Citation:

Hines, D., R. Conolly, AND A. Jarabek. A mechanistic approach for source-to-outcome risk assessment across human health and ecological endpoints using the Aggregate Exposure Pathway (AEP) and Adverse Outcome Pathway (AOP) frameworks. SOT RASS, MD, Annapolis, November 14, 2018.

Impact/Purpose:

This work expands a previous case study, which focused on integration of toxicity mechanisms using the Adverse Outcome Pathway (AOP) for a data rich chemical (perchlorate anion, ClO4-), and suggested linking this toxicity data to a hypothetical exposure model to demonstrate a quantitative source-to-outcome approach. Here, we extend the AOP component to do that by developing the Aggregate Exposure Pathway component using a fate and transport model to represent the movement of this chemical through a hypothetical wetland. We quantitatively describe external exposure under three different hypothetical contamination scenarios for humans, small mammals, and fishes. We then apply previously published, species-specific Physiologically Based Pharmacokinetic (PBPK) models to link external exposure predictions to internal exposure in these three groups of organisms, and demonstrate how these predictions, along with the integrated AEP and AOP frameworks, can inform a comprehensive source-to-outcome approach to CRA that includes both human health and ecological endpoints together.

Description:

Exposure to environmental contaminants can lead to adverse outcomes in humans as well as important non-human species. However, evaluating risk across multiple species in a contaminated environment can be challenging due to differences in relevant exposure pathways, behavior, physiology, and toxicity across organisms. In this work, we construct a conceptual model based on existing frameworks and develop mechanistic techniques to help address some of the challenges of integrating human health and ecological endpoints into risk assessments. Specifically, we combine the Aggregate Exposure Pathway (AEP) and Adverse Outcome Pathway (AOP) frameworks to link external exposures, toxicokinetics, toxicodynamics, and adverse outcomes for multiple species in an environment, and then present a case study of a hypothetical location that receives contamination through surface water, groundwater, and atmospheric inputs to demonstrate our approach. This case study focuses on the perchlorate anion, a well-documented inhibitor of iodide uptake into the thyroid across vertebrate species, and considers three organism groups: humans, fish, and small herbivorous mammals. The external exposure pathways contained in the AEP component of the case study were quantified using a fate and transport network that was constructed from a set of six ordinary differential equations, one for each environmental compartment considered (surface water, groundwater, soil, aquatic plants, terrestrial grasses, and terrestrial shrubs). We used a Monte Carlo approach to address uncertainty in parameter values and to provide a range of predicted perchlorate concentrations for each environmental compartment, as well as species-specific external exposures for the hypothetical site. Network flow analyses were applied to the AEP network to identify the relative source contribution of perchlorate for each species. Previously published species-specific Physiologically Based Pharmacokinetic (PBPK) models were then used to describe the toxicokinetics that link external exposures to internal concentrations within each organism. We applied the AOP framework to organize published dose-response data for mechanistic endpoints leading from toxicodynamic interactions to adverse outcomes in each species, and then linked the predicted exposures to dose-response data to evaluate species-specific effects. The case study considered three contamination scenarios: a mild contamination scenario with perchlorate concentrations lower than documented sites, a moderate contamination scenario with ten times the perchlorate inputs as the mild scenario, and a high contamination scenario, which simulated a groundwater spill and had 100 times the groundwater perchlorate input as the mild scenario. Despite similar dose-response data for some endpoints across species, results predicted no adverse outcomes for humans or fish in the mild contamination scenario for the hypothetical site, but possible effects in small herbivorous mammals, due to differences in external exposure pathways. Predicted effects increased in all species across scenarios as perchlorate contamination increased, but the magnitude of species-specific effects differed based on sources of perchlorate due to differences in exposure pathways. The case study presented in this work demonstrates that the joint AEP-AOP conceptual model can provide a source-to-outcome construct that is useful for 1) organizing mechanistic data for exposure and toxicity across species, 2) highlighting knowledge gaps, 3) quantifying uncertainties, and 4) evaluating the effects of contaminants on multiple species in an environment. We discuss the applications of this approach for informing cumulative risk assessment (CRA) and regulatory decision making.

Record Details: