Why are animal experiments still needed?

In principle the best way to determine, if a chemical is safe, is to rely on studies on human beings. This approach has two problems. It is too late, and it is unethical to expose people to risk simply for experimental purposes.

If we have to base our assessment on epidemiological studies or clinical findings, it is too late. By definition, this means that people have been unknowingly exposed to some hazard for a long time, and then it is revealed that at a certain exposure level some adverse effects have occurred. This has been the case with some occupational exposures to chemicals. Exposure to 1,2-dibromo-3-chloropropane caused male workers to become sterile, an unexpected adverse effect, and going back in history, 19th century chimney sweeps were found to have a high incidence of skin cancer of scrotum. Clearly we wish to avoid repeating such mistakes of yesteryear. Moreover, detecting chemicals which cause adverse effects in people is not very easy, because people are exposed to many things simultaneously.

Second, controlled studies are rarely possible in humans for ethical reasons. It is not possible to expose people deliberately to an unknown chemical even in a well organized study designed to reveal adverse effects. It is not considered ethical to cause harm to twenty people even if it is done in the name of ensuring safety for millions. Thus, we can administer drugs to humans only after they have been first demonstrated to be reasonably safe at the intended dose.

Why animals then?

It is 200 years this year (2009) since Charles Darwin’s birth. Darwin more than anybody made it clear that all animals are very similar, some are closer relatives than others, but all animals share many of the same basic biological characteristics. Subsequently, biochemical and genetic studies have confirmed this to an extent probably not even imagined by Darwin. Nonetheless, there are also differences and these must be appreciated, but in skilful hands, animal data can predict very well the possible risks to be encountered also in humans. Animal experiments do not guarantee that a chemical will be completely safe, but they can reveal most of the adverse effects, and provide also fairly useful information about the dose range in which they are likely to occur.

The human body is a very complicated organism. There are regulatory mechanisms such as the nervous system and the endocrine system that can change bodily functions to such an extent that chemicals act differently. They may also be targets of toxicity. There are also very delicately controlled systems such as the blood circulation which can be influenced by scores, perhaps hundreds, of different factors in the body. It is not possible to mimic these complicated systems in any artificial structure.

Systems safety is not the same thing as component safety. This can be illustrated by looking at traffic safety. Traffic safety depends on the safety of cars and their road worthiness. But there are many other factors affecting safety – road junctions, traffic lights, speed controls, quality of roads and streets, active supervision by police, and the number of other drivers on the road in the traffic and their behaviour. It is not possible to claim that we have tested traffic safety by only examining cars and their possible failures. It requires studies into the whole system, and even this is not perfect, but clearly a better guarantee of safety. In this respect, it is clear that studies on a much more complicated system as the human organism could never be simplified to studies on its single components?

How about ethics of animal experiments?

The ethics of animal experiments are not a question of doing or not doing them, but how they are performed. They can be performed in unethical ways, but that possibility has to be eliminated by training, legislation and inspections as well as by guaranteeing that independent assessors view every experimental plan. Often researchers are in a great hurry to perform some experiment and it is good to stop for a while to think, what is needed and perhaps what is not needed. Sometimes there may be alternative ways of achieving the same aims, although not in every case. It is clear and it has been long accepted by scientists as well that there must be no cruelty in animal experiments. It would be reasonable that animal rights activists would admit that there has been a huge improvement in animal treatment in the past 200 years.

The real philosophical problem of using animals is much wider than using animals for experiments. Is it right to grow animals for food? Is it right to deprive pet animals of their freedom? What is the ethics of keeping a cat just for fun, is it even less ethical than growing it for food? Is it right to make horses run to exhaustion just for sport and to provide excitement for spectators? Is it right to hunt and make an animal to suffer for minutes or even days, if the bullet does not kill it outright? It is easy to predict that a relatively high percentage of animals are only wounded by bullets, and these animals will suffer much more than would ever be the case in an animal experiment. What about hook and line fishing?

