Assessment of the Halle Register for classifying and labelling new chemicals according to the EU toxicity classes.

To reduce the number of animals used in acute oral toxicity testing, cytotoxicity data (IC50) can be used to predict the in vivo acute toxicity of chemicals by using statistical modeling approaches between the IC50 values and acute oral LD50 values in the Register of Cytotoxicity (RC).Here, we propose strategies for using in vitro data as a basis for classifying and labelling new chemicals representative of all 3 EU toxicity classes as well as the unclassified category and evaluate the probabilities of correct classification depending on the true LD50 and depending on the neighbourhood of class limits.Main topics will be the quality of data, in-vitro as well as in-vivo, and its impact on modelling approaches. IC50 (RC) and LD50 values (ICCVAM validation study using 72 chemicals) and associated 95% confidence limits were calculated for 347 resp. 72 chemicals. LD50 values of the RC, listed in the NIOSH Registry of Toxic Effects of Chemical Substances (RTECS) and LD50 values selected by ICCVAM will be compared.