For decades, in vivo models have been the standard in preclinical testing before a drug moves to first-in-human (FIH) trials. But a recent evolution in regulatory expectations has accelerated a steady shift away from in vivo testing, in favor of New Approach Methodologies (NAMs). This shift follows increased adherence to the principles of the Three Rs model (Replacement, Reduction, Refinement) and has been embraced by regulatory bodies worldwide.
A draft guidance document published by the U.S. FDA has underscored the urgency of this regulatory transition, and left sponsors scrambling to ensure they understand the magnitude of the change. Broadly speaking, the agency is pushing for a prohibition of in vivo testing in favor of NAMs, including 3D models, chemical reactivity testing, and in silico approaches.
However, new methodologies have their drawbacks that make them unlikely to fully replace in vivo testing. Each NAM is designed as a fit-for-purpose assay to address specific toxicological concerns. Use of a NAM should be in the context of the weight of evidence (WOE) and specific concerns that the NAM is designed to answer. This means developers must ensure they have a firm understanding of the emerging technologies in this space, their limitations, and the best strategies for building a hybrid approach that integrates in vivo and in vitro methodologies and achieves regulatory confidence.
An Evolving Regulatory Landscape
In 2022, regulators in the United States introduced the FDA Modernization Act 2.0, which stated that in vivo testing was no longer legally required if alternative approaches could demonstrate safety and efficacy. Since then, the development of new techniques to answer specific questions has increased greatly as the technology behind NAMs has improved.
In April 2025, further guidance was released in the United States, which described plans to increase the use of NAMs, especially when they can enhance human relevance and translational predictability. Regulators also published draft guidance on risk-based, science-driven approaches to justify in vivo testing, especially for certain biologics and monoclonal antibodies.
Changes to the regulatory landscape add several layers of complexity to the process of designing the in vivo and in vitro studies to support non-clinical development. Incorporating NAMs into a drug development program requires increased planning, greater emphasis on design responsibility, and adherence to the decades-old principles that started this shift: the 3Rs.
The 3Rs, Explained
Russell and Burch first defined the 3Rs over 50 years ago in their book The Principles of Humane Experimental Technique. Since then, the principles have been updated frequently but remain a guide for developers seeking to reduce their reliance on in vivo testing.
- Replacement: The original definition of this principle was to avoid or replace in vivo testing models when they would otherwise have been used. In updated guidance, developers are urged to accelerate the use of robust models and tools based on the latest technology and science.
- Reduction: Originally defined as reducing the number of test subjects required for in vivo models, this principle is now more focused on ensuring that this type of testing is appropriately designed and analyzed, robust and reproducible, and adds to the scientific community’s knowledge base.
- Refinement: This principle focuses on reducing the harm and suffering caused by in vivo testing, but its language has been updated more recently to ensure that developers use the latest in vivo technologies to reduce suffering and distress and understand the impact on welfare and scientific outcomes.
Real change in a major industry like drug development requires both push and pull. For decades, the principles of the 3Rs have driven efforts to reduce in vivo testing. Simultaneously, advances in NAMs offer more human-relevant, efficient techniques that are increasingly viable alternatives.
NAMs in Focus
In March 2026, U.S. regulators released draft guidance on the alternatives to in vivo testing models in drug development. The guidance listed NAMs that developers should consider using - each with its own advantages, nuances, and limitations.
3D Models
These include organoids, spheroids, and organ-on-chips, and are advanced culture systems that mimic 3D tissue architecture and, in some cases, organ-level function. They are often more relevant to humans at the tissue level and offer a better representation of cell-cell interactions and microenvironments than in vivo models. These techniques can also model human-specific disease biology, making them an exciting development for personalized medicine. However, challenges remain. 3D models typically offer insights into single cells or organs and cannot capture the complex interaction among multiple organs or the effects of a drug across the whole body.
Chemical Reactivity Studies
These assays assess how compounds or metabolites chemically interact with biomolecules. They’re extremely rapid, low-cost techniques for identifying early hazards and are mechanistically precise. However, they offer no biological context, cannot predict organism-level toxicity or exposure effects, and offer weak standalone predictive value.
In silico Modeling
In silico models are computer-based simulations that predict drug properties, behavior, and toxicity. They are one of the fastest and most cost-effective approaches and can be used to screen vast numbers of compounds prior to testing, with no ethical or resource constraints. But they rely heavily on data quality and assumptions and cannot fully replicate complex biology or unexpected effects.
Other NAMs mentioned by U.S. regulators include complex and 2D in vitro studies, as well as alternative-species research using zebrafish or C. elegans. Despite advances in technology, NAMs cannot be relied upon without rigorous methodological validation, which remains far from achievable in the current scientific testing environment. Because NAMs cannot meet all testing needs, sponsors must adopt a hybrid approach.
An Emerging Hybrid Approach
Drug developers and sponsors should already be preparing to pursue a hybrid approach, combining NAM data with translational modeling, targeted in vivo studies, and bioanalytical context to ensure the best possible testing and enhanced safety and efficacy.
A critical enabler of the hybrid approach is to integrate DMPK (drug metabolism and pharmacokinetics) studies, bioanalysis, and modeling capabilities so they work together to generate coordinated evidence. In hybrid scenarios, exposure, mechanism of action, and effect are interpreted together.
The challenge for developers will not just be to reduce reliance on in vivo studies, but also to enhance human predictability while maintaining IND confidence and regulatory fit.
As part of the FDA’s implementation roadmap and pilot program to increase the use of NAMs, regulators are responsible for tracking the correlation between NAM data and clinical outcomes, the overall reduction in in vivo testing, and the time to successful IND for NAM programs versus traditional methods. These will be key data points as the drug development industry moves forward with this new approach.
A Final Word
The FDA and CDER are working to provide a framework for integrating non-animal methods (e.g., in vitro, in silico, organ-on-chips) into regulatory submissions. Demonstrating human biological relevance, ensuring technical characterization, and showing the NAM is fit-for-purpose are critical when designing and implementing these technologies. Sponsors should not see this as a distant possibility. The shift is already underway and requires proactive adaptation. This means moving towards hybrid development strategies that integrate NAMs with targeted in vivo studies and translational modeling to maximize human relevance and regulatory confidence.
Reducing the use of animals is driving this push towards NAMs, particularly as regulators worry about human translation issues and welfare concerns become more prevalent. The scale of this transition should not be underestimated. It represents a fundamental reconfiguration of how evidence is generated, interpreted, and accepted in drug development.
There will be many challenges around validating and standardizing NAM technologies, but early investment and integrated capabilities will be critical. Sponsors that act now will gain a significant advantage, help shape the emerging regulatory framework, and play a crucial role in reducing the number of animal models used in the name of science. When sponsors engage with regulators to optimize the development and use of NAMs to support the safety and efficacy of new drugs, both drug developers and the regulatory agencies benefit.
About the Author
Kevin Denny, MBA, DABT, is an Executive Technical Director at WuXi AppTec’s Laboratory Testing Division and a board-certified toxicologist. He leads non-clinical and regulatory strategy for IND-enabling programs through late-stage submissions, with deep experience in GLP toxicology, global regulatory pathways, and cross-functional program leadership.