• Sustainability
  • DE&I
  • Pandemic
  • Finance
  • Legal
  • Technology
  • Regulatory
  • Global
  • Pricing
  • Strategy
  • R&D/Clinical Trials
  • Opinion
  • Executive Roundtable
  • Sales & Marketing
  • Executive Profiles
  • Leadership
  • Market Access
  • Patient Engagement
  • Supply Chain
  • Industry Trends

Resolving a Key Bottleneck in NASH R&D


Jan Lichtenberg, Ph.D, and Scott Friedman, MD, talk about the growing interest in nonalcoholic steatohepatitis (NASH) in the pharma industry, the complexities of modeling the disease, and what this means for drug development efforts.

An interview with Jan Lichtenberg, PhD, Chief Executive Officer, InSphero and Scott L. Friedman, MD, Chief, Division of Liver Diseases and Dean for Therapeutic Discovery at the Icahn School of Medicine, Mount Sinai.

Jan Lichtenberg, Ph.D, InSphero CEO and Scott Friedman, MD, recently spoke with Pharmaceutical Executive about the growing interest in nonalcoholic steatohepatitis (NASH) in the pharmaceutical industry, the complexities of modeling the disease, and what this means for drug development efforts.

InSphero, a pioneer in 3D-cell-based technologies for the pharmaceutical industry since 2009, has engineered a highly physiologically relevant 3D platform modeling the complexities of NASH to enable greater confidence in decision making in discovery and development. These comprehensive solutions for in vitro testing of novel drugs are used by major pharmaceutical companies worldwide to increase efficiency in drug discovery and safety testing.

Since 1986, Dr. Friedman has been supported by NIH grants for the study of the pathogenesis of liver fibrosis. He was instrumental in isolating and characterizing the hepatic stellate cell, which is the most important source of scar or extracellular matrix in chronic liver disease. Dr. Friedman serves as head of the Fibrosis Research Center at Mount Sinai, which was established to facilitate the development of novel diagnostic methods and treatments of liver fibrosis.

Pharm Exec: The pharmaceutical industry has an intense focus on NASH. What is driving this interest?

Scott Friedman

Scott Friedman: The uptick in interest is attributable to a few convergent events. One is that the prevalence of obesity is quite high and there is a growing recognition that in patients who have metabolic syndrome associated with obesity, there is a significant risk of underlying liver disease. The other factor is that there were several pharmaceutical companies in the liver space for many years, developing antivirals and other specialized treatments for viral hepatitis. With hepatitis C conclusively conquered, these companies are exploring new opportunities related to the unmet needs that exist across the liver disease spectrum. There is no question that NASH has become the most compelling.

Jan Lichtenberg: It’s no surprise that the Wall Street Journal cleverly described NASH as “a big fatty opportunity for big pharma”. There are currently no approved therapies and the market could reach $20-$35 billion by 2025. About 30-40% of the US population suffers from a precursor disease called non-alcoholic fatty liver (NAFL), closely associated with type 2 diabetes. Estimates are that 30% of these patients will go on to develop NASH. Left untreated, NASH can lead to cirrhosis and in some cases, the need for a liver transplant. Unfortunately, recent Phase III trial results have been disappointing. Compounds are failing or at best, showing mixed results where the impact in terms of NASH resolution is offset by potential liver toxicity. This is not only a setback for the companies involved, it is also a concern for the many patients out there waiting for an efficacious therapeutic intervention.

There are no FDA approved therapies for NASH and this is attributed, in part, to a lack of biologically relevant preclinical human models. Why is this the case?

Scott Friedman: NASH is a disease that’s much more complex, conceptually, than viral hepatitis. Viral hepatitis is an exogenous infection, not an endogenous disease. NASH is part of this systemic metabolic syndrome, and so the complexity is an order of magnitude greater. In NASH, there are many abnormalities, and we cannot conclusively establish a hierarchy for which are the most determinative in establishing or driving liver disease. Patients have hyperlipidemia; they have insulin resistance; they have an altered microbiome; they have increased oxidant stress; they have dysregulation of other hormonal signaling because of their obesity. We don’t know which starts the ball rolling, which are necessary, and which are sufficient. Modelling all the elements of the human disease is challenging in an animal model of NASH, compared to a humanized model, which you could infect with a human hepatitis virus.

