ICCVAM Biennial Report 2024-2025

PFAS are organofluorine chemicals that have been manufactured and used for decades in products and applications including firefighting foams, food packaging, and stain- and water-resistant fabrics and carpeting. PFAS are highly persistent and bioaccumulate in the environment, biota, and humans, but understanding PFAS hazards remains challenging. There is no consensus definition of PFAS, and their large variety of chemical structures and toxicological and physiochemical properties makes hazard prediction extremely challenging.

ICCVAM established a workgroup in 2023 to provide expertise in identifying and evaluating NAMs to predict PFAS toxicity. The workgroup includes members from nine ICCVAM member agencies. The group’s charges include:

• Evaluate current PFAS definitions and groupings
• Assess the utility of new approach methodologies (NAMs) currently being used by national and international regulatory agencies for testing and risk assessment of specific PFAS based on case examples of successful application of certain NAMs
• Identify research challenges and data gaps in the use of specific NAMs for testing and risk assessment of different PFAS
• Identify NAMs that have not been used for PFAS and explore opportunities for expanding their application for PFAS testing and risk assessment

The workgroup drafted a white paper that identifies specific challenges and data gaps that need to be addressed for NAMs implementation for PFAS testing. It notes that NAMs encompassing a range of in vitro and computational approaches will be important sources of data for initial characterization and prioritization of PFAS, and makes specific recommendations on how initial efforts should be directed. The white paper is being reviewed by participating agencies and will be published in 2026.

The ICCVAM PFAS Workgroup will host a workshop to address issues identified in the white paper, with findings from the workshop presented in a published manuscript.

Regulatory agencies use data from in vivo fish acute toxicity tests to assess the potential risk of substances to aquatic vertebrates. The ICCVAM Ecotoxicology Workgroup is preparing a review to help guide the prioritization of the development and application of alternatives to the fish acute toxicity test, focusing on regulatory use of fish acute toxicity data and alternatives. The review will be submitted for publication in 2026.

The Ecotoxicology Workgroup asked each agency:

• What does your agency do with acute fish toxicity data?
• What is your flexibility to use NAMs?
• What should the data submitter know about your agency’s process?
• Are there legal or regulatory impediments to the adoption of NAMs? For example, are live-animal data specifically called for in your agency’s regulations?

Responses were included in the draft manuscript, which will also address performance of the fish embryo toxicity test, an alternative test method described in OECD test guidelines.

A review by the DOI’s U.S. Geological Survey and other entities (Bean et al. 2024) critically examines current approaches to testing chemical effects on wildlife and outlines strategies to inform regulatory agencies in their assessments of ecological risks. At present, most regulatory frameworks rely on standardized laboratory tests using a small number of “model species,” such as quail, ducks, and rats, to represent a wide range of wildlife. However, the authors highlight significant limitations to this approach, noting that these species may not adequately capture the diversity of biological responses across other species, including bats, reptiles, and amphibians. Additionally, laboratory conditions often fail to reflect real-world environmental exposures, and there are currently no validated testing protocols for several important terrestrial species groups.

The review also notes that although existing tests use large numbers of animals, the resulting data are not always well-suited for modern statistical analysis or for the prediction of population-level impacts. Emerging technologies, such as computational models and cell-based assays, show promise for improving efficiency and reducing animal use. However, the authors conclude that these methods are not yet ready to fully replace animal models. Instead, they recommend a balanced approach that includes updating existing test guidelines to improve scientific rigor, developing standardized methods for field studies that observe wildlife in natural environments, and addressing key knowledge gaps about underrepresented species. By improving current methods along with continued investment in new technologies, this work aims to enhance the accuracy and relevance of wildlife risk assessments while gradually advancing toward more humane and effective testing strategies.

A comprehensive review from USGS and other entities explores how ecological risk assessments for wildlife can be modernized using emerging scientific tools to improve both efficiency and accuracy while reducing reliance on animal models (Rattner et al. 2024). Traditionally, ecological risk assessments have depended on measuring survival, growth, and reproduction in laboratory animals, which can be time-consuming, costly, and limited in ecological relevance. This review highlights several promising alternatives, including molecular and cellular methods such as high-throughput screening and gene expression analysis, that can detect early biological changes linked to toxicity. It also emphasizes the value of population modeling, which integrates multiple data sources to predict how chemical exposures affect entire wildlife populations rather than just individual organisms.

A key framework discussed is adverse outcome pathways, which connect molecular-level disruptions to broader biological and ecological effects to help scientists understand how harm progresses across biological scales. Advanced statistical tools such as Bayesian networks and improved dose–response models further enhance predictions by accounting for uncertainty and complex environmental stressors. The review demonstrates practical applications, such as modeling how behavioral effects of lead exposure influence population outcomes in herring gulls (Larus argentatus). These applications show how subtle, nonlethal effects can still have significant ecological consequences. Regulatory adoption of these advances remains limited due to the need for stronger validation linking early biological signals to population-level impacts. Overall, this review provides a roadmap to support integration of innovative methods into ecological risk assessments, with the goal of better protecting wildlife while supporting more humane and scientifically robust risk assessment practices.

