Alternatives to Animal Testing: A Tech Run-Down

One simple statistic points to the failure of reliance on animal models to predict the human response to novel pharmaceuticals: 95% of drugs fail in human trials despite some measure of efficacy and safety in animal studies.

It may be unsurprising, therefore, that new alternative methods (NAMs) for early research are being touted as more accurate in early-stage drug development.

“There are many new papers coming out showing how these non-animal methods consistently predict human response better than animal models,” Emily Trunnell, director of science advancement and outreach, laboratory investigations department at People for the Ethical Treatment of Animals (PETA), told In Vivo. “There are definitely areas where these methods can go ahead and replace animal testing.”

How Many Animals Are Used?

A clear picture of the magnitude of animal testing across the industry – much of which is arguably unnecessary – is obscured by the fact that the number of mice, rats, and birds bred for research do not need to be disclosed in the US. There is also a general lack of transparency when it comes to animal testing.

However, a 2021 paper in Scientific Reports sheds some light on the scale of animal testing. Based on information from 16 public and private life sciences research institutions (not biopharma companies) awarded National Institutes of Health grants, authors estimated that 111.5 million mice and rats (which consist of over 95% of animals tested) are used every year for research and testing in the US.

There have been several recent votes of confidence in favor of NAMs. On the regulatory side, President Joe Biden signed the Food and Drug Administration Modernization Act 2.0, lifting the requirement for animal testing and encouraging adoption of in vitro models (cell cultures and organoids) and in silico models. In the EU, regulators have also shown interest in shifting some aspects of testing to non-animal-based alternatives.

And the US Food and Drug Administration recently published a report examining approaches to drive integration of NAMs into regulatory decision-making.

For its part, some big pharmas are ploughing ahead with expanded use of NAM. Roche Holding AG, for example, hired organoid technology pioneer Hans Clevers to head its early research department. He is minting the Institute of Human Biology, which is expected to house 250 scientists and bioengineers dedicated to furthering organoid research and developing other advanced human model systems to improve the prediction of drug efficacy and safety.

Other big pharma companies have also stated their commitment to the “3 Rs” of animal testing: Replacing animal testing where possible, Reducing the number of animals used, and Refining techniques to minimize the suffering and stress of animals during testing. In the case of Pfizer, the company has adopted a 5R policy, adding Respect and Repay to the first 3 Rs, committing to further investments in NAMs to reduce animal needs.

Many Animal Models Ineffective

Despite government and industry taking promising steps to reduce use of animal testing, according to PETA, animal models are still being used for Alzheimer’s disease, sepsis, stroke, cancer, and neuropsychiatric diseases – areas where there is an abundance of literature showing they fail to mimic these diseases in humans.

“Many of the animal models aren’t giving us data that are meaningful, or worse, they are potentially very misleading,” emphasized Trunnell.

So, what technologies might be better suited to early research? The following is a selection of NAMs and companies providing these technologies for the biopharma industry.

Organoids

Pioneered in part by Roche’s Hans Clevers, organoids are 3D tissue constructs grown in vitro from stem cells or progenitor cells. The tissues replicate and self-organize into the architecture and functionality of the original organ. Organoids can differentiate into multiple cell types specific to a particular organ and can also be tailored to an individual, using patient-derived cells.

“Using patient-derived cells in many instances can help us answer questions that animal models are historically used for, because we have all the data from the patient, their history, and behavioral testing,” Trunnell said.

There are several vendors providing organoids for biopharma research. Dutch company MIMETAS employs organoid technology developed by Clevers. It is characterized by use of adult stem cells, as opposed to induced pluripotent stem cells (IPSC), explained CEO Paul Vulto.

“During our initial stages, we tried both the adult stem cell and IPSC approaches, and decided to go with adult stem cells because they maintain the epigenetic characteristics of the donor,” Vulto told In Vivo.

