CNTB / rademikibart-acute-copd-exacerbation — Rademikibart for acute exacerbations of chronic obstructive pulmonary disease

Program analysis · bpiq_drug_id 18937 · 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

Executive summary

• What it is (one sentence): A single injection of an anti-inflammatory antibody, rademikibart, given on top of the usual steroid-and-antibiotic treatment at the moment a COPD patient turns up at hospital with a flare-up, tested against placebo in 159 patients whose blood shows the specific inflammation the drug blocks.
• The event and when (as disclosed): Topline results from the Phase 2 Seabreeze STAT COPD study. The company said on 2026-08-12 that it expects to report topline from both Seabreeze STAT studies “in September 2026” [VERIFIED — exhibit 99.1 to Form 8-K, 2026-08-12]. No source names a day. The readout window used throughout this document is 2026-09-01 to 2026-10-31, most likely late September.
• The main reason it could work: The target is already proven to matter in exactly this patient group. Dupilumab, an antibody against the same receptor, cut COPD flare-ups by 30–34% over a year in patients with the same eosinophil threshold, and was approved on that basis. Rademikibart’s own Phase 2b asthma trial showed lung-function gains that began within Week 1 of the first subcutaneous dose, which is the specific property an acute setting needs.
• The main risk: Nobody has ever shown that blocking type 2 inflammation changes what happens in the 28 days after a flare-up has already started. Every piece of supporting evidence — dupilumab’s, mepolizumab’s, and rademikibart’s own — comes from months of continuous dosing to prevent future events, not from one dose given into an event in progress. On top of that, 159 patients is too few to settle a binary endpoint: the company itself says the study exists to “help determine a potential Phase 3 endpoint and sample size”, which is an admission that it is not powered to be definitive.
• What it means for the stock: The expected value is modestly positive — roughly +8% against a spot of $2.56 — and that is a much weaker case than the headline asymmetry suggests, because the probability of the primary endpoint being met is judged below even. The bigger caveat is that this readout cannot be traded on its own: the asthma study reads out in the same month, and the two toplines may arrive in one press release. Watch, do not chase.

0. Program-tier coverage — CLEARED

One row per mandatory tool in the program-tier table in 02-connectors.md, in the order that file lists them. Company-tier coverage is in ../company.md C.0 — CLEARED, swept 2026-08-13, reused here under 02-connectors.md § Company-sweep currency (swept the same day, no intervening earnings release or financing).

The regulatory tier was not called at all, which is correct rather than a gap: 02-connectors.md makes that tier conditional on a regulatory catalyst, and this catalyst is a clinical-trial readout.

Optional rows, recorded honestly. CT.gov search_investigators was called for the KOL block (02-connectors.md § KOL sources) and returned 23 investigators across 40 COPD-exacerbation trials — not one of them on NCT06940154. get_trial_details on that trial likewise returned every site with contacts: null and no overall official. That is a finding, not a failure, and A.5b records it as one. Europe PMC, CTIS, EDGAR full-text search and NIH RePORTER were not called; all four are optional and none carries a NOT CALLED state.

A. Scientific & clinical assessment

A.1 Mechanism of action and scientific rationale

What the drug is. Rademikibart is a human monoclonal antibody — a laboratory-made copy of one immune protein, engineered to stick to a single target and nothing else. Its target is the interleukin-4 receptor alpha chain (IL-4Rα), a docking point on the outside of cells. It is an IgG4 kappa antibody, given as an injection under the skin [VERIFIED — Wang et al., Clin Transl Sci 2023, DOI 10.1111/cts.13656]. ChEMBL confirms the identity and the class: CHEMBL5314768, molecule type Antibody, USAN stem -bart [VERIFIED — ChEMBL compound_search 2026-08-13].

How it works, in plain steps. Two inflammatory signalling molecules, interleukin-4 (IL-4) and interleukin-13 (IL-13), both have to dock at IL-4Rα to do anything. When they dock, they flip an internal switch called STAT6, and the cell responds by pulling eosinophils into the airway, making more mucus, and swelling the airway wall. All three narrow the airway. Rademikibart sits on IL-4Rα so that neither IL-4 nor IL-13 can dock. One antibody therefore blocks two signals.

Why this matters in a flare-up specifically. A COPD flare-up is a sudden worsening of that same narrowing. The premise of this program is that in the subset of patients whose flare-up is being driven by type 2 inflammation — identified by a blood eosinophil count of 300 or more per microlitre at the visit — switching the driver off at the moment of the flare-up should make the patient recover faster and relapse less over the following four weeks.

flowchart TB
  A["COPD airway with type 2 inflammation<br/>blood eosinophils 300+ per microlitre"] --> B["IL-4 and IL-13 released"]
  B --> C["Both cytokines signal through the shared<br/>IL-4 receptor alpha chain (IL-4R alpha)"]
  C --> D["Eosinophils recruited into the airway;<br/>mucus production rises; airway wall swells"]
  D --> E["Acute exacerbation: breathlessness,<br/>airflow drops, urgent healthcare visit"]
  E --> F["Standard care: systemic corticosteroid<br/>plus antibiotic and bronchodilator"]
  F --> G["Treatment failure within 28 days<br/>in a substantial minority of patients"]

  H["Rademikibart: one subcutaneous dose,<br/>added on top of standard care"] --> I["Binds IL-4R alpha and blocks<br/>STAT6 signalling inside the cell"]
  I --> C
  I --> J["Intended benefit: faster recovery of<br/>lung function and fewer treatment<br/>failures by day 28"]
  G -.->|"the endpoint the trial must move"| J

  style H fill:#e8f0fe,stroke:#4a6fa5,stroke-width:2px
  style I fill:#e8f0fe,stroke:#4a6fa5,stroke-width:2px
  style J fill:#e6f4ea,stroke:#3d8b5f,stroke-width:2px
  style G fill:#fce8e6,stroke:#b3453a,stroke-width:2px

How well is the target validated? In this exact population, unusually well for a Phase 2 asset.

