Estimating The Potential Success Of Late Clinical Stage NASH Treatments

by: Richelle Cutler-Strom

Caution should be used when assessing the promise of phase 2 results since some trials use biomarkers that are not conservative enough; e.g., NAS scores ≤ 5 with no other biomarker.

There are 5 NASH drugs in late-stage clinical testing and comparing details of their phase 2 trial data indicates one standout - Madrigal.

Madrigal's MGL-3196 shows high efficacy in reducing liver fat with significantly reduced liver enzymes in 125 disease-verified subjects over a 36-week dosing period, demonstrating efficacy with a robust safety profile.

Editor's note: Seeking Alpha is proud to welcome Richelle Cutler-Strom as a new contributor. It's easy to become a Seeking Alpha contributor and earn money for your best investment ideas. Active contributors also get free access to the SA PRO archive. Click here to find out more »

There is a worldwide epidemic of obesity and metabolic syndrome driving the need to find treatments for high cholesterol and nonalcoholic steatohepatitis (NASH), a non-alcoholic fatty liver disease (NAFLD). Currently, there is no approved drug for treating NASH, and the market for an effective treatment is huge.

The global non-alcoholic steatohepatitis market was valued at $1.2 billion in 2017, and is expected to reach $21.5 billion by 2025, growing at a compound annual growth rate of 58.4% from 2021 to 2025." - Business Wire, August 09, 2018

NASH clinical trials use biomarkers such as serum lipids and triglycerides, liver enzymes, fat, and histology to assess disease severity. The NAFLD activity score (NAS) is composed of three subjective histology taken from liver biopsy. Studies have shown that scores ≤5 have only a 60% or less accuracy in diagnosing NASH, so trials should only include subjects with NAS scores >5. The liver enzyme alanine aminotransferase (ALT) is a sensitive indicator of liver health and fat and its level is significantly correlated with NAS scores (p<0.0001). Thus, a decrease in ALT level may be a strong indicator of clinical significance.


Allergan's (AGN) phase 2b trial of Cenicriviroc included patients with NAS scores of ≥ 4. Cenicriviroc barely improved fibrosis by ≥ 1 (scale 0-4) with p = 0.04. ALT measures were worse or unchanged after 1 year on Cenicriviroc (buried in Table S5), and NAS improvement of ≥ 2 points was not met. Considering that ALT values are found to be significantly associated with fibrosis improvement, unchanged ALT further undermines the barely significant improvement in fibrosis. With such weak phase 2 results, efficacy should be a major shareholder concern with the release of phase 3 results.


The results from Genfit's (GNFTF) phase 2 trial of Elafibranor appear promising, but included patients with NAS scores ≥3, which are too low to adequately define NASH. Elafibranor improved NASH resolution with p=0.045 in subjects with NAS scores ≥ 3 and produced even stronger effects in subjects with increased baseline severity (NAS ≥4) (p<0.007). However, interestingly there was no significant difference in NAS scores after Elafibranor even within the increased baseline severity subgroup unless the analysis was restricted to a subset of patients from centers that enrolled at least 1 in each treatment arm. Furthermore, Elafibranor showed no significant difference in ALT values when compared to placebo, and tellingly, no report comparing patients with increased baseline severity. The modified NASH resolution criterion requires disappearance of ballooning and is indicative of decreased cell death. Thus, unchanged ALT values with disappearance of ballooning are unexpected but could be explained by the lenient ballooning inclusion score of ≥1. Given that tissue sampling variation could cause scores to vary from 1 (few cells) to zero, an inclusion score of 1 may be too low. A 72-week phase 3 clinical trial is underway and interim results are expected by the end of 2019, but subgrouping and unchanged ALT in phase 2 should bridle shareholders' optimism.


