The philosophical relationship between man and nonhuman animals has been predominantly shaped by religious philosophy and not scientific principles. Most if not all religions express not only principles of how humans should treat each other, but also how humans should treat animals. Western culture is steeped in Judeo-Christian principles that present animals as being bequeathed to Adam, to do with as he wished. Early principles of scientific inquiry were heavily influenced by the philosophy of Descartes in the 1500s, who argued that animals are automata that might look as though they are conscious, reasoning and reacting, but are not so. All animal behavior could be explained in mechanistic terms [3]. In the famous words of philosopher Nicholas Malebranche, “They eat without pleasure, cry without pain, grow without knowing it; they desire nothing, fear nothing, know nothing” [4]. Without consciousness, animals could have no interests, no experiences, and no moral standing.

Twentieth century religious scholars, however, were sorely challenged with regard to scientific theories and man’s relationship to the world and its other inhabitants. New scientific disciplines such as paleontology and the theory of evolution gained ground, suggesting that man not only was not truly different from the animals, but may have actually arisen from them. Such developments test the very structure of scientific inquiry, which relies heavily on the concept that man is separate from, and greater than his experimental subjects. In the words of primatologist Fran De Waal, “Our culture and dominant religion have tied human dignity and self-worth to our separation from nature and distinctness from other animals” [5]. He goes on to point out that scientists are biased against acknowledging uncomfortable similarities between human and nonhuman animals, because such acknowledgement weakens arguments for the moral superiority of man.

The scientific study of animals no longer supports the Cartesian paradigm, and few biologists seriously argue that animals can be clearly distinguished from humans on the basis of many traditionally held concepts. Rigorous research demonstrates that animal consciousness and abilities show startling parallels to those previously thought to be uniquely human. Tool manufacture and use is now well described in primate, nonprimate and even nonmammalian animals [6]. Great apes appear to be capable of using symbolic language [7], and nonhuman animals have been shown to work with numbers–both strong markers of abstract thinking [8]. Animals demonstrate “culture,” in which uniquely individual and adaptive behaviors are passed within social groups by observation rather than instinct or genetic programming [9]. “Episodic recall,” believed to be an important marker of sentience, is seen in meadow voles [10]. Cetaceans have knowledge of symbolic representations and exhibit an awareness of self [11], which has also been demonstrated in primates, elephants [12], and magpies [13]. Recent studies in dogs indicate an extensive abstract vocabulary, and the ability to extrapolate names [14].

Nevertheless, many scientists are still reluctant to relinquish Cartesian ideals. Bioethicist Bernard Rollin poses that scientists have evolved an “ideology” of their own–one that asserts, “science might provide society with the facts relevant to making moral decisions, but it steers clear of any ethical debate” [15]. This belief is self-evidently false: if there is no moral component to scientific efforts, then there would be no need to report the breach of moral principles with the cruel experimentation on research subjects. In fact, such an “ideology” asserts that scientific studies can be performed without the consideration for the treatment of its research subjects-be it human or animal. Thankfully, the idea that medical research does not answer to moral considerations has thoroughly been disparaged, as the Doctor’s Trial at Nuremberg after World War II clearly demonstrated.

Most bioethicists now agree that animals do have moral standing, but which animals and how much moral standing are subjects of debate. There is agreement that deliberately causing animal suffering is a moral harm to be prevented or mitigated and must be weighed heavily against any perceived benefits it might provide. Furthermore, Kantian philosophy holds that cruelty toward an animal is contrary to man’s duty to himself, because he who is cruel to animals will likely be cruel to humans as well [16].

There are many moral “theories” by which animal rights and moral standing are discussed, but it is useful to consider three basic types of arguments. The first two are based on the idea that “rights” belong to “persons” and “persons” are defined as any being that has sentience or consciousness.

The first theory is a simplified view of a “conservative”¹ moral philosophy that humans and only humans have sentience or consciousness and that therefore only humans have “rights.” Philosopher Christina Korsgaard suggests that humans not only have perceptions and desires (which she concedes animals do too) but also are uniquely able to reflect on those perceptions and desires, and that it is this reflection that sets us apart from other nonhuman animals [17]. Thus nonhuman animals do not necessarily have “rights.” However, even if they do not have “rights” per se, she points out that this does not imply that humans do not have explicit moral duties towards them. A second more “expansive” moral view is that morally important qualities such as rationality exist in some, if not most animals, and that therefore animals have moral standing in their own right. This point of view argues that some animals (e.g., great apes) have sentience and rationality, while some humans (e.g., infants and those with dementia) have only partial or potential rationality, and still other humans (e.g., permanently comatose individuals) lack rationality altogether. If rationality and sentience are the necessary and sufficient conditions for “personhood” and “rights,” then actually some animals do have “personhood” and “rights”, and some humans clearly do not. Finally, a third, “moderate” approach to moral obligations in our treatment of animals does not address the question of “personhood” and “rights” at all, but asks simply if it is ever morally acceptable to cause nonhuman animals to suffer. Peter Singer argues that animals have “interests” whether or not they have rights, and that all animals have an interest in avoiding suffering. Therefore, it is almost always morally wrong to willfully cause animals to suffer. He does not argue, for example, that using animals for food is wrong, so long as they are humanely treated and killed. However, using animals in ways that cause suffering must be seriously weighed against any human need it answers. He further argues that a great deal of animal research causes suffering without satisfying any direct or urgent purpose and answers only minor human interests when weighed against the animal’s more serious interest in avoiding suffering. In many (but not necessarily all) cases, Singer argues, we should end animal research [18].