• Dupilumab, an antibody against the same receptor, reduced moderate-to-severe COPD exacerbations by 30–34% against placebo over 52 weeks in patients with blood eosinophils of 300 or more per microlitre, in the BOREAS and NOTUS Phase 3 trials, and was approved by the US Food and Drug Administration on 2024-09-27 on that basis — the first biologic ever approved in COPD [WEB ESTIMATE — Sanofi press release 2024-09-27 and HCPLive/AJMC coverage, via web search 2026-08-13]. That is a randomised, placebo-controlled demonstration that blocking IL-4Rα changes COPD exacerbation biology in biomarker-selected patients. It is the single strongest fact supporting this program.
• Mepolizumab, targeting a different type 2 molecule (IL-5), cut annualised exacerbations by 21% in the MATINEE Phase 3 trial and was approved in COPD on 2025-05-22 [WEB ESTIMATE — GSK press release and AJMC/Pharmacy Times coverage, via web search 2026-08-13]. A second, independent confirmation that the type 2 pathway is causal in COPD, through a different molecule.
• The GOLD 2026 global COPD treatment guidelines formally added biologic therapy to the treatment pathway for patients on triple inhaled therapy who keep exacerbating with eosinophils at or above 300 [WEB ESTIMATE — cmetravelacademy.com summary of GOLD 2026, via web search 2026-08-13]. The mechanism is now guideline-endorsed, not speculative.

The exact scientific step this readout must prove, stated as narrowly as possible. Every one of the validating facts above is about chronic maintenance dosing over months to prevent future exacerbations. Not one of them says anything about giving the drug into an exacerbation that has already begun and measuring what happens over the next 28 days. That is the step this trial exists to take, and it is genuinely new. The scientific bet has two halves:

1. Speed. The drug has to do something clinically useful within days, not months. The direct evidence for this is rademikibart’s own Phase 2b asthma trial, where prebronchodilator FEV1 improvements “occurred rapidly during Week 1” and were sustained through Week 24 [VERIFIED — Kerwin et al., Am J Respir Crit Care Med 2025, DOI 10.1164/rccm.202409-1708OC]. That evidence is on the same subcutaneous route this trial uses, which matters: it is not borrowed from the intravenous formulation. The company has separately reported 200–400 mL FEV1 gains maintained to Day 29 in a Phase 1 intravenous study in asthma and COPD patients [VERIFIED — CNTB press feed 2026-03-30 and 2026-03-31, reiterated 2026-05-12; that study is pipeline rows 19715 and 19716 in ../company.md C.2], but that study was open-label and not in an acute setting, so it is supporting rather than decisive.
2. Relevance to the composite. Even if lung function improves quickly, the primary endpoint is not lung function. It is a composite dominated by whether the patient comes back. A large share of COPD exacerbations are triggered or sustained by bacterial and viral infection, and blocking type 2 inflammation does nothing about that in any patient, eosinophil-high or not. The drug can be pharmacologically successful and still fail to move this endpoint.

The honest scientific risk. The strongest reading is that this is the right target, in the right biomarker-selected patients, with a molecule that acts fast enough. The weakest reading is that 28-day treatment failure after an exacerbation has begun is largely determined by infection, airway remodelling and adherence — none of which this drug touches — and that the type 2 component was already suppressed by the systemic corticosteroid every patient in both arms receives as standard of care. That last point deserves naming plainly: the control arm is not untreated. Systemic steroids are themselves broad anti-inflammatories that suppress eosinophils, and they already reduce 28-day treatment failure by roughly 10 percentage points against placebo [WEB ESTIMATE — Cochrane review summary via TheNNT and the 1999 NEJM SCCOPE trial, web search 2026-08-13]. Rademikibart has to improve on a control arm that is already receiving a drug working partly through the same downstream pathway. That is the specific reason this analysis puts the probability of a positive primary-endpoint result below even, despite strong target validation.

One structural point in the drug’s favour, and its limit. A crystal structure of the rademikibart Fab bound to IL-4Rα at 2.71 Å, compared against dupilumab’s at 2.82 Å, shows the two antibodies bind at a 54.88° rotation from one another, and that rademikibart’s epitope overlaps more closely with the interface IL-4 and IL-13 use themselves; molecular-dynamics work adds a direct contact with a receptor loop (residues 145–153) absent in the dupilumab complex [VERIFIED — Shi et al., bioRxiv 2026-04-13, DOI 10.64898/2026.04.12.718052]. The limit, stated rather than glossed: this is a preprint, not peer-reviewed, and its senior author’s co-author list includes Connect Biopharma’s Chief Medical Officer. Higher binding affinity is a laboratory property; it has never been shown to translate into a better clinical outcome for this molecule against dupilumab in any head-to-head trial, and no such trial exists.

