Skip to content
Education
Isle of NAM
AOP

AOP

Table of Contents

Definition

An Adverse Outcome Pathway (AOP) is an analytical construct that describes a logical sequence of causally linked events at different levels of biological organisation, beginning with the interaction of a stressor with a biomolecule and culminating in an adverse health effect in humans or wildlife1 2. The framework models the propagation of a biological perturbation across successive levels of organisation, from the molecular level through cellular and tissue-level changes to organ, organism, and, in ecotoxicological contexts, population-level consequences1 3.

Structurally, an AOP consists of three principal components. The molecular initiating event (MIE) is the direct interaction between a stressor and a specific biological target, such as the binding of a ligand to a receptor or the covalent modification of a protein1 4. Key events (KEs) are measurable changes in biological state that are both essential to the progression of the pathway and detectable using defined assays1 5. Key event relationships (KERs) describe the causal linkages between adjacent KEs, articulating the biological plausibility and empirical evidence supporting the inference that perturbation of the upstream event results in perturbation of the downstream event4 5. By design, an AOP is stressor-agnostic and does not describe exposure, toxicokinetics, or any specific chemical; it represents only the sequence of biological events once an MIE has been triggered1 4.

Purpose

The primary value proposition of the AOP framework is that it converts dispersed mechanistic knowledge into a structured, reusable format that supports inference of cause-and-effect relationships without the generation of new apical animal data1 6. AOPs thereby provide the mechanistic scaffolding required for integrated approaches to testing and assessment (IATA), in which information from multiple sources, including in chemico, in vitro, and in silico methods, is combined to satisfy a defined regulatory endpoint6.

The framework improves upon legacy hazard characterisation models, which relied predominantly on whole-animal bioassays reporting apical endpoints with limited mechanistic interpretation1 3. Because an AOP is anchored in a defined MIE and measurable KEs, it permits the substitution of animal tests with mechanistically anchored new approach methodologies (NAMs) targeting individual KEs, provided that the targeted event is conserved between the test system and the organism of regulatory concern1 3 6. This anchoring underpins the human relevance of the framework: where an MIE or KE operates in human cells or tissues, in vitro measurements conducted in human-based systems carry direct mechanistic relevance to human health outcomes, in contrast to cross-species extrapolation from conventional animal models1 3. AOPs additionally strengthen read-across justifications, weight-of-evidence assessments, and the prioritisation of chemicals for further testing by making the mechanistic rationale explicit and auditable1 6.

Methodology

AOP construction follows a documented OECD methodology centred on the assembly and evaluation of KEs and KERs5 7. Development may proceed top-down, beginning with a well-defined adverse outcome and working backwards towards the MIE, or bottom-up from a characterised molecular interaction; however, the single-AOP principle dictates that one AOP describes one linear path from one MIE to one adverse outcome, with branching treated through the development of separate, overlapping AOPs4 7. Each KE must be a measurable biological state change, and each KER is documented with a weight-of-evidence evaluation addressing biological plausibility, empirical support, and essentiality, applying principles adapted from the Bradford Hill criteria for causal inference4 5 7.

Development is conducted collaboratively as a living document in the AOP-Wiki, the principal module of the OECD AOP Knowledge Base, where authors register AOPs, KEs, and KERs under structured templates, and where entries are versioned so that the endorsed state of an AOP remains distinguishable from subsequent edits3 8. AOPs developed in the AOP-Wiki may be submitted to the OECD AOP Development Programme for scientific review conducted according to the OECD guidance for the review of AOPs; AOPs that complete review and are endorsed are published as static, time-stamped versions in the OECD Series on Adverse Outcome Pathways2 9.

Performance

Adoption metrics for the framework are substantial. As of October 2026, following the release of AOP-Wiki version 2.9, the platform hosts 595 AOPs in various stages of development, together with their constituent KEs and KERs8. A smaller subset has completed OECD scientific review and been published in the OECD Series on Adverse Outcome Pathways, covering endpoints that include skin sensitisation, aromatase inhibition, androgen receptor agonism, and pathways initiated by the deposition of energy2 9. The rigour and reproducibility of the framework are governed by the standardised weight-of-evidence criteria applied to each KER and by the version-controlled, snapshot-based publication process, which distinguishes peer-reviewed and endorsed content from material that is subsequently added or modified7 8 9.

