FDA Shifts Away From Animal Testing: Impact on Rheumatology & Drug Development

The Food and Drug Administration is charting a course away from traditional animal testing in biomedical research, embracing new approaches like artificial intelligence and in vitro methods. This shift, initially announced in April 2025, continues to generate discussion about its potential impact on research timelines, drug development costs, and, crucially, animal welfare. While the move is being lauded by some as a scientific and ethical leap forward, questions remain about the readiness of these alternative methods to fully replace established practices.

A Paradigm Shift in Biomedical Research

The FDA’s decision focuses initially on monoclonal antibodies and other medications, signaling a desire for more “human-relevant evidence,” as described by Lena Smirnova, PhD, an assistant professor at Johns Hopkins University. This move is being implemented through a roadmap and bolstered by the FDA Modernization Act 2.0, signed into law in December 2022, which allows for preclinical drug testing using non-animal methods like cell-based assays and computer modeling.

Further legislative support arrived with the Senate’s approval of the FDA Modernization Act 3.0 in December 2025 (currently awaiting House action). This bill aims to update FDA regulations to reflect the new emphasis on nonclinical tests, removing language that implicitly requires animal testing in all cases. According to Smirnova, this regulatory alignment is a critical step in the process.

Scientific Relevance and the Limitations of Animal Models

Sarfaraz K. Niazi, PhD, adjunct professor at the University of Illinois and founder of Adello Biologics, frames the shift as a matter of “scientific relevance.” He points to growing evidence that animal testing often fails to accurately predict a biological drug’s effects in humans. For complex biological drugs, particularly those targeting the immune system, animal models frequently struggle to replicate the nuances of human disease and treatment responses.

The FDA’s roadmap emphasizes integrative combinations of new approach methodologies (NAMs) – such as organ-chip toxicity testing combined with pharmacokinetic modeling and AI-based predictions – to address the comprehensive assessment historically provided by whole-animal studies. AI and machine learning can help identify patterns and predict potential toxicities, but Smirnova stresses the importance of transparency, uncertainty quantification, and clearly defined applicability domains to ensure trustworthy results.

The Role of In Silico and In Vitro Methods

Beyond AI, in silico approaches – computational models and simulations – and in vitro methods – like tissue chips and organ-on-a-chip systems – are gaining prominence. Matthew R. Bailey, president of the National Association for Biomedical Research, highlights advances in cell cultures and computer modeling as paving new pathways for research. However, he emphasizes that these NAMs are not intended to completely replace animal studies, but rather to complement them within an integrated research ecosystem.

Bailey cautions that NAMs currently cannot fully replicate the complexity of a whole living organism. They are most effective when used in conjunction with in vivo research, accelerating drug discovery and safety testing, but not replacing it entirely. Current limitations include a lack of well-defined endpoints, the need for standardization across labs, and incomplete long-term chronic exposure data.

Navigating the Transition: FDA Workshops and Draft Guidance

The FDA and NIH hosted a workshop in July 2025 to explore collaborative strategies for reducing animal research. This was followed by a draft guidance issued in December 2025, outlining ways to reduce the employ of nonhuman primates in monoclonal antibody toxicity studies after an initial three-month testing period. These actions represent significant steps in implementing the FDA’s roadmap.

The agency is also encouraging the use of real-world human data, including international data where regulatory standards are comparable, as well as data from patient registries and post-market surveillance, to reduce redundant animal testing. However, Smirnova notes that real-world data also has limitations, such as confounding factors and potential biases, and should be used as part of a comprehensive, weight-of-evidence approach.

Implications for Rheumatology

The shift away from animal testing has particularly relevant implications for rheumatology, where many therapies target immune mediators. Niazi suggests that the FDA’s new approach may expedite access to novel biologics and biosimilars, enhance the predictability of immune-related adverse events, and promote more rational dose determination.

However, Bailey urges the rheumatology research community to carefully consider the safety implications of reducing nonhuman primate studies, particularly for monoclonal antibodies targeting rheumatic diseases. He notes that newer disease areas with limited research may not have sufficient toxicity data to forego nonhuman primate studies after the initial testing period. Artificial intelligence is also playing an increasing role in rheumatology, offering potential for improved diagnostics and treatment monitoring.

Looking Ahead: Integration and Standardization

The successful implementation of these changes hinges on continued integration of NAMs with traditional animal studies, as well as standardization and validation of NAMs across laboratories and countries. Addressing the limitations of current NAMs, such as their inability to fully represent a whole living organism and the need for better endpoint definition, will be crucial.

While the FDA Modernization Act 3.0 is not a complete prohibition on animal studies, it signals a clear commitment to reducing reliance on animal testing and embracing more human-relevant approaches. The ongoing process of refining and validating these new methods will shape the future of biomedical research and drug development, with the ultimate goal of improving patient safety and accelerating the delivery of innovative therapies.

For more information:

Matthew R. Bailey can be reached at [email protected].

Sarfaraz K. Niazi, PhD, can be reached at [email protected].

Lena Smirnova, PhD, can be reached at [email protected].