Seurat-1, the EU's largest collaborative project on alternative testing, has set out to demonstrate whether read-across can be strengthened by incorporating evidence from in vitro tests, -omics assays and computational models (CW 8 May 2014).
A high proportion of higher-tier toxicological studies in REACH registration dossiers include read-across reasoning, most often for repeated dose toxicity. “However, this does not mean that all read-across predictions satisfy the regulatory requirements,” writes a Seurat-1 team in Environmental Health Perspectives. “It rather shows that there is a further need to improve the methodology on how to perform read-across and categorisation to achieve the standard for regulatory acceptance.”
“Often the arguments used are not robust enough and then additional animal studies must be requested to complete the dossier,” says lead author Elisabet Berggren from the EU Joint Research Centre's Institute for Health and Consumer Protection (IHCP). “It would be helpful to have better guidance on how this could be done in a harmonised and structured manner, aiming at a confidence level that could reach regulatory acceptance,” she adds.
With this in mind, Seurat-1 has picked four different scenarios for a read-across case study:
- chemical similarity of compounds that exert a potential adverse human health effect, without needing to be metabolised;
- chemical similarity involving metabolism;
- chemicals with low or no toxicity; and
- distinguishing chemicals that are structurally similar, using a mode of action (MOA) hypothesis.
Seurat-1 scientists have come up with a chemical category to fit each scenario. For the case study, they have selected perfluoroalkyl acids for the first scenario, with PFOA as the data-rich source compound. For the second scenario, Seurat-1 is focusing on beta-unsaturated alcohols such as 1-propen-3-ol, which can cause liver toxicity by forming a toxic metabolite, acetaldehyde. Saturated alcohols, with 1-hexanol as source, fit the third scenario. Finally, the fourth scenario will cover short-chain carboxylic acids, including valproic acid and octanoic acid.
Seurat-1 is splitting the read-across exercise into two steps. The first will build a read-across case, based on structure, physico-chemical and molecular properties, and adverse outcome pathway knowledge. The second step will also include data from alternative methods, developed in Seurat-1 as well as from US high-throughput screening programme ToxCast (CW 19 December 2013).
The aim is then to judge whether the “new approach data” strengthen confidence in read-across for a selected category. How to assess such confidence is a “million dollar question that we have wrestled with for quite some time”, says co-author Catherine Mahony from Procter & Gamble, who is a member of Seurat-1's scientific expert panel.
“You really have to lay out clearly where the uncertainties lie and make some sort of quantitative assessment of that,” she explains. The Seurat-1 team is basing its approach to assessing uncertainty on a framework developed by scientists at Procter and Gamble, US.
The outcomes of the read-across case studies will be presented at the final Seurat-1 symposium, in December in Brussels.
Chemical Watch and PETA International Science Consortia are hosting a webinar on regulatory processes in the acceptance of non-animal tests on 10 June.