Barzolvolimab: Phase 2 evidence and the EMBARQ-CSU replication test
Clinical evidence review based on public sources. Unpublished results and unavailable details are identified; this is not the sponsor’s full clinical study report.
Biology / Target
Barzolvolimab targets KIT to reduce mast-cell activity through depletion. CSU is the relevant disease setting; positive results in induced urticaria or a negative result in another inflammatory disease are contextual evidence, not interchangeable CSU trial outcomes. [4]
A proximal pharmacodynamic effect needs to connect to itch, hives and patient function. Suppressing a biomarker can support target engagement without proving symptom relief. Conversely, symptom improvement that outlasts measurable drug exposure is interesting but needs controls before being called durable disease modification.
Molecule / Dose and PK–PD
Phase 2 tested subcutaneous 75 mg every four weeks, 150 mg every four weeks and 300 mg every eight weeks. Equal cumulative nominal dose does not imply equivalent peak, trough or target suppression. [1] [2]
The 150 mg and 300 mg schedules had similar mean symptom-score improvements but different complete-response proportions. This is not contradictory: a mean captures the whole distribution, whereas a threshold captures only how many cross a cutoff. A formal between-dose comparison is needed before claiming one schedule is superior.
Prior clinical evidence: randomized Phase 2
The study randomized 208 patients. The source distinguishes randomized from dosed participants: the 300 mg arm randomized 52 but its Week-12 table used 51. Do not call that difference a Week-12 dropout without the disposition record. [1] [2]
Lower UAS7 indicates fewer symptoms. The placebo-adjusted mean changes below are model estimates, not the difference between rounded group means. Complete control means UAS7=0; it should not be confused with the broader well-controlled category.
| Arm | Randomized / dosed N | Week-12 UAS7 change | Adjusted contrast (95% CI) | UAS7=0 |
|---|---|---|---|---|
| 75 mg Q4W | 53 / 53 | −17.06 | −6.60 (−10.71, −2.49) | 22.9% |
| 150 mg Q4W | 52 / 52 | −23.02 | −12.55 (−16.56, −8.55) | 51.1% |
| 300 mg Q8W | 52 / 51 | −23.87 | −13.41 (−17.47, −9.34) | 37.5% |
| Placebo / 安慰剂 | 51 / 51 | −10.47 | Reference / 参照 | 6.4% |
Week-12 complete symptom control
Current trial design / Statistics
EMBARQ-CSU1/2 enrolled 1,939 patients in aggregate, with topline guided to September/October 2026. Exact trial-specific and arm-specific counts must come from the corresponding reports; the combined enrollment must not be divided equally by assumption. [2] [3] [4]
The Phase 2 primary analysis used ANCOVA, multiple imputation and multiplicity adjustment. For Phase 3, inspect the prespecified hierarchy independently; an earlier analysis method cannot be assumed unchanged. Replication should be judged by effect magnitude, interval width and both studies, not merely whether one p-value crosses 0.05.
The Phase 2 placebo-controlled period ended at Week 16; subsequent switching means later response rates are not a year-long placebo comparison. Off-treatment analyses that select earlier responders also condition on a favorable outcome. Their persistence is informative but can overstate whole-population durability if the denominator is obscured.
In the readout, separate biologic-naïve and previously treated patients, rescue medication, diary completeness and missing-data sensitivity. A broad average benefit can coexist with important variation in prior-therapy subgroups; small subgroup samples cannot establish equivalence.
Safety and reversibility
At 12 weeks, reported active-treatment TEAEs included hair-color change (9%) and neutropenia/ANC reduction (8%). These are aggregate active-treatment rates; they do not provide each dose’s excess risk versus placebo. [1] [2]
Long-term dose switching complicates attribution. Distinguish events during the randomized period from events after switching, and distinguish transient laboratory changes from infections or discontinuations. Reversibility requires an observed recovery timeline, not only a mechanistic expectation.
Capital structure / Management
Celldex reported $717.6 million in cash, equivalents and securities and 78.5 million shares outstanding at June 30, 2026, after an April financing. The company projected planned operations through 2028. Outstanding shares are not a fully diluted count. [3]
Execution quality will be shown by synchronized disclosure of both Phase 3 studies, transparent dose-level safety and faithful follow-through on missing-data analyses. A fast enrollment milestone is operationally positive but is not evidence that blinded efficacy is positive.
September 22 readout / Phase 3 evidence
The September 22 disclosure, with a September 9 data cutoff, supersedes the former expected window. EMBARQ-CSU1 and CSU2 randomized 963 and 976 patients; active treatment continues through week 52. The company reported that both studies met the week-12 UAS7 primary endpoint and key secondary endpoints. A 2027 BLA is planned, not submitted or approved. [5]
The global double-blind studies assigned patients evenly to subcutaneous 150 mg every four weeks after a 300 mg loading dose, 300 mg every eight weeks after 450 mg loading, or placebo for 24 weeks. Exact per-arm analysis counts were not supplied; dividing total enrollment by three would not recover them. The reported standard errors describe estimated means, not patient variability or confidence intervals.
| Study / arm | Baseline UAS7 | Week 12 LS change (SE) | Versus placebo p |
|---|---|---|---|
| CSU1 placebo | 30.9 | −10.7 (0.65) | — |
| CSU1 150 mg Q4W | 30.6 | −20.2 (0.64) | <0.00001 |
| CSU1 300 mg Q8W | 30.3 | −20.5 (0.65) | <0.00001 |
| CSU2 placebo | 29.3 | −11.4 (0.64) | — |
| CSU2 150 mg Q4W | 29.7 | −20.2 (0.63) | <0.00001 |
| CSU2 300 mg Q8W | 29.6 | −19.7 (0.64) | <0.00001 |
Complete response / Remaining safety gaps
The reproducible placebo separation is supportive, but neither a cross-trial superiority claim nor a proven preference between doses follows from these tables. UAS7=0 means no itch or hives in the score interval. The angioedema analysis is limited to baseline AAS7>0; it must not be applied to all randomized patients. Missingness, estimand details and the full multiplicity hierarchy remain to be reviewed. [5]
The release described tolerability without numerical arm-level any-TEAE, grade 3 or higher, SAE, death or discontinuation tables. Those values remain unavailable, not zero. KIT-related monitoring and long-term exposure matter as treatment continues; the earlier Phase 2 safety record cannot substitute for Phase 3 arm-level reporting.
| Study / time | Placebo % | 150 mg Q4W % | 300 mg Q8W % |
|---|---|---|---|
| CSU1 UAS7=0 week 12 | 9.3 | 42.4 | 42.1 |
| CSU2 UAS7=0 week 12 | 12.6 | 45.7 | 44.0 |
| CSU1 UAS7=0 week 24 | 15.4 | 49.0 | 45.1 |
| CSU2 UAS7=0 week 24 | 17.6 | 54.0 | 48.4 |
| CSU1 AAS7=0 week 12; baseline >0 | 33.8 | 62.7 | 66.3 |
| CSU2 AAS7=0 week 12; baseline >0 | 33.7 | 74.3 | 66.2 |
Sources / References
For general information and research only; not investment or medical advice. Coverage and disclosed data may be incomplete. Verify primary sources before making decisions.