Detecting Human and Cross-species Drug Toxicities With the Liver-Chip

Geraldine A. Hamilton, PhD, President and Chief Scientific Officer of Emulate Inc., shares her perspectives on the challenges of using animal toxicity models and discusses how these can translate to late-stage attrition in drug development. She also highlights the advantages of using “organ-on-chip” technologies for toxicity testing and delves into the company’s recent findings related to the ability to reproduce human and cross-species drug toxicities using a liver-chip, published in Science Translational Medicine.

Laura Lansdowne (LL): What preclinical research is required by regulatory agencies to allow for a drug candidate to progress to clinical trials?

Geraldine Hamilton (GH): Liver toxicity is ubiquitous in drug development (as well as foods and consumer products), and so the Liver-Chip is one of the key Organ-Chips of interest to pharmaceutical companies. The fact that more than one pharmaceutical company collaborated in this study with the Liver-Chip shows that they collectively recognize the need for better and more predictive technologies beyond animal models. They also highlight a willingness to work together for the greater benefit so that we can develop industry-wide technologies to bring the best and safest medicines to patients. We see the Organ-Chips being used side-by-side with existing animal models to provide human relevant data to enable better predictions and translation to the clinic and reduce failure rates in drug discovery and development due to lack of safety or efficacy.

LL: What challenges are linked to the use of animal toxicity models and how do these translate to late-stage attrition in drug development?

GH: It is well known that there can be poor correlation between human and animal liver toxicity, with animal models being predictive of human toxicity only a percentage of the time. This can lead to failures of drugs in the clinic or withdrawal from the market.

This study with the Liver-Chip clearly demonstrates the ability of our product to predict human toxicity that was undetected in animal studies for drugs. By reenacting these real-world drug programs, we have shown the potential to predict human liver toxicity before clinical trials, thereby avoiding the time and money – as well as the toll on human lives – for drug candidates that ultimately were halted in clinical trials.

Also, the detailed insights that the Liver-Chip provides about biological mechanisms opens up ways to improve the drug development process by optimizing the design of efficacious drug candidates to avoid undesirable side effects.

We see the chips being used across the drug development process – from early drug discovery to enable target validation and disease modeling, through drug development in the clinic that enable better prediction of safety and efficacy.

LL: Can you summarize what is meant by “organ-on-chip” technologies? What advantages are there for using these models compared to “traditional” animal models?

GH: There are two main things that set our Liver-Chip – and all of our Organ-Chips – apart. The first is the human-relevant and exquisite detail with which our models recreate the key functions of a human organ, including all the true-to-life aspects such as: cell architecture, full range of key cell types, cell-to-cell and cell–extracellular matrix interactions, physiological-relevant mechanical forces, and a dynamic microenvironment/flow. In the case of the Liver-Chip, this human-relevant details also includes the full range of key cell types in the liver, key hepatic functions such as albumin secretion and drug metabolism, and the ability to measure hepatic clinically relevant biomarkers such as mir122 and albumin.

The biological complexity and relevance that we achieved enables us to translate human relevant data to the clinic. The second thing that sets our Liver-Chip apart is the fact that it works within a lab-ready platform, our Human Emulation System. Our platform consists of instruments, the biology (Organ-Chips), software, and all the supporting protocols and guidelines. Our instruments provide all the surrounding features to support the microenvironment akin to the human body, including rates of flow and stretch, for the cells to thrive in the Chips.  Importantly, the platform is automated so that any researcher in any lab can apply our technology in a robust and reproducible way. This is what Emulate calls democratizing our technology, so that it can be in the hands of any researcher.

LL: Could you summarize the recent findings published in Science Translational Medicine?

GH: The most important observation from our study is that our Liver-Chip could demonstrate different drug responses across species, showing clear differences in human liver toxicity from other animal species. Part of our findings were not surprising, because we expected that our human Liver-Chip would reproduce the toxicity that had occurred in drugs that were found to be toxic in human clinical trials. Unfortunately, in the past and without a human-relevant technology like our Liver-Chip, the human toxicity of these drug candidates was undetected in animal studies that were conducted at that time, and these trials were halted due to adverse effects and even death in patients. What was a remarkable finding in our study across all the drugs that we tested across species is that our Liver-Chip opened up insights as to the reasons for the cross-species differences. The chip was able to recreate the relevant mechanisms of action that drove the toxicity in the clinic.

Understanding of the way in which a drug causes toxicity enables better human risk assessment in safety testing. Because our Chips provide a high-fidelity window into the biological mechanisms, we were able to answer questions about “why” and “where” there are differences between the human and the animal models – well beyond giving a “yes” or “no” answer about toxicity. These deeper insights can offer solutions for improving or optimizing drug design, so important learnings can be drawn from drug programs. 

LL: Considering these findings and the increasing “popularity” of organ-on-chip models do you think the current preclinical regulatory requirements for drug developers will be reviewed in the near future?

GH: Indeed, through Emulate’s multi-year collaborative research with U.S. Food and Drug Administration (FDA) – which focuses on the Liver-Chip and cross-species toxicology (topics related to this published research) – we have an ongoing dialogue about our work in this area.

Collaboration with a range of stakeholders is such a key part of our strategy. Since the early days of founding Emulate, we have established collaborations with FDA, as well as with pharmaceutical collaborators (in addition to Janssen and AstraZeneca who co-authored the Liver-Chip publication, we also announced early collaborations with Roche, Merck and Takeda.) We believe it will take all players in the ecosystem to establish Organ-Chips as a new standard. It will be our pharmaceutical customers who ultimately submit data from our Organ-Chips.

Here are further details about our CRADA agreement with FDA:

Emulate has a multi-year Cooperative Research and Development Agreement (CRADA) with the U.S. FDA that is led by FDA’s Division of Toxicology, within the Office of Applied Research and Safety Assessment/Office of Foods and Veterinary Medicine. Under this CRADA, Emulate and FDA are collaborating to evaluate and qualify the use of Emulate’s Organs-on-Chips technology as a platform for toxicology testing to meet regulatory evaluation criteria for products – including foods, dietary supplements and cosmetics. The FDA has an onsite installation of Emulate’s platform for experimental testing at FDA. The near-term goal of the collaboration is to evaluate and qualify the human-relevant testing capabilities of the Human Emulation System, including correlation with existing cross-species toxicology data on human health effects.

Geraldine A. Hamilton, PhD shared her perspectives with Laura Elizabeth Lansdowne, Senior Science Writer for Technology Networks.