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

gantt
  title Rademikibart clinical development, with competitor precedent in COPD
  dateFormat YYYY-MM-DD
  axisFormat %Y-%m
  section Rademikibart
    Ph2b asthma n=322 (CBP-201-WW002)   :done,   a1, 2021-05-11, 2023-08-05
    Ph1 IV airway function, asthma/COPD :done,   a2, 2025-06-01, 2026-03-30
    STAT Asthma Ph2 n=160               :active, a3, 2025-08-08, 2026-07-17
    STAT COPD Ph2 n=159 (this program)  :active, a4, 2025-08-12, 2026-07-30
    STAT COPD registry completion       :milestone, m1, 2026-08-27, 0d
    Readout window, earliest to latest  :crit,   a5, 2026-09-01, 2026-10-31
    STAT IV Ph2 open-label n=40         :        a6, 2026-07-30, 2026-12-31
  section Competitor precedent
    BOREAS + NOTUS, dupilumab Ph3       :done,   b1, 2019-05-01, 2023-11-01
    Dupilumab approved in COPD          :milestone, m2, 2024-09-27, 0d
    MATINEE, mepolizumab Ph3            :done,   b2, 2019-01-01, 2024-09-01
    Mepolizumab approved in COPD        :milestone, m3, 2025-05-22, 0d

Trial milestones, where the registry discloses them. First patient in: 2025-08-12 (registry start_date). Last patient completing the primary-endpoint assessment: 2026-07-30 (registry primary_completion_date). End of the 56-day safety follow-up: 2026-08-27 (registry completion_date) [VERIFIED — CT.gov get_trial_details NCT06940154, read 2026-08-13]. Enrolment completed at 159 participants against a planned 160 [VERIFIED — exhibit 99.1 to Form 8-K, 2026-08-12: "The study has enrolled 159 participants globally with an eosinophil count of ≥300 cells/μL"]. Database lock and topline are not disclosed by any source; see the Readout section for how the window is built from what is disclosed.

Trials

Feasibility matrix

Resourcing sufficiency. The company has the people, the sites and the cash to reach this readout, and the question of whether it can fund what comes after is the live one. Cash, cash equivalents and short-term investments were $31.486M at 2026-06-30, against a recomputed burn of $5.321M a month — see ../company.md C.3, which also records that the company’s own filed statement (“at least one year”) and that recomputation disagree by roughly eight months and that this document does not resolve the disagreement. For the narrow question “does the cash reach a September readout?”, both readings answer yes with months to spare. For “can it fund a Phase 3 in COPD?”, the answer on current cash is plainly no: a registrational programme in acute COPD exacerbations would need several hundred patients per arm across two trials, and neither the $31.5M on hand nor the remaining Simcere milestones (up to approximately $99M, none of it committed to a date) covers that. A raise, a partner, or both, sits between this readout and Phase 3. That is not a criticism of the plan; it is the reason the run-up section below treats the runway as tight rather than comfortable.

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

A.3a Target label. Rademikibart as an add-on to standard of care for the treatment of an acute exacerbation of COPD in adults with type 2 inflammation, defined by a blood eosinophil count of 300 or more cells per microlitre at the time of the exacerbation.

A.3b Target product profile

A.3c Strategic Go/No-Go questions.

The set that matches this asset’s next decision is the Pre-Phase-III set: Phase 2 enrolment is complete, and the company has stated that the decision immediately after topline is an FDA meeting “to gain alignment on a Phase 3 program”. The Pre-Phase-II set is answered first and briefly, retrospectively, because those gates were passed rather than skipped.

Pre-Phase-II (Go-to-Phase-II) — answered retrospectively:

Pre-Phase-III (Go-to-Phase-III / registration) — the live set:

A.3d Regulatory designations.

None disclosed. No Breakthrough Therapy designation, Fast Track designation, Orphan Drug designation, Priority Review voucher or equivalent was found for rademikibart in this or any indication, in the BPIQ pipeline record, the 152-item company press feed, the ClinicalTrials.gov record, or the four web searches run this sweep [UNVERIFIED — searched and not found; the absence of a designation in these sources is weaker evidence than a company statement that there is none].

The one regulatory interaction on record is not a designation: the company announced a positive Type C meeting with the FDA for rademikibart on 2025-04-01 [VERIFIED — CNTB press feed 2025-04-01]. A Type C meeting is a general guidance meeting a sponsor can request at any point in development; it grants nothing — no faster review, no fee relief, no exclusivity — and confers only the FDA’s written advice on the questions asked. It is recorded here so a reader does not mistake it for a designation.

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.

A.5 Evidence quality and endpoints

Endpoints

A.5b Key opinion leaders.

Panel as of. 2026-08-13 — the date the investigator search (CT.gov search_investigators and get_trial_details) and the independent-voice search (PubMed search_articles on the molecule, then get_article_metadata on all 14 hits) were run.

Investigators

No investigators were returned for this program’s pivotal trial, and that is the finding. get_trial_details on NCT06940154 returned all 52 sites with contacts: null and no overall official named. search_investigators on COPD Acute Exacerbation analysed 40 trials and returned 23 named people, none of them on NCT06940154; the only Connect Biopharma trial it reached was the separate intravenous study, NCT07705737, where Radha Adivikolanu is listed as a site contact at Columbus Clinical Services. Connect Biopharma has chosen not to publish investigator names on the registry record for this trial, so there is no list to conflict-check.

