Mirum's Brelovitug Meets Phase 3 Endpoint in Hepatitis D

Mirum Pharmaceuticals cleared the Phase 3 bar in chronic hepatitis delta virus (HDV), the most aggressive form of viral hepatitis, with its investigational antibody brelovitug meeting the primary endpoint in the AZURE-1 study.

Brelovitug is a fully human monoclonal antibody that binds hepatitis B surface antigen (HBsAg). HDV can only replicate in the presence of HBV, so the drug attacks the virus indirectly by clearing the antigen it depends on. In the Phase 3 portion of AZURE-1, 153 treatment-naive patients were randomized 2:2:1 to brelovitug 300 mg once weekly, brelovitug 900 mg once every four weeks, or delayed treatment. At Week 24, 56% of the 300 mg weekly arm and 45% of the 900 mg monthly arm hit the combined primary endpoint of virologic response (at least a 2 log10 reduction in HDV RNA or undetectable) plus ALT normalization, against 0% in the delayed-treatment arm (p<0.0001 for both doses).

The individual components ran even stronger: virologic response reached 86% and 85%, with ALT normalization at 63% and 54%. The study enrolled a broad population, including patients with significant liver inflammation, cirrhosis and clinically significant portal hypertension. Safety was clean, with zero grade 3 or higher events and zero discontinuations in either brelovitug arm (one grade 4 myocardial infarction occurred in the delayed-treatment arm). The 48-week Phase 2b extension showed responses deepening over time, with ALT normalization reaching 55% in both arms.

Mirum shares (Nasdaq: MIRM) closed Friday at $89.70, a roughly $6 billion market cap. The readout landed before Monday’s open.

The combined endpoint matters because controlling HDV requires both suppressing the virus and calming the liver inflammation that drives progression to cirrhosis and liver cancer. Brelovitug would be only the second FDA-approved therapy for HDV, after Gilead’s bulevirtide (Hepcludex), approved in May. Mirum plans a BLA in the first half of 2027, with topline from the second registrational study, AZURE-4, expected later this year.

What to watch next: the AZURE-4 topline, and whether brelovitug’s self-administered subcutaneous dosing carves out an edge over the incumbent. (See our Mirum biliary atresia primer for the rest of the pipeline.)

What It Means for the Trade

HDV is not a large market, but it is a savage one — the most aggressive form of viral hepatitis, driving faster progression toward cirrhosis and liver cancer than HBV alone — and the approved toolkit has been one Gilead drug since bulevirtide’s approval in May. That approval created the category; AZURE-1 argues the category is about to get a better tool. The 0% response in the delayed-treatment arm makes the effect unambiguous, and 56% of the weekly arm clearing a combined virologic-plus-ALT endpoint at Week 24 is a strong result in a disease where the endpoint itself is demanding.

The commercial wedge is administration. Brelovitug is a self-administered subcutaneous antibody, and in a chronic infection treated for years, the gap between self-injection and provider-administered therapy shows up directly in compliance and payer math. The catalyst path is clear: AZURE-4 topline later this year, then a BLA planned for the first half of 2027. Mirum is already a commercial rare-disease company, not a pre-revenue biotech — a second franchise-quality asset changes the growth trajectory at a roughly $6 billion market cap.

Mechanism and Trial Context

The mechanism is elegant parasitology. HDV is a defective virus that can only assemble and propagate using hepatitis B surface antigen — no HBsAg, no HDV. Brelovitug binds HBsAg directly, clearing the envelope material HDV depends on, which suppresses the virus without attacking its own machinery. The endpoint design reflects how hepatologists judge HDV therapy: virologic response alone is not enough if the liver keeps burning, so the primary bar demanded both a 2-log10 HDV RNA reduction (or undetectability) and ALT normalization — and the virologic components ran at 86% and 85%. The 48-week extension showing responses deepening over time argues the effect consolidates rather than plateaus, and the enrolled population — including cirrhosis and clinically significant portal hypertension — is the real-world cohort, not a cherry-picked one.

Risks and What Could Go Wrong

1. AZURE-4 must confirm AZURE-1 before the 2027 filing — a second-trial miss strands the program one readout short of a BLA.
2. Gilead holds first-mover position and a full hepatitis commercial infrastructure; Mirum would be challenging a much larger machine with its first launch in the category.
3. Week 24 is an early look at a lifelong infection — payers will want durability beyond the extension data before underwriting chronic therapy.
4. HDV’s epidemiology bounds the market; launch execution and pricing matter more than headline response rates.