Gilead's (GILD) Selonsertib showed suggestion of fibrosis improvement but no significant change in liver health biochemistry in a 24-week phase 2 trial. Selonsertib selectively inhibits the apoptosis signal-regulating kinase (ASK1) to reduce stress response and cell death. Reducing cell death, however, may increase the risk of liver cancer, which increases with end-stage NASH. The FDA needs to ensure long-term safety before approving an anti-apoptotic drug. An interim report of phase 3 results could be released by mid-2019. However, until the 240-week event-free survival results are shown, shareholders should be on guard.

*Since this was written, Gilead released interim phase 3 results that showed Selonsertib failed to improve fibrosis by ≥1-stage at 48 weeks in patients with advanced fibrosis. Gilead's trial with Selonsertib in patients with bridging fibrosis is still ongoing.

Intercept Pharmaceuticals

Intercept's (ICPT) Obeticholic acid is another hopeful candidate for treating NASH, but there's a foreshadowing of toxicity issues. In 2018, the FDA issued a black box warning for Obeticholic acid in the treatment of cholestasis. The warning was issued after problems with liver toxicity. Aside from the concerning finding of significantly increased LDL cholesterol and insulin resistance after 72 weeks of Obeticholic acid treatment, there was a conspicuous lack of safety data in the trial report. Another report described the opposite desired effects with Obeticholic acid, which explains why the data safety and monitoring board recommended discontinuing treatment for the remaining patients. Intercept is expected to report phase 3 interim results in Q1 2019, but with Intercept insiders currently selling rather than buying, optimism is hard to muster.

Madrigal Pharmaceuticals

Madrigal (MDGL) recently reported extended phase 2 results showing MGL-3196 reducing liver fat by ≥ 30%, as measured by MRI-PDFF, in 77% of patients at 36 weeks and 75% of patients at 12 weeks, (p<0.001). MGL-3196 treatment also produced at least a 2-point reduction in NAS score in 61% of patients (p=0.02). Daily treatment with MGL-3196 in patients with elevated baseline ALT significantly reduced ALT compared to placebo (n=75), (p< 0.01).

Viking Therapeutics

Viking (VKTX) recently reported phase 2 result showing VK-2809 reduced liver fat by ≥ 30%, as measured by MRI-PDFF, in 90.9% of patients (n=11) at 12 weeks (p=0.002). In patients with elevated baseline ALT, every other day dosing with VK-2809 significantly reduced ALT (n=13) compared to placebo, (p< 0.02). However, significant ALT reduction is reported from week 16, which followed a 4-week off-drug phase. It would be nice to see ALT values at 12 weeks after daily dosing.

Comparing Madrigal and Viking

MDG-3196 and VK-2809 are both thyroid hormone receptor-beta (THR-β) agonists. MGL-3196 reportedly showed a 28-fold selectivity for THR-β over THR-α in a functional affinity assay. In contrast, VK-2809 displayed 16-fold selectivity for THR-β over THR-α in an affinity binding assay, however, VK-2809 is further targeted to the liver by liver-specific enzyme processing of its prodrug.

Toxicity issues sometimes don't appear until phase 3 trials. For instance, liver enzymes were mildly elevated in KaroBio's phase 2 trial of THR- β agonist, Eprotirome, but then failed its phase 3 trial due to liver damage. Considering that higher dosages of VK-2809 caused mild elevation in liver enzymes after only 14 days in trial 1B, off-target liver toxicity may explain why VK-2809 was also dosed at 10 mg every other day.

At first blush, VK-2809 appears to be more efficacious than MGL-3196. However, it's important to consider that Madrigal's trial included more test subjects (125 compared to Viking's 59) and used a more rigorous subject-inclusion criteria of heterozygous familial hypercholesterolemia (liver disease) and biopsy-proven NASH, whereas Viking's subjects were selected by a less-selective criteria, LDL >110 mg/dL. Thus, MGL-3196's efficacy may appear lower than VK-2809 because Madrigal's patients were much sicker. When VK-2809 is administered to sicker NASH patients in phase 3, efficacy and toxicity could be an issue.


