Scientific Skepticism and Animal Models of Human Disease

(2800 words)

Ray Greek MD Nathan Nobis, Ph.D. Niall Shanks, Ph.D. Jean Swingle-Greek DVM

Ray Greek, MD, is president of Americans for Medical Advancement (CureDisease.com). Nathan Nobis, Ph.D., is an Assistant Professor of Philosophy at Morehouse College. Niall Shanks, Ph.D., is the Curtis Gridley Distinguished Professor of History and Philosophy of Science at Wichita State University. Jean Swingle Greek, DVM, is a member of the Americans for Medical Advancement Scientific Advisory Board.

Are animal models predictive and reliable in modern-day medical research and drug development? Certainly, many discoveries in the past three centuries were made using animals. But times have changed and modern science, especially evolutionary theory, now provides reason to be skeptical of claims made by representatives of the animal model industry. For example, Frankie Trull of the Foundation for Biomedical

Research has stated, “Every major medical advance of this century has depended on animal research,” and Dr. Wise Young, neuroscientist from Rutgers University has said, “There's never been a (medical) therapy developed without animals.”

While claims like these are often made by scientists and scientific organizations, we will argue that there is not strong scientific evidence to support them. We are beyond the level of superficial similarities and are now studying disease at the molecular level. And it is at the molecular level that a rat becomes a rat and a human, a human. It is because of these differences, and scientific advances based on their recognition, that we have outstripped a mainstay of biologic research: animal models of human disease and drug response are ready to go the way of the sextant. We argue that: 1. the results from experiments on animals are not reliably predictive of what

will occur in humans;

2. the extrapolation of results from animal models harms human patients

indirectly by misleading scientists and consequently delaying life-saving

discoveries, and by wasting time, money, and personnel particularly in today’s

high-tech world. Further, using animal models directly harms patients by

allowing harmful treatments to be tried in humans because they were

considered safe and effective in some animals;

3. although discoveries have occurred using animal models, most of these past

discoveries could have occurred without animals. However, we acknowledge

that some could not. Be this as it may, the past is immaterial as we are

addressing a situation very different from the past. We are now in the 21st

century and the differences between humans and animals far outweigh the

similarities we were studying in the 19th century; and,

4. instead of using animals as models of humans, newer and better technology-

and human-based research should be better utilized and funded.

Safety In Numbers?

The scientific method does not allow any assumption to go unchallenged. It requires internal consistency and relies on hypothetico-deductive reasoning. However, when the hypothesis is that “animals and humans have so much in common that we can extrapolate the results from animals to humans”, will the results be disastrous or advantageous? Empirical and theoretical reasons show the former to be more likely.

Evolutionary biology lies at the heart of our arguments for the claim that animal models of human disease are scientifically untenable. One way of speaking about the results of evolutionary biology is to categorize life forms into groups known as species. Homo sapiens will have characteristics that are unique to it but it will also have characteristics that it shares with other species, like Drosophila melanogaster or

Pan troglodytes. With the advent of molecular biology, we have learned that what each member of the species in question will have in common with the others is, in part, a collection of genes. However, while different species may have many of the same genes, the way the genes are regulated will be different.

Speciation, the evolution of different species from a common ancestor, is both the reason why it appears that we can use animal models as well as the reason why we cannot. Because all mammals are derived from a common ancestor, it is not surprising that we all share certain characteristics; neither is it surprising that each species is unique. Although animal models may appear feasible when first viewed, closer examination of the differences between animals and humans reveals the shortcomings in this assumption. The question modern-day researchers must ask is,

‘do the similarities outweigh the differences?’ ‘Can we extrapolate the results of an experiment on one species to a different species?’ There is some evidence that we can.

For instance, since all mammals have hearts, lungs and immune systems, and we all share the same cell types and tissues, we might expect that results will reliably extrapolate between species.

