August 14, 2017 by Becky Marks


With the release of ECHA’s Read Across Assessment Framework (RAAF) for human health endpoints in 2015 and the subsequent 2017 update to include environmental endpoints, we have been putting the approaches into practice for a range of different substance types, including inorganics and UVCBs.

One of the objectives of REACH is to ensure the protection of human health and the environment whilst minimising animal testing. Read across between similar substances is one way of establishing the potential toxicity of a substance without conducting unnecessary animal testing. However, approaches need to be robust and well documented in order to ensure that adequate protection is maintained.

The aim of the RAAF is to provide a structured framework for justifying read across approaches, which allows ECHA and Member States to evaluate read across in a consistent manner. The RAAF outlines a number of scenarios, for both analogue and category approaches. The scenarios covered are ‘Different compounds have qualitatively similar properties’ (read across between structurally similar substances) and ‘(Bio)transformation to common compound(s)’ (read across between substances where the same compounds are expected based on hydrolysis, metabolism, or degradation). The category scenarios are also split depending on whether variation in properties is expected between category members, or not.

Scenario selection for the read across justification (ECHA 2017)

Each scenario is then split into a number of Assessment Elements (AEs), for both environmental and human health endpoints. The AEs assess considerations such as the impact of the composition of source and target substance (including impurities), the qualitative mechanism of action that is common to both substances and why this is considered to lead to quantitatively similar effects and an assessment of whether bias is likely to influence the predictions.

Although the principles outlined in the RAAF are not new, it provides a structure for documenting read across approaches and ensures that all relevant considerations are taken into account. We have found that there is a need for good quality composition information on both the source and target substance in order to justify a read across approach and this is particularly important if there are likely to be impurities that may influence the prediction, or if the substance is a UVCB. As the RAAF states ‘unambiguous substance identity for both the target and source substances is therefore a prerequisite for read-across assessment’. The composition of the source substance is equally as important as the composition of the target substance and this is something to consider, particularly if the source data is old and may not contain sufficiently detailed compositional information for the test item.

A key component of a read across justification is being able to demonstrate that structural similarity leads to similar effects (or lack of effects). ‘Anchor points’, where studies are available for both the source and target substance, are crucial for showing this, whichever scenario is being followed. We would recommend that for key endpoints test data are generated for both substances, to allow direct comparison and provide support for read across for other endpoints.

Our experience has shown that for simple organic substances, the RAAF provides a structured framework that can be easily followed and allows read across to be justified in a clear manner. The RAAF structure, along with the changes to IUCLID read across records that have been implemented in IUCLID 6, mean that it is clear to anyone reviewing the dossier where the key information justifying read across can be found. This limits the chances of an approach being rejected simply because the reviewer was not able to identify where the documentation for the approach was included in the dossier.

The RAAF does, however, need some adaptation for use with other types of substances. The RAAF has been developed primarily for organic substances. For metals, as acknowledged in the RAAF document, some of the criteria are not relevant (e.g. biodegradation) and others need to be modified in order to cover all of the metal-specific considerations (e.g. use of transformation/dissolution test results for assessing availability and consideration of oxidation states). However, with these adaptations the RAAF can be appropriate for the assessment of metal compounds.

UVCB substances also require special considerations, as recognised by the release of a separate guidance document specific to UVCB and multi-constituent substances. UVCB substances can require very different approaches depending on the number of constituents, the similarity of constituents and their variability between samples. In some instances, read across between different UVCB substances is possible, other times read across from key constituents or fractions is more appropriate. In many cases a combination of read across approaches is required, for different endpoints. The UVCB-specific guidance also notes that mixture toxicity should be considered, although there is still considerable debate about how this can best be achieved. Our experience of putting read across justifications together for UVCB substances has demonstrated that although the key principles of the RAAF can be followed, a substance-specific approach is needed. Consideration of the structure of the read across justification documentation is important, particularly if multiple read across approaches are used, so that read across is clearly justified for each individual endpoint.

wca have extensive experience of developing read across approaches and have been closely following the development of the RAAF. We are currently using the RAAF to develop read across justifications for a number of substances both for 2018 registrations and for updates of already registered substances. If you would like more information about these services, please contact us.