Madrigal, Viking, and Genfit are small clinical development companies that have yet to bring a drug to market. Madrigal and Viking have ≤11 employees, while Genfit has ~150 employees. Madrigal and Viking's drugs are licensed, while Genfit's drugs are proprietary. A license agreement can sometimes indicate the value of a drug. Under license agreement, Madrigal is required to pay a remainder of $10 million to Roche (OTCQX:RHHBY) on the commencement of its phase 3 trial and regulatory approval and single digit royalty of sales. Viking licensed 5 drugs including VK-2809 from Ligand Pharmaceuticals (NASDAQ:LGND) and Metabasis Therapeutics for a loan in aggregate of $2.5 million, with a 5% annual interest rate, and shares, however, in May 2018, Viking repaid the entire loan in cash. The patents for VK-2809 and MGL-3196 will expire in 2025 and 2026, respectively.

Currently, Viking has four drugs in their pipeline, Genfit has three, and Madrigal only two. According to the most recent US SEC filing, Madrigal has 15.4 million shares outstanding shares, out of 200 million authorized, and a market cap of 1.8 billion, while Viking has 71.5 million shares outstanding, out of 300 million authorized, and a market cap of 0.6 billion. Madrigal's book value is 490 million and Viking's is 300 million. Madrigal's monthly burn rate for last quarter reported was 7 million versus Viking's 5.6 million, the difference probably reflects higher expenditure from MGL-3196's advanced clinical development with a larger and extended trial. Madrigal has 38 million cash with another 525 million in marketable securities, while Viking has 169.5 million cash and 167 million in marketable securities.

Genfit is based in France and reported financials date from Dec. 2017. Genfit had 31 million shares outstanding, a monthly burn rate of 5.2 million, and cash equivalence of $310 million USD. Genfit's book value is 117 million but has a total of 163 million in debt from loans with interest. Considering debt and securities, Madrigal appears to be in better financial health but has higher risk with only two products.

The table below summarizes key findings from the phase 2 trials discussed above.





NAS score baseline - treat

Liver fat MRI-PDFF >30%



C-Chemokine antag


≥4 - NC




PPARγ agonist


≥3 - NC/NR




ASK1 inhibitor


>5 - NC



Obeticholic acid

FXR agonist


≥3 - 2 (45%)




THR-β agonist


4.9 - 2 (61%)

75% p<0.0001



THR-β agonist



91% p<0.002

NC- no significant change; NR- not studied; NAS score baseline -NAS reduction after treatment (percentage patients); (ASK1) selective inhibitor of the apoptosis signal-regulating kinase 1;


Some trial reports seem to infer treatment efficacy when results show only a suggestive trend. Tactics to coerce efficacy may include choosing subgroups for analysis, using too lenient criteria to categorize improvement, or maybe even a red herring, where in lieu of reporting treatment significance, a significant correlation between measures, such as collagen and fibrosis is reported. Based on the phase 2 reports, I was surprised that some companies decided a phase 3 trial was merited.

Shareholders should be cautious when trials report barely significant statistics. Too many failed clinical trials and non-replicable studies are causing the gold standard of significance, the p< 0.05 threshold, to be reconsidered. Accordingly, many argue that a stronger significance threshold of p< 0.005 would be more indicative of clinical significance.

The key to a successful NASH drug could be the drug's target since there is likely an optimal point of intervention in NASH pathogenesis. Drugs that only target the effects of fatty liver, e.g., inflammation and fibrosis, and not the fat itself, do not resolve all the effects of fatty liver. For instance, fatty liver causes mitochondrial dysfunction and impairs insulin secretion. Both MGL-3196 and VK-2809 activate THR- β to effectively reduce serum triglycerides and LDL and thereby ameliorates NASH. A drug for reducing serum triglycerides and LDL would also be prescribed for atherosclerosis prevention. There are at least a dozen more NASH medications currently in phase 2 trials. Among the many candidates, the first NASH treatment should be very profitable for shareholders.

Disclosure: I am/we are long MDGL. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.