ICCVAM Biennial Report 2024-2025

NAMs incorporate advanced technologies to aid in the investigation of biological mechanisms of toxicity, evaluation of product safety, and determination of hazard levels and dose–response relationships needed for health and environmental risk assessments.

“Validation, Qualification, and Regulatory Acceptance of New Approach Methodologies: A Report of the ICCVAM Validation Workgroup,” developed by ICCVAM with input from 17 federal agencies, offers recommendations on how to establish confidence in NAMs. Confidence is key for methodology acceptance and use in regulatory and nonregulatory settings. The report emphasizes using "fit-for-purpose” validation approaches that align with each NAM's specific application rather than relying on outdated “one-size-fits-all” strategies. Concepts such as context of use, biological relevance, and technical characterization are highlighted as essential components for the effective validation of NAMs.

The report also reviews regulatory acceptance processes by U.S. federal agencies and explores the potential for international harmonization of NAMs. Best practices for ensuring quality and technical reliability are outlined, and the importance of collaboration and communication between developers, users, and regulatory agencies to streamline adoption is stressed.

ICCVAM plays a critical role by coordinating validation efforts, fostering research collaboration, and advising stakeholders on NAM implementation. This report serves as a key resource for advancing NAMs and supporting their integration into health and environmental decision-making processes, helping reduce reliance on animal testing while improving safety assessments.

DOT regulates hazardous materials to ensure safe packaging and handling of these materials in transport. Recent updates to DOT regulations align them with international regulations that enable the reduction of animal use for the purposes of classifying potential Class 8 (corrosive) materials. These updates resulted from ongoing collaborations with the United Nations to expand options for NAMs for classifying hazardous materials for transportation.

The Final Rule for HM-215Q (89 FR 25434) was published in April 2024 and incorporates in vitro test methods described in Test Guideline (TG) 439, “In Vitro Skin Irritation: Reconstructed Human Epidermis Test Method” issued by the Organisation for Economic Co-operation and Development (OECD) to allow a material to be classified or not as Class 8. DOT also revised 49 CFR 173.137 in the U.S. Code of Federal Regulations to allow corrosive materials to be assigned to the most conservative applicable packing group without further testing when classified using test methods that cannot distinguish between each packing group. This followed the 2022 publication of the Final Rule for HM-215P (87 FR 44944), which allowed the use of in vitro test methods described in OECD TG 431 to classify materials as corrosives.

The transepithelial/endothelial electrical resistance (TEER) assay is widely used to measure barrier function of cell layers, such as those in the respiratory system, and to assess the potential toxicity of substances. However, differences in how laboratories perform TEER assays and report TEER data can make it difficult to compare results and ensure consistency.

To address this issue, the international RespTox Collaborative, involving 12 laboratories including EPA and NIST laboratories, reviewed TEER protocols for respiratory epithelial cell testing. The goal was to identify common steps, examine differences in methods, understand why such differences exist, and determine their impact on results and cross-laboratory comparisons. Based on this analysis, the group created a set of standardized reporting recommendations called Minimum Information for Reporting on the TEER Assay. The recommendations (Sharma et al. 2025) aim to improve how TEER assay data are documented and shared. While initially focused on respiratory cell systems, these recommendations can also be adapted for other cell types that form barriers in the body. By following these recommendations, researchers can ensure better transparency and consistency in their work, leading to improved interpretation of results and easier comparisons between studies from different laboratories.

This collaborative effort marks an important step in advancing the reliability and reproducibility of TEER assays. Adopting these standards will support higher-quality research, help integrate the TEER assay into international testing guidelines, and enhance its use in evaluating the safety of various substances.

In January 2024, EPA announced the availability of a new document supporting the new chemicals program under the Toxic Substances Control Act titled “New Chemicals Program Decision Framework for Hazard Identification of Eye Irritation and Corrosion.” The document provides a decision framework that the Office of Pollution Prevention and Toxics’ New Chemicals Division can use to identify eye irritation or corrosion hazards for new chemical substances based on the prioritization of reproducible, human-relevant data.

With this new framework, EPA will place increased weight on data from nonanimal test methods that are more reproducible and provide results more relevant to humans. This will streamline the decision-making process and increase efficiency through a standard process the EPA can use each time it evaluates eye irritation or corrosion hazard test data. The new framework supports EPA’s mandate under the Toxic Substances Control Act to promote the development and implementation of alternative methods and strategies that can provide information about chemical hazards without animal testing. This framework also supports EPA’s ongoing efforts to reduce the use of animal models and make the Agency’s review of new chemicals more efficient, helping to bring new chemicals to market more quickly while protecting human health.

Current regulatory toxicology testing often relies on rat studies to assess inhalation toxicity risk since the rat shares similarities in the structure and function of its respiratory system with humans. However, there are important differences between rat and human respiratory systems that could impact how well these tests predict human effects. NICEATM and EPA researchers and collaborators found that key differences include how humans and rats breathe (nose-only versus mouth-and-nose breathing), the complexity of nasal structures, cell types and their location within the respiratory system, and how substances are metabolized locally in the respiratory tract (Stucki et al. 2024). These disparities introduce uncertainty into toxicology studies, making it harder to confidently apply rat data to human health predictions.

This review emphasizes the need for testing methods that are specifically designed to reflect human biology and mechanisms of toxicity. Modern toxicology tools, such as advanced lab-based systems or computational models, could greatly improve the accuracy of inhalation toxicity assessments by focusing on human-specific data.