But there is also evidence to the contrary. For example, in chimpanzees, HIV reproduces slowly while in humans it does so rapidly. Because of differences in enzymes and metabolism a drug that causes cancer in male rats, such as saccharine, may be harmless to humans. Drugs such as Practolol, Opren, Fialuridine, Clioquinol,

Zelmid, Selfotel and others came to market, in part, because they tested safe in some animal species. They went on to prove dangerous in humans. It is still difficult to induce lung cancer in animals from cigarette smoke. Animals that were fed a high fat, high cholesterol diet failed to develop coronary artery disease, and so this diet was thought safe for humans. Asbestos, benzine, glass fibers and other environmental poisons were all proved ‘safe’ on animals then kept on the market long after epidemiological data proved them carcinogenic or dangerous (Greek and Greek, 2000,

2002, 2004).

Artificial heart valves, cyclosporin, beta-blockers, digitalis, the statins and other medications and treatments were kept off the market because animal models raised concerns that did not manifest in humans. Isoniazid and phenobarbital cause cancer in animals. Almost all currently used medications cause birth defects in some animal species (Greek and Greek 2000, 2002, 2004). In a study conducted at the

National Cancer Institute, most drugs tested that were known to be effective against human cancers were ineffective against the same human cancers that had been implanted in mice (Gura, 1997). Clearly we need to look beyond the empirical data to understand why different species diverge so greatly in their responses.

Analogies or Disanalogies?

One reason for the difference between species, vis-à-vis the spatial organization of the cells, lies within the genes. Genes can be divided into structural and regulatory genes.

The structural genes are responsible for the similarities. They are responsible for building the proteins of which the body is made. The regulatory genes turn the structural genes on and off thus affecting the development of the embryo and the organism as well as the physiology of the organism. They account for differences between species.

Hugh LaFollette and Niall Shanks (1996) argue that understanding the role of regulatory genes in evolution is “crucial to a proper understanding of biological phenomena. First, they focus our attention not merely on structural similarities and differences between organisms but also on the similarities and differences in regulatory mechanisms. Second, they illustrate an important fact about complex, evolved animal systems: very small differences between them can be of enormous biological significance. Profound differences between species need not indicate any large quantitative genetic differences between them. Instead, even very small differences, allowed to propagate in developmental time, can have dramatic morphological and physiological consequences.” (Emphasis added) (LaFollette and

Shanks, 1996)

A more concise way of explaining this would be to say that biological organisms are examples of a non-linear complex system and that explains why small differences between biological systems negate extrapolation. There are biochemical reasons for questioning the extrapolation of the results of experiments on animals to humans. Evolutionary biology supports and explains these reasons. Small differences between species lead to huge differences at the cellular level, which is where we focus when treating disease. This is the crux of our argument; that small variations on the molecular level not only define a species but also negate the ability of one species to

‘model’ another.

A theory or, in this case, a model, is reliable or scientific if it has predictive value. Researchers maintain that animals are causal analogical models (CAMs) and can be used to study human disease. Animal modelers such as Marilyn Carroll and

Bruce Overmier, in their book, Animal Research and Human Health, state that animals are causal analogical models and thus can be used to study human disease and predict human response: “When the experimenter devises challenges to the animal and studies a causal chain that, through analogy, can be seen to parallel the challenges to humans, the experimenter is using an animal model.” Causal analogies are a subset of analogy arguments in which causal assumptions arise based on the model. LaFollette and Shanks explain that the first condition that must be met in order for a thing to be considered a CAM is this: “X (the model) is similar to Y (the object being modeled) in respects {a…e}.” For instance chimpanzees and humans have a) an immune system, b) have 99% of their DNA in common with humans, c) contract viruses, etc. They continue, “X has additional property f.” HIV reproduces very slowly in chimpanzees. “While f has not been observed directly in Y, likely Y also has property f.” We therefore expect HIV to reproduce slowly in humans (LaFollette and Shanks, 1996). So if HIV replicates slowly in chimpanzees, animal experimenters reason by analogy that it will do the same in humans.

