Consideration of dosimetry in evaluation of ToxCast™ data

Abstract

The US Environmental Protection Agency (US EPA) Toxcast™ program has the stated goal of predicting hazard, characterizing toxicity pathways and prioritizing the toxicity testing of environmental chemicals through the use of in vitro high-throughput screening (HTS) assays. This analysis integrates data from biomonitoring and from in vivo toxicity and pharmacokinetic studies to examine the physiological relevance of the tested and responding in vitro concentrations for five case study chemicals: triclosan, 2,4-dichlorophenoxyacetic acid, perfluorooctanoic acid, monobutyl phthalate and mono-2(ethylhexyl)phthalate. This analysis also examines the ToxCast™ phase 1 data set for approximately 50 chemicals belonging to four 'common mechanism groups' which have been the subject of cumulative risk assessments by the US EPA for both the pattern of key responses and the relative potencies of included chemicals compared with the in vivo relative potencies. Responding concentrations in vitro were generally in the range of serum or plasma concentrations associated with no-observed to lowest-observed effect levels for the case study chemicals, while available biomonitoring data demonstrating actual exposures were generally lower. ToxCast™ assay endpoints related to acetylcholinesterase (AChE) inhibition had low sensitivity for detecting organophosphate pesticides but good sensitivity for detecting N-methyl carbamates. However, in vitro relative potencies did not correlate with in vivo potency. Both qualitative and quantitative predictive power is probably affected by the lack of comprehensive metabolic activity in most current in vitro systems explored in the ToxCast™ program, and this remains a fundamental challenge for high-throughput toxicity screening efforts.