Independent voices

None found, and the search that establishes that is worth stating. Three named people have published substantively on rademikibart’s respiratory or structural biology, and every one of them fails the independence test:

• Michael E. Wechsler, National Jewish Health, first author of the Chest 2026 eosinophil analysis — co-authored with two Connect Biopharma employees (Barry Quart and Raúl Collazo), and National Jewish Health is itself a listed site on NCT06940154. Two disqualifications, either of which alone would be enough.
• Edward Kerwin, Clinical Research Institute of the Allergy and Asthma Center of Southern Oregon, first author of the AJRCCM 2025 Phase 2b asthma trial — an investigator on the sponsor’s own trial, co-authoring with seven Connect Biopharma employees.
• Christopher G. Bunick, Yale School of Medicine, senior author of the crystal-structure preprint and of the JID 2025 comparative review — the preprint is co-authored with Connect Biopharma’s Raúl Collazo.

Four further author groups wrote about rademikibart with no sponsor co-author — Vieira et al. (Australas J Dermatol 2026), Cai et al. (Pharmacoepidemiol Drug Saf 2025), Babul et al. (Cureus 2026) and Obed et al. (Expert Opin Emerg Drugs 2024) — and all four are meta-analyses of atopic dermatitis efficacy. They are independent about the molecule and silent about this program’s endpoint, so none of them is a commentator on the endpoint at issue and none is recorded below. In a first-in-setting indication with one sponsor and no published trial, that outcome is ordinary rather than suspicious.

A limit on every conflict check above, stated because it changes what “checked” means here. PubMed’s get_article_metadata returns no conflict-of-interest field at all on any of the 14 articles. The independence assessments above are therefore made from author affiliations, which the connector does return in full, and not from the journals’ own disclosure statements, which it does not. Affiliation catches employment and it catches the trial-site relationship; it does not catch an advisory-board seat, a consulting arrangement or equity in a competitor. conflicts_checked records where the search was made, never what it failed to find (01-rules.md rule 40).

Judgement

Dissent

Empty, and correctly so: endpoint_supported is UNKNOWN rather than SUPPORTIVE_CONTESTED, so there is no stated position for a named voice to disagree with. Inventing a dissenter here would be inventing a view nobody holds.

B. Commercial assessment

B.0 Current treatment algorithm

What a COPD patient receives today, line by line.

1. Long-term maintenance, before any flare-up. Inhalers, escalating from one bronchodilator to two, then to “triple therapy” — two bronchodilators plus an inhaled steroid in a single device. This is the backbone for essentially every patient with symptoms.
2. Long-term maintenance, for patients who keep flaring despite triple therapy. Since 2024, a biologic added on top: dupilumab if eosinophils are ≥300/μL, or mepolizumab if ≥150/μL. GOLD 2026 formally added this step [WEB ESTIMATE — GOLD 2026 summary, web search 2026-08-13]. Ensifentrine and long-term azithromycin are alternatives at the same step.
3. The flare-up itself. A short course of systemic corticosteroids (typically five days), an antibiotic if the sputum suggests infection, and increased bronchodilators. Oxygen and ventilatory support if the patient is sick enough to need them. This step has not materially changed in three decades, and nothing in it is biologic.

Exactly where rademikibart would fit: step 3, and only step 3. It would be given once, at the urgent-care or emergency visit, on top of the steroid course the patient is already getting. It displaces nothing. It adds a line to a step that currently has no targeted option at all.

That positioning is the commercial point, and it cuts both ways. It means rademikibart is not competing with dupilumab — a patient could plausibly receive dupilumab as maintenance and rademikibart at a flare-up, and Sanofi’s own franchise is untouched by a positive result here. It also means the company is trying to create a market rather than take share in one, which is harder, slower and needs guideline change to reach scale.

B.1 Competitive landscape and positioning

Where this asset wins, and the single fact the thesis rests on.

It wins on timing within the patient journey, not on molecular superiority. The maintenance biologic market in COPD is occupied by two approved products from Sanofi/Regeneron and GSK, and a fourth-to-market anti-IL-4Rα antibody would have no route into it. The acute setting is empty. Rademikibart is the only asset visible in this sweep attempting it, and the mechanistic reason it can is speed: FEV1 gains inside Week 1 of the first subcutaneous dose.

The single fact the thesis rests on is therefore this: that blocking IL-4Rα produces a clinically meaningful benefit within days rather than months. If that is true, the acute indication is real, the maintenance incumbents are irrelevant to it, and the company owns a category. If it is false, this program has no reason to exist, because as a maintenance drug rademikibart arrives years behind two approved competitors with no demonstrated advantage over either.

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.

B.2 Addressable market

Launch markets. The United States first — it is where the trial’s largest site cluster sits, where the company is headquartered, where readmission penalties give payers a direct reason to fund an admission-avoiding drug, and where dupilumab and mepolizumab were approved first. The EU5 and Japan would follow. Greater China is excluded from every figure in this document: rademikibart is exclusively licensed there to Simcere for all indications, and Connect’s economics in that territory are milestones and royalties, not sales — see ../company.md C.3.

B.3 Value and feasibility

B.3a Expected peak sales.

