Conclusions and recommendations

Grouping of endocrine disruptors is possible and can generally be conducted like grouping of substances based on other endpoints. However, endocrine disruption is a complex endpoint which in some respects requires extra consideration. One example is that the occurrence of non-monotonic dose response curves, low dose effects, different effect profiles depending on exposure windows and times of investigation and some substances acting through several mechanisms and modes of action at the same time leads to complex hazard profiles, which should be given special attention when creating a data matrix.

Regulatory identification of endocrine disruptors requires information about both endocrine activity, adversity and a link between the two, and grouping exercises are conducted with different focuses and aims. The main focus of this report was on grouping based on structural similarities and mechanistic information ultimately aiming to read-across adversity information from data rich to data poor substances. Supporting this work, ECHAs assessments of regulatory needs for groups of substances and the continued development and application of (Q)SARs are valuable efforts and tools.

Another approach to grouping is exemplified in the EFSA CAGs, where adversity information is used to group substances for combined exposure assessments, independently of their mechanisms and modes of action (at the first levels). These groups, defined by adversity information, could in future be broadened by including a mechanistic approach based on structural similarities and mechanistic information, leading to read-across of adversity information and thereby inclusion in the relevant CAGs. A third approach to grouping is exemplified by the PFAS proposal, in which a specific molecular moiety (the carbon-fluorine bond) is identified as decisive for the concern. Comparative studies, docking analyses and (Q)SAR modelling are valuable tools for exploring future applications of this approach for endocrine disruptors.

Main findings of the report are that:

EDs can be grouped like other chemicals
Complex hazard profiles of EDs require special attention
Groupings should start with the broadest group possible
In the grouping process, data-rich substances are valuable source substances and should be included regardless of their regulatory status and whether they are in use
(Q)SARs should be used to guide, challenge, support and revise the choices made.

Through our analysis, we have identified some recommendations for grouping, which we find of general value, and some which are more specific to grouping of endocrine disruptors. Both the OECD GD 194 and the ECHA RAAF provide valuable recommendations and frameworks for grouping and can be used as the basis for the grouping of endocrine disruptors. We have developed a simplified workflow (Figure 4) based on these frameworks. The workflow can be used as an overall guide for grouping of endocrine disruptors, as it at each step incorporates both the general recommendations and the recommendations specific for endocrine disruptors, we have identified in this project.

Step 1 Define purpose and hypothesis
Define the regulatory purpose Define the grouping hypothesis

Step 2 Include relevant substances
Start with the broadest group possible. Use (Q)SARs to check if additional substances should be included. Include substances under regulatory scrutiny (For EDs: Regulatorily identified EDs) or registered in high tonnages under REACH for use as source substances if possible Use well-argued inclusion and exclusion criteria, refrain from excluding substances based on current regulatory status or use. Make sure that all substances are characterized in detail, including impurity profiles.

Step 3 Produce a data matrix
Include in silico, in vitro, in vivo data Include relevant metabolites and ADME data For EDs: Consider complex hazard profiles, including different effects depending on timing of exposure and investigation, low dose and non-monotonic effects etc.

Step 4 Redefine the group and define sub-groups, as appropriate
For EDs: Consider that different molecular moieties may be relevant for different mechanisms and thereby modes of action and adverse effects. Some overlap may occur, but this should ideally be analyzed and described. For EDs: Consider that many endocrine disruptors act through more than one mechanism/mode of action and that several different mechanisms/mode of actions can lead to the same adverse effects. Consider that if in vitro data are used to define sub-groups, ADME must be considered and that if ADME/in vivo information is considered, the sub-groups may have different boundaries. Use QSARs to check if more substances could be added to the (sub-)group.

Step 5 Read-across to fill data gaps

Step 6 Conclude on results, revisit the hypothesis, as apppropriate

Figure 4 Recommended workflow for (sub-)grouping of endocrine disruptors with the ultimate aim of filling data gaps by read-across. General recommendations are provided at relevant steps in the workflow. Recommendations specific for endocrine disruptors are highlighted in bold.