Continuing Education

Chair(s):
Alexandre Borrel, Sciome; and Kamel Mansouri, NIEHS/NICEATM.

New approach methods (NAMs), such as quantitative structure-activity relationship (QSAR) models and read-across (RAx) approaches, are intended to rationalize and reduce the need for further experimental testing on animals. Both these approaches rely on the principle of molecular similarity for assessing property profiles of untested chemicals. Software tools to visualize chemical structures and their experimental and predicted properties employing principles of molecular similarity can make the process of reliably assessing the profiles of untested compounds efficient. This course will explore the concept of chemical similarity and chemical space and their application for toxicity assessment.

The first speaker will introduce the concept of similarity, covering both supervised and unsupervised methods. The speaker will discuss the context of toxicity studies, detailing common errors and best practices for effectively using similarity approaches.

The second speaker will introduce OrbiTox and ChemMaps.com, presenting them as two powerful tools designed for exploring chemical space. The speaker will explain how these platforms provide access to toxicity data and QSAR models, facilitating the reporting of analogue profiles to support decision-making.

By the conclusion of this course, participants will possess a better understanding of chemical similarity and chemical space in relation to toxicity assessment. They will be equipped with the necessary expertise to efficiently navigate ChemMaps and OrbiTox for effectively exploring chemical structure and property landscapes. This course will empower attendees with practical skills and insights, fostering informed decision-making and enhancing proficiency in utilizing cutting-edge tools for chemical analysis and toxicological assessment.

Chair(s):
Fenna Sillé, Johns Hopkins University; and Emanuela Corsini, Università degli Studi di Milano, Italy.

The developing immune system is particularly vulnerable to environmental exposures during critical windows from prenatal development through adolescence. These exposures can lead to persistent immune dysfunction, including increased risk of allergy, autoimmunity, infection susceptibility, and chronic inflammation. However, traditional in vivo testing methods for developmental immunotoxicity (DIT) are resource-intensive, lack human relevance, and fail to capture nuanced mechanisms across key developmental windows. This course offers an in-depth and timely look at emerging new approach methodologies (NAMs) that address these issues and have the potential to replace or reduce reliance on animal testing for protection of the developing immune system.

Participants will strengthen existing knowledge on how in vitro platforms, including stem cell–derived hematopoietic assays and immune organoids, are being developed and optimized to model early immune development. Novel technologies, such as bone-marrow-on-a-chip and lymphoid organoids, provide physiologically relevant systems to assess developmental immune outcomes, offering higher resolution and mechanistic insight. Molecular frameworks such as adverse outcome pathways (AOPs) are also gaining traction in the field of DIT and serve as important organizational tools for integrating mechanistic knowledge and guiding regulatory strategy.

This course will highlight critical components necessary to strengthen existing knowledge of NAMs into validated tools for hazard characterization and identification and regulatory use. These include the development of reliable functional endpoints, the identification and use of appropriate reference compounds, and consensus criteria for assay readiness. Attendees will also gain insights into the evolving regulatory landscape that increasingly emphasizes human relevance, transparency, and scientific confidence in NAMs for health risk decision-making.

The course will include a brief concluding panel discussion, with an interactive Q&A with all speakers and the audience on remaining challenges, harmonization, and how the field should advance beyond 2026.

By bringing together international leaders in toxicology, immunology, and regulatory science, this course aims to build a shared understanding of where the field stands today and where strategic investment is needed to ensure that DIT NAMs are accepted and implemented to improve public health protection.

Chair(s):
Zhoumeng Lin, University of Florida; and Nicole Kleinstreuer, NIH.

Machine learning and artificial intelligence (AI) approaches have significantly contributed to progress in many scientific disciplines, including toxicology, risk assessment, and environmental health. A number of machine learning models have been published to predict the absorption, distribution, metabolism, excretion, and toxicity (ADMET) of chemicals; the degree of environmental pollutions, such as air quality and water quality; and the human health outcomes due to environmental chemical exposures. Development of machine learning models to study toxicology and environmental health issues requires multidisciplinary expertise, including mechanistic toxicology, pharmacokinetics, environmental health, computer science, programming, and machine learning algorithms. However, there are not many researchers who have all the relevant expertise. It is common and essential for computational scientists and mechanistic toxicology and environmental health scientists to collaborate to complete machine learning–based projects. However, computational scientists often find it challenging to grasp the ADMET mechanisms of xenobiotics, just as mechanistic toxicologists may struggle to comprehend and apply machine learning models. Additional teaching materials and training opportunities are needed to address this challenge. To address this need and further leverage machine learning approaches to advance toxicology and environmental health, a group of scientists recently completed a dedicated textbook entitled Machine Learning and Artificial Intelligence in Toxicology and Environmental Health. This book introduces the fundamental concepts and principles of widely used machine learning and AI techniques relevant to toxicology and environmental health. It provides practical examples of the applications of machine learning and AI methods to predict ADMET properties of chemicals, air quality, water quality, food safety, antimicrobial risk assessment, and chemical risk assessment. It also includes case studies, sample code, and interactive computer exercises to support the practical implementation of these methods in toxicology and environmental health. This course aims to facilitate the dissemination of the scientific knowledge of this textbook to the field of toxicology and environmental health. Note that this book will only be used to relay and support the science of this course, and it is not a requirement or recommendation of speakers for attendees to purchase. This course will start with a brief overview of the book and an introduction on advances in applying machine learning and AI in the regulatory landscape, followed by individual presentations. The first presentation will introduce how to develop and apply machine learning–based quantitative structure-activity relationship (QSAR) models to predict ADME properties of chemicals. The second presentation will introduce how to build machine learning–empowered physiologically based pharmacokinetic (PBPK) models for chemicals and nanoparticles. The third presentation will demonstrate how to integrate machine learning with cell painting to predict bioactivity based on in vitro assays. The fourth presentation will discuss how to develop machine learning–based QSAR models to predict developmental toxicity. The fifth presentation will explore how machine learning models function in human health risk assessment. The sixth presentation will discuss how to apply generative AI for research translation in toxicology and environmental health and the ethical considerations. Most of these presentations will cover both a theoretical introduction to model development and practical examples to demonstrate each application. This course will introduce both theory and many practical example applications of machine learning approaches in different areas of toxicology. It is expected that this course will be beneficial for a diverse population of SOT attendees, including experimental toxicologists and computational toxicologists at different career stages (e.g., students, postdocs, and senior scientists) across various sectors, including academia, industry, and regulatory agencies.

