- Next-Generation Read-Across: Improved Understanding of Relevant Uncertainties
-
Stars at Night Ballroom B2‚ Convention Center
Chair: Gina Hilton, PETA Science Consortium International e.V., Germany
Co-Chair: Donna Macmillan, International Collaboration on Cosmetics Safety
Primary Endorser: Regulatory and Safety Evaluation Specialty Section
Endorser(s): In Vitro and Alternative Methods Specialty Section
Endorser(s): Computational Toxicology Specialty Section
Read-across is a technique that infers unknown toxicological endpoints of a chemical based on its similarity in properties or chemical structure to other substances with known data. Regulatory agencies have strategically embraced this paradigm as a cornerstone of modern chemical assessment, offering a scientifically defensible pathway to reduce animal testing burdens, maximize utilization of existing knowledge, and enhance cost-efficiency in safety evaluation. Numerous guidelines have been published in recent years to establish the best practices in the field, reflecting continuing efforts to align approaches and methodologies for performing read-across. One key aspect of read-across is handling uncertainty, both how to describe it and how to reduce it using new approach methodologies (NAMs). NAMs data may be used at multiple stages of the read-across process, particularly to support the similarity hypothesis, inform toxicokinetic and toxicodynamic considerations, and address data gaps. They provide mechanistic and biological evidence that complements structural and physicochemical similarity within a weight-of-evidence framework. In this context, NAMs play a key role in reducing uncertainty. By contributing directly to uncertainty analysis, they increase confidence in the read-across conclusion when results are coherent and biologically plausible. Specifically, NAMs provide: 1) Contribution to uncertainty characterization: the approach requires a systematic analysis of uncertainty at each step of the read-across. NAMs data help identify key assumptions (e.g. similarity in metabolism or biological response) and test them using mechanistically relevant evidence, and 2) Reduction of uncertainty related to data gaps: NAMs can address specific data gaps, particularly where apical animal data are missing or limited, where existing data are heterogeneous or difficult to interpret, or where ethical or regulatory constraints restrict new in vivo testing. Specifically, mechanistically based NAM data can support toxicodynamic similarity justification, in vitro NAMs for clearance and metabolism can support similarity in toxicokinetics. The sessions will include discussion of approaches for evaluating uncertainty associated with in vitro data, including assay variability, reproducibility, and confidence in interpretation, alongside frameworks for establishing biological relevance and translating in vitro findings into meaningful toxicological and regulatory conclusions. Agencies like the European Chemicals Agency (ECHA) publish decisions on testing proposals and compliance checks (including use of read-across methods) on a publicly available website. However, the extent to which information on regulatory decision-making is publicly available varies across agencies, which can limit opportunities to better understand regulatory information thresholds and approaches for fulfilling data requirements to protect human health and the environment. To improve confidence in a read-across assessment, we need to fully understand the types of uncertainty and how to address them. The presentations in this session will focus on practical insights regarding the different types of uncertainty, prioritization of which uncertainties to address, and how to achieve tolerable levels of uncertainty. The first presentation will provide a scheme to quantify and refine read across uncertainty, defining a practical level of tolerable uncertainty based on criteria such as chemical similarity and class, and data quality. The second presentation outlines the European Food Safety Authority’s (EFSA’s) stepwise approach to identifying, characterizing, prioritizing, and mitigating uncertainty throughout the read across workflow—using weight of evidence, new approach methodologies (NAMs), and targeted data generation to assess remaining uncertainty for transparent regulatory decision making. The third presentation will highlight a real-world case study leveraging in vitro assays, in silico models, and toxicokinetic similarity to help address uncertainty in read-across in the cosmetic industry for a complex mixture. The last presentation will discuss practical strategies for optimizing read-across submissions in light of documented ECHA decision patterns, balancing scientific thoroughness with regulatory pragmatism. The session will conclude with a panel discussion to explore the impact of uncertainty in read-across applications in an industry context.Session and Activity Type
Enter Note
Go to previous page in this tab
Event
