Reliability of the TTC approach: Learning from inclusion of pesticide active substances in the supporting database

Data on pesticide active substances were used to assess the reliability of the Threshold of Toxicological Concern (TTC) approach. Pesticides were chosen as a robust test because of their potential for toxicity. 328 pesticide substances were classified on the basis of their chemical structure, according to the generic scheme proposed by the European Food Safety Authority. 43 carbamates and organophosphates were allocated to the group for neurotoxicity alerts, and 279 substances to Cramer structural Class III. For Class III, the 5th percentile value as calculated from the cumulative distribution curve of the no-observed-effect levels (0.20 mg/kg bw per day), was slightly higher than that determined by Munro (0.15 mg/kg bw per day) from his original database. The difference is explained by the inclusion of carbamates and organophosphates in Munro's Class III. Consideration of the acceptable daily intakes and their underlying toxicity data showed that the TTC approach is conservative for 96.2% of the substances. Overall, this analysis gives added support to the utility of the generic scheme of application of the TTC approach for hazard assessment of substances for which few or no experimental toxicity data are available. A convenient alternative to the Cramer decision tree is proposed.