Obviously, to weigh the morality of animal research, we need to attempt to quantify the role of animal research in the safety and efficacy of medical treatments and whether animal research is beneficial.

When human birth defects first began to appear in the offspring of women who had ingested thalidomide during pregnancy, researchers pointed out that animal studies had failed to demonstrate any teratogenic potential of thalidomide in rats and that therefore thalidomide could not be the culprit. In fact, as J.L Schardien has observed, “in approximately 10 strains of rats, 15 strains of mice, 11 breeds of rabbit, 2 breeds of dogs, 3 strains of hamsters, 8 species of primates, and other such varied species as cats, armadillos, guinea pigs, swine and ferrets in which thalidomide has been tested, teratogenic effects have been induced only occasionally” [19]. Ultimately, similar birth defects to those seen in humans were found when high doses of thalidomide were administered to the New Zealand White rabbit after the fact, but only to a lesser extent than in other strains of rabbits. Schardein goes on to say, “Trying to justify the performing of animals studies, however, is a somewhat frustrating exercise in futility. Despite the extensive testing in animals prior to human use, the labeling of marketed drugs does not fully reflect the results of such studies, nor does it indicate the clinical experience with a drug that may have had many years of apparent safe use…” [19].

The United States largely escaped the tragedy of thalidomide compared to other Western countries. Only 17 cases were ever reported [19]. Why? Because Frances Oldham Kelsey–a reviewer for the Food and Drug Administration (FDA)–on her very first assignment in her first month with the agency, had grave doubts about thalidomide that were based ironically on neuropathic side effects shown in human studies, but had been largely ignored. She halted the approval of the drug in the United States (U.S.) only one day before approval was to automatically go into effect [20].

Is the thalidomide story a unique example of how animal studies fail? Unfortunately, the answer is no.

How much do animal experiments contribute to medical advancements? Do animal experiments accurately predict human responses?

In many cases, medical treatments commonly cited as depending for their success on animal studies were in fact actually impeded or contradicted by animal research. Even the Center for Disease Control (CDC), on its “Overview of Animals in Scientific Research Fact Sheet” lists smallpox vaccine (cow), and penicillin (mouse), as among the “tangible benefits” of animal research to medicine [21], yet Jetsy and Jenner actually each tested their smallpox vaccines in human subjects and not cows, and Fleming decided based on rabbit experiments that penicillin wouldn’t work. He shelved the drug and only found it to be effective years later when he used it to treat a human infection because he had run out of all other options. Only after the human success did he test the drug in mice. We are all fortunate that Fleming did not choose to test it in guinea pigs–in which it is toxic [22].

Possibly two of the most spectacular recent failures of animal studies were the infamous trials of TGN1412, a monoclonal antibody developed to treat leukemia and autoimmune disorders, and Fialuridine, an antiviral drug with activity against Hepatitis B.

Animal studies in mice, rabbits and monkeys supported TGN1412’s safety. But when the drug was administered to six human volunteer subjects during 1 day in March of 2006 in a dose 1/500th of that used in animals, all volunteers suffered catastrophic reactions within minutes [23]. All volunteers survived the initial experience after receiving intensive medical support, but one required the amputation of all fingers and toes and a second volunteer was diagnosed within several months with a rare hematologic cancer believed to be secondary to receiving TGN1412 [24].

In the case of Fialuridine, preclinical testing in mice, rats, dogs, monkeys and woodchucks indicated that the drug was safe in doses that were hundreds of times higher than those to be used in humans. However, the administration of the drug during human Phase II clinical trials lead to the deaths of 5 healthy volunteers due to liver failure. Two other volunteers survived, but only after undergoing liver transplantation [23].

Many other well-known examples of research in which animal studies may have mislead researchers or impeded scientific advances include studies of the role of asbestos exposure in lung disease, AIDs research, and cancer treatments [25].

Simply stating “research involving animals has played a vital role in virtually every major medical advance of the last century” does not make it so. Nor is the report of a few admittedly spectacular failures sufficient to discount the value of animal research. Just as other elements of medical practice must be based on real evidence, so too must our understanding of the role that animal research plays now and that it should play in the future be based on sound evidence and not conjecture. Disciplined analysis of the translatability of published animal subjects research is a relatively new, but growing field of inquiry in the medical literature, and these reviews pose unsettling, but important questions about the true utility of animal research.

In a recent analysis of systematic reviews of animal studies to determine if they had informed clinical research in humans, Pound and colleagues [26] found the following problems:

Roberts et al. systematically reviewed 44 randomized controlled trials of fluid resuscitation experiments in traumatized animals [27]. Three large trials could not be included in the analysis because they had no control group, multiple studies compared different blood pressure targets, mortality was not reported in 2 of the 44 trials, and reanalysis of trial data using appropriately sophisticated statistical techniques to account for confounding variables demonstrated statistical heterogeneity [27]. The authors suggest that continuing systematic reviews of animal studies are needed to ensure that animal experiments do not set out to answer questions that have already been answered, and that they will actually provide generalizable results.