Animals are used as causal analogical models and the reasoning process used is called causal analogical reasoning. LaFollette and Shanks state that, “CAMs must satisfy two further conditions: (1) the common properties {a,…,e} must be causal properties which (2) are causally connected with the property {f} we wish to project – specifically, {f} should stand as the cause(s) or effect(s) of the features {a,…,e} in the model.” (LaFollette and Shanks, 1996)

The causal/functional asymmetry theory implies that causal mechanisms may differ between species. Causal disanalogies compel caution in extrapolating data between species. The use of animal CAMs also suffers from the systemic disanalogy argument. Since systems (organs, tissues etc.) may differ in subtle and unknown ways, identical exposure to a given compound will often cause different reactions in different species. In other words, for a CAM to be predictive, “there should be no causally-relevant disanalogies between the model and the thing being modeled.”

(LaFollette and Shanks, 1996). Considering our knowledge of evolutionary biology, this is arguably impossible without total knowledge of both the model (animal) and thing being modeled (human).

Only by comparing the results from testing each given substance or procedure in an animal species with human-based data can we determine whether the animal is sufficiently similar to humans to allow extrapolation. We can only know which animals mimic humans after we study the human data. Since animal studies are not predictive, we see no need to use them.

There are areas of research which offer results that are much more reliable than animal models, since they are human-based and hence are far superior methods of doing research. Scientific non-animal based methods would be for example: epidemiology, in vitro research; clinical research; autopsies; mathematical and computer modelling; post-marketing drug surveillance; research with human tissue; basic science research in the fields of physics and chemistry and other human-based and technology-based research methods such those using positron emission tomography, functional magnetic resonance imaging, magnetoencephalography, magnetic resonance imaging, transcranial magnetic stimulation and single photon emission computed tomography.

Animal modelers will insist that animals, notwithstanding their lack of isomorphism and inability to be CAMs, are still necessary because without animals researchers could not evaluate the drug or procedure in an intact system. We agree that life processes are interdependent, that, e.g., the liver influences the heart, which in turn influences the brain, which in turn influences the kidneys and so on. Thus, the response of an isolated heart cell to a medication does not confirm that the intact human heart will respond as predicted by the isolated heart cell. The liver may metabolize a drug to a new chemical that is toxic to the heart whereas the original chemical was not toxic.

We also concede that cell cultures, computer modelling, in vitro research etc., cannot replace the living intact system of a human being. But the question remains,

“Does the intact animal model do better than the non-animal methods?” The evidence suggests that it does not (Greek and Greek, 2000, 2002, 2004; LaFollette and Shanks,

1993). While animal models may be intact they still suffer from the systemic disanalogy argument which invalidates their use as intact systems.

Paradigm Shifts

It is very disturbing to talk to physicians who say they question their career choice because they look back on their practice with regret. Problems of illness and side effects due to interventions associated with the medical care they gave and poor results from medical research they were involved in are demoralizing for them. Much of this regret is due to the use of animals in biomedical research.

Science is so successful because new theories are formed and old ones abandoned based on evidence. From our examination of the philosophy of science and animal models of human disease and drug response, we conclude that the use of animal models is not beneficial to humans today. The very small differences between animals and humans at the cellular and sub-cellular levels make extrapolation between species dangerous. Our theoretical explanation for this is that causally relevant disanalogies exist between species.

Researchers who use animals are operating under the paradigm that says all animals are more similar than different. Modern evolutionary biology (Shanks 2002;

Greek and Shanks, 2006) reveals that the differences are today far more important than the similarities with regards to how the organism operates at the cellular level, the level where disease occurs. The animal model paradigm appeared viable in the

19th century when we knew so little. On the gross level all animals were similar: dogs had hearts, so did humans; cats had electrical activity in their brains, so did humans.

But today we are studying things on the very level that defines the species’ as being different – the cellular level. It is unreasonable to assume that at this level what we learn about one species will apply to another.

We conclude that a paradigm shift in biomedical research is absolutely vital for medicine to advance. Paradigm shifts, such as occurred in physics in the early 20th century do occur. It is important to remember that when a paradigm shift occurs it does not mean that everything that went before it was wrong. Modern physics did not refute or negate Newton’s laws but only modified them. Newton’s laws are still used daily but Newton’s physics cannot explain objects at speeds near that of light nor can it explain the effect gravity has on light. Hence a paradigm shift occurred. Likewise, animal models that were successfully used to explain basic areas of physiology or anatomy cannot explain what occurs at cellular and molecular levels where disease is currently being studied. Hence another paradigm shift is needed.

