The pharmaceutical industry is at the threshold of a historic transformation in how medicine safety is evaluated. For decades, animal models were regarded as the “gold standard” for supporting efficacy and safety prior to human clinical trials. However, the clinical failure rate for new therapies remains high, with approximately 90% of drugs failing during clinical development—often due to a lack of efficacy or unforeseen toxicity in humans. This realization has driven a global movement toward New Approach Methodologies (NAMs): a suite of in vitro, in silico, ex vivo, and in chemico tools designed to improve predictive toxicology by centering on human biology.
The Current Regulatory Landscape: From Mandates to Modernization
Regulatory bodies are no longer just observing NAMs; they are actively building the infrastructure to accept them as pivotal evidence.
1. The U.S. FDA: A Landmark Pivot The regulatory environment in the United States shifted fundamentally with the passage of the FDA Modernization Act 2.0 in late 2022. This legislation clarified that non-animal testing data—including cell-based assays and computer models—can support an Investigational New Drug (IND) application or a Biosimilar Biologics License Application (BLA). Following this, the FDA released a strategic roadmap in April 2025 to reduce reliance on animal testing, particularly for monoclonal antibodies. By March 2026, the FDA issued a comprehensive Draft Guidance outlining the general considerations for the use of NAMs, established a validation framework to facilitate their broader integration into regulatory packages.
2. The EMA and the 3Rs Strategic Goal The European Medicines Agency (EMA) has similarly institutionalized the “3Rs” (Replacement, Reduction, and Refinement) as a core strategic goal. Horizon scanning reports from February 2025 indicate a massive upward trend in NAM-related research, with publications growing from just 182 between 2009–2012 to over 16,700 between 2021–2024. The EMA utilizes its Innovation Task Force (ITF) as a primary entry point for developers to discuss NAM innovation early in the pipeline, with liver, brain, and heart tissues being the most frequently represented models in recent requests.
3. International Harmonization Because most new medicines are developed for a global market, international cooperation through the ICH and OECD is essential. The International Medicines Regulators’ Working Group on 3Rs (IMRWG3R)—which includes the FDA, EMA, Health Canada, and Japan’s PMDA—is currently working to align regulatory acceptance criteria globally to prevent a “validation gap” that could hinder drug approvals across different regions.
The Pillars of Regulatory Acceptance: A Framework for Industry
The FDA’s 2026 guidance establishes a rigorous validation framework that every pharmaceutical company should internalize. Scientific confidence in a NAM is built upon four essential pillars:
• Context of Use (COU): This is the foundation of any regulatory submission involving a NAM. Developers must clearly define the intended regulatory purpose—such as supporting patient monitoring, supporting dosage selection, identifying off-target effects, or justifying the omission of an animal species that adds no value.
• Human Biological Relevance: The model must demonstrate a clear relationship to human physiology. This involves justifying the cell types used (e.g., hepatocytes, Kupffer cells), the cellular architecture (e.g., organoids vs. 2D cultures), and ensuring the model mimics the specific human biological mechanisms or developmental windows being investigated.
• Technical Characterization: Regulators require evidence that the platform is robust, reliable, and reproducible. This includes defining assay stability over time, characterizing biological variability (such as donor cell phenotypes), and ruling out technical interference, such as drug absorption into device materials.
• Fit-for-Purpose: A NAM is fit-for-purpose if it assists in regulatory decision-making. It must either offer a viable alternative to traditional methods, fill a data gap where animal models are insufficient (e.g., human-specific mechanisms), or complement existing findings to support a Weight of Evidence (WOE) approach.
A Mature Case Study: PBPK Modeling
Physiologically Based Pharmacokinetic (PBPK) modeling serves as a prime example of a NAM that has already reached mainstream regulatory status. Between 2020 and 2024, PBPK models were included as pivotal evidence in 26.5% of FDA-approved new drug applications. It is primarily utilized for predicting drug-drug interactions (DDI) (81.9% of cases) and informing dose recommendations for special populations, such as pediatrics. The success of PBPK modeling stems from its ability to establish a “complete and credible chain of evidence” from in vitro data to clinical predictions.
What the Industry Should Do: A Strategic Roadmap
The transition to a NAM-centric model requires more than just new technology; it requires a shift in R&D strategy.
1. Engage Early and Often: Both the FDA and EMA strongly encourage “early interaction”. Developers should not wait for a complete data package but should instead utilize the EMA’s Innovation Task Force or request pre-IND meetings with the FDA to discuss indication-specific NAM strategies.
2. Define Context of Use Specially: Regulators prioritize specificity over broad claims of versatility. Before investing heavily in a new platform, define exactly what drug development question it is intended to answer.
3. Adopt a Weight of Evidence (WOE) Approach: NAMs are currently most successful when they complement existing knowledge. Use them to provide mechanistic insights that animal models cannot offer, thereby strengthening the overall safety case.
4. Invest in Parallel Validation: Where animal data are still deemed necessary, industry should submit NAM data in parallel. This “dual data-package” approach allows regulators to build confidence in the new methods by comparing them to classical results.
5. Prioritize Diversity in Human Models: One of the greatest opportunities of NAMs is the ability to represent diverse human populations. Ensure your cell-based models include diverse genomic backgrounds to avoid the medical inequalities often found in traditional, less-representative animal strains.
6. Leverage AI-Integrated Toxicology: The synergy between NAMs and Artificial Intelligence (AI) is expected to revolutionize drug safety. AI-based computational models of toxicity are already being used to predict side effects based on molecular composition, and as AI continues to fill data gaps in modeling structures (like PBPK), the path to First Time in Human (FTIH) trials will become faster and more ethical.
Conclusion
The era of relying solely on traditional animal models is concluding. Global regulatory bodies have provided the roadmap; it is now up to the industry to adopt these human-relevant tools. By focusing on rigorous validation, clear contexts of use, and early collaboration, we can create a future where drug development is safer, faster, and fundamentally more human.