Mirum’s Phase 3 AZURE-4 study met its primary endpoint, the company announced on November 10, 2026. That is meaningful progress for an investigational hepatitis D treatment. It is also an opportunity to examine what sits inside a reassuring phrase: “virologic response.” A response can represent substantial suppression while falling well short of undetectable virus, durable control after treatment, or cure. [1]
A treatment connected to hepatitis B, addressing a different infection
Hepatitis D virus, or HDV, depends on hepatitis B virus. This biological relationship makes HDV relevant to people living with HBV, but success against HDV cannot be described as a cure for hepatitis B. WHO distinguishes management of these infections and notes that nucleoside analogues used for HBV have no direct benefit in controlling HDV. [3]
Brelovitug is an investigational monoclonal antibody targeting hepatitis B surface antigen, which is also present on the HDV envelope. The developer describes neutralization and clearance of viral particles as its intended action. Targeting an external structure does not itself establish elimination of the intracellular sources of persistent infection. [1]
AZURE-4 eligibility also required patients to be taking or willing to take HBV nucleos(t)ide treatment. A “single-agent” development strategy for HDV should therefore not be interpreted as an instruction to abandon management of underlying HBV. [2]
The endpoint contains two conditions
The company reported 113 randomized participants across weekly 300 mg brelovitug, 900 mg every four weeks, and delayed treatment. At week 24, 59% of the weekly group’s 58 participants achieved the combined endpoint of virologic response and ALT normalization; the corresponding proportion was 44% among 27 participants receiving the less frequent regimen. The rounded percentages should not be reverse-engineered into exact responder counts. [1]
ALT is a blood marker of liver inflammation. Normalization provides information about disease activity, not proof that all existing liver damage has reversed. The virologic component also admits more than one depth of improvement, including an HDV RNA decline of at least 2 log units. A two-log reduction means a fall to roughly one hundredth of the starting viral concentration; detectable virus may remain. It does not mean a 99% reduction in cancer or mortality risk. [1,2]
The same disclosure illustrates this distinction: in the weekly group, virologic response was 95%, while HDV RNA target-not-detected was 14% at week 24. Both findings can be true because their thresholds differ. Neither percentage is a cure rate. [1]
The comparison has two different time points
The announcement contrasts active-treatment results with a 0% response in delayed treatment. The trial registration adds a crucial detail: the prespecified primary comparison is week 24 in brelovitug arms against week 12 in the delayed-treatment arm. Those assigned to delay begin brelovitug after the 12-week waiting period. They are not an untreated control followed for 24 weeks. [2]
That does not automatically invalidate the study. Randomization and prespecification still matter. It defines the question answered. The result should not be rewritten as a same-time comparison of 24 weeks of treatment versus 24 weeks without treatment.
The trial is open-label. Laboratory measures are more objective than expectations alone, but the full analysis still needs to clarify measurement, missingness and the interpretation of the specified comparison. The registration retains an estimated enrollment of 80 in its June 2026 update; the later company report provides the actual 113. A planned denominator should not replace the reported results population. [1,2]
Longer suppression raises another question
Mirum also highlighted a separate early-study cohort of 15 patients with cirrhosis, reporting 100% combined response following up to 108 weeks of treatment. This is a small, longer-term cohort, not all patients in the Phase 3 trial. Continued-treatment responses cannot establish the outcome of stopping therapy. [1]
The company reported no new safety signals. Flu-like illness occurred in 17.2% and 18.5% of the weekly and four-weekly AZURE-4 groups, respectively, mostly after the first dose. These observations inform tolerability; they do not rule out uncommon harms or establish safety across every stage of liver disease. [1]
The clinically important next step is to connect suppression and inflammatory improvement with outcomes people experience: less liver decompensation, fewer serious complications, and a sustainable life with treatment. Even where lifelong control would be valuable, it creates different burdens from a finite course followed by durable control off treatment.
What would change the judgment?
Our interpretation is that AZURE-4 strengthens a registration case, while leaving the duration and ultimate benefit of treatment open. Mirum plans a U.S. application in the first half of 2027; that is not an approval. We have not verified an approval or formal generic Chinese name for brelovitug in mainland China. [1]
Reliable off-treatment durability would strengthen the case for a finite therapeutic strategy. Evidence of fewer liver-related complications would make the link from laboratory response to patient benefit more direct. If benefit requires sustained dosing, persistence, monitoring, affordability and access become central parts of the treatment’s value.
The open question is not simply whether a laboratory result improved. It is what patients can reasonably expect that improvement to change—and for how long.
Sources
[1] Mirum, November 10, 2026: company-reported AZURE-4 topline and AASLD long-term data; complete peer-reviewed Phase 3 publication and statistical analysis plan were not obtained for this article. https://ir.mirumpharma.com/news/news-details/2026/Mirum-Pharmaceuticals-Announces-Primary-Endpoint-Met-in-AZURE-4-Phase-3-Study-and-Highlights-Long-Term-Brelovitug-Data-Presented-at-AASLD/ https://www.sec.gov/Archives/edgar/data/1759425/000175942526000064/mirm-20261110.htm
[2] ClinicalTrials.gov, NCT07298330, last posted update June 29, 2026. The prespecified primary outcome compares active-arm week 24 with delayed-arm week 12. https://clinicaltrials.gov/study/NCT07298330 https://clinicaltrials.gov/api/v2/studies/NCT07298330
[3] WHO, Hepatitis D fact sheet, July 28, 2026. https://www.who.int/news-room/fact-sheets/detail/hepatitis-d