An antisense agent, GSK-3228836, significantly reduced
hepatitis B surface antigen levels in previously untreated people with chronic
hepatitis B after three injections, an early phase trial reported this week at
the Digital International Liver Congress.
GSK-3228836 (formerly known as ISIS-505358) (proposed name
bepirovirsen), is an antisense oligonucleotide, a chain of nucleic acid designed
to cut hepatitis B RNA, preventing the transcription of viral proteins.
The drug was developed by the biotechnology company Ionis
and has been licensed to GlaxoSmithKline (GSK) for development, along with a second
antisense oligonucleotide, GSK3389404. GSK aims to develop a functional cure – long-term virus suppression after a fixed course of treatment – using a
combination of agents.
GSK-3228836 was selected for development based on evidence
from pre-clinical studies that showed dose-dependent reductions in hepatitis B
surface antigen. The phase 2a study presented this week evaluated the antiviral
activity and safety of GSK-3228836 dosed over a 4-week period followed by 6
months of nucleoside/nucleotide analogue (NRTI) treatment with tenofovir or
The study recruited 18 NRTI-naïve and 6 NRTI-experienced
people with chronic hepatitis B in Hong Kong and South Korea, with hepatitis B
surface antigen levels above 50 IU/ml at screening. The study excluded people
with cirrhosis or co-infection with hepatitis C, hepatitis D or HIV.
Dr Man-Fun Yueng of Queen Mary Hospital, University of Hong
Kong, reported on treatment responses in the 300mg arm of the study (previous
results of the phase 2a dose-ranging study were presented at the AASLD Liver
Meeting in November 2019).
GSK-3228836 was administered by subcutaneous injection in
six doses at days 1, 4, 8, 11, 15 and 22.
The NRTI-naive group had a mean age of 42, two-thirds were
women and the mean hepatitis B surface antigen was 3.9 log10 IU/ml. Half were
HBeAg positive. Twelve were assigned to active treatment and six to placebo. Three
out of 12 had hepatitis B surface antigen reductions > 3 log10 IU/ml. Two
were below the lower limit of quantification (LLOQ) (0.05 IU/ml) by day 29. The
mean reduction in HBsAg was – 1.55 log10 IU/ml at day 29 compared to no change in
the placebo group (p = 0.001).
The NRTI-experienced group had a mean age of 48 years, four
out of five were male, the mean hepatitis B surface antigen level was 2.8 log10
IU/ml and all were HBeAg negative. Four were assigned to active treatment and
two placebo. Three out of four who remained on study treatment had hepatitis B
surface antigen reductions > 3 log10 IU/ml and two were below the lower
limit of quantification (LLOQ) (0.05 IU/ml) by day 36. The mean reduction in
HBsAg was – 2.51 log10 IU/ml at day 29 (non-significant compared to the placebo
group, p = 0.454).
Three participants in the NRTI-experienced group had ALT flares
that began during the GSK-3228836 treatment phase. In two cases the ALT increase
was 3 x upper limit of normal (ULN) and commenced around day 20 of treatment. In
the third case, ALT increased to 5 times ULN from day 20. All cases were
asymptomatic and coincided with a reduction in HBsAg, and ALT returned to the normal
range by day 120.
ALT flares occurred in both participants in the treatment-naïve
group who had hepatitis B surface antigen clearance (< LLOQ), also
self-limiting and asymptomatic.
In both study groups, ALT flares were associated with
greater hepatitis B surface antigen reduction.
Study treatment was well tolerated. Injection site reactions
lasting at least two days were observed in three of 12 treatment-naïve participants
and reactions lasting at least four days in two of five NRTI-experienced
GSK-3228836 is moving forward in a phase 2b study evaluating
the effects of 12 or 24 weekly injections of two doses in NRTI-naïve and -experienced