Experimental therapy improves liver fat and fibrosis in people with NASH

GS-0976, an acetyl-CoA carboxylase (ACC) inhibitor being developed by Gilead Sciences, led to significant reductions in liver fat accumulation and fibrosis in people with non-alcoholic steatohepatitis (NASH), according to phase 2 study results presented at the 2017 AASLD Liver Meeting last week in Washington, DC.

NASH, and its less severe form, non-alcoholic fatty liver disease (NAFLD), refer to the build-up of fat in the liver in people who do not drink heavily. Fatty liver disease, which is often associated with obesity and metabolic syndrome, is a common cause of chronic liver disease worldwide. Over time, fat accumulation in the liver (steatosis) and the accompanying inflammation and build-up of scar tissue (fibrosis and cirrhosis) can interfere with normal liver function and lead to liver cancer. While weight loss and exercise can improve fatty liver disease, to date there are no approved therapies for NASH.

Rohit Loomba of the University of California at San Diego and colleagues evaluated the safety of efficacy of GS-0976 for people with NASH in a phase 2 clinical trial.

Acetyl-CoA carboxylase plays a key role in de novo lipogenesis, or conversion of carbohydrates into fatty acids in the liver, which are stored as triglycerides to provide energy when food is unavailable, Loomba explained as background. If food is plentiful, de novo lipogenesis can contribute to steatosis, NAFLD and NASH.

GS-0976, which blocks the first chemical step in the process (conversion of acetyl-CoA to malonyl-CoA), inhibits de novo lipogenesis. It was shown to improve liver steatosis, inflammation and fibrosis in pre-clinical studies, and it was associated with reduced de novo lipogenesis and liver fat accumulation in a previous proof-of-concept study of NASH patients, Loomba said.

The phase 2 trial enrolled 126 people with NAFLD at 41 sites in the US. Participants had to have a liver fat proportion of at least 8% according to magnetic resonance imaging (known as MRI-estimated proton density fat fraction or MRI-PDFF), liver stiffness of at least 2.5 kPa according to magnetic resonance elastography (MRE) or a liver biopsy showing NASH and mild to moderate fibrosis (stage F1-F3). However, people with liver cirrhosis (stage F4) were excluded.

About two-thirds of participants were women, the median age was approximately 56 years, about 85% were white and more than a third were Latino/Hispanic (a group with a high rate of fatty liver disease). The median body mas index was 33 (considered obese), about 60% had diabetes and about 40% had advanced fibrosis.

Study participants were randomly assigned to receive GS-0976 at daily oral doses of 5mg or 20mg, or else a placebo, for 12 weeks.

The main study endpoints were a reduction in MRI-PDFF of at least 30%, an MRE liver stiffness reduction of at least 15% or a reduction in liver stiffness by FibroScan. The researchers also looked liver enzyme levels including ALT and AST and the enhanced liver fibrosis (ELF) score, a biomarker index that includes serum levels of tissue inhibitor of metalloproteinases 1 (TIMP-1), amino-terminal propeptide of type III procollagen (PIII-NP) and hyaluronic acid.

During the 12-week treatment the amount of liver fat by MRI-PDFF fell by a median of 29% from the baseline level in the 20mg GS-0976 group, by 13% in the 5mg group and by 8% in the placebo group. Nearly half (48%) of people in the 20mg group, 23% in the 5mg group and 15% in the placebo group saw at least a 30% reduction. The decline in the 20mg group – but not in the 5mg group – was statistically and clinically significant, Loomba reported.

Liver stiffness according to FibroScan fell by 11% in the 20mg group, 8% in the 5mg group and 3% in the placebo group. Conversely, MRE stiffness showed the least improvement in the higher-dose GS-0976 group (6%, 10% and 13%, respectively). However, the differences in liver stiffness by either method were not statistically significant.

Although ALT, TIMP-1 and PIII-NP decreased more in the 20mg GS-0976 group than in the 5mg or placebo groups, only the TIMP-1 decline was statistically significant, and the composite ELF score did not fall significantly in any group.

In an analysis restricted to those who had at least a 30% MRI-PDFF response, liver stiffness by MRE, liver enzyme levels and some fibrosis markers fell more than they did in MRI-PDFF non-responders, with some of these differences reaching statistical significance.

Treatment with GS-0976 was generally safe and well tolerated, and almost all participants completed treatment. Serious adverse events and discontinuations due to adverse events were rare and did not appear to be drug-related. The most frequently reported adverse events were nausea, abdominal pain and diarrhoea, which were more common in the 20mg group than in the 5mg or placebo groups.

The laboratory abnormality of most concern was increased triglycerides and, interestingly, elevations were greater among those receiving the lower GS-0976 dose. Triglyceride levels rose by 11% in the 20mg group and by 13% in the 5mg group, while falling by 4% in the placebo group. Ten per cent of people in both the 20mg and 5mg groups developed grade 3 triglyceride elevations (500-1000 mg/dl), while 4% and 8%, respectively, had grade 4 elevations (over 1000 mg/dl). No one in the placebo group had seriously elevated triglycerides. People with triglyceride levels over 250 mg/dl at baseline and higher MRE liver stiffness were more likely to see substantial triglyceride elevations.

Among 16 people with asymptomatic triglyceride elevation who continued on GS-0976, the four people who took fish oil or fibrates (medications to reduce blood lipid levels) all saw their levels fall below 500 mg/dl at week 12. Seven of the 12 who were not treated with lipid-lowering therapy still spontaneously fell below the 500 mg/dl threshold by week 12.

People treated with 20mg GS-0976 in this study saw a "significant improvement in hepatic steatosis" by MRI-PDFF, a significant improvement in TIMP-1 and dose-dependent reductions in ALT and PIII-NP, the researchers summarised.

Although they concluded that treatment with GS-0976 was "safe and well tolerated," they said that its effect on triglyceride levels requires longer-term follow-up.

"In this first randomized, placebo-controlled, phase 2 study of an ACC inhibitor in NASH, the data suggest that GS-0976 has the potential to play an important role in treating patients with this disease," Loomba said in a Gilead press release. 

Future studies of GS-0976, both alone and in combinations, are planned, according to Loomba.