Keywords

REACH, read-across, RDT, systemic toxicity, NAM, uncertainty, mode of action


Objective

In-vivo evidence indicates that short branched carboxylic acids –such as the drug valproic acid (VPA)– induce microvesicular liver steatosis in animals and humans. Would an analogue compound induce the same adverse outcome? To investigate this, we investigated a category of branched carboxylic acids by characterizing their toxicology profile with an in-vitro test battery. This study aimed to explore how far biological data from NAMs can be used in a read-across scenario to prove a shared mode of action and similar toxicokinetics.


Testing Strategy

Based on available in-vivo data of three structurally related analogues, the target compound 2-ethylbutyric acid (2-EBA) was assumed to be a liver toxicant with special concern for hepatic steatosis.
Toxicokinetics and toxicodynamics properties of 2-EBA were analysed and compared to a category of structurally highly similar carboxylic acids that vary only with regard to their aliphatic side chain length.
To characterize the toxicodynamics, published signalling pathways/AOPs leading to steatosis were compiled from literature and described in an AOP network. Two high-throughput reporter gene assays (CALUX and stress pathway responses in HepG2) were used to measure some of the described molecular initiating events (MIEs). Furthermore, three liver models were used to measure lipid accumulation as a direct in-vitro surrogate for in-vivo steatosis. It was shown that the number of activated MIEs and the induction of lipid accumulation increased with the side chain length of the tested carboxylic acids, whereas short-chain analogues like 2-EBA remained inactive up to the highest tested dose in vitro. The low activity of short-chain analogues was in good agreement with known in-vivo animal data.
To characterize the toxicokinetics, a PBPK model was developed and parameterized with in-vitro ADME data. Bioavailable concentrations were predicted and QIVIVE extrapolation was performed using PBPK-based reverse dosimetry.
Overall, the NAM data proved the initial read-across hypothesis to be valid by showing a consistent trend concerning toxicokinetics and toxicodynamics within grouped compounds.


Publications

van Vugt-Lussenburg et al. 2018 [link]; Leist et al. 2017 [link]; Tolosa et al. 2017 [link]; Fisher et al. 2018 [link]; Gadaleta et al. 2018 [link]; Wink et al. 2018 [link]; Niemeijer et al. 2018 [link]; Escher et al. 2019 [link]


OECD IATA report

Report [link]; Annex II [link]; Annex III [link]


Contributors

ITEM, UL, UM, RISE, INERIS, BDS, TNO, DC, IRFMN, SYMCYP, VUA, UOF, UNIVIE, HULAFE, UCPH, CRX, IS, UHEI, L’OREAL, LHASA.