Multistate models of developmental toxicity: Application to valproic acid-induced malformations in the zebrafish embryo

For the determination of acute toxicity chemicals in zebrafish ( Danio rerio ) embryos, OECD test guideline 236, relative to Fish Embryo Toxicity Test (FET), stipulates a dose-response analysis four lethal core endpoints and quantitative characterization abnormalities including their time-dependency. Routinely, data are analyzed at different observation times separately. However, observations given time strongly depend on previous effects should be jointly with them. To solve this problem, we developed multistate models for occurrence developmental malformations live events embryos exposed eight concentrations valproic acid (VPA) first five days life. Observations were recorded daily per embryo. We statistically infer model structure parameters using numerical Bayesian framework. Hatching probability rate changed compared forms its time-dependence; constant rate, piecewise fixed hatching 48 h post fertilization, variable time, as well Hill Gaussian form. A function adequately described data. The other transition rates conditioned embryo body concentration VPA, obtained physiologically-based pharmacokinetic model. VPA impacted mostly malformation hatched non-hatched embryos. Malformation reversion lowered by VPA. Direct mortality was low tested, but increased linearly internal concentration. makes full use gives finer grain teratogenic than OECD-prescribed approach. discuss obtaining toxicological reference values suitable inter-species extrapolation. general result is that complex can efficiently evaluated numerically. • Multistate improve They coupled PBPK chemical exposure. inference conducted values. exercised dose-time-response acid. integration.