From Principle to Practice: A Decision Tree for Acute Oral Toxicity

How a structured, evidence led framework can help make animal testing a last resort

Regulators are reframing their position on the significance of vertebrate animal studies for acute oral toxicity assessments.  Historically, these studies informed chemical hazard classification, labelling, and risk assessment. Now, these studies are increasingly expected by regulatory agencies and industry standards only when existing information, adaptations, and New Approach Methodologies (NAMs) cannot adequately answer the scientific or regulatory question being asked.

A newly published paper in the Regulatory Toxicology and Pharmacology Journal, titled A Decision Tree for Acute Oral Toxicity: Implementing the Last Resort Requirement in Practice, addresses the practical challenge behind this expectation. Co-authored by Clive Roper and additional industry experts, the paper outlines a structured framework for deciding whether a new animal study is necessary.

The importance is not simply a shift from one test to another, but from a test led mindset to an evidence led decision process focused on defining the question, evaluating what is already known, identifying any remaining uncertainty, and selecting the most appropriate way to address it.

Why the Last Resort Principle Matters

Under the EU REACH Regulation, testing on vertebrate animals should only be used as a last resort. This is both an ethical obligation and a scientific discipline. A new animal study should not be commissioned automatically because it is familiar, nor should existing data be overlooked because they do not arrive in the form of a traditional standalone test.

Applying the principle properly requires an assessor to demonstrate that relevant evidence and permissible adaptations have been considered first. It also requires confidence that avoiding a study will not compromise protection of human health or the environment.  The action is not a choice between safety and animal welfare. It is to determine what evidence is necessary for a defensible safety decision and to generate no more data than that decision requires.

Why Acute Oral Toxicity Decisions Can be Difficult

Acute oral toxicity has traditionally been associated with in vivo studies.  Although testing strategies have evolved, acute systemic toxicity remains challenging because a substance may affect multiple organs and biological processes, and no single non-animal method necessarily captures every relevant outcome.

Simultaneously, assessors may already have access to a substantial evidence base which can include physicochemical properties, existing human or animal data, repeated dose studies, information on structurally similar substances, computational predictions, and in vitro findings. The difficulty is often deciding how the different sources fit together, whether they are relevant and reliable, and whether the combined evidence is sufficient for the decision at hand. 

Regulatory contexts also differ. The information needed to support hazard classification may not be identical to that needed for another risk management decision. Data quality, applicability domains, and exposure considerations need to be addressed individually.  Without a common process, assessors can review similar evidence yet reach different conclusions or default to a new study. 

What a Decision Tree Adds

The proposed Decision Tree provides a transparent sequence for working through these questions. It is grounded in the OECD concept of an Integrated Approach to Testing and Assessment (IATA). Rather than starting with a prescribed animal test and asking whether it can be avoided, the process starts with the decision that must be made.

The framework prompts organisations to ask:

·    What is the specific regulatory or scientific question?

·    What relevant information is already available?

·    Is that information adequate and sufficient, individually or as part of a weight-of-evidence assessment?

·    Can an accepted adaptation, read-across approach, computational model or existing in vitro method address the remaining need?

·    What uncertainty remains, and is it material to the decision?

·    If additional evidence is required, what is the most targeted and proportionate way to obtain it?

This structure does not predetermine a non-animal outcome. Its value is that it makes the reasoning visible. When the evidence supports a conclusion, the basis for avoiding a new animal study can be documented. When the evidence is insufficient, the unresolved gap is clear, allowing further work to focus on that need rather than repeating information unnecessarily.

Using Existing Information and New Approach Methodologies (NAMs) Effectively

The greatest value of NAMs is realised when they are used as components of an integrated evidence strategy rather than treated as isolated replacements for an animal test.

The Decision Tree encourages assessors to examine relevance, reliability, limitations, and uncertainty, and then determine whether the combined evidence answers the question with sufficient confidence. This is more scientifically meaningful than treating any one method as universally decisive.

Why the Framework Matters for the Toxicology Community

A consistent framework can support better study planning, reduce avoidable animal use, and prevent unnecessary cost and delay. It can also improve the quality of regulatory submissions by providing transparent workflows that can make evidence packages more consistent and reviewable while preserving case by case scientific judgement.

For NAM developers, the Decision Tree provides important context. A method becomes useful in practice not simply because it performs well in validation, but because users understand which decision it informs, the boundaries within which it can be applied, and how its output integrates with other evidence. Clearer decision pathways can then help developers focus validation and communication on regulatory needs.

For the wider toxicology community, the framework offers a common language for discussing evidence, uncertainty, and next steps. That shared structure is important if modern methods are to move from promising technologies to routine safety assessment.

Turning a Principle into Routine Practice

This paper does not claim that every acute oral toxicity question can be resolved without animal data. Its immediate contribution is more practical in that it provides a disciplined way to determine when an animal study is or is not needed.

Scientific progress depends not only on developing new methods, but also on building the decision frameworks, confidence, and regulatory pathways that allow those methods to be used appropriately. By making the evaluation process more consistent, transparent, and evidence led, the Decision Tree helps translate the last resort requirement from a statement of intent into a working scientific practice.

Read the publication - Reference

Pallocca, G., Bishop, P., Browne, P., et al. (2026). A Decision Tree for Acute Oral Toxicity: Implementing the Last Resort Requirement in Practice. Regulatory Toxicology and Pharmacology, 172, 106222. DOI: 10.1016/j.yrtph.2026.106222 Full Author List: Giorgia Pallocca, Patricia Bishop, Patience Browne, Marie Darracq-Ghitalla-Ciock, Fabian A. Grimm, Laura Holden, Jay Ingram, Kamel Mansouri, Hans Raabe, Clive Roper, Katherine Santizo, Barbara G. Schmitt, Thomas A. Ward and Catherine Willett.

#AcuteOralToxicity#RegulatoryToxicology#EUVertebrateAnimalTesting