Rats! To Animal Models
Kenneth P. Stoller, M.D.
The continual development of new drugs and chemical products in the United States along with their potential for causing adverse human health effects have led to the presumptive need for preclinical and premarket testing on animal models in the hope of accurately predicting potential human responses. The major problem facing regulatory agencies is one of risk. However, there is no certainty that a substance that has been found safe for an animal model will also be safe for humans. Indeed, the sensitivity and specificity of animal models for acute and chronic toxicity testing has been widely criticized.1-5
The value of animal testing has been challenged on several levels. For example, testing protocols often lack common sense. When saccharin was evaluated, rats were given the human daily equivalent of the saccharin contained in about 1100 cans of diet soda. Also, there are fundamental anatomical and physiological differences between rats and people. For example, the metabolism of the rat is at least several times faster than that of humans; rats have no gallbladder, whereas humans do; rats synthesize their own ascorbic acid, and humans, of course, do not. Human beings have been reproductively isolated for millions of years, and numerous metabolic differences have developed. Consequently, rats and humans may respond very differently to the same agent.
Furthermore, there are important differences within human populations, and an individual animal model cannot correspond to all of them. For example, there is much variability in the human species due to genetic, cultural, dietary, and other factors, and it is difficult to know for whom an animal study may be relevant.
The tests themselves are fraught with difficulties. A lifetime toxicity/ carcinogenicity bioassay using rodents may last longer than two years, during which time several batches of diet will be used. The variation in diet composition, coupled with inadequate diet desccription, makes nutrient-toxic interactions possible, and these are difficult to recognize.6 Salsburg notes, "...the lifetime feedipg study has never been subjected to proper validation as an assay jor human carcinogens ... it appears to lack acceptable specificity and specificity."7
The laboratory environment also affects data. J.R. Fouts states:
We do indeed recognize that the environment affects our experiment results and that effects of environment in our animals can be both dramatic and subtle. What we acknowledge is not, however, translated often into practice; and many of our data are confounded by environmental effects that we do not control, are unaware of, or for which we make no provision.2
Exposure to many carcinogens is unavoidable. The goal of toxicity testing is to determine how much exposure will result in a significant health risk to humans. But, deriving quantitative risk evaluations from animal tests assumes that the animal model responds in a manner identical to that of humans. In reality, animal models are often chosen for such non-scientific reasons as cost and ease of handling, not their ability to predict human toxicity. The lack of quantitative metabolic data on the model species and especially on humans precludes direct comparisons. Joshua Lederberg, Nobel Laureate and President of Rockefeller University, writes:
The one or two or three hundred millions of dollars a year that we're now spending on routine animal tests are almost all worthless ... it is simply not possible with all the animals in the world to go through new chemicals in the blind way that we have at the present time, and reach credible conclusions about the hazards to human health. We are at an impasse. It is one that has deep scientific roots, and we had better do something about it.9
Many of these millions are spent on experiments or testing protocols using rats. Can one expect the highly refined and modified laboratory rat that has been shaped to such a great extend by the unnatural selection of domestication to be a good predictive model for humans? The white rat seems to be strikingly modifiable -- rats of a given appellation in 1970 may be quite different from those of the same name today. That the process of domestication has affected behavior is obvious, because unacceptable behaviors were bred out. How susceptibility to toxicological agents was affected by domestication is not known.
Horwitz, of the FDA Bureau of Goods, observes:
...the biologist, the chemist and the statistician are all working far beyond the region of reliable measurements in the safety assessment process. When the chemist ventures beyond the limits of reliably measurement, he is punished by obtaining irreproducible results, false positives and false negatives. These manifestations of method overextension are verifiable. When this behavior occurs with humans, it is called emotional illness (nervous breakdown); when it occurs in experimental animal colonies, it is called biological variability. (This is) then manipulated by statisticians as if they were reproducible measurements ... Many of them make the classical mistake of poor operations research. They apply mathematical formulae to unreliable data and then ascribe the preciseness of the mathematics to the final product. Manufacturers who persist in producing a product that cannot be kept in statistical control go bankrupt; but toxicologists, chemists and statisticians who deal with similar situations saturate the literature with contradictory findings; each claims to be absolutely correct, but all are equally invalid because the systems they describe are not in statistical control! ... In analytical chemistry, as concentrations go lower, measurements go out of statistical control. There is no reason to think that biological measurements are an exception to this predicament.10
Inherent problems such as biological variability, false positives, dose-response interpretations, and experimenter or observer bias come to a focus when one attempts to transfer results across species lines and interpret the results in terms of human risks. Judging relevance or irrelevance requires detailed knowledge of comparative metabolism and pharmacokinetics of the material. Such data are usually lacking, expensive, and difficult to obtain. In truth, bad math and arbitrary legislation are used to cover missing biology.