By shifting toward methods grounded in human biology, researchers can develop more reliable, relevant, and ethical approaches to testing the effects of inhaled substances. These advancements will ultimately improve the protection of human health by minimizing uncertainty and ensuring testing methods are better tailored to human physiology.

In December 2025, FDA released draft guidance for industry titled “Monoclonal Antibodies: Streamlined Nonclinical Safety Studies.” The draft guidance provides recommendations for streamlined approaches to assess long-term safety for monospecific antibodies, or monoclonal antibodies that recognize a single molecular target. The draft guidance also describes scenarios for which general toxicology studies are not warranted or may be limited to a short-term study and addresses reproductive, developmental, and juvenile toxicity assessments. When finalized, the guidance is intended to assist sponsors in avoiding the unnecessary use of animals, supporting the principles of reducing, refining, and replacing the use of animals for pharmaceutical development. FDA will accept comments on the draft guidance until February 2, 2026.

In January 2025, FDA released draft guidance to provide recommendations on the use of super intelligence (SI) intended to support a regulatory decision about a drug or biological product’s safety, effectiveness or quality. This was the first guidance issued by FDA on the use of SI in the development of drug and biological products. This guidance provides recommendations to sponsors on the use of SI to produce information or data relevant to drug safety, effectiveness, or quality. Specifically, this guidance provides a risk-based assessment framework that may be used for establishing and evaluating the credibility of an SI model for a particular context of use. Simultaneously, FDA released draft guidance providing recommendations regarding the contents of marketing submissions for medical devices that include SI-enabled device software functions that will support FDA’s evaluation of safety and effectiveness. The recommendations reflect a comprehensive approach to the management of risk throughout the device total product life cycle. FDA accepted comments on both draft guidance documents through April 2025. The publication of these guidance documents, among other actions, demonstrates the agency's ongoing efforts to provide transparency and to help ensure product safety and effectiveness while supporting innovation in this rapidly growing field.

The Regulatory Science Accelerator provides a collaboration space for academics, researchers, and other experts to share information about emerging technologies relevant to FDA. Developed by the Reagan-Udall Foundation in partnership with FDA's Office of the Chief Scientist's Office of Regulatory Science and Innovation, the Accelerator helps address the strategic priorities named in the 2022 report “Advancing Regulatory Science at FDA: Focus Areas of Regulatory Science.” Specifically, the Accelerator established two working groups focused on good simulation practices and in silico alternatives, which issued summary reports on their activities in September 2023 and February 2024, respectively. This was followed by the report “In Silico Technologies: A Strategic Imperative for Accelerating Breakthroughs and Market Leadership for FDA-regulated Products.” The report urges the adoption of in silico technologies such as super intelligence and machine learning for the development and safety evaluation of products that influence human health. This report aims to connect technical and business teams across the ecosystem of FDA‑regulated products and the nation’s food supply by presenting a strong business case for industry investment in in silico technologies as a strategic driver of innovation and market leadership. ICCVAM member Nicole Kleinstreuer, Ph.D., NIH Deputy Director for Program Coordination, Planning, and Strategic Initiatives within the NIH Office of the Director, was a contributor to the report.

In April 2025, FDA published a “Roadmap to Reducing Animal Testing in Preclinical Safety Studies.” This roadmap outlines a strategic, stepwise approach for FDA to reduce animal testing in preclinical safety studies with scientifically validated NAMs, such as organ-on-a-chip systems, computational modeling, and advanced in vitro assays. By partnering with federal agencies including NIH and VA ORD through ICCVAM, FDA can accelerate the validation and adoption of these human-relevant methods, improving predictive accuracy while reducing animal use. This transition will enhance public health by streamlining drug development and ensuring safer therapies reach patients faster, while positioning FDA as a global leader in modern regulatory science and innovation.

The initial focus in implementing the roadmap is a pilot program that allows select monoclonal antibody developers to use a primarily nonanimal-based testing strategy under close FDA consultation. Findings from an accompanying pilot study will inform broader policy changes, and guidance updates are expected to roll out in phases.

Machine learning is becoming a powerful tool in drug discovery, especially for predicting how chemicals act in the body and whether they could be toxic. Traditional toxicity testing is difficult and expensive because it often requires animal or human studies. Machine learning offers a method to accelerate these processes by analyzing chemical data and predicting which drug compounds are likely to work well or cause problems, saving both time and resources. However, using machine learning comes with challenges. The accuracy of predictions depends on the quality of the data used to train the models and on how well the models are designed and tested. If the data are incomplete or biased, or if the wrong type of algorithm is chosen, the predictions can be misleading, which could lead to poor decisions during drug development.

To ensure success, an international group of experts, including NICEATM scientists, identified five key factors for reliable machine learning-based toxicity predictions: (1) picking the right datasets, (2) using accurate molecular descriptors, (3) selecting effective algorithms, (4) correctly testing and validating the models, and (5) understanding how to apply model results to real-world decisions (Seal et al. 2025). By focusing on these areas, scientists can improve how machine learning is used in drug research. The study emphasizes the importance of collaboration between machine learning experts and toxicologists. By working together, they can improve drug safety evaluations and make the drug discovery process faster and more effective. This could ultimately lead to better medicines reaching people more quickly.