infohep is no longer being updated. Visit for HIV and hepatitis news.

Experimental hepatitis B combination suppresses key marker but fails to achieve functional cure

Keith Alcorn
27 June 2022
Dr Kosh Agarwal discusses results of the REEF-2 study at the International LIver Congress. Photo © Steve Forrest & Andrew McConnell / EASL.

People with hepatitis B who received a combination of experimental drugs as well as their standard treatment for 48 weeks were significantly less likely to have high levels of hepatitis B surface antigen (HBsAg) or hepatitis B flares after discontinuing treatment in a clinical trial compared to people who received standard treatment alone, Dr Kosh Agarwal of London’s Kings College Hospital reported at the International Liver Congress.

But the study failed to show that the experimental combination could clear hepatitis B surface antigen in a substantial proportion of participants, the primary outcome of the study.

The study also reported that one patient, a middle-aged man who had been taking tenofovir for eight years, suffered such an extreme flare-up of hepatitis B after stopping treatment that he needed a liver transplant. The serious adverse event highlights the potential risks involved in studies of experimental hepatitis B treatments that require people with apparently well-controlled virus to stop treatment.

The REEF-2 study was designed to test whether discontinuing treatment after 48 weeks of combination treatment with JNJ-3989 and JNJ-6379 and an NRTI resulted in an increased rate of HBsAg loss 48 weeks after treatment discontinuation compared to NRTI treatment alone. Achieving HBsAg loss after stopping treatment is considered a ‘functional cure’, meaning that no further treatment for hepatitis B is required.

HbsAg loss after discontinuing treatment is rare in people treated with a nucleotide or nucleoside analogue (NA), the standard form of treatment for hepatitis B. Drugs that interfere with hepatitis B in several ways are being tested in early-stage and mid-stage clinical trials to assess their safety and impact on hepatitis B DNA and HBsAg levels.

JNJ-3989 is a silencing RNA (siRNA) that targets production of all hepatitis B proteins. JNJ-6379 is a capsid assembly modulator. It promotes the formation of empty non-infectious viral shells that lack DNA or RNA.

The REEF-1 study reported that JNJ-3989 reduced levels of hepatitis B surface antigen after 48 weeks of treatment in combination with a nucleotide or nucleoside reverse transcriptase inhibitor (NRTI).

Dr Kosh Agarwal of the Institute of Liver Studies at King’s College Hospital, London, presented 24-week interim results of REEF-2.

REFF-2 recruited people with hepatitis B who had received NA treatment for at least two years, had hepatitis B DNA levels below 60 IU/ml, HBsAg levels above 100 IU/ml at screening and ALT < 2 x upper limit of normal. All participants received an NA (tenofovir disoproxil (TDF) or tenofovir alafenamide (TAF) or entecavir) and were randomised in 2:1 proportion to receive JNJ-3989 and JNJ-6379 or placebos for 48 weeks.

After discontinuing treatment, participants were mandated to re-start NA treatment if they experienced HBeAg reversion, or a post-treatment increase HBV DNA above 20,000 IU/ml or a post-treatment increase in HBV DNA above 2,000 IU/ML coupled with an ALT increase to more than five times the upper limit of normal, on at least two occasions four weeks apart.

The study recruited 130 participants (85 to the active arm and 45 to the placebo arm. Two-thirds of participants were male, two-thirds were white and the median age was 46. Participants had been taking NRTIs for a median of 8 years (38% were taking etravirine, the remainder TDF or TAF) and 78% had an HBsAg level above 1,000 IU/ml.

The primary endpoint of the study was the proportion of participants who achieved clearance of HBsAg after treatment discontinuation. No participants in either study arm achieved the primary outcome.

However, participants in the active treatment arm were significantly more likely to achieve an HBsAg level below 100 IU/ml by week 48 of treatment and to maintain this level 24 weeks after stopping treatment. At week 48, 71.1% of the active treatment arm had an HBsAg level below 100 IU/ml compared to 2.4% of the placebo arm. By week 24 after discontinuation, 67.1% in the active arm had an HBsAg level below 100 IU/ml compared to 10.3% in the placebo arm.

Fifty-seven per cent of participants in the active arm maintained a stable or declining HBsAg level after discontinuing treatment.

Hepatitis B DNA levels tended to rebound higher in people in the placebo arm after stopping treatment. Just over a third (34.1%) of those in the placebo arm had an HBV DNA level above 20,00 IU/ml during the 24-week follow-up period compared to 6.5% in the active treatment arm. Conversely, nearly 30% of participants in the active arm had an undetectable HBV DNA and an HBsAg level below 100 IU/ml at week 24, compared to 6.5% in the placebo arm.

Flares in ALT, signalling reactivation of hepatitis B virus, were more common in the placebo arm after discontinuation. Twenty-three per cent in the placebo arm experienced an ALT increase to at least three times above the upper limit of normal by week 24 compared to 2.4% in the active treatment arm.

Nineteen per cent of the placebo group resumed treatment by week 24 due to hepatitis B DNA and ALT increases compared to one patient in the active treatment arm. The patient resuming treatment in the active treatment arm did so as a consequence of a diagnosis of cholangiocarcinoma.

The most common serious adverse event during the treatment phase was reduction in kidney function (measured by eGFR) associated with JNJ-6379 in the active treatment arm (7%). In the discontinuation phase, the most common serious adverse event was a severe ALT increase in 9% of the placebo group.

One patient in the placebo arm, a 54-year-old Asian male who had received tenofovir treatment for eight years, experienced a life-threatening episode of liver failure after discontinuing treatment. The participant experienced a severe increase in HBV DNA and ALT eleven weeks after discontinuation accompanied by jaundice, coagulopathy and sub-acute liver failure. HBV DNA peaked at over 6 billion IU/ml and HBsAg rose to 124,000 IU/ml. This participant underwent an urgent liver transplant and recovered.

Dr Agarwal said that the speed and magnitude of the HBV DNA rebound in this patient has implications for future studies that use treatment discontinuation as a means of evaluating efficacy. In this study, treatment resumption criteria were revised to require re-starting of treatment if HBV DNA levels rose above 5 log, regardless of ALT levels.

In practical terms, this case of rebound has implications for “the turnaround time for labs and the speed of restarting of NAs”, said Dr Agarwal.

The study findings also raise questions for the design of future studies, not least the duration of active treatment and the number of drugs used during the active treatment phase.


Agarwal K et al. Efficacy and safety of finite 48-week treatment with the siRNA JNJ-3989 and the capsid assembly modulator JNJ-6379 in HbeAg negative virologically suppressed chronic hepatitis B patients: results from the REEF-2 study. International Liver Congress, London, abstract, 2022.