Predicting Toxicity in Drug Development

Toxicology studies are employed in drug development to evaluate the safety and viability of potential drug candidates, and the data from such studies are used to support regulatory submissions. As it is widely known that toxicity is a contributing factor to drug candidate attrition and with demand for faster and cheaper drug development increasing, researchers are looking at ways to perform toxicology studies as efficiently as possible.

To learn more about toxicology studies, including how accelerated timelines impact these studies, the expectations and requirements of regulatory bodies, issues surrounding animal models, and possible alternative assays, Pharmaceutical Technology spoke with Steven Bulera, PhD, DABT, corporate vice-president, global head of Toxicology, Charles River Laboratories.

Patient protection

Pharm Tech: Could you elaborate on the importance of toxicology studies?

Bulera (Charles River): When developing a new therapeutic, the new therapy must demonstrate two simple criteria. First, the therapy must be efficacious: it must treat the disease or improve the quality of life of the patient. Second, the therapy must be safe and not cause additional and/or unacceptable harm to the patient. Toxicology studies are designed and conducted to identify and characterize any potential adverse events or toxicities prior to the new drug being administered to patients or volunteers participating in clinical trials (1). To accomplish this task, a variety of nonclinical toxicology studies are performed in non-human in-vivo and in-vitro models. Additionally, these studies are required for regulatory submissions around the world and support ‘first-in-human’ clinical trials and clinical trials needed to bring the drug to the market and to patients in need.

In addition to characterizing potential adverse events, toxicology is important in that the studies conducted help identify and remove potentially ‘bad’ drugs from development, which protects patients from potentially harmful drugs and removes drugs from development before large financial investments are made in drugs that have limited or no potential to be a successful new therapy.

Types of studies

Pharm Tech: What type of toxicology studies are performed in drug development?

Bulera (Charles River): The types of studies run include general toxicology (various durations), genetic toxicology, safety pharmacology, reproductive toxicology, and carcinogenicity bioassays. Additionally, other specialty studies may be run to identify abuse/seizure liabilities, musculoskeletal toxicities, neuro-toxicities, ocular toxicities, and/or phototoxicities that could potentially be caused by the new therapeutic. The in-vitro studies are run in a variety of cell types, and the in-vivo studies are typically run in two species (a rodent and nonrodent species). In general, regardless of the new drug modality (small molecule, monoclonal antibody, gene therapy, oligonucleotide, protein, etc.), the study designs and endpoints are similar.

Where differences are seen are with the specific biomarkers that are included in the study and are based on the specific type of molecule and the target of the new drug. It is important to mention and acknowledge that while many of the toxicology studies use animals, the studies are designed with the 3Rs (reduce, refine, and replace) in mind and that the animals are treated humanely with the highest standards of animal welfare in mind.

Accelerated timelines

Pharm Tech: How might accelerated development timelines impact toxicology studies?

Bulera (Charles River): When there is an unmet medical need, the question that is always on a researcher’s mind, is ‘how fast can we get this drug to the patients that need it?’ A common trend in drug development is ‘how do we go faster?’ To support accelerated programs, toxicology has developed faster ways to generate, analyze, and interpret data, as well as faster ways to create the reports needed for regulatory submission. Additionally, some of the best practices that can be called upon involve strong partnerships between clients and contract research organizations (CROs) so that both parties are working together to remove white space from study initiation and reporting timelines. Milestones can then be reached more rapidly.

Also, companies have developed paradigms that require decisions to be made at risk or in parallel with the reporting process to move to the next steps faster. Often, companies wait for final reports to make decisions. To accelerate the toxicology program, companies are encouraged to make decisions on data, and then start the next study/phase before the previous study report is finalized. This [method] can remove weeks if not months from a toxicology program.

Regulatory expectations

Pharm Tech: What are the regulatory expectations and requirements for toxicology studies in drug development, both for a regular timescale and an accelerated one?

Bulera (Charles River): The vast majority, if not all, toxicology studies used for regulatory submissions are run under good laboratory practice (GLP) regulations. These are a series of regulations required by countries that dictate how studies must be conducted to ensure that the studies were conducted with the highest scientific quality. Additionally, these regulations have documentation expectations so that regulatory agencies reviewing the toxicology study package can know how the studies were conducted and that they can accurately recreate the study from what was documented.

Regardless of whether the new drug was conducted under a standard or accelerated timeline, countries have the expectation that the studies were conducted under the auspices of the GLPs. Where there is some potential flexibility for the toxicology program, is what individual studies are needed and the duration of those studies that are required to move the drug to the next development milestone. There are examples for rare and orphan disease therapies, where regulatory agencies have been open to alternative toxicology study designs and customized toxicology packages so that these drugs can get to patients that need them faster.

