Efficacy and safety of the siRNA JNJ-73763989 and the capsid assembly modulator JNJ-56136379 (bersacapavir) with nucleos(t)ide analogues for the treatment of chronic hepatitis B virus infection (REEF-1): a multicentre, double-blind, active-controlled, randomised, phase 2b trial

Background: JNJ-73763989 (JNJ-3989), a small interfering RNA, targets all hepatitis B virus (HBV) RNAs, reducing all HBV proteins. JNJ-56136379 (JNJ-6379; also known as bersacapavir), a capsid assembly modulator, inhibits HBV replication. We aimed to evaluate the efficacy (ie, antiviral activity) and safety of these therapeutics in combination with nucleos(t)ide analogues in patients with chronic hepatitis B.

Methods: The REEF-1 multicentre, double-blind, active-controlled, randomised, phase 2b study was done at 108 hospitals or outpatient centres across 19 countries in Asia, Europe, and North and South America. We included patients aged 18-65 years with chronic hepatitis B (defined as HBsAg positivity at screening and at least 6 months before screening or alternative markers of chronicity [eg, HBV DNA]), including those not currently treated, virologically suppressed, HBeAg positive, and HBeAg negative. Patients were randomly assigned (1:1:2:2:2:2) via permuted block randomisation according to a computer-generated schedule to receive oral nucleos(t)ide analogues once per day plus placebo (control group); oral JNJ-6379 250 mg daily plus nucleos(t)ide analogues (JNJ-6379 dual group); nucleos(t)ide analogues plus subcutaneously injected JNJ-3989 at doses of 40 mg (JNJ-3989 dual 40 mg group), 100 mg (JNJ-3989 dual 100 mg group), or 200 mg (JNJ-3989 dual 200 mg group) every 4 weeks; or JNJ-6379 250 mg plus JNJ-3989 100 mg every 4 weeks plus nucleos(t)ide analogues (triple group) for 48 weeks followed by a follow-up phase. An interactive web response system provided concealed treatment allocation, and investigators remained masked to the intervention groups until the primary analysis at week 48. The primary endpoint was the proportion of patients meeting predefined nucleos(t)ide analogue-stopping criteria (alanine aminotransferase <3 × upper limit of normal, HBV DNA below the lower limit of quantitation, HBeAg negative, and HBsAg <10 IU/mL) at week 48. All patients who received at least one dose of study drug were included in the analysis population used for primary efficacy assessment, excluding those who withdrew because of COVID-19-related reasons, withdrew before week 44, or had no efficacy data (ie, the modified intention-to-treat population). Safety was assessed in all participants who received at least one dose of study drugs. This trial is registered with ClinicalTrials.gov, NCT03982186. The study has been completed.

Findings: Between Aug 1, 2019, and April 26, 2022, 470 patients (310 [66%] male and 244 [52%] White) were randomly assigned: 45 to the control group, 48 to the JNJ-6379 dual group, 93 to the JNJ-3989 dual 40 mg group, 93 to the JNJ-3989 dual 100 mg group, 96 to the JNJ-3989 dual 200 mg group, and 95 to the triple group. At week 48, five (5%; 90% CI 2-11) of 91 patients in the JNJ-3989 dual 40 mg group, 15 (16%; 10-24) of 92 in the JNJ-3989 dual 100 mg group, 18 (19%; 13-27) of 94 in the JNJ-3989 dual 200 mg group, eight (9%; 4-15) of 94 in the triple group, and one (2%; 0-10) of 45 in the control group met nucleos(t)ide analogue stopping criteria. No patients in the JNJ-6379 dual group met stopping criteria. 38 (81%) patients who met nucleos(t)ide analogue-stopping criteria at week 48 were virologically suppressed and HBeAg negative at baseline. Ten (2%) of 470 patients had serious adverse events during the treatment phase, and two patients (one each from the JNJ-3989 dual 200 mg group [exercise-related rhabdomyolysis] and the triple group [increase in ALT or AST]) had serious adverse events related to study treatment. During follow-up, 12 (3%) of 460 patients had a serious adverse event; one (<1%), a gastric ulcer, was considered to be related to nucleos(t)ide analogues and occurred in a patient from the JNJ-3989 dual 200 mg group. 29 (6%) of 460 patients in the treatment phase and in ten (2%) of 460 patients in the follow-up phase had grade 3 or 4 adverse events. Five (1%) of 470 patients discontinued treatment due to adverse events, and there were no deaths.

Interpretation: Although treatment with JNJ-3989 led to a dose-dependent response for meeting nucleos(t)ide analogue-stopping criteria, it rarely led to HBsAg seroclearance. However, most patients treated with JNJ-3989 had clinically meaningful reductions in HBsAg that might contribute to a liver environment conducive to better immune control and, in turn, might improve the response to immune-modulating therapies.

Funding: Janssen Research and Development.