Episode 14 - Mice or men: Beyond animal testing

Intro (00:00:10:00)

Science on the Menu, a podcast by the European Food Safety Authority. 

Ed (00:00:20:21)

Hello and welcome to another episode of Science on the Menu. My name is Ed Bray and I work in the Communications team at EFSA. And joining me today is Jean-Lou Dorne, who is a Senior Scientific Officer working on Scientific Methodologies in Risk Assessment. Welcome to the podcast, Jean-Lou. 

Jean-Lou (00:00:39:21)

Thank you very much Edward.

Ed (00:00:41:11)

Today we're going to be talking about animals, but not the animals that are on our farms that are producing food and drink. We're going to be talking about animals that we think about less in the food safety system, and that's the animals that are tested on to ensure that substances in the food system are safe. We're going to talk about some more recent scientific developments that could help us to move beyond animal testing altogether. So first, Jean-Lou, what is the purpose of animal testing?

Jean-Lou (00:01:15:21)

Historically, animal testing exists since the 1960s after the famous thalidomide incident discovered in Australia.  

Ed (00:01:26:06)

Can you tell me more about that?

Jean-Lou (00:01:28:15)

Thalidomide was taken as a drug, as an antidepressant or as a sleeping pill, and pregnant women took it. The consequence of thalidomide intake was basically children with no limbs. A medical doctor discovered that thalidomide itself was so-called teratogenic, so it basically blocked the development of limbs in the foetus because it went through the placental barrier.

From that, the legislation was set to then test, very briefly, all those chemicals for their potential to have adverse effects. For example, on the foetus, but also on any organ of the body: liver, lung, and so on, to be very brief.  

Ed (00:02:24:00)

So essentially, it became mandatory to test on animals before certain substances could come to market?

Jean-Lou (00:02:30:06)

It became mandatory. So, it is set in stone within the legislation, put it that way.  

Ed (00:02:35:15)

And let's be clear: we, the European Food Safety Authority, don't have labs at EFSA for animal testing. This is done elsewhere around the EU. Is that right? How does that work?  

Jean-Lou (00:02:49:20)

Depending on if the company has a regulated chemical, in the sense that industry submits the dossier to put this compound on the market, then we're talking about regulated chemicals such as food additives, feed additives, pesticides and so on. Or if the chemical is an environmental contaminant, then of course, because it is a contaminant, it doesn't belong to any industry. So, we would take for this contaminant the historical data we have, dioxins and so on, for example. When it is a compound that is about to reach the market, or an application for this compound to reach the market, let's take a fungicide or a food additive, the legislation requirements would ask for data on toxicity of that chemical on selected test species. We are talking about rat and mice, dogs, and rabbits.  

Ed (00:03:43:07)

Often in toxicology we talk about: the dose makes the poison, correct? So, I guess essentially, we are testing to see at what dose a substance becomes unsafe firstly for animals and then it's converted into something for humans. Is that more or less right?  

Jean-Lou (00:04:05:05) 

Yeah, absolutely. You have what we call dose response, which is Paracelsus (the dose makes the poison) from that dose-response we derive a no-effect level or, if we do modelling, we derive what we call a benchmark dose limit. From that we apply uncertainty factors to convert the dose from animal data to human health, in the sense that there is a so-called uncertainty factor of a hundredfold: tenfold to take into account the differences across species and tenfold to take into account the human variability from the baby to the elderly.

Ed (00:04:44:22) 

Because basically humans are obviously different from animals and therefore there needs to be a conversion factor essentially to convert it into something meaningful for humans. But also, a margin to ensure there's even greater safety. Is that correct? 

Jean-Lou (00:05:01:07)

Yes, absolutely.

Ed (00:05:02:01)

We know that in society there's a lot of concern, rightly so, about the ethical value of testing on animals. EFSA is doing what it can to move in a direction where we don't need animal testing and we adhere at EFSA to the principles of the three Rs. So that's to replace, reduce and refine. Can you talk about this? The replace, I imagine, is to move to something, a system where we are not testing on the animals, reducing - to reduce the number of experiments while getting the results we need, and refining - to refine the experiments themselves to reduce suffering. Is that right?  

Jean-Lou (00:05:45:09)  

Yeah. And on top of that, in the refinement with the new methodologies that we've been working on for the last four decades, we can integrate some understanding of the processes that lead to toxicity, the toxicokinetics, which is what the body does to the chemical, such as metabolism and excretion, and what the chemical does to the body, which is the toxicity itself.

Sometimes metabolism will bring something which is toxic or sometimes the metabolism will bring something which is not toxic anymore, and it's eliminated in the urine for example. So, using these approaches, we can combine historical data from in-vivo testing, which is animal testing. But also, we can look at cellular systems that allow you to look at a number of given organs that would be targeted by that chemical such as the liver, the kidney, lung.