Juvenile animal studies are traditionally used to evaluate the safety of drugs for children during the drug development process. However, as NAMs advance, it is important to examine the roles and outcomes of these studies in pediatric drug development and explore how NAMs could complement or improve these safety assessments. A retrospective study by FDA researchers and collaborators reviewed the data of 74 approved pediatric drug products that included juvenile animal study findings in their prescribing information (Abulwerdi et al. 2025). Among the products reviewed, 83.8% used one animal species for testing, with rats being the most commonly studied. Using guidance from the S11 Nonclinical Safety Testing for Pediatric Pharmaceuticals, researchers analyzed key factors, such as the youngest age group targeted and the clinical treatment duration. The findings showed that most drugs (64.9%) were intended for children and adolescents, while nearly half (51.4%) were for short-term use or single treatments. Notably, the results of juvenile animal studies rarely led to safety warnings in drug labels; only 8.1% (6 out of 74) of the reviewed medications included warnings related to pediatric use based on juvenile animal study findings. This suggests that while juvenile animal studies play a role in safety assessments, their impact on labeling decisions may be limited. NAMs approaches such as developmental genetics, microphysiological systems (or organ-on-a-chip models), and advanced computational tools offer innovative alternatives to strengthen preclinical safety testing for pediatric drug development. These methods can provide deeper insights into developmental toxicity while reducing the reliance on animal models. This study highlights the importance of NAMs as complementary tools to improve the developmental safety of pediatric drugs.

Systemic toxicity refers to the effect of a toxic substance after it is absorbed and distributed throughout the body. Traditionally, information on the systemic toxicity of a chemical has been obtained from animal experiments assessing nonlocalized toxicity, or toxicity to organ systems that are not the point of exposure. To better understand the existence and availability of NAMs to assess systemic toxicity, NIEHS and the International Collaboration on Cosmetics Safety conducted a landscape analysis by searching the available public literature. The landscape analysis focused on these questions:

• What are the current and past efforts to develop and refine NAMs for systemic toxicity? 
• What are current opportunities to validate mature systemic toxicity NAMs to support their regulatory use and market adoption? 
• What NAMs have been developed in support of decisions specifically addressing the question of human health risk?

An initial query was developed from a list of terms considered relevant to systemic toxicity NAMs. This query returned nearly 200,000 relevant papers when used to search PubMed, Web of Science, and SCOPUS for publications between 2019 and 2024. Supervised and unsupervised clustering approaches and keyword analysis were used to identify 8,418 studies for further screening. This screening employed a machine learning process that predicted what references were relevant for human review, yielding a set of 3,048 references that were predicted to represent at least 75% of available studies that were relevant to the landscape focus.

Key conclusions from the review of the final reference set are as follows: 

• The development of NAMs has predominantly targeted key systems associated with systemic toxicity: hepatic, developmental, respiratory, nervous, and reproductive systems. NAMs assessing toxicity in these systems are well-represented in the literature.
• Despite their relevance to systemic toxicity, gastrointestinal, cardiovascular, and renal systems are less frequently addressed in current NAMs.
• In vitro assays commonly utilize human-derived cell types, underscoring a strong emphasis on human relevance. However, many studies also incorporate nonhuman cells, and how these are being applied needs to be better understood.
• Many studies do not explicitly reference traditional animal tests as direct replacements, indicating that NAMs are often developed as complementary tools rather than one-to-one substitutes.

The International Collaboration on Cosmetics Safety will develop a paper on the landscape analysis in 2026.

The NIH Common Fund’s Complement Animal Research in Experimentation (Complement-ARIE) program was established in 2024 to speed the development, standardization, validation, and use of human-based NAMs. To ensure that the program focuses on areas of science with the greatest need and the best opportunities for human-based model development, NICEATM conducted a landscape analysis to collect information on ongoing efforts in the NAMs field. The analysis included a survey of in vitro, in chemico, and in silico approaches that have the potential to improve understanding of human health and disease mechanisms, reduce reliance on animal models, and improve the efficiency of animal use. To ensure a rapid and comprehensive approach, generative super intelligence (SI) and other computational methods were leveraged, supplemented with subject matter expertise. Use of generative SI was estimated to reduce the labor hours and overall time required to complete the project by 90–95% compared to traditional approaches. Data associated with and generated by NAMs were also evaluated to determine the extent to which they supported findability, accessibility, interoperability, and reusability (FAIR) principles and were SI-ready. This evaluation included considerations for a suitable data ecosystem and analysis of currently available infrastructure, including existing data centers and repositories, that can be leveraged.

