A new paper co-authored by XCellR8’s Dr. Tom Ward has set out a practical Decision Tree designed to help safety assessors determine when animal testing for acute oral toxicity is genuinely necessary.
We were delighted to see its acceptance for publication in Regulatory Toxicology and Pharmacology. The paper, titled A decision tree for acute oral toxicity: Implementing the last resort requirement in practice, was developed through the Animal-Free Safety Assessment Collaboration (AFSA), bringing together scientists and regulatory experts from industry, academia and organisations including OECD and NICEATM/NIEHS/NIH.[1]
The framework addresses a deceptively simple problem. Under REACH, vertebrate animal testing is already supposed to be used only as a last resort. But stating that principle is easier than applying it consistently.
Before an animal study is commissioned, what evidence should have been considered? When can existing data, computational approaches or in vitro methods provide enough information? And how should those different sources be combined into a transparent regulatory decision?
The new Decision Tree is intended to provide a clearer route through those questions, turning the idea of “last resort” into something that can be applied more systematically in practice.
In our latest blog, we look at the thinking behind the new framework, why acute oral toxicity is such an important test case, and how clearer decision-making could help make better use of non-animal evidence before an animal study is considered.
A structured route through the evidence
The framework is based on the OECD concept of an Integrated Approach to Testing and Assessment, or IATA.
Rather than beginning with a prescribed test and asking whether it can be avoided, an IATA begins with the regulatory question, considers the information already available and identifies what, if anything, is still missing.
For acute oral toxicity, the new Decision Tree brings together existing information, in silico approaches, in vitro evidence and weight-of-evidence assessment within a structured workflow.[1]
Where the available evidence is sufficient to support the required hazard decision, the framework provides a route to reach that conclusion without automatically progressing to a new animal study. Where uncertainty remains, it makes the information gap clearer so that any additional testing can be targeted towards what is actually needed.
Its purpose is therefore not simply to catalogue alternatives, but to make the decision-making process itself more systematic and traceable.
Why acute oral toxicity needs this approach
Acute oral toxicity has historically been closely associated with animal testing and, in particular, the LD50 — the dose expected to cause death in 50% of a test population.
Testing practices have changed considerably since the original LD50 methods were introduced, but acute systemic toxicity remains one of the more difficult areas for full replacement because the effects of a substance can involve multiple organs and biological processes.
At the same time, relevant information can already come from many places. Existing studies, physicochemical properties, read-across, computational predictions and in vitro data can all contribute to the assessment of a substance.
The challenge is therefore not simply whether non-animal evidence exists, but how it is brought together, when it is sufficient, and when further information genuinely needs to be generated.
That is the gap the new Decision Tree is designed to address.
Making “last resort” measurable
REACH already requires vertebrate testing to be undertaken only as a last resort, and registrants are expected to consider existing information and appropriate adaptations before generating new animal data.[2]
But a legal principle does not automatically create a consistent scientific workflow.
If different assessors approach the same evidence in different ways, or if established animal studies remain the default because the route to using alternatives is less clearly defined, the last-resort requirement may not always be applied as effectively as possible.
A structured Decision Tree turns that principle into a sequence of questions that can be documented: what do we already know, is it sufficient for the decision, what uncertainty remains, and what is the most appropriate way to address it?
Importantly, the framework does not assume that every substance can follow the same non-animal route. Some may have extensive existing data; others may lend themselves to read-across or computational prediction; some may benefit from in vitro evidence; and in other cases significant uncertainty may remain.
The purpose is not to predetermine the answer. It is to make sure the available alternatives have genuinely been considered before an animal study becomes the conclusion.
A step towards practical implementation
This is increasingly important as the conversation around animal-free safety assessment moves from developing new methods towards integrating them into regulatory practice.
New assays are essential, but they only change testing practice if assessors have confidence in where they fit, what questions they can answer and how their results should be interpreted alongside other information.
Frameworks such as the new acute oral toxicity Decision Tree provide that wider architecture, helping move the discussion from “what can replace this animal test?” towards “what evidence do we need to make this safety decision?”
XCellR8 is thrilled to have contributed to the publication through Tom’s involvement as a co-author, alongside the wider group of scientists and regulatory experts involved through AFSA.
The paper does not suggest that every acute oral toxicity question can now be answered without animal data. What it offers is something more immediately practical: a structured way to determine whether an animal study is actually required in the first place.
If animal testing is to remain a last resort, frameworks like this help define what should happen before that last resort is reached.
References
[1] 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.
[2] European Chemicals Agency (ECHA). How to avoid unnecessary testing on animals. Guidance for REACH registrants.
Email info@x-cellr8.com or call +44 (0)1925 607 134.
———
Contact us if you’d like to learn how you can support our mission, become a corporate sponsor or donate to our crowdfunder.
Email info@x-cellr8.com
Call +44 (0)1925 607 134.
Follow Us!