Animal Models Are Seeing Upgrades to Tackle Complex Molecule Testing

Animal models are widely used in the preclinical testing phase of new active ingredients. Novel approaches for the design and development of in-vivo models has become more relevant to predicting drug safety in patients. However, to optimize an animal model’s ability to predict a corresponding human situation, the models need to be clearly defined, controllable, and as “humanized” as possible (1). Strategies such as using novel technologies—for instance, small interfering RNA for mediating gene knock-down—and systems that enable tight and reliable control of human (trans) gene expression in a time- and tissue-specific manner, are approaches that can optimize animal model prediction (1).

Furthermore, there have been new approaches to designing and developing safety assessments using human in-vitro systems. These in-vitro systems have become increasingly important for both biopharmaceuticals and pharmaceuticals. These tests likely combine human stem cell cultures, or co-cultures, with animal disease models (1).

To further explore the state of animal model approaches in preclinical testing for bio/pharmaceuticals, BioPharm International spoke with Andrew Reaume, president and CEO of Melior Discovery, a US-based provider of in-vivo pharmacology services.

Preclinical importance

BioPharm: What is the importance of animal models in preclinical studies for bio/pharmaceuticals?

Reaume (Melior Discovery): Animal models are generally regarded as providing the highest quality of preclinical proof-of-concept towards efficacy.For most therapeutic indications, positive activity in an animal model is regarded to be a necessity, prior to entering the clinic.

Given the tremendous cost and potential risk towards human subjects, animal models provide not only the confidence that a candidate therapeutic will have efficacy towards a given indication, but even help guide clinical trial design by providing answers to questions such as:What patient population should be initially targeted? What dose regiment will provide optimal therapeutic effect? What primary endpoint is most likely to illustrate therapeutic effectiveness? And so on. These are all examples of questions that may not be answered on a theoretical or mechanistic basis and in which activity in an animal model may provide insights.

BioPharm:What are the main differences between using in-vivo animal models and in-vitro models, especially as pertains to the type of data generated?

Reaume (Melior Discovery): Animal models, as opposed to reductionist in-vitro models, represent a fully intact biological system with all of the complexities inherent in evaluating test articles in human subjects. For example, in psychotherapeutic active indications, such as schizophrenia, depression, and attention deficit disorder, in-vitro models simply do not provide a modeling of the complex neuronal connections and associated pathologies of their disorders needed for any kind of efficacy evaluation.

Even for disease states that are more readily evaluated with in-vitro systems, such as the anti-proliferative effects on human cells in culture for an oncology candidate, in-vitro systems do not provide a holistic picture of test article effects that factor in pharmacokinetic/pharmacodynamic relationships of a drug, effects of the drug on a target cell population in the context of surrounding tissue effects, and toxicity effects, among other important factors.

Data and innovations

BioPharm:What meaningful data are generated from in-vivo and in-vitro animal models, and how are the data applied to the development program of a new drug candidate (specifically a large-molecule drug)?

Reaume (Melior Discovery): Cell culture models can be used to reveal target engagement—that the test article is modulating its intended target in the way that it was intended to.

Animal models provide a bigger picture view of test article effect—that the test article is not only modulating the intended target, but that this in turn is having the downstream therapeutic consequences that are intended. Moreover, they provide a view of the potential therapeutic benefit in the context of a safety and tolerability profile (i.e., are the therapeutic effects provided without undo toxicity effects in other organ systems).

BioPharm: What are some of the latest innovations (technology, methodology, approach) in animal model studies, specifically applied to large-molecule preclinical studies?

Reaume (Melior Discovery): Advances have been made recently in the area of immune-oncology, which also represent some of the latest innovation in animal model studies. Animal models involving patient-derived xenografts (PDX) and immune modulatory therapeutics, such as chimeric antigen receptor T cell (CAR-T) and programmed cell death ligand 1 (PDL-1) are examples of the types of studies which we are able to only now routinely approach but were almost unheard of 10 years ago.

Many humanized genetic mice models have been developed to evaluate immunotherapy in PDX models.For example, humanized CD34+ mice (hu-CD34) are a robust in-vivo platform for analyzing the safety and effectiveness of potential new drugs to modulate specifically the human immune system. Also, models engrafted with human cord blood-derived hematopoietic stem cells develop multi-lineage engraftment and display robust human T-cell maturation and human T-cell dependent inflammatory responses. In addition, an improved human myeloid and natural killer lineage development is demonstrated in humanized NSG-SGM3 and NSG-Tg(IL15) mice.

Alternatives to animal models

BioPharm: Are virtual models, either of animals or of human tissues/organs, comparable to in-vivo/in-vitro animal studies? In other words, could virtual models be an alternative for animal models?

Reaume (Melior Discovery): It is our firmly held belief that, given the tremendous complexity of biology, there is far more that we do not know about living systems than what we do know, in spite of our great advances in understanding the physiological and biochemical underpinnings of many disease states. As such, we will not be able to model disease adequately with artificial intelligence (AI) in the foreseeable future. For example, the challenge of computationally predicting whether a drug will be effective in a given disease far, far, exceeds that of having a computer autonomously and safely drive an automobile through the real-world environment.Therefore, animal models will remain a core tool necessary for advancing therapeutics into the clinic for many years to come.

BioPharm: Are there other alternatives to animal models that would generate meaningful preclinical study data for a new drug development program (large molecule or small molecule), and if so, what are these alternatives?

Reaume (Melior Discovery): Not really. For the reasons stated earlier, computational approaches such as AI are not yet able to model the immense complexity of intact biological organisms. While AI approaches are used, the predictive quality towards efficacy of a test article almost invariably will not match that of an in-vivo system. The same can also be said of in-vitro (cell culture) model systems.

In some instances where a terminal disease involves antibodies to human-specific antigens (i.e., the therapeutic test article does not cross-react with a target in animals), there may not be a practical way to evaluate the test article in an animal model.In these instances, given the life-and-death nature of a disease in which no alternative therapeutic candidate is available, test articles may bypass the use of animal models.

Reference

1. J.T. Koppele and R. Witkamp, “Use of Animal Models of Disease in the Preclinical Safety Evaluation of Biopharmaceuticals,” in Preclinical Safety Evaluation of Biopharmaceuticals: a Science-Based Approach to Facilitating Clinical Trials, J.A. Cavagnaro, Ed. (John Wiley & Sons, Hoboken, New Jersey, 2008), pp. 293–308.

About the author

Feliza Mirasol is the science editor for BioPharm International.