STAT COPD: fewer treatment failures, with important endpoint boundaries

September 30 readout / evidence reviewed October 4

The COPD result has occurred. This dated update supersedes the earlier September readout expectation; the original dossier is retained below as historical evidence. The September 15 asthma disclosure remains a separate event. The overall interpretation is mixed: a favorable treatment-failure signal with a failed key lung-function endpoint, rather than success across all outcomes. [7]

Biology and acute-dose interpretation

IL-4Rα blockade addresses type 2 inflammation. Enrichment by eosinophils makes the biological hypothesis more specific, but limits extrapolation to unselected COPD. Analysis: preventing deterioration after an exacerbation and reversing fixed airway damage are different claims; the observed clinical composite should not be presented as proof of structural disease modification. [7]

A single 600 mg subcutaneous administration was added to standard care. Acute exposure-response and receptor-occupancy thresholds were not provided. The maintenance-asthma dose studies retained below offer mechanistic context, not a substitute for this acute COPD experiment. Standard care is part of both arms; the comparison does not support replacing rescue treatment.

Design, population and denominators

NCT06940154 / CBP-201-207 is randomized, double-blind and placebo-controlled; 325 screened and 159 randomized, 78 active and 81 placebo, with eight-week follow-up. Treatment failure includes death, hospitalization, urgent reassessment or intensified medication. Eosinophils were at least 300 cells/µL. A full protocol/SAP was not recovered; planned enrollment and missing-data rules are not independently verified. [7] [8]

The release describes adults, while the design slide says adolescents and adults; the baseline age ranges are 40–80 and 45–80 years. The slides also contain references to asthma and Study 206 within the COPD deck. These internal inconsistencies are preserved as source limitations, not silently corrected into new trial facts.

Clinical results / different time windows

The ±3-day visit window and the strict 28-day interval yield different analyses. The secondary exacerbation finding loses conventional statistical significance in the strict interval. Without the complete testing hierarchy, secondary p-values should be treated as nominal. Neither a favorable symptom score nor the 70 mL Week-4 difference rescues the unsuccessful Week-1 key secondary endpoint. [8]

The visit-window absolute treatment-failure difference is approximately 11.0 percentage points, calculated as 11/81 minus 2/78. This descriptive calculation lacks a source-reported confidence interval. Zero hospital/ED returns in 78 treated participants does not imply the true future risk is zero, and the small number of events makes replication important.

Safety / preserve the late-death distinction

The table reproduces the presentation’s safety population. No death occurred within the primary endpoint window, but a separate slide records a Day-37 death of unknown cause, with a history of ventricular arrhythmia and a fall with head injury. Its treatment column could not be reliably recovered from flattened text; it is not assigned to either arm here. This prevents an unsupported claim of zero deaths over the whole study. [8]

Analysis: fewer acute disease events may lower the all-cause AE count without proving a generally safer antibody. Exposure is short and the trial is too small to define uncommon infection, ocular or hypersensitivity risk. Trial withdrawal, treatment discontinuation and later death are distinct outcomes and must remain separate.

Capital and development interpretation

The dated June 30 financial snapshot retained below remains historical, not an October cash balance. The result release describes contingent Greater China milestones of up to approximately $99 million and tiered royalties; contingent proceeds are not available cash. Management’s proposed FDA discussion is a planned next step, not agreement on a registrational design or an approval. [7] [9]

The execution test is whether a larger program reproduces the clinical composite under a transparent estimand while resolving the lung-function and safety uncertainties. Selecting only the most favorable visit window for the next development narrative would overstate certainty. No success probability, financing certainty or price target is inferred from these preliminary data.

Biology / Target

Rademikibart targets IL-4Rα. The STAT trials select eosinophils ≥300 cells/μL. [4]

Analytical judgment: target plausibility is not proof of rescue efficacy. During an exacerbation, a useful effect must arrive soon enough to change the clinical course on top of standard care. A maintenance-asthma lung-function signal supports biological activity, but cannot answer that acute-care question.

The main external-validity question is the selected inflammatory phenotype. A positive result in this enriched population should not be extrapolated to every asthma or COPD exacerbation. The two diseases should be evaluated separately before proposing a common commercial indication.

Molecule / PK–PD

The maintenance study used a 600 mg loading dose followed by 150 or 300 mg every two weeks. [1]

Analytical judgment: repeated-dose results cannot establish adequate exposure after a single acute dose. The missing bridge is time-resolved exposure, target engagement and clinical benefit under acute-care conditions. Faster lung-function improvement would be supportive, but not a substitute for the prespecified clinical endpoint.