Scope. The build below is for the acute COPD exacerbation indication only — the indication this catalyst is about. It excludes the acute asthma indication (a separate trial and a separate readout, analysed at ../rademikibart-acute-asthma-exacerbation/), any maintenance COPD or asthma indication (no trial running), Greater China entirely (licensed to Simcere), and the atopic-dermatitis franchise. Every figure is conditional on approval.

A third-party figure declined, and why. The company’s own market research forecasts peak sales “upwards of $5 billion” globally for rademikibart, and at least one sell-side initiation has cited peak-sales potential above $5 billion [WEB ESTIMATE — company commentary and sell-side initiation coverage via web search 2026-08-13]. That figure is not passed through, for a specific reason rather than general scepticism: it covers acute and chronic asthma and COPD together, which means it embeds rademikibart competing head-to-head with dupilumab in maintenance — a market this readout does not address and in which the company is running no trial. Quoting it against a readout about single-dose acute treatment would answer a different question from the one asked (01-rules.md rule 6).

Which input is least defensible: the price. Patient and event numbers can at least be bounded by epidemiology, and penetration by analogy to other injectables in urgent settings. There is no drug priced for a single acute dose in this setting anywhere in the world, so the $2,000–$6,000 range is an inference from a maintenance list price and an avoided-admission cost, not a benchmark. A reader who disagrees with the price should scale every row in the table proportionally. The event pool is the second-weakest input, for the reasons given in B.2.

Against market value. Enterprise value is $129.72M — recomputed as 62,974,309 shares (EDGAR, 2026-07-31) × $2.56 (the 2026-08-13 close from the committed price cache), less $31.486M of cash and short-term investments, with no debt; see ../company.md C.3 and C.4 for the underlying figures and why BPIQ’s own market_cap and enterprise_value fields are discarded. The base case of roughly $0.35B of annual peak sales for this one indication is about 2.7× enterprise value, and the low case about 0.2×. Those are ordinary multiples for a Phase 2 asset, not remarkable ones. Read plainly: the stock is not obviously mispriced against a risked view of this indication alone. What makes the setup interesting is the optionality stacked around it — the asthma readout in the same month, the intravenous presentation, and the China royalty stream — not a valuation gap on this program.

B.3b Feasibility

B.4 Product-development risk

Framing questions.

1. Does the development plan support the target product profile claims? Partly. The plan supports the efficacy and safety claims — a randomised, double-blind, placebo-controlled trial with an objective composite primary endpoint is the right instrument for them. It does not yet support the payer-value claim, because no health-economic endpoint and no formal quality-of-life instrument beyond EXACT-PRO is in the trial, and the argument that the drug pays for itself by avoiding admissions will eventually need evidence the current plan does not generate.
2. Will the identified risks affect the target product profile? The dominant risk — statistical power — does not change the profile the drug could have; it changes the chance this particular trial demonstrates it. A 159-patient study can miss a real effect. The risk that would change the profile is the endpoint risk: if the FDA declines to accept 28-day treatment failure as a registrational endpoint, the target label itself has to be rewritten around whatever endpoint it will accept, and the whole commercial case in B.3a is rebuilt from a different base.
3. If a risk cannot be mitigated, is the asset still differentiated from competitors? In the acute setting, yes — nothing else is there. In any other setting, no: as a maintenance drug rademikibart is a fourth-to-market anti-type-2 antibody with no clinically demonstrated advantage over dupilumab. The differentiation lives entirely in the indication, which is why a failure here removes most of the argument for the equity rather than redirecting it.

Readout

Sources

The modelled estimate. CT.gov’s primary_completion_date for NCT06940154 is 2026-07-30. Adding the stated default lag to database lock and analysis of two to four months (01-rules.md rule 34) gives 2026-09-30 to 2026-11-30. The tag is [UNVERIFIED — modelled, default lag] rather than benchmarked because data/benchmarks/readout-lag.json holds zero observations — no comparable trial’s lag has been confirmed in this repository yet, so there is nothing to benchmark against. Two observations about how this range sits against everything else in the table, stated so the window judgement below can be argued with: it opens at the very end of the month the company named, and it runs two months past where the company expects to finish. That is what a generic default does to a trial with an unusually short endpoint. The default is calibrated on studies where database lock follows long, complex follow-up; here the primary endpoint is a 28-day binary composite in 159 patients and the entire safety follow-up closed on 2026-08-27. The window below therefore leans on the company’s guidance for its early edge and borrows only partially from this range for its tail.

Window

Basis. The earliest edge is the first day of the month the company itself named in a filed exhibit one day before this analysis was locked, with the trial’s 56-day safety follow-up already closed on 2026-08-27 — nothing operational stands between the lock date and a first-week announcement. The likeliest date is set in the second half of September rather than the middle, which is where this program differs from its asthma sibling: the company’s 2026-06-17 wording put the asthma readout in “early September” and said the COPD topline would come “soon after”, and the COPD trial’s own primary completion (2026-07-30) is thirteen days later than the asthma trial’s (2026-07-17). The latest edge sits one month past the guided month rather than at the modelled range’s 2026-11-30 tail, because that default lag is calibrated on trials with far longer follow-up than a 28-day endpoint whose safety window has already closed — while this program’s two prior slips are real and argue against treating the company’s month as a hard ceiling. Precision is MONTH because a source names a month and none names a day. Confidence is MEDIUM rather than HIGH: enrolment is complete, the endpoint is short and fixed, and the company reaffirmed one day before the lock date, but the date has slipped twice and the answer is still a month rather than a day.

