2026-08-11, taken from the company press feed. The current named, dated cluster for context: H.C. Wainwright Hold $9 (2026-08-11), BTIG Buy $23 (2026-08-10), CCORF Buy $20 (2026-08-07), Roth MKM Buy $20 (2026-07-16), Freedom Capital Buy $27 (2026-07-02), Citigroup $8 (2026-03-04), and Wall Street Zen at Sell (2026-08-09) — a dispersion from $8 to $27 that is itself the finding.