Animal testing is the first preclinical step in the development of any drug; the safety and efficacy of a treatment must be evaluated in animal models before it can move on to trials in humans. Despite this, it seems that animals’ usefulness as models for human disease is limited; studies have found that over 90% of investigational drugs fail in clinical trials, despite yielding positive results in preclinical animal studies.
Researchers’ reliance on animal testing is fraught with the potential for error. Animal immune systems function very differently to humans, and what is safe and effective for an animal may not be for a human – and vice versa. A growing number of pharma and biotech groups are exploring the potential of alternative research methods – and CytoReason, which counts pharma giants Pfizer and GlaxoSmithKline among its partners, is one company leading the way.
CytoReason CEO David Harel speaks about the drawbacks of animal models, and the potential that data testing holds for the future of drug research and development.
Darcy Jimenez: Can you outline how CytoReason’s platform works, and what makes it unique?
David Harel: What CytoReason does is model the human body in different diseases; we’re building a simulator of how the body behaves under certain conditions, and it’s accurate enough to be used for drug development, among other things. When you have a simulator, you can test it and generate predictions of how the body would react to these drugs relative to existing treatments, and so forth.
Animal models are called animal models because they’re modelling certain behaviours of the human body – what CytoReason does, and our biggest strength and our uniqueness in the market, is that we are building computational models of humans using human data rather than animal data.
DJ: Animal testing is the first stage in developing a drug, but what are the disadvantages of this approach?
DH: Curing an animal and curing a human are not the same thing. I think the biggest trend that we see in pharmaceutical discovery, in this day and age, is the focus on trying to tackle the disease mechanisms and get a better understanding of the disease. The reality is that animals that show the same phenotypes are likely to have different mechanisms of disease than humans. So by building a model that is built on human data, we are replicating the disease mechanism that exists in humans, rather than curing animals – and that would basically suggest that the probability of success is going to get much higher.
Nine out of ten genes that change in an animal do not change in a human, and the other way around. So, the fact that you saw your drug working on an animal doesn’t mean necessarily that it’s going to work on a human. And worse than that is the other way around – the false negatives. You saw your drug not affecting an animal and killed the project, and gave up on a drug that could make very positive changes in the lives of millions.
DJ: How does CytoReason’s technology bypass the issues associated with animal models, while still ensuring accurate data on new drugs?
DH: If we focus on efficacy, we have a mechanistic model of the disease in the computer, and we generated that using data from humans that was aggregated throughout clinical trials. This model can simulate what the response would be to a hypothetical drug, relative to what you see with other drugs in the market, the standard of care, successful drugs, or failed drugs. And once you go into the clinic, the data from your clinical trials can be used to model the response in either subpopulations or in other indications to expand the market.
Safety is lagging a little bit behind efficacy with computational models for two reasons. One is that the regulators are more strict on safety than they are on efficacy, and the other is that the transition from animal models to computational models is something that is going to take a long time.
If you think about how drugs are going to be developed 100 years from today, there are two things that we probably all can agree on: there will be fewer animals and more technology. But it’s not going to be in in a big bang, it’s going to be gradual changes, with very small victories over a long period of time. If you want to predict the safety profile of a drug you need to know what it does in the heart muscles, in the brain, in the lymph nodes, in different ages and phases of different diseases – so you need to model everything in one shot.
DJ: What milestones has the CytoReason platform achieved so far?
DH: Pfizer is using our disease models in a wide variety of diseases for target identification. They are looking at our model, trying to target specific pathways and specific mechanisms in a wide range of immune-mediated diseases, and immune oncology. That’s the biggest pharma company in the world using the models right now, in addition to animal and traditional methods of target ID.
DJ: Aside from non-animal approaches being potentially more accurate, would you say pharma companies have a responsibility to seek out more ethical drug testing methods?
DH: When we are thinking about ethical development, obviously we all would like to minimise animal models. Firstly, because we care about animals, and secondly, because it’s not working as good as we want, or we need, as mankind.
At CytoReason, we are committed to minimising the use of lab animals, but we recognise the complexity and risk involved in drug development. Pharma companies need to help people as much as possible, while taking into consideration the wellbeing of animals.
DJ: What role have animal-free methods played in the development of Covid-19 treatments?
DH: When the Covid pandemic started, CytoReason developed a disease model of a disease called acute respiratory distress syndrome (ARDS); the severe inflammation in the lung that is caused by the virus, and is causing the cytokine storm that is killing most of the patients. We modelled this disease in the lung, and we gave all those models for free to all our customers.
Two drugs entered clinical trials using our models; the FDA accepted the data from our model, under emergency application, as substantial enough to start clinical trials in Covid treatment. And today we’re still involved in Covid therapy development, although the global attention is on the vaccines rather than the treatments. But very much like most diseases, prevention and treatment are two sides of one strategy that mankind needs to take in order to overcome these situations.
DJ: Are you optimistic about the future of data testing in drug research?
DH: I think the big change that we’re going to see in the future, with regard to R&D technologies, is the reliance on models. Basically, the model is software data; it’s a tool that allows you to generate predictions of what’s going to happen in reality. And those models, we started doing those four years ago, and they were very well accepted by the pharmaceutical industry. But my expectation is that the models will continue to improve and include more aspects of the human body as we continue. Drug development is going to have a lot more technology, and a lot fewer animal trials, and the probability of success will be lifted by getting rid of animal models.