INSIGHTS: Animal research is a failed research paradigm and not morally permissible

By Guest Opinion

As a pediatric intensivist, I struggle daily with patients who depend on research to improve their survival and quality of life. In the past, I am ashamed to say, I believed animal research was necessary for the medical advances my patients needed. Then I looked into the animal research paradigm. What I learned I aim to explain, however briefly, in the space below.  

An IJB investigation about research on dogs being hidden by a “code of silence” at an Ontario hospital is a moral concern. This publicly funded research with sentient animals, capable of consciously experiencing suffering, pain, pleasure, and happiness, was harmful, non-therapeutic and non-consensual. The dogs had nothing to gain, and everything to lose.

The moral outrage expressed across Canada is appropriate for two reasons. Animals have proven to be extremely poor models of human disease, with disappointing translation rates for human benefit. Even if there were great benefits to people, painful and harmful research with sentient animals is not morally permissible. 

The idea of animal research is to model human disease so that findings predict human results. This prediction has not borne out in reality. In safety testing (e.g., toxicology, fetal congenital malformations, and carcinogenicity) the match between animal research and human results is 40 to 50 per cent. There is so much inaccuracy that harmful drugs have been used in humans, and safe drugs have been thrown away. 

Treatment findings from animal research translate to humans at even lower rates. At least 95% of treatments that test safe and effective in animals fail in human studies. Even of drugs that show great promise in animals, enough to be invested in by pharmaceutical companies, only 7% are successful in humans. This lack of prediction includes studies on cancer, neurodegenerative diseases, stroke and brain trauma, spinal cord injury, cardiac arrest, heart failure, septic shock, and HIV vaccines.

The reason animal research cannot predict human results is simple: animals are complex biological systems with many interacting functions. In such complex systems, small differences in initial conditions can result in dramatic differences in response to the same intervention. There are many differences in initial conditions between animal species. For example, the genes expressed in response to burns, trauma, and severe infection in rodents are close to random in matching their human gene counterparts. Even non-human-primates are not susceptible to many human diseases, including atherosclerosis, Alzheimer’s, and epithelial cancers of the breast, lung, pancreas, stomach, colon, ovary, or prostate.

Animal research has proven a failed paradigm of research for human benefit, inherently unable to achieve higher translation rates because the response of one complex system—species—to an intervention poorly predicts the response of another complex system—humans. 

Even if animal research predicted human results, it would remain morally impermissible. The assertion that it is justified, because there are no alternatives, is missing a key premise: the notion of human exceptionalism. This notion is said to justify using animals in research instead of using humans in the same research, despite the much greater benefits that would be expected in using humans. Human exceptionalism is based on some unique human capacity such as autonomy or moral agency—the ability to know right from wrong. But the human exceptionalism paradigm fails. 

First, I doubt we truly believe it. Which statement makes more sense: 
A) “It is wrong to harm me for no good reason because I am able to make moral decisions”, or 
B) “It is wrong to harm me for no good reason because that would be harmful for me –  that would be harmful for me because I am able to experience things like pain and suffering.” 

The second statement explains the vulnerability of both humans and sentient animals to harm and why it is wrong to harm either. Even if you disagree, do you think that having superior abilities (e.g., moral agency) justifies a human actively harming those with inferior abilities? Could you harm an infant or an elder with dementia?

Second, the capacities that define human exceptionalism are not present in all and only humans. The moral irrelevance of these capacities is made clear by the fact that some humans lack them, and we know that, while they may not be moral agents, they are certainly moral patients (i.e., they can experience pain and suffering) and therefore cannot be used in harmful research. 

Some animals have at least as much sentience and are just as much moral patients that can experience pain and suffering. This shows that human exceptionalism is speciesism—a prejudice against another based on the morally irrelevant property of species. That dogs were used aroused outrage because we are familiar with the sentience of dogs; however, not to extend the same concern for other mammals, such as the millions of mice and rats used in labs every year, is another example of speciesism.

Some might find reassurance that the dog research unveiled by the IJB investigation was approved by an ethics board in accordance with Canadian standards. This is misplaced trust in regulation. Regulatory commitment to weighing benefits and harms of animal research would not have led to the following problems. Regulation has not resulted in animal research that meets expectations for animal welfare, methodological quality, or human benefit and is often overseen by other animal researchers who share a speciesist mindset. 

There are alternative human-based research methods that are not used because they are inconvenient or require new training and breaking old habits. 3-D human cell-derived organoids that mimic complexity of a human organ, human cell-based organs-on-chips that mimic interactions among organs, sophisticated computer models, and novel in-human clinical studies are all modern advances that make animal research unnecessary. 

Some will say, surely animal research has led to advances in human medicine, such as dog studies that led to the discovery of insulin. Indeed, if there was a recent medical advance, there surely was an animal model used. There are problems with this argument. 

First, a model’s ability to predict human outcomes is a prospective ability, and does not refer to a retrospective look back for rare examples of concordance between animal and human findings. Second, by law, since 1930, any medical advance must have been tested in an animal model prior to use in humans. If you generate enough hypotheses, some, however rarely, might prove useful. There are many instances of no medical advance despite a great many animal models. Third, we have new approach methodologies available. 

Fourth, I am unconvinced that research on dogs was necessary for the discovery of insulin. I challenge anyone reading the book “The Discovery of Insulin” to explain how dog experiments were essential, particularly when human research had already shown that the pancreas produced the protein essential for glucose homeostasis. Finally, harmful research on sentient animals remains morally impermissible, unless the same research is permissible when done with sentient humans.