Interpretation of fish biomarker data for identification, classification, risk assessment and testing of endocrine disrupting chemicals

Chemical induced changes in fish biomarkers vitellogenin (VTG), secondary sex characteristics (SSC), and sex ratio indicate modes/mechanisms of action (MOAs) of EAS (estrogen, androgen and steroidogenesis) pathways. These biomarkers could be used for defining MOAs and the causal link between MOAs and adverse effects in fish for the identification of endocrine disrupting chemicals (EDCs). This paper compiled data sets of 150 chemicals for VTG, 57 chemicals for SSC and 38 chemicals for sex ratio in fathead minnow, medaka and zebrafish. It showed 1) changes in fish biomarkers can indicate the MOAs as anticipated; 2) in addition to EAS pathways, chemicals with non-EAS pathways induced changes in fish biomarkers; 3) responses of fish biomarkers did not always follow the anticipated patterns of EAS pathways. These responses may result from the interaction of chemical-induced multiple MOAs and confounding factors like fish diet, infection, culture conditions, general toxicity and stress response. The complex response of fish biomarkers to a chemical of interest requires EDC testing at multiple biological levels. Interpretation of fish biomarker data should be combined with relevant information at different biological levels, which is critical for defining chemical specific MOAs. The utility of fish biomarker data for identification, classification, PBT assessment, risk assessment, and testing of EDCs in the regulatory context was discussed. This paper emphasizes the importance of fish biomarker data in the regulatory context, a weight of evidence approach for the interpretation of fish biomarker data and the need for defining levels of evidence for the identification of EDCs. •Chemicals interfering with EAS pathways induced changes in fish biomarkers.•Chemicals interfering with non-EAS pathways induced changes in fish biomarkers too.•Confounding factors (diet, infection, culture conditions, stress) were discussed.•Use of biomarker data for identification, classification, risk assessment, testing was discussed.