{"id":203,"date":"2026-05-07T16:07:51","date_gmt":"2026-05-07T16:07:51","guid":{"rendered":"https:\/\/sites.wp.odu.edu\/alaaelhaimeurcom\/?p=203"},"modified":"2026-05-07T16:07:51","modified_gmt":"2026-05-07T16:07:51","slug":"strengths-and-limitations-of-animal-models-in-biomedical-research","status":"publish","type":"post","link":"https:\/\/sites.wp.odu.edu\/alaaelhaimeurcom\/2026\/05\/07\/strengths-and-limitations-of-animal-models-in-biomedical-research\/","title":{"rendered":"Strengths and Limitations of Animal Models in Biomedical Research"},"content":{"rendered":"\n

Alaa Elhaimeur<\/strong> | Biomedical Science Major at Old Dominion University<\/p>\n\n\n\n

This article explores the role of animal models in biomedical research, including their advantages, limitations, and the reasons behind the FDA\u2019s movement toward reducing reliance on animal testing. It also examines the scientific, ethical, and practical considerations while exploring emerging alternatives that may improve future research methods.<\/h4>\n\n\n\n

Introduction<\/strong>:<\/p>\n\n\n\n

Animal models have been used in biomedical research for a long time. They are used to understand mechanisms of disease, screen drug efficacy and toxicity. They are also used to gain knowledge about biological systems that cannot solely be gained from studying cell cultures alone. <\/p>\n\n\n\n

As the FDA contemplates a reform of the regulatory processes that reduce the amount of animal testing that currently form the basis of drug approvals, an examination of animal models as research tools brings to light their strengths and weaknesses. Animal models have multiple strengths that explain their historical use.<\/p>\n\n\n\n

Yet the limitations of animal models have contributed to the search for alternative strategies. Animal models are physiologically whole organisms. Many responses to drugs and diseases affect multiple organ systems. Animal models enable the study complex interactions that exist between biological systems. Animal models enable the research of controlled living organisms in experiments that researchers are unable to ethically perform on human beings (Hooijmans & Ritskes-Hoitinga,2013). This ability makes for a very high level of internal validity, where many aspects of the environment of the animals can be stringently controlled in ways that are, as yet, still not possible for human beings.<\/p>\n\n\n\n

Many animal models display high genetic similarities to human beings. Rodents, zebrafish and other mammalian animals are used for a wide range of studies as they can be genetically engineered to exhibit the characteristics of humans in relation to the disease that is being studied. Genetically engineered animal models have been shown to be invaluable in studies of diseases such as cancer, metabolic disorders and autoimmune diseases. Toxicity tests on animals are also a prerequisite of regulatory bodies, such as the FDA, before any testing is performed on human participants. Animal models thus form the foundation of all preclinical toxicity tests, pharmacokinetics studies, and safety tests before subjects even get on to the research radar or regulatory agencies for the commencement of trials on human beings.<\/p>\n\n\n\n

Animal models are limited in numerous ways. Studies have shown that animal experiments are a poor predictive model for humans (Van Norman, 2019). This translatability limitation is critical in the field of drug development, when a large percentage of promising therapies discovered using animal models eventually fail once they enter human clinical trials. Findings in one study showed no genomic translation between the human and animal species (Seok et al., 2013). The inflammatory response of model organisms such as mice did not even reflect the inflammatory diseases of humans on a genomic level. The inflammatory response elicited in mice did not activate relevant pathways that were remotely similar to those activated during human diseases. Conclusively, in this case, the animal model was ineffective, therefore invalidating any findings of the study that used it as a model.<\/p>\n\n\n\n

Translational limitations between species are especially prevalent in cancer research trials. Currently, less than 8% of cancer treatments that work in animal studies end up working in human clinical trials (Mak, Evaniew, & Ghert, 2014). When drugs eventually fail after being subjected to effective but time -consuming and costly animal testing for years. The conclusions are obvious: they are indeed poor predictive models. However, they do come at a cost.<\/p>\n\n\n\n

Apart from presenting limitations of a scientific nature, animal models also raise several ethical issues. Greek & Menache (2013) state that animal models are limited in their scientific translatability, but also raise ongoing ethical concerns. Even though researchers may comply with guidelines set out by regulatory authorities such as the FDA, many of the studies still involve inflicting pain on the animals involved. It is this inflicted suffering that many people have become concerned about.<\/p>\n\n\n\n

Animal research is also an expensive and time consuming process. Animal testing agencies have to maintain a relatively complex research facility for the animals that are used to undergo tests in research on drugs, prior to the use of them on humans. Furthermore, studies can either be long-term or require several rounds of tests; this results in additional expenses being placed on research agencies. Perhaps even modern alternative animal models lack the current complexity required to replace animal models for biomedical research.<\/p>\n\n\n\n

