FDA adopts groundbreaking thinking

In a landmark shift, the US Food and Drug Administration (FDA) has announced plans to phase out traditional animal testing requirements in drug development [1]. This decision reflects a growing consensus that alternative methods—collectively termed new approach methodologies (NAMs)—can offer more ethical, efficient, and potentially more human-relevant means of evaluating drug safety and efficacy. These methods encompass a range of innovative techniques.

Traditional animal testing has long been the cornerstone of preclinical drug evaluation. However, concerns have mounted regarding its ethical implications, cost, time consumption, and translational relevance to human biology. Studies have shown that animal models often fail to predict human responses accurately, leading to high attrition rates in clinical trials [2]. Moreover, the maintenance and care of laboratory animals contribute significantly to research expenses. The rising costs of preclinical animal studies have been driven by increasing ethical scrutiny, supply chain constraints, and soaring prices of non-human primates. For example, the cost of a single cynomolgus monkey—often used in toxicology and pharmacology studies—has surged from approximately $5,000 in 2019 to over $30,000 in 2024 due to shortages and heightened global demand. These expenses inflate significantly when you add housing and experimentation, escalating the cost of the early stages of drug development [3]. The standard animal test costs $1.4 million [4]. This has prompted pharmaceutical companies to seek alternatives such as in vitro systems and AI-driven models. The escalating financial burden underscores the urgency for validated NAMs to reduce reliance on animal models.

Recognising these challenges, the FDA has initiated a paradigm shift. This month, the agency announced its plans to reduce, refine, or potentially replace animal testing requirements with NAMs, particularly for monoclonal antibody therapies and other drugs [1]. This initiative aims to enhance drug safety, reduce development costs, and expedite the availability of treatments while minimising animal use. The FDA will encourage the submission of non-animal safety data for human trial applications and offer incentives like streamlined reviews for robust non-animal evidence. 

New approach methodologies

These methodologies will comprise a suite of cutting-edge techniques designed to assess drug safety and efficacy without relying on animal data. They will include:

• In vitro cell cultures and organoids: Laboratory-grown human cells and organoids (miniature, simplified versions of organs) provide platforms to study drug responses in a controlled environment. These systems can mimic human tissue architecture and function, offering insights into drug toxicity and metabolism [5].

• Organ-on-a-chip technology: Microfluidic devices that simulate the physiological responses of human organs. These chips can replicate blood flow, mechanical forces, and cellular interactions, enabling the study of complex organ-level responses to drugs [6].

• Computational models and artificial intelligence: Advanced algorithms and machine learning techniques can predict drug behaviour, toxicity, and efficacy based on chemical structure and biological data. These models can rapidly screen compounds and identify potential risks before clinical testing [7].

AI-Based computational models

One of the most promising developments in NAMs is the application of AI to generate synthetic toxicology data. The FDA’s National Center for Toxicological Research has developed AnimalGAN, a generative adversarial network (GAN) model that creates synthetic animal study data. By training on existing animal data, AnimalGAN can predict toxicological outcomes for new compounds, potentially reducing the need for live animal testing. Studies have demonstrated that synthetic data from AnimalGAN can be used in toxicity assessments, mechanistic studies, and biomarker development, offering a viable alternative to traditional methods [8]. 

Organs-on-a-chip and organoids

Organ-on-a-chip and organoid technologies have gained traction as physiologically relevant models for drug testing. These systems can replicate the microarchitecture and functions of human organs, allowing for the study of drug effects in a human-like context. For instance, liver-on-a-chip devices have been used to assess drug-induced liver injury, a leading cause of drug withdrawal from the market. By providing more accurate predictions of human responses, these technologies can improve the safety and efficacy of new drugs.

Regulatory and Legislative Support

The FDA’s shift towards NAMs is supported by legislative changes. The FDA Modernisation Act 2.0, signed into law in December 2022, permits the use of alternative testing methods, including cell-based assays, organoids, and AI models, for drug safety assessments. This legislation signifies a major shift, allowing preclinical studies to adopt methods that may more accurately predict human responses based on the latest scientific advances rather than being solely dependent on animal testing. 

Challenges and Future Directions

While NAMs offer numerous advantages, challenges remain in their widespread adoption. Key issues include:

• Validation and standardisation: Ensuring that NAMs produce reliable and reproducible results is crucial for regulatory acceptance. Standardised protocols and validation studies are needed to establish confidence in these methods.

• Complex biological interactions: Some aspects of human biology, such as immune responses and chronic toxicity, are challenging to replicate in vitro or in silico. Further research is needed to develop models that can capture these complex interactions.

• Regulatory harmonisation: Global regulatory agencies must align their guidelines to facilitate the acceptance of NAMs across different jurisdictions. International collaboration and consensus-building are essential for the global implementation of these methods.

Conclusion

The FDA’s decision to embrace NAMs marks a transformative moment in drug development. By reducing reliance on animal testing, the agency aims to enhance the ethical standards, efficiency, and human relevance of preclinical studies.

This change aligns with the FDAs projected a significant increase in the approval of gene and cell therapy products over the next 5 years. This forecast is based on the substantial growth in investigational new drug (IND) applications for regenerative medicines, with the agency receiving over 800 INDs and expecting more than 200 additional submissions each year. In these cases, while animal models have previously predicted pharmacodynamic effects, they can fall short in forecasting human-specific toxicities, especially for immunomodulatory and gene therapies [9]. To accommodate this surge, the FDA plans to expand its review staff and issue new policy guidance to expedite the development and approval processes for these innovative treatments.

While challenges remain, the integration of AI-based models, organoids, and organ-on-a-chip technologies holds great promise for the future of medicine. Continued investment in research, validation, and regulatory harmonisation will be key to realising the full potential of these innovative approaches.

References

1. US Food and Drug Administration. AnimalGAN Initiative. [https://www.fda.gov/about-fda/nctr-research-focus-areas/animalgan-initiative](https://www.fda.gov/about-fda/nctr-re

2. Bailey J, et al. Limitations of animal studies for predicting toxicity in clinical trials: is it time to rethink our current approach? J Transl Med. 2020;18(1):1-12. doi:10.1186/s12967-019-2368-6

3. Nonhuman Primate Models in Biomedical Research: State of the Science and Future Needs. https://www.ncbi.nlm.nih.gov/books/NBK593002/?utm_source=chatgpt.com

4. In situ with Thomas Hartung. https://www.chemistryworld.com/culture/thomas-hartung-i-am-not-a-funny-guy/4012692.article

5. Matsui, T, Shinozawa, T. (2021). Human Organoids for Predictive Toxicology Research and Drug Development. Frontiers in Genetics, 12, 767621. https://doi.org/10.3389/fgene.2021.767621 

6. Ronaldson-Bouchard K, Vunjak-Novakovic G. (2018). Organs-on-a-chip: A fast track for engineered human tissues in drug development. Cell Stem Cell, 22(3), 310-324. https://doi.org/10.1016/j.stem.2018.02.011

7. Badwan BA, et al. (2023). Machine learning approaches to predict drug efficacy and toxicity in oncology. Cell Reports Methods, 3(2), 100413. https://doi.org/10.1016/j.crmeth.2023.100413

8. Chen X,  et al. A generative adversarial network model alternative to animal studies for clinical pathology assessment. Nat Commun 14, 7141 (2023). https://doi.org/10.1038/s41467-023-42933-9

9. Polson AG, et al. (2012). The successes and limitations of preclinical studies in predicting the pharmacodynamics and safety of cell-surface-targeted biological agents in patients. British Journal of Pharmacology, 166(3), 823–846. https://doi.org/10.1111/j.1476-5381.2012.01916.x