You are here:
Integrative Approaches to Evaluating Neurotoxicity Data for Risk Assessment.
Citation:
BOYES, W. K., P. J. BUSHNELL, AND V. A. BENIGNUS. Integrative Approaches to Evaluating Neurotoxicity Data for Risk Assessment. Presented at Joint meeting of the 20th International Conference on Epidemiology in Occupational Health & 10th International Symposium on Neurobehavioral Methods and Effects in Environmental and Occupational Health, San Jose, COSTA RICA, June 09 - 13, 2008.
Impact/Purpose:
We have developed several quantitative methods for approaching these problems.
Description:
Risk assessment classically has been based on single adverse outcomes identified as the Lowest Observable Adverse Effect Level (LOAEL) or the highest dose level in a credible study producing a No Observable Adverse Effect Level (NOAEL). While this approach has been useful overall, it can cause difficulties when a single study is identified as the critical study. For example, it is often unclear whether the critical study contained the most sensitive outcome because it had the most sensitive methodology and best experimental design, or because of a statistical Type 1(11)error that identified an effect where none actually existed. In addition, the selection of a single outcome from a single study fails to benefit from the total data available, Toxicologically important considerations are external to the calculation, such as dose-response relationships or the consistency and replication of outcomes across studies. When benchmark values are calculated, the shape and variability of the dose-response curve are quantitatively integrated, but alternative approaches to synthesize more of the available data and describe additional quantitative relationships for extrapolations across factors such as dose, exposure duration, species and chemicals with similar actions are potentially quite useful. We have developed several quantitative methods for approaching these problems. Research has shown that the momentary brain concentration of volatile organic solvents is sufficient to predict their acute neurotoxic actions. This information has allowed quantitative models to be developed that predict doses giving rise to adverse effects across exposure durations and across species. In addition, by modeling the results from multiple studies as a function of internal dose metrics, it was possible to conduct meta-analysis of the effects of exposure across the body of available literature, thus creating dose-response functions not dependent on single "critical" studies. These models, based on internal concentrations, also provide a platform to integrate results from in vitro experiments to those of in vivo assessments based on common target tissue concentrations. Development of successful integrative models is dependent on several factors including; (1) identification of the internal dose metric that is predictive of toxicity (peak concentration, area under the curve, etc), (2) verified pharmacokinetic models capable of predicting internal dose, (3) reports in the literature that are sufficiently documented to be able to derive estimates of internal doses, and. (4) comparable neurotoxic outcomes across studies, or the ability to transform different outcomes into suitably comparable dependent measures. This set of integrative approaches can be used where the data are of sufficient quantity and quality to obtain quantitative risk predictions that were not previously possible.