Conclusion

The knowledge gained from the combination of the fields of molecular biology, evo- devo, complex systems analysis, philosophy of science and others has revealed that the differences between species are more important than the similarities when applied to using animals to model humans. This is why researchers cannot study a pig heart and extrapolate the results to humans suffering from coronary artery disease. The new physics of Bohr and Einstein explained everything Newton’s physics did plus much more. Modern-day biomedical research methodologies, such as those listed earlier, could likewise have been used to discover everything that animal models were used for in the past but they also give us data and discoveries that animal models never could and never will.

We realize that our claims are controversial, particularly among those whose livelihoods depend upon animal models, but our arguments are straightforward. If our arguments are unsound, they should be easy to refute. Here is how one would do this:

1. Explain why animals, when used in biomedical research for the study of

human disease and to test drugs are not used as CAMs.

2. Show that animal models, when used as CAMs are successful far more often

than not. This can be accomplished by comparing the results of drug toxicity

studies in animals with studies in humans or by comparing the results of

induced diseases in animals with the same disease in humans. The data for

these studies exists but seems to support our arguments, not the claims of

animal modelers.

We have not been able to find data contradicting our arguments, and none of our critics have been able to present this data either. One hypothesis that explains this is that there is no such data. We argue this is the best hypothesis, and thus that animal models remain in vogue not for scientific reasons, but for non-scientific reasons, e.g., economic, legal, and social reasons. Those who have an interest in social policy being guided by science have good reason to demand that science prevail and, thus, that society turn more of its attention to more fruitful methods of research.

References and further reading

Barnard, N.D. and Kaufman, S.R. (1997) “Animal Research Is Wasteful and Misleading,” Scientific American February, pp. 80-82.

Carroll Marilyn E. and J. Bruce Overmier (Eds) (2001). Animal Research and Human Health. American Psychological Association, p. 5. Greek R. and Greek J. (2000) Sacred Cows and Golden Geese: The Human Cost of Experiments on Animals. Continuum International Publishing.

Greek R. and Greek J. (2002) Specious Science. Continuum International Publishing.

Greek J. and Greek R. (2004) What Will We Use If We Don’t Experiment On Animals? Trafford Publishing.

Greek R. and Shanks N. (2006) Animal Models in Light of Science. Rodopi.

Gura, T. (1997) “Systems for identifying new drugs are often faulty,” Science, 278:1041-2.

LaFollette H. and Shanks N. (1993) “The Intact Systems Argument: Problems with the Standard Defense of Animal Experimentation,” The Southern Journal of Philosophy, XXXI: 323-333.

LaFollette H. and Shanks N. (1993) “Animal Models in Biomedical Research: Some Epistemological Worries,” Public Affairs Quarterly, 7:113-130.

LaFollette H. and Shanks N. (1994) “Animal Experimentation: The Legacy of Claude Bernard, International Studies in the Philosophy of Science, 8:195-210.

LaFollette H. and Shanks N. (1995) “Two Models of Models in Biomedical Research,” Philosophical Quarterly, 45:141-160.

LaFollette H. and Shanks N. (1995) “Utilizing Animals,” Journal of Applied Philosophy, 12:13-25.

LaFollette H. and Shanks N. (1996) Brute Science. Routledge.

Reines, B.P. (1991) “On the locus of medical discovery,” The Journal of Medicine and Philosophy,16: 183-209.

Shanks, N. (2002). Animals and Science. A Guide to the Debates. Controversies in Science Series. ABC-Clio.

Trull, F. Animal Models: Assessing the Scope of Their Use in Biomedical Research. Charles River, Mass.: Charles River;1987, pp. 327-36

Young, Wise. As quoted in Scott LaFee, “Crucial or cruel?”San Diego Union Tribune, March 16, 2005.