Chair(s):
Vijay Kale, Bristol Myers Squibb; and Herve Lebrec, Sonoma Biotherapeutics.

Impurities in biotherapeutics include residual host cell proteins (HCPs), high and low molecular weight species, Protein A, host cell DNA and RNA, lipids, aggregates, process-related contaminants, and product-related variants. Due to their intricate nature, potential impurities in biotherapeutics can have a significant potential impact on safety and efficacy profiles.

This course will offer a comprehensive overview of risk assessment of impurities in biotherapeutics. The introductory talk will provide an overview of impurities in biologic products, emphasizing the complex nature of these contaminants and the critical importance of understanding analytical methods, testing strategy, and regulatory expectations. This talk will establish a foundational understanding for analytical considerations, impurity risk assessment, and mitigation strategies. One of the speakers will focus on highlighting sophisticated techniques, like high-performance liquid chromatography (HPLC), mass spectrometry (MS), and capillary electrophoresis, essential for detecting and quantifying impurities from various sources, such as host cells and raw materials. This presentation will outline the current state of analytical technology and methods for impurity testing, including challenges and limitations, as well as advancements that are driving continuous improvements in impurity testing. This course will further discuss the framework for assessing biologic impurities of concern with emphasis on the significance of controlling HCPs as critical quality attributes through comprehensive risk assessments. To understand current industry practices in risk assessment of HCPs, the findings of an IQ DruSafe Impurities Safety Working Group industry-wide survey will be discussed in this course. This survey highlights the current challenges, strategies, and regulatory interactions associated with HCPs in biotherapeutic development and commercialization. Concluding the session, interactive case studies will engage the audience in practical impurity assessment scenarios using polling tools, fostering dialogue and sharing learning among participants.

Overall, this course will provide the audience with a clear understanding of the types of impurities encountered with biotherapeutics, fit-for-purpose analytical methods, and the conduct of risk assessment for those impurities.

Chair(s):
Robert Nocco, Consultant; and Nadia Moore, JS Held LLC.

Chemical emergencies may have adverse impacts on public health as evidenced by landmark incidents: the 1984 Bhopal disaster, the 1995 Tokyo subway attack, and the 2023 East Palestine train derailment. Toxicology plays an essential role in successfully managing the associated risks of these emergencies, which could include outcomes from terrorist (CBNR), military, transportation, and chemical plant incidents.

This course will cover (1) the evolution and history of public health emergency exposure guidelines; (2) an overview of current guidelines; (3) public health exposure values; (4) the application of toxicology disciplines to emergency response decisions; (5) the role of toxicologists in the incident management system (NIMS); (6) remedies and treatment; (7) exposure measurements; and (8) risk assessment and risk tolerance criteria.

Chair(s):
Bette Meek, University of Ottawa; and Alexandra Schaffert, Tampereen Yliopisto (Tampere University), Finland.

Adverse outcome pathways (AOPs) provide convenient integrating organizational constructs for assembling and evaluating mechanistic information at different levels of biological organization in a form designed to support a range of regulatory applications. These include the development of integrated approaches to testing and assessment (IATAs) and chemical-specific assessment to inform predictive inference and mode-of-action (MOA) analysis. There has been much experience gained in the last 13 years since the introduction of the AOP development program by the Organisation for Economic Co-operation and Development (OECD). Guidance and an associated handbook support developers in the description and evaluation of AOPs via a publicly available knowledgebase. The program also includes formal peer engagement in the form of coaching for AOP development, external scientific review, and endorsement by parent OECD committees on testing and assessment. Increasing experience in AOP development and application contributes to the continuing evolution of the methodology to meet regulatory needs. This course addresses aspects of the application of systematic review tools in the development and application of AOPs.

Areas of evolution include the more systematic consideration of supporting evidence, extension of qualitative weight of evidence considerations to quantitation of key event relationships, the integration of emerging data such as ’omics, and increasing experience in application. This course builds on training developed within the OECD program on best practices of documentation and assessment to support regulatory implementation, focusing particularly on the consideration of systematic workflows and tools in AOP development and practical examples of application. Practical exercises include the application of topic modeling in AOP development. The course is designed to promote critical evaluation of data and familiarity with emerging approaches supporting transparent, systematic AOP development for regulatory and research applications.

Following a brief introduction of the course outline and learning objectives, the first presentation will update attendees on how the available guidance and tools for AOP development have evolved to reflect experience with increasing and expanding content and biological space. The second presentation will emphasize the importance of transparently and effectively documenting evidence collection and evaluation by sharing tools and best practices with prospective AOP authors and users. The third presentation outlines methodology adopted to systematically consider available data in the development of an AOP as a basis for an IATA by the European Food Safety Authority, while the fourth presentation addresses the application of ’omics in the context of AOPs. The fifth presentation considers relevant approaches to systematic evidence mapping in AOP development, and the final presentation addresses case studies to assess biological plausibility in hazard assessment integrating systematic review methods in AOP development and consideration. A hands-on exercise addresses the application of topic modeling in AOP development.