The company offers colon, kidney and liver organoids that include vasculature integrated into a proprietary microfluidic platform – i.e., organ-on-a-chip technology – enabling users to layer cell types and perfuse with growth medium as well as immune cells.

MIMETAS’s OrganoPlate platform can support up to 96 organoids (as well as other 3D cell structures, such as spheroids) on a single plate, with microfluidic networks running throughout to mimic blood flow and a membrane-free environment to enable free migration and self-organization of cells and tissues.

Swiss biotech InSphero offers TrueCardium, a cardiac organoid that closely mimics the human heart’s cellular architecture and functionality and includes cardiomyocytes, fibroblasts, endothelial cells and neurons.

Another company in the organoid space, Organovo Holdings, Inc., has developed a 3D bioprinting process and built a small therapeutic pipeline based on its use of organoid technology in early research.

In a recent In Vivo podcast, Organovo’s CEO and co-founder Keith Murphy explained how the company has employed its organoid-based disease models for Crohn’s disease and ulcerative colitis.

“We take cells from individual patients that have disease, we build a tissue – in this case an intestinal model – in 3D that has a nice thickness to it and actually demonstrates the disease, because we can keep the cells interacting as they do in the body,” said Murphy.

“We think we’re creating the gold standard for disease, because animal models are not only a species gap, but then you’re inducing disease somehow,” and it is usually dissimilar from the native disease, as evidenced by the high failure rate of drug development, said Murphy.

The organoid technology was key to Organovo’s development of FXR314, its lead asset for inflammatory bowel disease, which has completed Phase IIa study, and the company is developing additional therapeutics for other disease states using organoid technology.

Organs-On-A-Chip

Organs-on-a-chip are microfluidic devices engineered to mimic the functional units of an organ. They are made using microengineering techniques and seeded with cells from the desired tissue layer. They often incorporate mechanical actuators.

Dutch startup chiron has developed several organs-on-a-chip that can be used to monitor the effects of mechanical or chemical stimuli. The company’s product line focuses on arthritis and cancer, and their catalogue includes cartilage and synovial membrane-on a chip. However, co-founder and director Dustin Dopsa told In Vivo that chiron is moving beyond these areas and focusing on making their platforms “more user friendly” as well.

California-based Frontier Bio has a suite of on-chip offerings. In autumn 2024, the company successfully 3D bioprinted lung microtissue, which combines the organoid format with blood vessels. The company can create custom projects to replace animal studies, model disease states and generate personalized models.

One of the most mature companies in the organ-on-a-chip space is Emulate. The company was founded by scientists at Harvard’s Wyss Institute for Biologically Inspired Engineering, where on-chip technology was pioneered. Among the models the company has created are brain, colon, duodenum, kidney, liver and lung-on-a-chip.

The research applications of organs-on-a-chip vary. The colon-on-a-chip, for example, can be used to investigate mechanisms of colon inflammation and intestinal barrier disruption and to evaluate the efficacy of anti-inflammatory drug candidates, while the kidney and liver-on-a-chip can be used to study drug toxicity.

Notably, Emulate has studied its models extensively. For example, in a 2018 publication in Clinical Pharmacology and Therapeutics, Emulate and Janssen Pharmaceuticals validated the company’s blood vessel–on-a-chip, showing it could predict thrombosis caused by an anti-CD154 therapeutic antibody that Janssen had stopped developing in late-stage trials. Nearly two decades prior, the drug proved safe in animals but later caused thromboses in clinical trials in people with autoimmune disease.

In a separate journal publication studying 27 approved drugs, some of which were known to be hepatotoxic in humans despite proving safe in animals, Emulate showed its liver-on-a-chip was 87% sensitive and 100% specific in predicting drug-induced liver injury (DILI).

Emulate’s liver-on-a-chip is also the first organ-on-a-chip to be accepted into the FDA’s Innovative Science and Technology Approaches for New Drugs (ISTAND) pilot program for innovative drug development tools (DDT), paving the way for it to potentially become acceptable as a testing methodology in new drug applications.

