How tech could spell the end of animals in drugs testing


guinea pig. Image shot 06/2012. Exact date unknown.

We are in a pharmaceutical ice age where fewer drugs are being developed. But new computer modelling and bioengineering techniques could revitalise the industry – and make animal testing obsolete.

Read my latest piece for The Observer/ The Guardian in full below or by following this link. It was a double page colour feature in The Observer’s New Review and a featured story on The Guardian’s home page all day. It was shared 5900 times on Facebook.

Think of a world where more new medicines are available more quickly, more safely and with few, if any, animals used in research. Think of a world where you know for sure that the drug you are prescribed is safe for you. This is what researchers at some of the world’s leading universities are trying to bring about.

Academics from disciplines as far apart as computer science and biologyare growing mini-human hearts in a lab or creating a model of the human biological system on a common computer chip. Their aim is to bring to an end to what has been called the “pharmaceutical ice age”, an era of ever more research but fewer medicines reaching the shelves of Boots.

According to the Association of the British Pharmaceutical Industry, it can take up to 12 years and £1.15bn before a new medicine reaches the high street. For that one medicine, 25,000 chemicals will have been tested, 25 potential drugs will have had human clinical trials, perhaps five will have been passed by the regulators and only one will have been marketed by a drugs company. Even after that, a good number of drugs have to be withdrawn owing to previously undetected side-effects.

For some, the fault for this freeze is due to the pharmaceutical industry’s obsession with mergers and acquisitions or a culture that inhibits innovation. For others, it is a sign that perhaps there just aren’t many chemical compounds that can interact with the body and have a positive effect; or even that the tools scientists use during discovery and early-stage testing mean the wrong drugs go to human trials.

While animal testing is a regulatory requirement for any new drugs, it remains a fairly unpredictable guide to how a drug will react in humans. Testing drugs on cells in a petri dish allows much higher dosages than with an animal, but is an unreliable guide as to how they might interact with the human body. Virtual computer models of the likely impact of a drug have also, up to now, largely been based on the reactions of a single cell.

Now scientists are trying to use new technology to make drug testing faster, cheaper – and more accurate – by modelling more closely how a chemical interacts with the human body. They hope to do this to the point where a drug could be declared safe for a certain section of the population or even personally safe for you.

“Over the past 15–20 years, there has been a dramatic increase in results from biomedical research,” says Richard Seabrook, head of business development at the Wellcome Trust. “However, this hasn’t led to an equally dramatic increase in new medicines approved.It has been pretty static in Europe and the US. This might be because it is a relatively rare event for a small chemical to interact with the human body and have a positive effect.”

Dr Kathryn Chapman, head of innovation and translation at NC3Rs, the National Centre for the Replacement, Refinement and Reduction of Animals in Research, says: “There has been a pharmaceutical ice age over the last 20–25 years. So people have been trying to find new ways of doing things, as the old ways don’t work, and the money is there for this. The issue with all these technologies is how to add on the complexity.”

“The problem has been that the discovery of new drugs hasn’t been given as much like-by-like value by analysts as the development of drugs that can shift the share price,” says Professor S John Lyon of Warwick Business School. “The development of new drugs also has to follow strict regulations laid down by bodies like the FDA.”

The tag line for the Wyss Institute at Harvard is “biologically inspired engineering”. Professor Donald Ingber, the institute’s founding director, is fulfilling this aim by being one of the pioneers of the concept of an “organ on a chip”, which involves placing the human cells and fluids necessary to mimic the complexity of human physiology on a single, tiny piece of see-through, flexible silicone rubber the size of a USB stick. In his new startup, Emulate, these chips will be mass-produced for the first time.

The chips have fluid flowing through them, so you can connect them to each other, creating a “human on a chip” where lungs, livers, intestines, skin, kidneys and eyes can be integrated to simulate how a whole body would interact with a new drug.

“You could write a 2,000-word article just on this pharmaceutical ice age alone,” laughs TED speaker Dr Geraldine A Hamilton who works with Ingber to develop the technology and commercialise it, and who move to become president and CSO of Emulate in the coming months.

“The structures you find yourself working in can stifle innovation. Every time companies merge, productivity goes down. Then there has been a problem with the tools they use during testing. Often, cells are put in a simple petri dish, but they live in a complex dynamic organic environment with each other. Animals can fail to predict in humans how drugs will work. Different animal species may give you different answers as to whether a drug is toxic for humans. Which one do you trust? Many good drugs are lost at that point.

“So we recreate the conditions the cells find themselves in by using micro-engineering to provide them with all the cellular integration, fluids and even mechanical forces that they are used to, such as breathing in and out. We can also connect them in our human bioemulation platform to model and better understand diseases, and to study how humans respond to drugs.”

Until now, the complexity of chips meant they had to be made by hand, limiting their numbers. It also introduced variability, whereas they need to be reproducible and robust. So, in partnership with Sony, the team has worked on ways of scaling up the manufacture. This is the key to translating this technology to industry, Hamilton believes.

Illustration by Dale Edwin Murray.Illustration by Dale Edwin Murray.