A numerical advantage at the higher dose is not itself a proven dose response. A formal dose comparison, exposure-response model and uncertainty estimates would be needed. This dossier does not have a validated acute PK/PD dataset; no receptor-occupancy threshold is assumed.

Prior clinical evidence

The randomized, double-blind, placebo-controlled Phase 2b maintenance-asthma trial enrolled 322 participants over 24 weeks. Its primary endpoint was week-12 pre-bronchodilator FEV₁. [1] [2]

The table distinguishes randomized enrollment from the primary-endpoint table counts reported in the initial release. Those smaller counts must not be described as withdrawals without a participant-disposition table and the analysis specification.

The confidence intervals exclude zero for both placebo comparisons. That supports an average lung-function effect in this trial, not a guarantee of benefit for each patient. FEV₁ is the volume forcibly exhaled in one second; it is not a direct count of hospitalizations avoided.

Safety and conflicting source versions

The 2023 release counted 25 placebo exacerbation events versus 11 and 13 in the active arms. The 2025 update reported 26 placebo events and 24 pooled active events. The reason for the placebo-count revision is not resolved here. [1] [3]

Events are not necessarily unique patients: repeated events invalidate a simple patient-risk calculation from these counts. Use the final prespecified event-rate analysis, exposure time and confidence interval. Do not combine one version’s numerator with another version’s denominator.

The publication update described mostly mild or moderate treatment-emergent adverse events and no treatment-related serious adverse events. Complete arm-level safety denominators and event rates were not extracted into this report. “Not available here” does not mean “zero events.”

A useful full readout must tabulate any adverse event, serious adverse events, discontinuations, deaths and relevant mechanism-related events by arm, with n/N, exposure and follow-up. A reassuring press-release adjective cannot replace that table.

Current trial design / Statistics

STAT asthma enrolled 160; COPD enrolled 159. Both are randomized, blinded, placebo-controlled adjunct-to-standard-care trials. Primary: 28-day treatment failure. Key secondary: week-1 post-bronchodilator FEV₁. The company guided September 2026 topline results. Actual per-arm counts and results are not provided in that update. [4]

Readout interpretation: report failures as n/N in each randomized arm before calculating absolute risk reduction. For control risk Pc and treated risk Pt, ARR = Pc − Pt. NNT = 1/ARR only when ARR is positive, clinically interpretable and accompanied by uncertainty; it must use the same follow-up and endpoint definition.

Check the treatment-failure components, missing-outcome rules, intention-to-treat population, multiplicity hierarchy and rescue therapy. A favorable lung-function secondary endpoint does not rescue a failed primary endpoint. One positive disease trial and one negative trial require separate interpretations, not a pooled headline.

Strong scenario: a credible primary effect, consistent components and acceptable safety. Mixed: nominal significance with fragile missing-data assumptions or one indication only. Negative: primary failure or benefit offset by harm. These are analytical scenarios, not assigned probabilities or share-price forecasts.

Capital structure / Management

June 30, 2026 cash and short-term investments: $31.486m. Q2 R&D: $16.273m. Management projected at least one year of runway from its August release. [4]

Analytical judgment: expense is not cash burn. Do not divide cash by one quarter of R&D and label the result a reliable runway. Reconcile the cash-flow statement, collaboration receipts, working capital and the cost of a larger trial before relying on management’s estimate.

Execution diligence remains open: verify financing facilities, outstanding options/warrants, contractual commitments and the final trial reports. A source-backed record of guidance versus actual delivery is more informative than an unsupported management-quality score.

September 15 update: asthma reported, COPD still pending

The original combined September program window above is preserved as a historical expectation. Connect has now disclosed the asthma study, which is recorded in a separate History event and complete report linked in the references. Its primary treatment-failure endpoint missed statistical significance (2/79 versus 6/81; p=0.153); the day-seven FEV₁ difference was +130 mL (nominal p=0.023). The COPD study remains Upcoming, with September 15 company guidance for results later this month. This record now tracks only that outstanding COPD milestone; it does not count the asthma result twice. [5] [6]

Sources / References

1. Connect Biopharma · Phase 2b topline release · December 12, 2023
2. Rademikibart Treatment for Moderate-to-Severe Uncontrolled Asthma: A Phase 2B Randomized Clinical Trial · 2025
3. Connect Biopharma · Publication update · March 31, 2025
4. Connect Biopharma · Q2 business update · SEC Exhibit 99.1 · August 12, 2026
5. Connect · asthma results; COPD still expected later in September · September 15, 2026
6. Independent History report · STAT Asthma September 15 results
7. Connect · STAT COPD results · September 30, 2026
8. SEC Exhibit 99.2 · Preliminary STAT COPD presentation · September 30, 2026
9. Connect · Q2 results · August 12, 2026