An injected hepatitis B drug produced a functional cure in about one in five

Chronic hepatitis B is usually controlled, not cured. The standard nucleoside or nucleotide antiviral pills suppress the virus for as long as a patient keeps taking them, often for life, but they rarely clear it. So the headline from two new phase 3 trials — that an injected drug called bepirovirsen produced a "functional cure" in about one in five patients after a fixed course — is a genuine step. The numbers underneath it, and who paid for the trials, are worth reading carefully [s1].

Bepirovirsen is an antisense oligonucleotide, a short strand of engineered genetic material designed to bind and destroy the virus's messenger RNA before it can be made into viral proteins [s1]. A functional cure was defined stringently: at least 24 weeks of sustained hepatitis B viral DNA below the limit of quantification, plus loss of the hepatitis B surface antigen, after therapy had stopped [s1].

What the trials did

The programme comprised two replicate, double-blind trials, B-Well 1 and B-Well 2 — running the same design twice is a deliberate strength, because a result that reproduces is far more convincing than one that appears once [s1]. Adults with non-cirrhotic chronic hepatitis B were assigned in a 2:1 ratio to subcutaneous bepirovirsen (300 mg weekly) or placebo for 24 weeks [s1].

Every participant was already on stable nucleoside or nucleotide analogue therapy and had a hepatitis B surface antigen level of more than 100 to 3,000 IU per millilitre — a mid-range group, not those with the highest viral burden [s1]. Eligible patients then discontinued their background antiviral at 48 weeks, and the primary outcome was a functional cure at week 72 [s1]. The trials were funded by GSK, which is developing the drug, and registered on ClinicalTrials.gov [s1][s2]. That industry sponsorship is not a disqualifier, but it is the frame within which every efficacy and safety number below should be read.

What they found

Both trials hit their primary endpoint, and by similar margins. A functional cure at week 72 was reached in 127 of 650 patients (20%) on bepirovirsen versus none of 328 on placebo in B-Well 1, and in 106 of 570 patients (19%) versus none of 286 on placebo in B-Well 2 [s1]. That the placebo groups produced no cures at all makes the contrast clean: the effect is the drug's [s1].

It is also worth stating plainly what "about one in five" means the other way round: roughly four in five patients did not achieve a functional cure, and then, having stopped their background antiviral, would need it restarted or close monitoring. This is a real advance over a near-zero baseline, not a cure for most.

The safety numbers

Bepirovirsen was not gentle. In a pooled analysis at 72 weeks, adverse events were reported in 91% of patients on the drug versus 73% on placebo, and serious adverse events in 7% versus 4% [s1]. During the treatment period, grade 3 or higher adverse events occurred in 16% of the bepirovirsen group versus 3% of the placebo group [s1].

The most common serious-grade problem points at the liver itself: rises in alanine aminotransferase, a liver enzyme, were the most frequent grade 3 events with bepirovirsen, in 6% of patients [s1]. Some degree of enzyme flare can accompany the immune clearance of the virus, but a 6% rate of grade 3 elevations is a signal that demands monitoring rather than a shrug.

How to read it

The authors' conclusion is measured: in two phase 3 trials, a functional cure after stopping background antiviral therapy "was reported in significantly more patients treated with bepirovirsen than in those who received placebo" [s1]. That is true, reproduced, and clinically meaningful against a disease that standard therapy almost never cures.

The honest qualifiers are the denominator and the design. The benefit accrued to about a fifth of a selected, mid-range population; the trials were funded by the drug's maker; and the safety burden, including liver-enzyme flares, was real [s1]. Independent replication and longer follow-up — to see whether cures hold and whether any serious harms emerge later — will decide how large a role this drug plays. Related coverage has examined the evidence on hepatitis C screening and cure, Europe's off-track hepatitis-elimination goals, and a US vote on the hepatitis B birth-dose vaccine.

What to watch

The questions ahead are whether regulators judge a roughly 20% functional-cure rate, with this safety profile, sufficient for approval; how the drug performs outside the enrolled surface-antigen range; and whether the cures seen at 72 weeks prove durable [s1].

This article describes trial results and is not medical advice. Nothing here should be used to start, stop or change any treatment for hepatitis B.

Sources

• Phase 3 Results of Bepirovirsen Treatment for Chronic Hepatitis B Virus Infection — New England Journal of Medicine, 28 May 2026
• B-Well 1 trial record, NCT05630807 — ClinicalTrials.gov