Meanwhile, in the labs of Abertay University, Dundee, Professor Nikolai Zhelev has mass produced thousands of miniature hearts 1mm across that can beat 30 times a minute. “This is the first time it’s been possible to induce diseases in them, so it’s an exciting area of research,” says Zhelev. “After they start beating, we use chemicals to make the cells become hypertrophic – enlarged – and then we test different compounds and drugs on them to see if we can prevent them from becoming hypertrophic or even everse the damage.’

Biosensors or markers enable his team to track individual molecules to see which cause the hearts to become enlarged so they can target new drugs just at these molecules. The implications of this are important. For example, they are developing a cancer drug that is already in human trials and initial tests suggest it could work on full-size hearts as well.

“Mini hearts are much better than using animals to test drugs – they are closer to the real thing and they should speed up the time it takes to move towards clinical trials,” says Zhelev.

“Our mini hearts are a real biological system providing relevant answers to drug screening efforts. Virtual drug testing faces enormous challenges in such complex systems. However, this is an area which is advancing rapidly with the development of mathematical and computational modelling approaches.”

Far away from the petri dishes and lab coats, but just around the corner from Oxford University’s animal testing laboratory and the protesters outside it, researchers in the Department of Computer Science believe that they have made a breakthrough in the use of virtual computer models to simulate heart cells and eventually skin, brain cells, and whole organs, for the faster and more accurate testing of the toxicity of drugs at an early stage. Their research to develop the Virtual Assay software has been funded by the Engineering and Physical Science Council.

Rather than moving their research into the private sector, they have decided to stay within the university and will release their software free for academics in September and in due course it may be licensed as an app-style product to the pharmaceutical industry.

“It is exciting to feel that you are exploring a genuine frontier,” says PhD student Oliver Britton, one of five finalists in the Transforming Society category of the 2014 UK ICT Pioneers competition and on whose research this project is based.

Virtual computer models comprise a large collection of equations that describe how components of a cell interact with one another and with physical quantities such as the voltage of the cell’s membrane or concentrations of chemicals such as calcium and potassium.

“We solve these equations using computers and doing this allows us to simulate how a heart cell might behave in different conditions,” says Britton.

“A lot of scepticism about the accuracy and usefulness of virtual models comes from a misunderstanding of how they should be used,” saysBlanca Rodriguez, Wellcome Trust senior research fellow in basic biomedical sciences. “If you ask a model to do something and it can’t then you will be disappointed.” In the past, she explains, a model might use one cell so when you did the computational experiment you got an output that corresponded to just one cell. “If you did an actual experiment, you would never trust the outcome of just one cell because there is a high degree of variability.”

“So we came up with this idea of using a population of varied models,” says Britton. “This was the key step, as we can achieve more accurate results by first generating lots of different models of a human heart cell, then simulating how they would behave in control conditions and then discarding those that don’t fit the range of experimental data we have available.”

“The challenge is to understand the limitations of the model such as in terms of what it can predict and then start to refine it,” says Rodriguez. “The success of it will depend on how easy it is to use.”

Seabrook warns that it is “a long way off” before any of these technologies are accepted instead of animals. “There needs to a lot of very good evidence that the result of using these technologies is predictive of what is happening in the human body.”

The problem, he says, is that predictions about the impact of a new drug are built on existing evidence. “So you don’t know what you don’t know, and you don’t know how a chemical will behave in a body until you do it.”

“If you think about it, a pill is swallowed to get into the intestine, absorbed through the gut and then has to get to where the disease is, whether in the brain, eye or lungs. So during this process chemicals are exposed to proteins, enzymes and cells so it is hard for any technology to predict reliably how it will interact with human body.”

He accepts these technologies mean we should get there quicker and so the more technology is available the quicker the process will be, and the easier it will be, to put your effort into promising candidates for new medicines. “If you are going to fail, fail early,” he says.

The drive to melt the ice age has also encouraged the development of new approaches such as those that use DNA and human proteins. Kathryn Chapman agrees that the technology is as yet unproven and years of data will be needed before you can use it to really start to reduce the number of experiments on animals.

“There is a powerful cultural mindset about how you do drug development. It is the courtroom argument. Would you be happy standing up in a court of law to explain why you hadn’t tested this drug on animals?

“Still, if the process changes and allows the technology to be taken up then the outcome will hopefully be better drugs and faster.”

In June, the US Food and Drug Administration (FDA) proposed changes to how molecular entities are assessed that could open the door to new tools such as VirtualAssay. Ultimately, new techniques may allow the personalisation of medicine, so “you may even be able to use your own stem cells to test side-effects on you. That will be a massive shift,” says Seabrook.

For Bina Rawal, research director at the Association of the British Pharmaceutical Industry, this is all part of a “tremendous acceleration” in new approaches to testing. This is indicative of “a cultural shift within pharma too,” she says. “The model has been adapted to be all about collaborations, working together and open innovation. This industry is traditionally very competitive, but now in the pre-commercial stage there is more sharing to stop everyone going up a blind alley.”

Perhaps the ice age is ending.



Leave a Reply