Animal models are currently an invaluable tool in biomedical research. They enable researchers to examine whole organisms on whom drugs have been tested, how diseases affect organisms, and they form the foundation of all preclinical research tests. However, as valuable as they are to present-day medical science and its research methodologies, they are not without limitations. The differences between species compromise their effectiveness at predicting human outcomes. Animal models are poor predictive means to translate results obtained from tests and trials on them to human beings. They have detrimental limitations with relation to translatability and ongoing ethical and cost related issues. As the FDA contemplates reforming its reliance on animals to regulatory agencies such as the European Medicines Agency request new alternative methods, perhaps in future, the medical world can find middle ground. Like, adopting protocols that follow the ethical guidelines that many people seek while remaining reliable and accurate in predicting how humans may respond to newly developed\/researched drugs, trials, and treatments.<\/p>\n\n\n\n

New Approach Methodologies in Biomedical Research<\/h1>\n\n\n\n

New approach methodologies (NAM’s) are strong alternatives to animal testing, utilizing human-based systems to more accurately reflect the human body’s response to drugs. Technologies such as organs-on-chips and organoids use induced pluripotent stem cells (IPSCs), which are human cells that can be programmed to change into other types of human cells. Researchers can use these cells to mimic human organs. IPSCs are programmed into the cells of the human heart or liver, which are used to test the effect of drugs on those human organs (Kwon, 2026). These results will be more relevant to human patients rather than animals, whose biological systems can differ from humans.<\/p>\n\n\n\n

Organoids are three-dimensional cultures of human cells that mimic the properties of the body’s tissues. Organs on a chip introduce movement to these cultures to mimic the movement of blood through the human body. For instance, the use of a “liver on a chip” device that has allowed researchers to test the effects of drugs on the human liver, simulating the conditions of the body in which the liver functions (Kwon, 2026). These models have been able to detect harmful drugs that were identified as not harmful by animal models of the drugs’ effects. These systems are considered more predictive because they better replace human tissue structure and function compared to traditional models (Low et al, 2021).<\/p>\n\n\n\n

Computational models and artificial intelligence are alternative approaches to testing drugs. These systems use data sets of the effects of drugs on human cells and populations to predict the effect of a new drug on the body. For example, AI models have been developed to recognize the risk that a drug will cause liver damage or cause skin allergies. One model, in particular, utilized data from hundreds of chemicals to successfully predict whether a chemical would cause skin sensitization without the need for testing the chemicals on humans (Kwon, 2026). Ai-based toxicology models have been shown to improve prediction accuracy by analyzing large biological data sets (Zhu et al., 2020).<\/p>\n\n\n\n

Models like generative AI models named AnimalGAN<\/em> work in a slightly different way from other modes. These models use biological simulations instead of testing on humans to stimulate the responses of thousands of individuals to a certain drug. This model is beneficial because it reduces the length of time and the amount of effort required to test drugs on human populations. However, the accuracy of the models depends upon the data used to train the models. Thus, the outcome of the models can vary if the data set is limited or biased.<\/p>\n\n\n\n

One of the major limitations of these models is their reductionist nature. Most models use only specific types of cells rather than the entire body of a human being. These models are beneficial in that they focus on specific organs to test for specific types of problems. However, it is difficult to simulate the interaction of one organ with others in the body with these models. For instance, although a liver-on-a-chip model can test for the harmfulness of specific drugs to the liver, it cannot simulate the effect of the liver on other organs of the body. Other processes, such as the regulation of hormones in the body are also beyond the scope of these models. The Nature article also notes that processes like aging, hormone regulation, and other whole-body systems are outside the scope of these models without the use of animals to test for such effects (Kwon, 2026).<\/p>\n\n\n\n

In addition, these models are limited to a few types of human cells. Organs are comprised of many types of cells, all interacting with one another. Thus, it is difficult to simulate organs accurately with these models. Additionally, many of these models cannot simulate the long-term processes of the human body. Most models have been tested over short periods of time with human populations, but the effects of drugs over long periods are outside the capabilities of these models. Additionally, the models often have issues with the stability of the cells in the model over time. The outcomes of these models are also not consistent across different laboratories. Finally, these models have difficulties with validation. Researchers must validate these models to ensure that they are reliable and accurate in their predictions of the effects of drugs on humans prior to their use as alternatives to animal testing methods.<\/p>\n\n\n\n

Finally, as beneficial as these models are, they are still based upon human biology. Organoids and cells from human origins are used in these models. However, these models do not account for the effect of the environment, lifestyle, and other factors on human health. Additionally, the way in which humans respond to drugs can vary between individuals. Thus, accurately predicting the effect of a drug upon the human body without making errors remains a challenge with these models. New approach methodologies are a great development in the field of scientific research. Their benefits include improving accuracy in measuring drug effects on humans and reducing ethical concerns by limiting the need for human subjects, making these systems very valuable to scientific research. However, they are not yet complete alternatives to using animals in testing for drugs. The complexity of simulating processes of the entire human body, the interactions of each of its organs, and the long-term effects of drugs on those humans remain beyond the scope of these models. Thus, even with the development and improvements of these models to simulate the human body, the use of a combination of approaches to research is likely to remain necessary in scientific research and development of new drugs.<\/p>\n\n\n\n

References<\/p>\n\n\n\n