These are difficult questions and there are no immediate or easy answers. My main point is, however, that animal experiments cannot and should not be seen separately from other similar activities utilising animals in our present societies. Most animals in animal experiments do not suffer any more and probably a great deal less, than animals used for other purposes in our society. So a more reasonable attitude would be to improve the conditions of animals and to control any unnecessary suffering of animals rather than try to limit all use, when similar use is allowed in other sectors of life.

A carcinogenicity study is done to prevent chemicals which could cause cancer from ever reaching the market. In a carcinogenicity study, perhaps 10 rats out of 50 controls will contract cancer during their two-year lifetime even when they are only given food and water. In the experimental group typically 15-20 animals will grow a cancer, if the studied substance is indeed a carcinogen. If it is not, also in this group about 10 rats will contract cancer. A similar increase in cancer risk is seen in pets that are overfed. Just as in humans, the cancer risk of pets increases if they eat too much and gain weight. Why then is overfeeding pets not considered as cruelty?

How reliable are cell studies (in vitro studies)?

The fundamental question between whole animal studies and cell studies is not reliability, but the fact that these studies answer totally different questions. In vitro studies are most useful for finding out mechanisms of toxicity, whole animal studies provide an assessment of overall toxicity. There are thousands of different functions in the body, and it would require thousands of different cell tests to screen all of these functions. Even then, it would not be clear how they interact and function together in the whole organism. These kinds of properties can be studied only in a complete organism.

Activists demand that animal studies must be “replaced” by in vitro studies. This is a clear misunderstanding of the issue. Measuring body weight change, which is one of the most sensitive measures of toxicity in repeated-dose studies, cannot be “replaced”, but it may be possible to understand many of the mechanisms leading to reduced growth by conducting in vitro studies. Therefore some in vivo measures might become unnecessary, if we had a more complete understanding of mechanisms. But it is silly to think of a one-to-one replacement of an in vivo study by in vitro study except in a very few relatively simple cases such as tests for skin or eye irritation.

How about computer modelling?

Computer models are nice, but have one peculiar property. You have to understand the effect very well indeed before you can construct a model. This means that you can only create a model for a purpose about which you already have very good understanding. So it is useful in cases where you know exactly what you want, and you have access to basic data to allow modelling. In practice, this means that it is possible to rapidly widen the scope of predictions to similar but broader use, but the basic information must exist, and usually this is information from animal experiments. Therefore like in vitro assays, computers widen the scope, but they do not “replace” anything or create anything new.

Administrative errors

Unfortunately administrators and politicians in many countries have little understanding of what animal experiments are all about. To put it simply, they are done to guarantee the safety of their people. This lack of understanding has been very clearly shown by recent decisions from the European Union and European Parliament. The REACH program that is very important for chemical safety in Europe, strives at the same time to decrease the number of animal experiments. These two aims simply are not compatible. One must choose either we will decrease the number of animal experiments or we will try to improve the safety of chemicals. If one wants the latter, then you have to accept the fact that this inevitably will significantly increase the number of animal experiments. As such, both aims are defensible, but it is plain dishonesty to claim that both can be achieved at the same time.

Another ill-advised piece of legislation is the prohibition of the use of animal experiments in developing cosmetics products. Administrators will sooner or later bitterly lament this decision when young people contract cancer after colouring their hair with chemical dyes. In addition it appears as if many people do not understand what is included in the list of cosmetic products: soaps, shampoos, toothpastes, skin lotions, sunlight protection ointments etc. It is truly a group of chemicals that everybody is using daily directly on their skin or in their mouth. This is certain to cause much more exposure than an average industrial chemical. Therefore cosmetics should be investigated better, not worse, than most other chemicals.

Many people do not realise what they want when they seek to ban animal experiments. In simple terms, it would mean exposing people to potentially risky chemicals without trying to gain as much useful information as possible beforehand.

Notes and references

One level up: Here a risk, there a risk, everywhere risks, risks!

Previous chapter: We breathe oxygen; it can’t be dangerous, can it?

Next chapter: Are electric lines and transformers dangerous?