The results demonstrate that NAMs work has mainly focused on in vitro and in silico studies, with fewer in chemico studies or general development of methods such as high-throughput screening assays. Within the in vitro category, stem cell and 3D cell culture were the best represented categories. In chemico analyses have focused on biochemical assays or structure/identity analysis methods. In silico studies were generally focused on the development of SI/machine learning models and molecular simulation. The topics with the highest representation examined both cell and tissue type, human physiology considerations, method information including detailed descriptions of new or improved technologies, and technological challenges. Diseases, metabolic capabilities, and method limitations were also addressed, but to a lesser extent. Areas that were seldom reported or addressed in these publications included regulatory application, workforce or ethical considerations, population diversity and variability, and economic considerations. The review of databases for adherence to FAIR principles indicated that databases that incorporated curation/harmonization steps, detailed reporting criteria or forms tended to be more FAIR-compliant, indicating that establishment of metadata reporting guidelines or standardization will enable more FAIR resources.

A study by researchers in the FDA Center for Drug Evaluation and Research (CDER; Dao and Sadrieh 2026) reviewed the extent of NAMs submissions in drug applications submitted to the FDA over the past 15 years. The investigation used a custom search tool to analyze Module 4 of the Electronic Common Technical Document and focused on five NAM categories: in vitro (lab-based tests), in silico (computer-based models), in chemico (chemical interaction testing), nonhuman in vivo (testing on species such as zebrafish [Danio rerio]), and other combined methods.

The results showed that most NAM submissions fell into two categories: in silico (49%) and in vitro (44%). In vitro tests, including stem cell-based models, cocultures, and sandwich cultures, constituted the majority of in vitro NAM study reports, while 3D models and advanced organ-on-a-chip systems were less common. In silico tests were widely used to study metabolites and impurities and predominantly consisted of QSAR studies. By comparison, in chemico methods and nonhuman in vivo tests such as zebrafish studies were represented less frequently. It should be noted that the number of NAMs study reports submitted to CDER over the past 15 years represented around 0.5% of the total number of study reports, which is a very small fraction of the nonclinical studies in drug applications. These findings highlight the limited extent to which NAM data have been submitted to the FDA. Encouraging further NAM submissions could help build scientific confidence in these approaches, speeding up validation and regulatory acceptance. The study supports CDER’s goal to prioritize NAMs with significant regulatory impacts to achieve the goals of the 2025 FDA NAMs Roadmap, ultimately replacing animal studies when possible and feasible.

The FDA, through the CDER Office of New Drugs (OND), is promoting the use and refinement of NAMs in drug development. In the context of nonclinical testing, FDA considers NAMs to include in vitro, ex vivo, in silico (computational), or in chemico (chemical-based) studies with the potential to improve the predictivity of clinical outcomes, utilize phylogenetically lower animals, or otherwise help replace and reduce animal use and refine nonclinical drug development programs. Studies using NAMs submitted to CDER/OND that are designed to predict drug safety or potential clinical effectiveness will provide important momentum to continue evolving toxicological assessments of new drugs while reducing reliance on animal studies, in line with the principles of replacing, reducing and refining use of animals.

FDA has previously discussed the potential of nonclinical tests using NAMs to improve drug development, identified gaps in traditional testing methods, and highlighted challenges with NAM integration into nonclinical development programs. An OND review (Yao et al. 2026) showcases how some NAMs have already been validated and adopted into standard drug assessment practices, becoming internationally recognized as alternatives to animal testing for regulatory decision-making. The publication also highlights some recent examples of products in which studies using NAMs were integral for approval. Perhaps most importantly, in this publication, OND shares insights about reviewer experiences evaluating studies using NAMs submitted by sponsors for regulatory applications, highlighting strengths and limitations of the studies, and the impacts of these studies on regulatory decision-making at the agency. Although NAMs are promising, the publication emphasizes the importance of specifically designing studies to address scientific and regulatory needs, adequately describing contexts of use, justifying biological relevance, including control and reference chemicals, and validating the data. The insights and recommended considerations for the conduct of studies using NAMs aim to help sponsors improve their submissions and align with FDA statutory requirements for drug development. The FDA remains committed to continuing regulatory acceptance of scientifically justified and appropriate studies using NAMs in drug development programs as the science and technology of NAMs advances. By receiving more scientifically robust submissions using NAMs, the FDA aspires to reduce, refine, and, eventually, replace the need for animals in nonclinical testing while protecting public health by ensuring new drugs are both safe and effective. This collaborative effort highlights the FDA’s dedication to fostering innovation and confidence in adopting new and advanced approaches to drug testing.