Abstract
BACKGROUND: The European Centre for the Validation of Alternative Methods (ECVAM) designed the Embryonic Stem Cell Test (EST) as a tool for classifying developmentally toxic compounds. An in vitro tool to assess developmental toxicity would be of great value to the pharmaceutical industry to help with toxicity‐associated attrition. METHODS: ECVAM's EST protocol was used, but employing a different mouse embryonic stem cell (ESC) line and an alternative differentiation medium. A subset of the compounds used to validate the EST assay along with a number of in‐house pharmaceutical compounds plus marketed pharmaceutical compounds were used to assess the EST performance with receptor‐mediated compounds. RESULTS: Our results with ECVAM compounds mirrored ECVAM's. Compounds that were developmentally toxic in vivo were classified by the EST as moderate risk. Overall, the accuracy was 75% with the current set of data and the predictivity of low‐, moderate‐, and high‐risk compounds was 90, 71, and 60% while the precision was 59, 86, and 100%, respectively. Interestingly, a number of the non‐developmentally toxic compounds had values for the 3T3 IC~50~ values, which were lower than the ESC IC~50~ and ID~50~, a situation not taken into account by ECVAM when designing the EST algorithm. CONCLUSIONS: The assay as currently constructed has a significant false‐positive rate (∼40%), but a very low false‐negative rate (∼7%). Additional moderate‐ and high‐risk compounds need to be assessed to increase confidence, accuracy, and understanding in the EST's predictivity. Birth Defects Res (Part B), 2008. © 2008 Wiley‐Liss, Inc.