Jan Lichtenberg

Jan Lichtenberg: Development of novel therapeutics is also impeded by the lack of predictive in vitro models which reflect the complex mechanisms underlying NASH initiation and progression. Liver tissue is heterotypic and many cell types interact during initiation and progression of this disease including hepatocytes, stellate cells, sinusoidal endothelial cells and Kupffer cells, the resident macrophages. NASH develops slowly; the stages of progression from steatosis to NASH and eventually fibrosis – scarring of the liver –occur in humans over an average of more than 10 years. An in vitro model has to recapitulate this in an accelerated timeframe to be effective and practical.

Historically, in vitro models of liver fibrosis have been 2D cultures of stellate cells, which play a key role in the initiation, progression, and regression of liver fibrosis. This approach, however, ignores the contribution and interplay of other cell types and the overall liver microenvironment and it comes with a somewhat ironic, yet, serious liability: stellate cells become spontaneously activated upon attaching to plastic surfaces in culture.

Where do you see 3D disease models having an impact?  

Scott Friedman: In preclinical testing. Several key questions must be addressed animal models:  Can we replicate all the features of the disease faithfully? Is the animal’s response to a particular drug indicative of a likely response in patients? And can you do that quickly and screen large numbers of drugs or compounds, and even combinations? This last requirement is where the animal models start to fall short, because if you want to know if a drug has a high likelihood of working in humans, you want to use an animal model that’s faithful, but how many different drugs and doses can you realistically test? Screens with animal models take months, and we do many of them so we are aware of these limitations.

The other challenge has been that rodents in general are more responsive to drugs than humans. And there is some interesting data that speaks to that. There was a terrific study from Barbara Rehermann at NIH in Cell in which she showed that inbred mouse strains used in drug development tend to have a relatively simpler microbiome, and in some way that’s linked to a higher severity of disease and presumably higher response to therapies. If you study mice that have a wildtype or wild microbiome, they tend to be much more resistant, at least in a couple of different disease models. It may well be that our mouse models are ratcheted up to develop more disease than what we actually see in humans, and therefore are more responsive to therapies.

If 3D systems like InSphero’s can provide the kind of speed and predictability of drugs and combinations of drugs, they could have a real niche in the market. Keep in mind that fatty liver disease is part of a systemic disease, and when you isolate a liver spheroid and you’re just working with liver or liver cells, you’re excluding the potential impact of circulating hormones that are derived from abnormal metabolism. But that’s a limitation that I think most could live with if it’s accurate. The rationale for these 3D systems is pretty clear and compelling, as long as it translates into efficacy in vivo.

Jan Lichtenberg: Most everyone, whether in industry or academia, agrees that moving to 3D models makes a lot of sense from a physiological perspective and it has a lot of technical and practical advantages. And there is a unison of understanding that this is a powerful addition in our toolbox to create more predictive assays. With NASH, we have a perfect opportunity to contribute to the quest for new drugs with a technology that fulfills two important key requirements: it is predictive, which we could already demonstrate with numerous reference compounds, but it is also fast and scalable. Especially the latter will be instrumental to look into potential synergies of anti-NASH or otherwise metabolically relevant compounds in a combinatorial fashion. If we want to go one step further by combining liver and other tissues relevant to this complex disease, Akura™ Flow, our organ-on-a-chip technology, enables us to develop these types of multi-tissue networks for preclinical testing.

What advice do you have for pharma companies interested in moving to use of 3D models?

Jan Lichtenberg: The first step is always to identify the real problem you need to solve. That should be the starting point of every discussion. This is one of the reasons InSphero started in liver toxicology – it was clear this was an area in need of new technology. Hepatocytes grown in 2D culture for tox studies lose their liver-specific functionality after two days and as a result, you only see really acute aspects of toxicology. If you want to take a more physiological stance and do it for 7 days or 14 days, there was no way of doing it. This challenge led us to develop 3D liver microtissues for toxicology studies.

Coming back to NASH, there are very specific needs and customer problems requiring a new solution. NASH will have a substantial impact on healthcare cost in the decades to come but currently, there is no pharmaceutical intervention. There is currently no proper in vitro model to understand what’s going on, how drugs work – alone or in combination – and which drugs work well or not. So in this case, the risk of using a 3D model is fairly low. We already know that 2D models and animal models have limitations in terms of translation, and for animal models, the added problem is the time required to get data – four to six months compared to three weeks using a properly designed in vitro model.


Related Videos
Related Content