Disagreement. CONSISTENT. Every dated source points at September 2026: BPIQ’s synthesized 2026-09-30 sits at the end of it, the company names it outright, and the registry’s completed primary-completion and completion dates are both consistent with a September announcement. The modelled range runs later than the rest but overlaps them rather than contradicting them, and its divergence is explained by the generic lag rather than by a conflicting disclosure. No source was picked over another and none was averaged.

Date slippage. Two slips across nine dated statements, parsed from the BPIQ note field on row 18937 and cross-checked against the company press feed. Both occurred while enrolment was the binding constraint, which it no longer is.

Attribution

Status. CONTAMINATED

Conflicts

Note. The conflicting row is the Seabreeze STAT Asthma study (NCT06940141, n=160), and this is not a coincidence of two placeholders touching. The gap is zero days — the two windows overlap outright — and the conflict is confirmed because both sides rest on a disclosed date: each program’s own readout.window carries MONTH precision.

The contamination is not merely arithmetic, and understating it would be the easy mistake here. The company’s own words in a filed exhibit are that it expects “to report topline results from both Phase 2 Seabreeze STAT studies in September 2026”, and it frames the pair as one inflection. The two studies share a sponsor, a molecule, a mechanism, a primary endpoint definition, a key secondary endpoint, an eosinophil threshold and many of the same clinical sites; the two toplines may well appear in the same press release. Consequence, stated rather than glossed: no price move in this window can be read as belonging to the COPD result alone. A positive COPD result paired with an asthma miss would not deliver the positive range below, and a COPD miss paired with a positive asthma result would not deliver the miss range.

Two further points a reader needs in order to size that correctly. First, the correlation between the two results is high but not perfect: same drug, same mechanism, same endpoint, so both-positive and both-negative are the most likely joint outcomes — but COPD patients here are 40–80 years old and are by definition smokers or ex-smokers with at least ten pack-years, whom the asthma trial excludes, and type 2 inflammation explains a smaller share of COPD than of asthma. Second, the asthma study is the larger commercial prize on the sponsor’s own framing and the stronger prior for this molecule, since rademikibart has a positive randomised asthma trial behind it and nothing in COPD. The scenario ranges below are therefore written for the two-study package weighted toward the COPD result, and both ends are set wider than an attributable single-program readout would need.

Computed: lib/clustering.mjs attributionFor('CNTB', companyRecord, programs, 6, 18937), run over this ticker’s full pipeline table on 2026-08-13 and transcribed rather than judged by eye. The two other has_catalyst: true rows on this ticker (19715 and 19716) contribute no window because their catalyst_date is null, so they never enter the comparison — see ../company.md C.8 row 10, which records that both flags are stale against a delivered 2026-03-30 event.

Market and timing for this event

• Plain takeaway. A 19-day wait to the opening of a month-wide window, on the second of two toplines that will land in the same month and possibly in the same release. The stock sits at 51% of its 52-week range after rising 29% since mid-June, short interest has nearly doubled since June and then gone flat, one director has bought in the open market above today’s price, and the options chain cannot price anything. The expected value is positive but modest — about +8% against spot — which is a far weaker case than the headline range widths suggest, because the probability of this endpoint being met is judged below even.
• Months to this catalyst. 0.6 months — 19 days from 2026-08-13 to readout.window.earliest of 2026-09-01, measured from the earliest edge and never from catalyst_date (01-rules.md rule 23). Said plainly, as the template requires: readout.precision is MONTH, not DAY, so this is the distance to the opening of a month-wide window, not to a known event date. The likeliest date is 39 days out and the latest edge 79 days out.
• Expected move around this event. A bracket, not a point estimate, because ../company.md C.6 records the chain as unusable — three strikes ($2.50, $5.00, $7.50), no strike at or below spot, 22 contracts of daily volume, one leg of the money unquotable and two implied-volatility artifacts present. The bracket carried there is roughly ±35% to ±65%, inferred from the 2026-09-18 $2.50 call quoted at $0.35–$0.60. Two cautions specific to this program: that bracket was inferred against a $2.33 reading and this document’s spot is $2.56, which moves the $2.50 strike from 7% above the money to 2% below it; and the 2026-09-18 expiry covers this window’s earliest and likeliest edges but not its latest one — the first expiry that covers the whole window is 2026-12-18. Used as a sanity check on the scenario ranges below and never as an input to them.
• Nearest comparable past reaction. ../company.md C.7’s 2024-05-22, +11.73% — Phase 2b asthma data on this same molecule presented at ATS, showing lung-function improvement to Week 24. It is the closest analogue by subject matter, and it is only partly comparable: it was a presentation of data already known to be positive, not a blind binary readout, so it measures the market’s appetite for good news on this molecule rather than its reaction to news that could go either way. The second-closest, 2025-06-13 at +13.21% (EAACI 2025), has the same limitation. Nothing in C.7 is a comparable negative reaction at all — the one large decline, −19.89% on 2026-03-30, was bundled with a $20.2M private placement struck below the prior close — so the downside side of this ticker’s record is unpopulated rather than reassuring, and a reader should not take the absence of a big down day as evidence this stock does not fall on bad news.
• Materiality. Material and shared — the larger commercial prize of the two, the weaker prior — the wording recorded for this program in ../company.md C.2. COPD is the bigger market of the two acute indications and the one with approved-biologic precedent, but this molecule has a positive randomised asthma trial behind it and nothing at all in COPD. The stock-direction call below is made consistent with both halves per 01-rules.md rule 27: direction up, confidence low, with the attribution caveat carried explicitly in its basis rather than left to be inferred.
• Date slippage. Two slips across nine dated statements (see Readout § Date slippage). Both occurred while enrolment was the binding constraint, which it no longer is.