The framework nonetheless carries well-defined technical boundaries. AOPs are qualitative by design and contain no dose-response, temporal, or exposure information; quantitative prediction requires the addition of toxicokinetic modelling and quantified KERs4 10. Because an AOP is stressor-agnostic, it says nothing about which chemicals can trigger the MIE or about absorption, distribution, metabolism, or excretion4. Extrapolation across taxa, life stages, and exposure durations remains a source of uncertainty, and because the framework is a living-document paradigm, there is no objective definition of a complete AOP; the maturity of an AOP is instead judged by the strength of the evidence supporting its constituent KERs4 5 10.

Applications

The principal research applications of the AOP framework include the organisation of mechanistic toxicological knowledge, the identification of data gaps and critical uncertainties, and the rational selection of mechanistic endpoints for method development1 4. In industrial and regulatory practice, AOPs serve as the mechanistic basis for IATA, for read-across justifications, for the interpretation of NAM results, and for the prioritisation of chemicals for hazard assessment6. The framework is standardised within the OECD testing-of-chemicals programme, where guidance documents define its use in developing IATA and its scientific review process6 9.

Deployment is most mature in the area of skin sensitisation. The skin sensitisation AOP, proceeding from protein binding as the MIE through dendritic cell activation and T-cell proliferation to allergic contact dermatitis, underpins the defined approaches on skin sensitisation codified in OECD Test Guideline 497, which was first adopted in 2021 and permits hazard and potency conclusions to be reached using in chemico, in vitro, and in silico information without animal testing9. More broadly, endorsed AOPs are in regulatory use in ecotoxicology, including endocrine-mediated pathways such as aromatase inhibition leading to male-biased sex ratio, and the framework underpins ongoing NAM-based assessment strategies across multiple jurisdictions2 6. Quantitative applications remain at an earlier stage of readiness, as quantified KERs and integrated toxicokinetic-toxicodynamic models are still required for dose-response prediction10.

Variations

  • AOP Networks: The assembly of multiple individual AOPs that share common KEs into a connected network, enabling the representation of the complexity of systems toxicology and supporting network-level inference and analytics11 12.
  • Quantitative AOPs (qAOPs): The extension of a qualitative AOP with mathematical models describing KERs, dose-response, and time-course behaviour, permitting predictive hazard and risk assessment rather than hazard identification alone10.
  • Non-Chemical Stressor AOPs: AOPs whose MIE is triggered by physical or non-chemical stressors rather than chemical-biomolecule interaction, exemplified by the endorsed OECD AOPs describing the deposition of energy leading to adverse outcomes such as cataracts, bone loss, learning and memory impairment, lung cancer, and abnormal vascular remodelling2.

  1. Ankley GT, Bennett RS, Erickson RJ, et al. Adverse outcome pathways: A conceptual framework to support ecotoxicology research and risk assessment. Environmental Toxicology and Chemistry, 2010.
    This foundational paper introduced the AOP framework and formalised its core structure of a molecular initiating event, key events, and key event relationships traversing multiple levels of biological organisation. The authors articulated how AOPs support mechanistically based risk assessment and identified the framework as a means of linking mechanistic data to apical outcomes for both ecological and human health assessment. ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎

  2. OECD. OECD Series on Adverse Outcome Pathways. OECD Publishing, accessed 20 October 2026.
    This OECD publication series constitutes the official repository of AOPs that have completed scientific review and been endorsed under the OECD AOP Development Programme. It publishes time-stamped, static versions of endorsed AOPs, including the deposition of energy pathways for non-chemical stressors and endocrine-mediated pathways such as aromatase inhibition and androgen receptor agonism. ↩︎ ↩︎ ↩︎ ↩︎ ↩︎

  3. Vinken M, Knapen D, Vergauwen L, Hengstler JG, Angrish M, Whelan M. Adverse outcome pathways: a concise introduction for toxicologists. Archives of Toxicology, 2017.
    This review provides a compact introduction to the AOP framework and its structural components, situating it within the transition of toxicology from an observational, animal-based discipline towards a mechanistic, pathway-based science. It emphasises the framework’s reliance on measurable KEs and its role in supporting mechanistically anchored, human-relevant assessment. ↩︎ ↩︎ ↩︎ ↩︎ ↩︎

