The use of toxicokinetic and toxicodynamic data in risk assessment: an international perspective.

Risk assessment of chemicals is a process which is usually based on data derived from animal testing in which the exposure of animals results in toxicological effects. By extrapolation, the dose/exposure in humans, which will not result in toxicological effects ('safe dose', 'safe exposure'), is estimated. Traditional approaches use 'safety factors' or 'uncertainty factors' to extrapolate from animal to man and from the 'mean' subject to the general population, including sensitive subgroups. Traditionally, a default factor of 10 has been used to account for interspecies variation. It is proposed that this factor be subdivided into a subfactor to address the toxicokinetic aspects and a second subfactor for the toxicodynamic aspects. Likewise, a default factor of 10 with subfactors is proposed to account for the intraspecies variability. In the framework of the International Program on Chemical Safety's (IPCS) project on the Harmonization of Approaches to the Assessment of Risk from Exposure to Chemicals, an activity has been initiated to provide guidance to risk assessors on the use of quantitative chemical specific data to account for interspecies variation and interindividual variability in risk assessment. To address the toxicokinetic aspects, the active species, the relevant internal exposure and the adequate metrics must be considered. Data quality and availability, in vitro or in vivo, the route of administration and the relevant dose level are relevant information for interspecies extrapolation. The availability of experimental data, including the relevance of the population studied, the number of subjects and/or samples obtained in the relevant group allow one to estimate the population distribution, e.g. difference between central tendency and given percentiles. In a similar fashion, the toxicodynamic data must be addressed. In addition to the identification of the active chemical species, the relevant endpoint must be determined. In extrapolation from animal to man, in most of the cases, the definitive endpoint (e.g. anemia) is lacking. It can be substituted by in vitro data (e.g. in vitro hemolysis) if it is a key event and relevant for the toxicity in animal as well as in humans. In extrapolating from animal to man, the dose-effect relationship plays an important role. To account for the toxicodynamic variability in the human population, similar aspects have to be taken into consideration, which have been discussed for toxicokinetics. The IPCS document is available at the IPCS website and risk assessors are invited to use the framework and report back their experience with it to enable revision and improvement.