Spot. $2.56, the close on 2026-08-13, read from the committed price cache data/prices/CNTB.json through lib/prices.mjs closeOnOrBefore [VERIFIED — data/prices/CNTB.json, fetched 2026-08-13, source Yahoo Finance chart API v8].

A disagreement with the company document, named rather than smoothed over. ../company.md C.1 records $2.33 for the same day, from BPIQ’s last_price field read during the company sweep earlier on 2026-08-13. The two figures are both correct about different moments: the cache’s bar for 2026-08-13 opens at $2.33 and closes at $2.56, with a high of $2.64, so BPIQ’s “last price” was the day’s opening print, read mid-session before the company’s price cache existed for this ticker. This document uses the close, for one concrete reason rather than a preference: 05-prediction-protocol.md requires a run-up settlement to be read from the committed price cache and never from a fresh fetch, so an entry price that does not appear in that cache could never be settled against it. C.1’s figure is left unchanged because the sibling program locked its own prediction against it, and rewriting a figure another locked record cites would corrupt that record rather than correct it. See ../company.md C.8 row 8, updated by this run to record that the cache now exists.

Scenario prices

Expected value. Applying the locked 30% probability to the midpoint of each range: 0.30 × $5.50

• 0.70 × $1.60 = $2.77, which is +8.2% against the $2.56 spot. This is arithmetic, not advice, and it is not a price target. Read it with two things in view: the +8% is the residue of a wide positive range multiplied by a below-even probability, so it is highly sensitive to the probability estimate — at 25% it falls to +2%, at 35% it rises to +15% — and the downside path runs to $1.10, a 57% loss, which no expected-value figure conveys on its own.

Run-up

Entry and exit

Why T-2 rather than T-5 or T-1. The window is MONTH-precision, so readout.window.earliest is the opening of a month rather than a known event date, and the risk being managed is that the company announces on the first trading day of September. T-5 would close the position a full trading week before a window that is itself a month wide, giving up most of the drift the call is about; T-1 leaves no margin at all if the readout arrives a day early. T-2 is the compromise, and it is stated as a rule rather than a date so that it keeps resolving correctly if the window moves.

exit resolves to null at lock time, and that is the expected state rather than an omission: lib/runup.mjs resolveExit('T-2', {earliest: '2026-09-01'}, bars) returns null because the committed price cache ends on 2026-08-13 and therefore has no trading history at or after the window’s opening to count back from.

Predicted move and peak

Priority score drivers

Priority score. The seven drivers above combined by lib/runup.mjs priorityScore, transcribed rather than hand-computed:

Priority score 20 · formula_version 1.0.0

How to read a 20. This is a low score, and the arithmetic says plainly why: two genuinely attractive drivers (unmet need 85, value uplift 70) are multiplied down by a MONTH-precision date gate at 48 and then hit by the maximum possible clustering penalty at 100. The score is a statement about tradeability, not about the science — a program can be scientifically interesting and a poor run-up candidate at the same time, and this is one. What would raise it most is not a better result: it is the company disclosing a day for the readout, and the asthma topline landing separately rather than in the same release.

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.

Verdict

What I would do. Watch.

Why. The target is validated in this exact patient group by two approved competitors, and rademikibart has the one property an acute indication needs — a lung-function effect that starts inside the first week of subcutaneous dosing. But 159 patients is too few to settle a binary composite endpoint, the control arm already receives systemic steroids that work partly through overlapping biology, and no biologic has ever been shown to change what happens in the 28 days after a COPD exacerbation has begun. On top of the science, the trade itself is poor: the asthma topline lands in the same month and possibly the same release, so no price move here is attributable to this result, and the expected value of about +8% against a downside path to $1.10 does not compensate for that.

What would change this. Three specific observables, in descending order of weight. First, the company disclosing a day for the topline, or announcing the two studies will report separately — either would move readout.precision toward DAY and break the contamination that pins the run-up’s worst driver at 100, and the second would make this program tradeable on its own for the first time. Second, a positive asthma topline arriving before the COPD one, which would both de-risk the shared mechanism and let the COPD result be read cleanly. Third, on the science rather than the trade: any disclosure of the dose used in Seabreeze STAT COPD, which no source discloses today and which is the missing link between the asthma exposure–response and this trial.

What to watch.

