Emricasan and selonsertib fail to improve fibrosis in people with NASH

A pair of experimental therapies for non-alcoholic steatohepatitis (NASH) failed to offer notable improvements in controlled clinical trials, despite having shown promising effects on liver biomarkers in earlier studies.

Dr Stephen Harrison of Pinnacle Clinical Research in Texas, who presented results from both studies this week at the AASLD Liver Meeting in Boston, said that although the results were negative, they will help researchers understand more about this difficult-to-treat condition and inform the design of future trials.

Non-alcoholic fatty liver disease (NAFLD) and its more severe form, NASH, are responsible for a growing proportion of advanced liver disease. The build-up of fat in the liver triggers cell death and inflammation, which over time can lead to fibrosis, cirrhosis and liver cancer.

Linked to obesity and diabetes, NAFLD/NASH is increasingly recognised as a manifestation of the metabolic syndrome. Currently there are no effective medical therapies for the condition and weight loss remains the mainstay of treatment.

Emricasan is a pan-caspase inhibitor that interferes with the activity of caspase protease enzymes that play a role in apoptosis (programmed cell death) and inflammation. Over the past few years, researchers have reported that the drug improved biomarkers of liver health in people with cirrhosis, though it did not significantly improve portal hypertension in people with NASH-related cirrhosis.

Dr Harrison presented findings from a randomised, placebo-controlled trial, sponsored by Conatus, that compared emricasan to placebo in people with NASH and fibrosis. The primary endpoint was fibrosis improvement by at least one stage without worsening of steatohepatitis (inflammation related to fat accumulation).

The study included 318 participants at nearly 90 sites in Europe and the US. Over half (56%) were women, more than 90% were white and the median age was 54. Three-quarters were classified as obese, half had diabetes, about 60% had metabolic syndrome and two-thirds had hypertension. Study entry criteria included an NAFLD activity score (NAS) of 4 or higher and mild to advanced fibrosis; about 21% had stage F2 fibrosis, 42% had stage F2 and 37% had stage F3.

Participants were randomly assigned to receive 5mg or 50mg of emricasan or a placebo twice daily for 72 weeks. Liver biopsies were done at baseline and again at week 72.

At the end of treatment, the primary endpoint was not met: just 11% in the emricasan 5mg group and 12% in the 50mg group had at least a one-stage fibrosis improvement without worsening steatohepatitis – about half as many as in the placebo group (19%). In fact, emricasan recipients were more likely than placebo recipients to experience worsening fibrosis (41%, 38% and 20%, respectively).

The secondary endpoint of NASH resolution was also less likely in the two emricasan groups compared with the placebo group (4%, 7% and 11%, respectively). This was also the case for NASH response, defined as at least a 2-point decrease in the score (10%, 9% and 18%, respectively) and the components of inflammation and ballooning of liver cells.

Treatment with emricasan was generally safe and well tolerated. Adverse events were similar in the emricasan and placebo groups and just 2% discontinued the study for this reason.

Harrison noted that ALT and caspase levels did decrease during treatment, indicating that emricasan was engaging the target as expected, but this appeared to result in worsening clinical outcomes rather than improvement.

"Further evaluation of the mechanisms underlying this finding may provide important insights into the understanding the role of necro-apoptosis in NASH pathophysiology," the researchers concluded.

Harrison also reported disappointing findings from two phase 3 studies of selonsertib (formerly GS-4997). This drug inhibits apoptosis signal-regulating kinase 1 (ASK1), which promotes inflammation, liver cell injury and fibrosis.

After an earlier phase 2 trial showed that a substantial number of NASH patients who took selonsertib experienced at least a one-stage decrease in fibrosis and a reduction in liver fat, Gilead Sciences initiated a pair of larger phase 3 trials, STELLAR-3 and STELLAR-4.

Together these studies enrolled 1679 participants at 450 sites in 27 countries. STELLAR-3 included 802 people with bridging fibrosis (stage F3) while STELLAR-4 included 877 people with compensated cirrhosis (stage F4).

Across both studies, about 60% were women, most were white and the median age was 59 years. More than half were classified as obese and three-quarters had diabetes. All had biopsy-confirmed NASH and more than 80% had NAS scores of 5 or higher.

Participants were randomly assigned to received selonsertib at doses of 6mg or 18mg taken by mouth once daily or a placebo for 48 week, at which point they received a second biopsy.

The trials were originally planned to last for five years, but they were halted at week 48 after biopsy results at that point showed a lack of efficacy, as Gilead announced earlier this year.

In STELLAR-3, 12% of participants treated with the lower dose of selonsertib and 10% of those who received the higher dose had at least a one-stage improvement in fibrosis without NASH worsening, Harrison reported. These percentages did not differ significantly from that of the placebo group (13%).

Likewise, 13% of STELLAR-4 participants randomised to receive 6mg of selonsertib and 14% of those treated with the higher dose achieve the primary endpoint, again not significantly different from the placebo group (13%).

Looking at other histological endpoints in STELLAR-3, 4% in the low-dose group and 5% in the high-dose group experienced NASH resolution without worsening fibrosis, compared with 9% in the placebo group. Both improvement in fibrosis (14%, 13% and 16%, respectively) and progression to cirrhosis (16%, 13% and 16%) occurred at statistically similar rates across all three groups.

The same pattern was seen in STELLAR-4. Fibrosis improvement (17%, 19% and 16%, respectively) and NASH resolution without worsening fibrosis (4%, 2% and 4%) were again statistically similar across all groups.

What's more, selonsertib had no beneficial effects on liver biochemistry, non-invasive tests or other biomarkers, according to Harrison. Overall, 3% of participants experienced liver-related clinical events during about 16 months of follow-up; these were numerically more frequent in the selonsertib groups compared with the placebo group, but differences were not significant.

As with emricasan, biochemical analyses revealed that selonsertib did inhibit ASK1, as intended, but again this did not lead to the desired improvement in clinical outcomes.

Selonsertib monotherapy was also safe and well tolerated. Approximately 15% of participants across treatment groups in both studies experienced grade 3 or 4 adverse events. Discontinuation due to adverse events was rare in all treatment arms (3% or less).

Given the number of different biological processes that play a role in the development of fatty liver disease, optimal treatment may involve drugs that work by different mechanisms. Agents that do not work well on their own but show positive effects on biomarkers may still have potential as components of combination therapy, Harrison said.

Gilead is currently testing selonsertib in combination with the nonsteroidal farnesoid X receptor (FXR) agonist cilofexor (formerly GS-9674) and the acetyl-CoA carboxylase inhibitor firsocostat (GS-0976) in the phase 2 ATLAS study (NCT03449446). FXR regulates bile acid synthesis and plays a role in lipid metabolism, while ACC is involved in de novo lipogenesis, or conversion of carbohydrates to fatty acids in the liver.