Screening chemicals for thyroid-disrupting activity: A critical comparison of mammalian and amphibian models

In order to minimize risks to human and environmental health, chemical safety assessment programs are being reinforced with toxicity tests more specifically designed for detecting endocrine disrupters. This includes the necessity to detect thyroid-disrupting chemicals, which may operate through a variety of modes of action, and have potential to impair neurological development in humans, with resulting deficits of individual and social potential. Mindful of these concerns, the consensus favors in vivo models for both hazard characterization (testing) and hazard identification (screening) steps, in order to minimize false negatives. Owing to its obligate dependence on thyroid hormones, it has been proposed that amphibian metamorphosis be used as a generalized vertebrate model for thyroid function in screening batteries for detection of thyroid disrupters. However, it seems unlikely that such an assay would ever fully replace in vivo mammalian assays currently being validated for human health risk assessment: in its current form the amphibian metamorphosis screening assay would not provide capacity for reliably detecting other modes of endocrine-disrupting activity. Conversely, several candidate mammalian screening assays appear to offer robust capacity to detect a variety of modes of endocrine-disrupting activity, including thyroid activity. To assess whether omission of an amphibian metamorphosis assay from an in vivo screening battery would generate false negatives, the response of amphibian and mammalian assays to a variety known thyroid disrupters, as reported in peer-reviewed literature or government agency reports, was critically reviewed. Of the chemicals investigated from the literature selected (41), more had been tested in mammalian studies with thyroid-relevant endpoints (32) than in amphibian assays with appropriate windows of exposure and developmental endpoints (27). One chemical (methoxychlor) was reported to exhibit thyroid activity in an appropriate amphibian assay in the absence of corresponding activity in a suitable mammalian assay, whereas none of the chemicals surveyed were reported as thyroid active in mammalian assays but reported as negative in an appropriate amphibian assay. Given the need for one or more in vivo mammalian assays for screening chemicals for detecting (anti-)estrogenic/(anti-)androgenic activity and effects on steroidogenesis, inclusion of an in vivo amphibian assay specifically to detect thyroid disrupters would likely be as an addition to these mammalian assays. As there is little convincing evidence that an amphibian screening assay would detect significant numbers of thyroid-active chemicals not picked up by mammalian assays, the routine use of an amphibian metamorphosis assay at screening level could introduce unnecessary and unjustified redundancy into chemical safety assessment programs, when there is pressure to reduce animal use in toxicity testing.