In vitro micromass teratogen test: Results from a blind trial of 25 compounds

The results obtained by Huntingdon Research Centre participating in a blind trial of the micromass assay for the prediction of teratogenic potential are presented. Twenty-five coded compounds were tested without S-9 mix using a pre-agreed protocol; three compounds were later tested with S-9. The data were assessed for sensitivity, specificity and accuracy using three separate sets of criteria based on either concentration (the <500 mug/ml rule (i) and the <50 mug/ml rule (ii)) or specific inhibition of cell differentiation at relatively non-cytotoxic concentrations (the 2-fold rule (iii)). The best in vivo/in vitro correlation was obtained using the 2-fold rule; the <500 mug/ml rule was the most sensitive but gave a high false positive rate and the <50 mug/ml rule was of low overall accuracy (60%). It is suggested that selective inhibition of differentiation of one cell type and cytotoxicity at low dose levels may also indicate risk of embryo-foeto toxicity, a factor to be considered with the pharmacokinetics of the compound. The teratogens procarbazine, methotrexate and caffeine were not detected; diphenhydramine and furazolidone initially classified as non-teratogens in vivo, were predicted as teratogens by the micromass assay.