Other companies, like TissUse, also provide on-chip technology, in its case offering a platform with multiple interconnected organ models, allowing researchers to mimic biological systems.

Spheroids

In addition to providing organoid technology, InSphero’s main focus is on creating spheroids. These are spherical aggregates of human organ microtissues. While not as architecturally sophisticated as organoids, the 3-dimensional structure of spheroids is superior to 2-dimensional cell cultures in terms of providing a more physiologically relevant model and promoting cell-cell and cell-matrix interactions. Like organoids, spheroids are composed of stem cells.

InSphero studied its 3D InSight Human Liver Microtissues, which employ specialized 384-well microplates with a spheroid of multi-donor primary human hepatyocytes cultured in each well, representing various ages, ethnicities and sexes. The study was done with the US FDA and evaluated 152 FDA-approved compounds that had been approved after conventional preclinical animal testing. Some of these drugs were later withdrawn because they caused liver injury. The technology correctly identified DILI in 80% of the withdrawn compounds.

“This represents a promising step toward reducing attrition and improving safety testing,” InSphero CEO and co-founder Jan Lichtenberg told In Vivo.

Beyond liver toxicology, InSphero is developing spheroid models of primary tumors “that provide insights into cancer progression and treatment to creating physiologically relevant metabolic disease models that bridge the gap between preclinical research and clinical translation,” Lichtenberg added.

AI Simulation

Drug discovery and development company VeriSIM Life’s platform, BIOiSIM, is a computational tool that can help companies identify which animal tests will most accurately reflect the human reaction to a given medicine. Additionally, the platform’s Translational Index uses chemical and biological modeling along with AI and machine learning (ML) techniques to account for multiple elements of a research program (clinical toxicity, preclinical toxicity, hepatoxicity and cardiotoxicity) and reduce reliance on animal testing.

As the company’s CEO and founder Jo Varshney told sister publication Pink Sheet, VeriSIM Life has used BIOiSIM for its own drug development.

“We identified a molecule and now in less than two years, we’re putting together a whole IND package for the drug,” Varshney said.

Another company in this space is Boston-based startup Quris-AI. Quris’s BioAI platform combines machine learning with organ-on-a-chip technology, linking multiple miniaturized organs-on-a-chip – including liver, heart, blood-brain barrier and brain – to test drugs. The organs-on-a-chip can be derived from up to nine genomically diverse stem cell lines, presenting the option of doing what they call clinical trials on a chip that represent diverse populations.

The drugs are tested in a high throughput fashion and nanosensors monitor the miniaturized organs for metabolites and other biomarkers to detect toxicity. Multiple drugs can also be administered at the same time, to test for drug-drug interactions.

The company has partnered with Merck KGaA to predict liver toxicity risk for a selection of drug candidates, hoping to integrate Quris’s technology to “work towards an AI-enabled IND process that reduces the reliance on animal testing.”

Data Analytics

Elsevier’s PharmaPendium combines US and EU regulatory data with analytics tools to help predict drug success or failure. Olivier Barberan, director of translational medicine solutions at Elsevier, said the tool can serve as “one piece of the puzzle” when developing a drug.

“The tool provides researchers a way to analyze past studies with similar drug indications to choose the right animal model the first time around, rather than working through a list of animals,” he told In Vivo. “If you’re trying to get a best-in-class drug on the market where a first-in-class already exists, by looking at what people have done in the past, you can see if they have used a mouse, rat or dog model, and go directly to the species that proved most sensitive.”

PharmaPendium’s Safety Margin Tool calculates the risk of off-target adverse drug reactions based on past in vitro results and supports the total replacement of animal testing at certain stages of secondary pharmacology.

“PharmaPendium can provide past precedent insights about drugs, clinical trials, and animal trials, which you can leverage in discussions with regulators to potentially reduce the need for an animal study or reduce the length of the study,” Barberan said.