Animal rights groups have long propagated the myth that animal research could be replaced tomorrow by a plethora of “alternatives,” from computer modeling and micro-dosing to MRI scanning and in vitro testing.
So, why the quotation marks on “alternatives”? The fact is that such methods currently tend to complement, rather than replace, the use of animals in research. For instance, computer modeling is a useful tool in the development of a new drug; however animal research will often be crucial in the creation of the computer program itself. Before looking at some of the reasons that these “alternatives” are not yet sophisticated enough to replace animal research it is important to consider the following two facts:
- 1. Animal research is very expensive. Animals must be housed, fed and cared for by trained animal welfare technicians and veterinarians. By comparison, replacement methods tend to be much cheaper and thus academic researchers (who compete for limited available funds) and pharmaceutical companies (who are profit-seeking) will prefer to use these cheaper replacements wherever possible. The fact that animal research continues is testament to the fact that for some research there are no alternatives to the use of animals.
- 2. Replacements to animal research must be used, by law, wherever possible. The Animal Welfare Regulations [2.31] explain the role of Institutional Animal Care and Use Committee (IACUC – every research institute must have one) in ensuring that researchers have considered (and will use) any replacement method available:“The principal investigator [researcher] has considered alternatives to procedures … and has provided a written narrative description … to determine that alternatives were not available” The fact that animal research continues is once again testament to the fact that there are no alternatives to animal research.
“Computers can do amazing things. But even the most powerful computers can’t replace animal experiments in medical research.” – Professor Stephen Hawking, quoted by Seriously Ill for Medical Research in 1996
Computer modeling plays an important part in the research process however its capacity to replace the use of animals is limited. Before one can program a computer model to reflect an aspect of our physiology, an understanding of the physiology being modeled is needed. This knowledge tends to come through research using animals. So animals are needed before we even get to the computer.
Computers are also limited by their processing power. A recent simulation of just half a mouse’s brain required the use of the world’s fastest supercomputer – Blue Gene/L (300% faster than the second fastest super computer). However the simulation was far from a perfect representation:
“The vast complexity of the simulation meant that it was only run for 10 seconds at a speed ten times slower than real life – the equivalent of one second in a real mouse brain.
The researchers say that although the simulation shared some similarities with a mouse’s mental make-up in terms of nerves and connections it lacked the structures seen in real mice brains.” (link)
Most scientists do not have access to supercomputers on the scale of Blue Gene/L, which are needed to attempt more complex simulations.
Computer simulations of organs have some use, but, unlike in vivo research, they are generally forced to focus on major interactions at the cost of minor ones. A simulation of a heart may appear to reproduce the movement of muscles used in pumping blood, but will likely be at the cost of minor reactions and interactions going on within an individual cell.
Professor Dennis Noble, who was part of the team, at Oxford University, designing a virtual heart said:
Because hundreds of millions of differential equations are simultaneously being solved, it may take 30 hours just to do a few beats of the heart.
But denied that his research could completely replace animal research
“I would say the real benefit of the model is that it can do a preliminary filter of your compounds, and that can replace some of the very early stages in animal experimentation”.
Read more on the limits of computer simulations in medical research.
Micro-dosing is an exciting new technique for measuring how very small doses of potential new medicines move around the body. It should be possible to use micro-dosing in humans to reduce the numbers of animals needed to study the effectiveness of new compounds.
However, micro-dosing has limitations. By its very nature, it cannot predict toxicity or side effects that occur at higher ‘therapeutic’ doses. It is an unrealistic hope, and a false claim, that micro-dosing can replace the use of animals in scientific research wholesale. This was confirmed recently by the respected organization FRAME (Fund for the Replacement of Animals in Medical Experiments), which stated in this context: “animal studies will still be required“. *
In vitro testing means testing things in a test tube, cell culture flask, petri dish or multi-well plate, and includes a vast range of cell and tissue culture techniques, as well as cell-free methods. Dr Phil Stephens has pioneered an in vitro test for ulcer treatments based on genetic manipulation. He says:
“There are a number of different animal models out there, but they are not really good models for these wounds. So, we began developing an in vitro system.”
This is a great example of replacement. Scientists always want a better model for their experiments so as to get better (more accurate) results. If a non-animal method can work better than an animal method, great! Not only does it yield better results, it’s a lot cheaper, too. However, Dr Stephens also notes:
“The in vitro system is not going to replace the animal models, but it will enable a vast number of pre-screens to be undertaken, hopefully vastly reducing the number of animal experiments that go on.”
Again, although the aim is to refine the models and reduce the number of animal experiments, Stephens notes that in vitro testing cannot replace animal testing altogether. The reasons for this are fairly straightforward: a drug might work fine on a cell in a test tube, but how will it work in a body? A test tube has no blood circulatory system, no liver, no brain, and no nervous system at all. A test tube cannot feel pain or get pregnant. We just don’t know whether it would work for sure until we try it on a living creature. And again, it’s either animals, or us, that we have to trial the drugs on next.
Professor Chris Higgins uses MRI imaging to reduce the number of animal experiments needed:
“One area we are looking at is what controls appetite and satiety. To do this in the traditional way, we would have to dissect the animal brain, but to avoid this we use in vivo imaging to look at the areas of the brain related to hunger and satiety.” (link)
Rapid advances in technology have allowed us to get to the stage where scientists can use scanning to see how certain parts of the brain “light up” under certain conditions, giving us clues about what parts of the brain control different aspects of our bodies, thoughts, cravings, and so on, and clues about how the brain works. However, Prof Higgins goes on:
“The one thing that is difficult to do is to understand the genetic and the underlying molecular basis of obesity, and for this we need to use animals, mainly mice, if we are going to develop more effective therapies.”
So again, although this ‘alternative’ can fulfill a useful role and help reduce the number of animals used, it cannot replace animal research altogether. Watching how the brain works can help us understand part of the problem, but it also occurs on the genetic and molecular level, which MRI scans cannot show us.
MRI scans may show us a problem in the brain, but animal research is likely needed to fix the problem. We cannot alter a human brain between MRI scans in an attempt to find a cure, so we must use animals first, to ensure the methods safety.
Read more about the limits of MRI, and specifically fMRI.
•* ATLA, (2005) Human microdosing reduces the number of animals required for pre-clinical pharmaceutical research, 33, 439.