  4. Villeneuve DL, Crump D, Garcia-Reyero N, et al. Adverse outcome pathway (AOP) development I: strategies and principles. Toxicological Sciences, 2014.
    This article establishes the foundational principles of AOP development, including the single-AOP principle, the requirement that KEs be measurable and essential, and the deliberate separation of the AOP, as a stressor-agnostic sequence of biological events, from exposure and toxicokinetic considerations. It also explains that AOPs are living documents with no objective definition of completeness. ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎

  5. Villeneuve DL, Crump D, Garcia-Reyero N, et al. Adverse outcome pathway development II: best practices. Toxicological Sciences, 2014.
    This companion article codifies best practices for constructing AOPs, including the documentation of KERs and the application of weight-of-evidence evaluation addressing biological plausibility, empirical support, and essentiality. It provides the practical methodological standards that were subsequently adopted into the OECD AOP development guidance. ↩︎ ↩︎ ↩︎ ↩︎ ↩︎

  6. OECD. Guidance Document for the Use of Adverse Outcome Pathways in Developing Integrated Approaches to Testing and Assessment (IATA). OECD Series on Testing and Assessment No. 260, OECD Publishing, 2017.
    This guidance document defines how AOPs are to be used as the organising scaffold for IATA, in which information from in chemico, in vitro, and in silico methods is combined to satisfy defined regulatory endpoints. It provides the systematic framework for characterising the biological relevance of alternative methods relative to KEs along a pathway. ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎ ↩︎

  7. OECD. Users’ Handbook supplement to the Guidance Document for developing and assessing Adverse Outcome Pathways. OECD Series on Adverse Outcome Pathways No. 1, OECD Publishing, 2018.
    This handbook provides focused, practical instructions for AOP developers and reviewers, describing the identification, organisation, and evaluation of key information on KEs and the linkages between them. It operationalises the evaluation criteria for AOP development in the AOP-Wiki and constitutes the standard reference for the construction of an AOP. ↩︎ ↩︎ ↩︎ ↩︎

  8. AOP-Wiki. Collaborative Adverse Outcome Pathway Wiki, version 2.9. aopwiki.org, accessed 20 October 2026.
    The AOP-Wiki is the collaborative platform of the OECD AOP Knowledge Base and the primary global repository of AOPs, their KEs, and their KERs. As of October 2026, following the release of version 2.9, the platform lists 595 AOPs in various stages of development, maintained as version-controlled living documents. ↩︎ ↩︎ ↩︎

  9. OECD. Guidance Document for the scientific review of Adverse Outcome Pathways. OECD Series on Testing and Assessment No. 344, OECD Publishing, 2021.
    This guidance document defines the process by which AOPs developed in the AOP-Wiki are submitted for scientific review under the OECD AOP Development Programme. It describes the standards for the review of AOP structure and evidence and the pathway to endorsement and publication in the OECD Series on Adverse Outcome Pathways. ↩︎ ↩︎ ↩︎ ↩︎ ↩︎

  10. Perkins EJ, Antczak P, Barros S, et al. Building and Applying Quantitative Adverse Outcome Pathway Models for Chemical Hazard and Risk Assessment. Environmental Toxicology and Chemistry, 2019.
    This article establishes the methodology for extending qualitative AOPs into quantitative adverse outcome pathways, describing the coupling of toxicokinetic and toxicodynamic models to quantified KERs. It identifies best practices for model construction and documentation and the requirements for regulatory confidence in qAOP-based predictions. ↩︎ ↩︎ ↩︎ ↩︎

  11. Knapen D, Angrish R, Fortin M, et al. Adverse outcome pathway networks I: development and applications. Environmental Toxicology and Chemistry, 2018.
    This article defines the concept of the AOP network, in which individual AOPs that share common KEs are assembled into connected structures. It describes the development of AOP networks and their applications for representing the complexity of systems toxicology and for supporting more comprehensive mechanistic assessment. ↩︎

  12. Villeneuve DL, Angrish R, Fortin M, et al. Adverse outcome pathway networks II: network analytics. Environmental Toxicology and Chemistry, 2018.
    This companion article describes analytical methods for interrogating AOP networks, including metrics characterising network topology and the relative importance of individual KEs. It sets the stage for tools and case studies facilitating the use of AOP network-based predictions in regulatory applications. ↩︎