Use of anatomical and kinetic models in the evaluation of human food additive safety.

Toxicological testing in animals is relied upon as a surrogate for clinical testing of most food additives. Both animal and human clinical test results are generally available for direct additives when high levels of exposure are expected. Limited animal studies or in vitro test results may be the only sources of toxicological data available when low levels of exposure (microg/person/day) are expected and where no effects of the additive on the food itself are desired. Safety assessment of such materials for humans requires mathematical extrapolation from any effects observed in test animals to arrive at acceptable daily intakes (ADIs) for humans. Models of anatomy may be used to estimate tissue and organ weights where that information is missing and necessary for evaluation of a data set. The effect of growth on target tissue exposure during critical phases of organ development can be more accurately assessed when models of growth and known physiological changes are combined with pharmacokinetic results for test species. Kinetic models, when combined with limited chemical property, kinetic, and distribution data, can often be used to predict steady-state plasma and tissue levels of a test material over the range of doses employed in chronic studies to aid in interpretation of effects that are often nonlinear with respect to delivered dose. A better understanding of the reasons for nonlinearity of effects in animals improves our confidence in extrapolation to humans.