And then combining that within algorithms which allow you to predict both dimensions: the fate, so the kinetics of the chemical, or the effect and dynamics.  

Ed (00:06:59:10)

That's where we are now. We're already moving in that direction. The replacing of the in-vivo testing on the animal to two other options. You talked about in-vitro in taking, essentially, cells of an organ and testing on those rather than the animal. But there are also new methods where we can do it with the data, the algorithms, modelling of the data. And I understand that you've been working on a very specific model called TK plate which is moving us in the direction of reducing animal testing even further. Is that right? 

Jean-Lou (00:07:39:07)

Very briefly, TK plate is an open access platform that allows you to model the kinetics of chemicals, what the body does to the chemical, but also the dynamics of the chemicals. And it's relevant to human health, animal health, but also to species of ecological relevance, so species that are in the environment. 

Ed (00:08:02:01)

So, for example, a pesticide that is tested on to see what effects they might have on birds or insects, or other species. 

You say it's open source, so anyone could use it, any scientists could come in, input the data and at the beginning choose the target organ and then the outcome, is that right?

Jean-Lou (00:08:26:00)

Of course, you can register and then you would have complete access to the platform. The first thing you would do is to select the model itself for a given species and here we're talking about having models for humans. All the test species - the rats, the mice, the dog, the rabbit and farm animals, pigs, cattle, sheep, and chickens for example and we are planning to put mammals inside that. You would choose your model, you would choose the chemical. There is a link to a database, but you can put your own chemical if it's something which we don't have much data on. From there you can select your exposure patterns. Let's put it in simple terms; the exposure to that chemical could be an acute exposure which is, for example, just one dose, or it could be a chronic exposure, which is a dose you would have every day. And from there we could, for example, simulate in-vivo studies without testing for 90 days which is the standard, or for several years. You can simulate the kinetic of the chemical and go to the dose response to then calculate your safe level without testing on any animal.

Ed (00:09:49:17)

Is this being used already or does something need to change for us to use it systematically in risk assessment? 

Jean-Lou (00:09:57:17)

When we published the database, we took quite some care to, first of all, produce a user guide which is available to users, but also concrete case studies to take the user through the different steps on how we can really apply it to contexts of relevance. Let's say a food additive, a pesticide and so on. You generate the data sets from the simulation. We've taken quite some care to take the user through every step all the way from inputting the data, choosing the model, all the way to downloading the data and generating what we have, which we call an automated report.

Ed (00:10:44:13) 

I imagine the more data we gather, the more can be put into the tool, the more established it becomes and therefore the more useful. 

Jean-Lou (00:10:55:13) 

Absolutely. And we gain experience and, of course, these models are so-called generic models because they can be applied to any chemical for a given species. However, these models evolve as we understand the biology of organisms because they get more refined.

Ed (00:11:14:21)

Which could also mean greater safety for us? We know more about the substances and we're able therefore to understand the effects better as well. 

Jean-Lou (00:11:26:00)

Absolutely. The consequence of metabolism can be that the compound is eliminated from our body, or it can be bio-activated, which means that the consequence of metabolism is that the compound becomes more toxic. 

If we understand the difference between these two processes, we can determine if it's what we call the chemical itself, the parent compound, or the metabolite that will exert its toxicity in a given target organ such as the liver. 

Ed (00:12:00:00)

And could you imagine this changing the whole process of risk assessment altogether in the future if, as you said, that there were legislative changes that happen, that we could actually have a world without any animal testing at all, could that be possible?

Jean-Lou (00:12:16:11)

Well, I think it's already possible to reduce it, I would say at least by a significant amount. You could do targeted testing by identifying again using cell systems, even in animal cell systems and human cell systems, you could identify the targets, be it the liver, the thyroid, the lung, the kidney, and you could test that within human cell systems and let's say on test species’ cell systems such as rats. Eventually if there is a need, a legislative need, to get an in-vivo study done then instead of having different studies, you could have a very targeted study and it could be just one. So that would be a reduction already of 90% of the testing. 

Ed (00:13:08:03)

Wow, it's potentially a very exciting development in how risk assessment is done and will be done in the future.

Ed (00:13:28:03) 

Well, I think that's all we have time for now. Thank you very much for joining us Jean-Lou and explaining about this exciting development in scientific methodology. Thanks to our listeners for joining us for another episode of Science on the Menu. We hope we'll see you again next time. Please subscribe to our social media channels where you can also find more material about our episodes.

But for now, it's goodbye from me, it’s goodbye from Jean-Lou.

Jean-Lou (00:13:55:00)

Thank you so much, Edward. 

Ed (00:13:57:00)

Hope to see you again soon on Science on the Menu, bye.