Abstract
Current efforts to assess human health response to chemicals based on high-throughput in vitro assay data on intra-cellular changes have been hindered for some illnesses by lack of information on higher-level extracellular, inter-organ, and organism-level interactions. However, a dose-response function (DRF), informed by various levels of information including apical health response, can represent a template for convergent top-down, bottom-up analysis. In this paper, a general DRF for chronic chemical and other health stressors and mixtures is derived based on a general first-order model previously derived and demonstrated for illness progression. The derivation accounts for essential autocorrelation among initiating event magnitudes along a toxicological mode of action, typical of complex processes in general, and reveals the inverse relationship between the minimum illness-inducing dose, and the illness severity per unit dose (both variable across a population). The resulting emergent DRF is theoretically scale-inclusive and amenable to low-dose extrapolation. The two-parameter single-toxicant version can be monotonic or sigmoidal, and is demonstrated preferable to traditional models (multistage, lognormal, generalized linear) for the published cancer and non-cancer datasets analyzed: chloroform (induced liver necrosis in female mice); bromate (induced dysplastic focia in male inbred rats); and 2-acetylaminofluorene (induced liver neoplasms and bladder carcinomas in 20,328 female mice). Common- and dissimilar-mode mixture models are demonstrated versus orthogonal data on toluene/benzene mixtures (mortality in Japanese medaka, Oryzias latipes, following embryonic exposure). Findings support previous empirical demonstration, and also reveal how a chemical with a typical monotonically-increasing DRF can display a J-shaped DRF when a second, antagonistic common-mode chemical is present. Overall, the general DRF derived here based on an autocorrelated first-order model appears to provide both a strong theoretical/biological basis for, as well as an accurate statistical description of, a diverse, albeit small, sample of observed dose-response data. The further generalizability of this conclusion can be tested in future analyses comparing with traditional modeling approaches across a broader range of datasets.
Highlights
Identification of chemicals that may pose a health risk following chronic decadal exposure to extremely low doses is a challenge [1], complicated further by possible toxicological interactions among chemicals and other health stressors
One approach proposed for setting regulatory standards based on high throughput screening (HTS) data has been systems biology-based modeling to determine concentrations that would likely lead to excessive perturbation of intracellular pathways, physiologicallybased pharmacokinetic (PBPK) modeling to assess concentrations that would cause adverse effects in humans [4]
The general form of the linearized multistage model is based on the assumption that “the time from cancer initiation in a single cell until an observable cancer develops in a tissue is . . . functionally independent of the dose rate” [29], and more generally on the assumption of independent numbers/sizes of initiating events, or causes, along a mode of action (MOA) through its basis in the original multistage model [31]
Summary
Identification of chemicals that may pose a health risk following chronic decadal exposure to extremely low doses is a challenge [1], complicated further by possible toxicological interactions among chemicals and other health stressors. Traditional high-dose animal tests have been expensive in terms of time, expense, and animal subjects, may induce extraneous responses such as cytotoxicity, and in any case require extrapolation to low doses of regulatory interest To address these issues, the U.S National Research Council [1] recommended assessment based on high-throughput in vitro assays targeting intra-cellular processes, and as a result such testing is producing large databases of high throughput screening (HTS) data [2,3]. Functionally independent of the dose rate” [29], and more generally on the assumption of independent numbers/sizes of initiating events, or causes, along a mode of action (MOA) through its basis in the original multistage model [31] These are important constraints not typically satisfied in complex systems [32]. Log-log linear probability distributions must be truncated or otherwise
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
