Route to route extrapolation factors for regulatory risk assessment - a probabilistic approach

In human risk assessment any relevant exposure route has to be addressed. Consequently, for numerous chemicals no-adverse effect-levels have to be determined for all three routes (dermal, oral and inhalation), leading to testing that demands a large number of animals. On the other hand, the 3R principles (reduction, refinement and replacement of animal testing) shall be implemented in risk assessment according to several regulations, e.g. REACh. A reduction of animal numbers could be achieved by using route-to-route (R2R) extrapolation factors (EF) to derive the relevant route-specific no-adverse-effect-level. One prerequisite for this approach is the use of scientifically sound extrapolation factors, which are based on a high quality and broad data basis. In our project, the toxicological database (DB) RepDose (including approx. 2900 studies on approx. 850 chemicals) served as the basis to derive R2R-EF. New peer-reviewed studies, also addressing the dermal route, were added to the DB to evaluate the distribution functions in a probabilistic approach and derive R2R-EF for all routes (oral, inhalation, dermal). Prior to the actual analysis, the consistency of the datasets was checked by analysing the effects of differences in study design on the EF. The distribution functions for the EFs inhalation to oral and dermal to oral were evaluated by a probabilistic approach. The final investigations aimed at the derivation of R2R-EF for the entire datasets and subcategories grouped on the basis of shared physico-chemical properties, structural similarity or a similar toxicological pattern.