Animal toxicology models

Pharm Tech: Could you run through some of the issues surrounding animal models for toxicology studies and what possible solutions there may be in the pipeline or already being used that could benefit industry for wider adoption?

Bulera (Charles River): Some of the common reasons cited as to why whole animal toxicology models are not appropriate to use in drug development are that animal models do not reliably predict human outcomes, drugs that appear to work well in animal models fail to work when given to humans, and having to use animal models and run animal studies delays the drug development process. Numerous articles exploring the limitations of these models have been published (2–4).

Alternatives to the use of animal models are also continually being discussed and developed. We have seen advances made using in-vitro and ex-vivo models (1D, 2D, 3D cell cultures, tissue slices, organs-on-a chip). As an example, for many years chemical irritancy was tested by administering compounds to the eyes of rabbits (Draize eye irritation assay). However, that assay has been replaced by the ex-vivo Bovine Corneal Opacity and Permeability (BCOP) assay that only uses animal corneas that normally would have been discarded.

Also, researchers continue to look at alternative models for testing the toxicity of compounds. Khabib et al. outline the advantages and disadvantages of using alternative models such as brine shrimp, zebrafish, daphnia, copepods, molluscs, and so on (4). These in-vitro, ex-vivo, and alternative animal models are useful for identifying if a compound is toxic, can be used as screens to rank order compounds from more to less toxic, and can be used to determine if a compound impacts a specific cell type or a specific cellular pathway. However, they currently do not capture the complex interactions with organ pathways (i.e., communication pathways between cells and organ systems) that are linked together in a whole animal model.

While the predictiveness of animal toxicology models may not be as predictive of human outcomes as researchers would like, history has shown that the testing performed using these whole animal toxicology models has and continues to protect humans/patients from harmful drugs. In addition, the more that is learned about cellular and biochemical mechanisms, and how organ systems operate and communicate, the predictiveness of our models will also improve. Also, as researchers continue to learn about how to better care for and manage research animals, the reproducibility of results will also improve. Even as the predictability and reproducibility of whole animal models improves, researchers will continue to look for assays/methods to replace animal toxicology models in an effort to reduce animal usage (3Rs), bring drugs to ‘in need’ patients faster, and to appreciate the value that these alternate methods bring to drug development.

Alternative assays

Pharm Tech: Might the use of computer simulations (or other disruptive technology) become the industry standard for toxicology studies in the near future?

Bulera (Charles River): As new data are obtained about cellular pathways, organ systems, and cellular communication networks, and are combined with information about the new drug (structure data, similarity to previously tested drugs, absorption, distribution, metabolism, and excretion data, in-vitro and ex-vivo data, etc.), it is conceivable to envision a time when it will be possible to employ artificial intelligence and machine learning algorithms to characterize new drugs completely in silico. There are numerous examples in the literature, where in-vitro, in-vivo, and/or in-silico methods have been used as accepted replacement/substitutes for whole animal models or for examining specific biomarkers. However, as in-vitro, ex-vivo, and in-silico models are developed as replacements/substitutes for whole animal toxicology models, they will need to be thoroughly evaluated before they are accepted and implemented, and the associated whole animal toxicology model discontinued.

One of the main challenges facing the use and implementation of these alternate assays/methods is the validation/predictiveness of these alternate assays/methods. Acceptance of the assays/methods as a substitute by regulatory agencies and by the public will be based on researchers being able to demonstrate that the alternative assays/methods are equivalent or better at predicting human toxicity then the current whole animal toxicity models and that the outcomes of these assays/methods can be used to make drug development decisions that will protect patients and not jeopardize human health and safety.

Additionally, with the concerns about rising drug costs and cost of developing new therapies, researchers will need to be aware of the costs of implementing these alternate assays/methods. Current estimates to develop a drug range from $944 million to as high as $4.54 billion (5). If the cost of these alternative assays/methods substantially increases the cost of a drug’s development, companies will need to acknowledge these costs and still be supportive of these alternative assays/methods.

In the end, the development, implementation, and use of these alternative assays/methods will be a balance of how these assays/methods will decrease drug development timelines, how much faster can needed medicines get to patients, how much better they are at predicting human safety, and the value that these investments/costs bring to the drug development process.

References

1. M.A. Dorato and L.A. Buckley, Curr. Protoc. Toxicol., 31 (1) 19.1.1–19.1.35 (2007).
2. G.A. Van Norman, J. Am. Coll. Cardiol. Basic Trans. Science, 4 (7) 845–854 (2019).
3. G.A. Van Norman, J. Am. Coll. Cardiol. Basic Trans. Science, 5 (4) 387–397 (2020).
4. M.N.H. Khabib et al., Toxicology, 465, 153053 (2022).
5. M. Schlander et al., PharmacoEconomics, 39 (11) 1243–1269 (2021).

About the Author

Felicity Thomas is the European editor for Pharmaceutical Technology Group.