Locked prediction

• Outcome-direction (will the readout succeed on its primary endpoint?): miss — probability 30%, band 20–42% [UNVERIFIED — modelled]. The probability is that the pre-registered positive definition below is met. It sits below even for two reasons that are about the trial rather than the target: 159 patients gives well under half the conventional chance of detecting a 10-percentage-point absolute reduction in a binary composite, and the control arm already receives systemic corticosteroids acting partly through overlapping biology. Working against those, and the reason the band’s top reaches 42%: the target is validated in this exact biomarker-selected population by two approved competitors, and rademikibart’s own Week-1 lung-function effect is documented in a peer-reviewed randomised trial on the same route of administration.
• Stock-direction (which way do the shares move?): up — confidence low — window 2026-08-13 to 2026-11-14, basis: from the lock date through two weeks past readout.window.latest, so the window covers the full readout window plus the settling period after a late announcement. Materiality is cited from ../company.md C.2, which records this program as material and shared — the larger commercial prize of the two, the weaker prior; the call is made consistent with that (rule 27) by taking the direction from the expected-value arithmetic while holding confidence at low. Attribution is CONTAMINATED: the Seabreeze STAT Asthma topline lands in the same month with a zero-day gap and may appear in the same release, so this call is explicitly not presented as attributable to the COPD result alone.
• Scenario prices: positive $4.00–$7.00 · miss $1.10–$2.10
• Expected value: $2.77, +8.2% against spot $2.56 (arithmetic, not advice)
• Run-up: entry $2.56 on 2026-08-13, exit rule T-2 trading days before readout.window.earliest — predicted move +5% to +25%, predicted peak +10% to +35% around 2026-08-31 — priority score 20, formula_version 1.0.0
• Settles on: the pre-registered positive definition is a statistically significant reduction (two-sided p < 0.05) in the primary endpoint — treatment failure rate over 28 days — for rademikibart plus standard of care versus placebo plus standard of care, in the Phase 2 Seabreeze STAT COPD study (NCT06940154), as reported in Connect Biopharma’s own topline press release or the corresponding Form 8-K. A release that reports the endpoint as numerically favourable without reaching significance settles as a miss. A release that reports only the key secondary (Week-1 post-bronchodilator FEV1) without the primary settles as a miss, because the primary is what the definition names.
• Locked: yes · Settled: no

Program data-quality flags

1. Open Targets is BLOCKED, and it is a standing block rather than a transient one. Verbatim on four attempts: Rate limit exceeded for client: global. 02-connectors.md already records this as a shared platform throttle. Consequence for this document: no independent human-genetics validation of IL-4Rα in COPD, and A.5’s “Target validation” row is scored on trial evidence alone. Low cost here, because two approved drugs are stronger evidence than a genetic association, but recorded rather than waved away.
2. The PubMed authors-returns-null bug did not reproduce on this sweep. 02-connectors.md records that on 2026-08-13 all 14 rademikibart articles returned complete metadata with not one author on any of them, and calls it the most expensive bug on that page because authorship is exactly what the KOL block needs. On this sweep, the same 14 articles returned full author lists with affiliations. That is what made the independence assessment in A.5b possible at all. The bug is therefore intermittent rather than standing, and an empty authors field should be re-attempted rather than accepted.
3. ChEMBL compound_search succeeded where the same-day sibling sweep recorded HTTP 500. 02-connectors.md records three failed attempts on this ticker (twice on rademikibart, once on CBP-201) with a verbatim upstream 500. This sweep’s single call on rademikibart returned CHEMBL5314768 immediately. Server-side and transient, not a standing block.
4. CT.gov discloses no investigator for the pivotal trial. All 52 sites on NCT06940154 return contacts: null and the record names no overall official. This is a sponsor choice about what to publish, not a connector failure, and it is why A.5b’s Investigators table is empty.
5. PubMed returns no conflict-of-interest field. The connector’s article metadata carries affiliations but no disclosure statement, so every independence judgement in A.5b rests on affiliation alone. Affiliation catches employment and trial-site relationships; it does not catch advisory-board seats, consulting arrangements or equity.
6. The dose used in Seabreeze STAT COPD is not disclosed by any source read this sweep. The registry lists the intervention as “Rademikibart in prefilled syringe” with no dose, and no release names one. This blocks any exposure–response reasoning about the acute regimen (A.3c) and is a real gap rather than a detail.
7. Source conflict left open: the spot price. ../company.md C.1 records $2.33 for 2026-08-13 from BPIQ’s last_price; the committed price cache closes the same day at $2.56. Both are correct about different moments — the cache’s bar opens at exactly $2.33 — and neither is averaged or picked over the other. This document uses the close, and says why, in Market and timing § Spot. The consequence a reader should carry: this program’s prediction and its sibling’s are locked against different spot prices on the same day, so their percentage figures are not directly comparable.
8. The rapid-onset evidence and the trial’s route of administration match, and an earlier reading of this that they did not is corrected here. The company’s most-quoted rapid-onset data point (200–400 mL FEV1 gains to Day 29) comes from the intravenous Phase 1 study, while Seabreeze STAT COPD uses a subcutaneous prefilled syringe. That mismatch would be a serious problem if the IV study were the only rapid-onset evidence. It is not: the peer-reviewed Phase 2b asthma trial reports FEV1 improvements occurring “rapidly during Week 1” on subcutaneous dosing [VERIFIED — Kerwin et al., AJRCCM 2025]. Recorded because the IV framing is what the company’s own releases lead with, and a reader working only from the press feed would draw the wrong conclusion about which route the speed claim rests on.
9. No health-economic model and no payer research exist for a per-event biologic in this setting. The entire payer argument in A.3b and B.2 is an inference from avoided-admission costs, not evidence. Named because B.3a’s price input — the least defensible of the four — depends on it.