Better Science II. PROBLEMS WITH IN VIVO ANIMAL-BASED METHODS Alternatives have played a critical role in the advancement of biomedical research and modern medical practice. By any objective measure the Nobel Prizes in Physiology and Medicine represent the best and most significant accomplishments in the biomedical sciences. An analysis of the specific projects for which these awards were given since their inception in 1901 documented that more than two-thirds of them were for work that was either partially or entirely based on the use of alternative methods. This percentage is even higher for the past few decades due to the increasing importance of in vitro and mathematical techniques. Nobel Prizes are frequently awarded for the development of major new experimental techniques, not for new animal models. Enders received his Nobel Prize in 1954 for creating an in vitro means, utilizing human cells, for growing the poliovirus. This new method is widely acknowledged as the key event leading to the first successful polio vaccine. Researchers using hundreds of thousands of monkeys to study polio did not receive similar recognition. Historically, the concept of "animal models" of human health problems was formulated in response to legitimate concerns about infectious diseases. The basic assumption was that if animals used in laboratories experimentally contracted an infection and were cured, there was a high probability of stopping the same disease in humans. Although a useful concept at the time, such uses can now be replaced in most instances by available alternatives and clinical studies of naturally-occurring diseases in human and non-human animals. The traditional animal "model" approach to studying human illness rapidly collapses and is most questionable when the focus switches from introduction of a common disease-causing organism to species-specific health problems such as psychopathology, cancer, drug addiction, Alzheimer's and AIDS. As originally conceived, to be a valid model of human health concerns, the animal disease must have the same biological mechanisms, symptoms and responses to treatment as the theoretically similar human counterpart. Failure to meet one or more of these criteria invalidates the animal "model." It is not sufficient to artificially produce a condition in an animal in a laboratory that only mimics, resembles, imitates or is similar to the so-called human equivalent. The current epidemic of iatrogenic (disease or injury caused by medical treatments) diseases -- one of the leading causes of death in the United States -- is partly the result of using inappropriate animal "models" to predict human responses to drugs and other treatments. Patients then have unexpected reactions or die from exposure to these supposedly safe drugs and chemicals. In an attempt to overcome the severe limitations of traditional animal "models," researchers now are genetically engineering animals by either removing or adding genes believed to be related to specific human diseases. The underlying assumption here is that these new genetically constructed animals will be more human-like. The fact that existing animal models need to be genetically "improved" is further evidence of their original lack of biological and/or clinical relevance. The concept of animal models becomes even more tenuous when it is applied to the fields of toxicology and risk assessment. After exposure to potentially toxic or dangerous substances, both the inter- and intra-specific (between and within a species of animal) differences in morphology (anatomy), physiology and biochemistry between humans and the species commonly used in such tests introduce multiple significant biasing factors which cannot be avoided. The data derived from such experiments are not scientifically relevant to the purposes of the tests. Consider that in some carcinogenicity (cancer promotion) studies there is no effective correlation between the results for mice and rats (closely related rodents), let alone relevance to evolutionarily more distantly related humans. Although seldom mentioned, essentially all of the in vivo animal safety and toxicity tests currently in use were never validated and would be unlikely to pass present scientific validation procedures. These in vivo tests continue to be used for reasons of familiarity, tradition and checkbox/six-pack regulatory schemes. They are not used because they are the result of proven relevance and reliability. In vivo tests are subject to a series of basic biasing factors that simply do not exist for their in vitro and in silico (computer) replacements. Differences in lifespan and maturation processes between humans and rodents are significant. There are meaningful contrasts between processes that develop naturally over the course of time versus accelerated laboratory tests of induced, unnatural levels and routes of exposure. Commercial in vivo safety testing usually sacrifices accuracy and relevance for speed and cost. These problems are especially applicable to chronic (long-term) toxicity testing, the results of which may be no more accurate than simply flipping a coin. Because of the multiple, well-documented differences in responses, the use of non-human species in toxicity testing requires the application of often complex mathematical equations to extrapolate the results to potential human exposure. Major differences are associated with simple differences in body size. Extrapolations between species are not and should not be based on such simplistic criteria as length or weight differences. The husbandry conditions under which animals are typically bred, raised and housed seriously biases any data derived from their use. This is true for even the best state-of-the-art laboratory animal facilities. Recent studies suggest that much, if not all, of the research and testing done utilizing captive laboratory species in traditional cage environments may be so biased as to be useless, even if it can be replicated. In vitro replacement alternatives, especially with regard to safety and toxicity testing, have a number of positive characteristics: They were scientifically validated and proven to be relevant to the desired endpoints. They allow multiple, simultaneous tests under a range of concentrations and controlled conditions. They allow larger numbers of tests in shorter periods of time. They are easily adapted to high throughput (high volume and high speed) conditions that cannot be replicated by in vivo methods. They are logistically simpler and economically less costly. For example, several decades ago the National Cancer Institute adopted an in vitro replacement for their standard animal-based procedures to identify potential anti-cancer compounds. This single decision dramatically increased the number of tests conducted; significantly reduced the per unit cost of the program; and saved more than a million rodent lives every year. << Back to Contents | Next Section >>

Other NEAVS Fact Sheets: Benefits of Non-Animal Tests | Xenotransplants | Animal Welfare Act | Limitations of Animal Tests | Non-Animal Product Safety Test Alternatives << Back to Better Science  | Home