Struggles for equivalence: in vitro developmental toxicity model evolution in pharmaceuticals in 2006.

Our group has been using the ECVAM Embryonic Stem Cell assay to predict developmental toxicity. In order to improve the separation of non-teratogens from weak teratogens, we have employed measures of gene expression, and different statistical methods from those originally used to develop the test. These approaches have fundamentally not improved the discrimination of 'weaks' from 'nons'. A realization that a very low value for cytotoxicity IC50 would drive a final result for the test in ways that were inappropriate for pharmaceuticals has led us to re-examine the cytotoxicity component. Our current efforts are focused on other, perhaps more sensitive, measures of cytotoxicity, combined with gene expression changes in mouse stem cells in an attempt to correctly identify weak teratogens and non-teratogens.