S1: Test guidelines, AOPs and IATA
Test guidelines, AOPs and IATA - current and future
Swiss Center for Applied Human Toxicology (SCAHT)
Regulatory acceptance of toxicology data for human health chemical risk assessment is based on adherence to internationally accepted guidelines for data reliability (currently usually Good Laboratory Practice GLP; OECD or national equivalent) and scientific relevance (currently usually OECD Guidelines for the Testing of Chemicals, or national equivalent).
The old apical toxicity endpoints (infertility, cancer) were clear and apparently relevant to humans; new readouts (various omics, high content imaging) and test systems (in vitro, in chemico, in silico) are a challenge for regulators, but the fundamental requirements have not changed; the data need to be both reliable and relevant.
Comprehensive specific guidance on reliability of in vitro data is given in the Good In Vitro Methods Practice (GIVIMP) guidance, developed by the EURL ECVAM for the OECD (adoption scheduled 2018). This GIVIMP guidance will likely be very important for regulatory aceptance. However, guidance for in silico modelling (reliability and relevance) is currently lacking (SCCS 2016).
Relevance is addressed by the Adverse Outcome Pathway (AOP) framework, which is based on the WHO/IPCS mode of action (MOA) approach. This is a pathway-based analysis in which a causal chain of necessary and measurable key events leads from chemical exposure to a single given chemical to an adverse outcome. The MOA approach has been applied successfully to many chemicals, mainly to demonstrate whether a given animal adverse effect is or is not predictive for human health effects.
AOPs delineate a causal chain of key events from a "Molecular Initiating Event" (MIE) to an Adverse Outcome (AO). In contrast to MOAs, AOPs are chemical-agnostic; they encompass all stressors which induce a given MIE. Key events are shared by multiple AOPs, thus turning linear pathways into networks.
Integrated Approaches to Testing and Assessment (IATAs) are intended to integrate and weigh all relevant data (including in vitro and in silico) so that they are acceptable in regulatory hazard and risk assessment.
The AOP framework provides clarity and confidence in developing IATAs, by establishing whether given data plausibly represent key event readouts in an AOP, thereby validating those data for human health risk assessment.