Thalidomide illustrates the unreliability of rodent and other animal tests for teratogenicity. Lasagna writes, "Thalidomide is a teratogen in a few rabbit breeds and in seven species of primates. It is not a teratogen in at lest 10 rat strains, 15 mice strains, 11 rabbit breeds, two dog breeds, three hamster strains, and eight species of primate.11
There are many examples of substances with a considerable history of human use, such as aspirin, insulin, epinephrine, and certain antibiotics, which are all known to cause malformations in laboratory animals.12 But since the results of teratogenicity tests sometimes have little effect on marketing, it becomes difficult to understand the scientific basis for carrying them out in the first place. The inescapable conclusion is that the tests are performed not for scientific purposes but for political and legal reasons. Smithells notes:
... the emotive picture of a deformed baby looking at a pharmaceutical juggernaut with wide, reproachful eyes provides constant joy to the media and is not a matter of total disinterest to the legal profession ... The extensive animal reproductive studies to which all new drugs are now subjected are more in the nature of a public relations exercise than a serious contribution to drug safety. Animal tests can never predict the actions of drugs on humans.13
Melmon puts the entire issue in perspective:
In most cases, the animal tests cannot predict what will happen when the drug is given to man. Standards for toxicology are often set by officials, such as Federal regulators, who are responding to the pressures of ill-advised but obviously well-intended legislators or consumer groups who may or may not be aware of the futility of increasing the amount of testing required when some tests often have no bearing on bow man will respond to the drug. The multiplication of tests in animals, often invalid tests and possibly performed in the wrong species, can only add to the cost of drug discovery and can only limit the range of discovery. The result is not only a waste of animals but also a waste of limited scientific resource; the loss is compounded by the fact that human life will not benefit from drugs whose release is unnecessarily delayed.14
In summary, the reliance on rats as models for human toxicity, carcinogenicity, and teratogenicity reflects the poor scientific bases of toxicological testing in animals. Whether the concern is absorption, tissue distribution, biliary excretion, intestinal flora, enterohepatic circulation, mechanisms of conjugation, or other aspects of metabolism, there are profound differences between the values of the rat and those of humans. Despite the biochemical and physiological differences between the rat and human, regulatory bodies readily extrapolate results obtained from tests on rats to humans. There seems to be disproportionate amount of confidence in any risk evaluation made, especially once the value is forever written in the Federal Register or other official-looking parchment.
Whenever regulations are to be established, proof of the validity of the toxicologists's assertions should be demanded, but this is not typically the case with quantitative risk assessment. We need valid mechanisms to assess human toxicity, including post-market surveillance of drugs and chemicals and in vitro human cell and tissue culture tests. Until appropriate models are used to predict human toxic responses, countless individuals will become victims of our chemically-dependent society.
References
1. Balls M, Riddell RJ, Worden AN (eds): Animals and Alternatives in Toxicity Testing. New York, Academic Press, 1983.
2. Zbinden G, Flury-Roversi M: Significance of the LD5O-test for the toxicological evaluation of chemical substances. Arch Toxicol 1981;47:77-99.
3. Lewis PJ: Animal tests for teratogenicity, their relevance to clinical practice. In, Hawkins DF (ed): Drugs and Pregnancy: Human Teratogenesis and Related Problems. Edinburgh, Churchill Livingston, 1983.
4. Sharpe R: The Draize test -- motivations for change. Fd Chem Toxic 1985;23:139-144.
5. Jansen JD: The predictive value of tests for carcinogenic and mutagenic activity. In Deichmann WB (ed): Toxicology and Occupational Medicine. Amsterdam: Elsevoir/ North Holland, 1979.
6. Conner MW, Newberne PM: Drug-Nutrient interactions and their implications for safety evaluations. Fund Appl Tox 1984;4:S341-S356.
7. Salsburg D: The lifetime feeding study in mice and rats - an examination of its validity as a bioassay for human carcinogens. Fund Appl Toxicol 1983;3:63-67.
8. Fouts JR: Fed Proc 1976;35:1162-1165.
9. Lederberg J: A challenge for toxicologists. Chem Engin News March 2, 1981, p.5.
10. Horwitz W: Effects of scientific advances on the decision-making process: analytical chemistry, Fund Appl Tox 1984;4:5309-317.
11. Lasagna L: Regulatory agencies, drugs, and the pregnant patient. In Stern L (ed): Drug Use in Pregnancy. Science Press, 1984.
12. Freidman L: Tox Appl Pharmacol 1969;16:498-506.
13. Smithells RW: Drug teratogenicity. In Inman WH (ed): Monitoring for Drug Safety. Philadelphia: JB Lippincott, 1980.
14. Melmon KL: The clinical pharmacologist and scientifically unsound regulations for drug development. Clin Pharmacol Ther 1976;20:125-129.