How well can in vitro data predict in vivo effects of chemicals? Rodent carcinogenicity as a case study.

A recent research article by the National Center for Computational Toxicology (NCCT) (Kleinstreuer et al., 2013), indicated that high throughput screening (HTS) data from assays linked to hallmarks and presumed pathways of carcinogenesis could be used to predict classification of pesticides as either (a) possible, probable or likely rodent carcinogens; or (b) not likely carcinogens or evidence of non-carcinogenicity. Using independently developed software to validate the computational results, we replicated the majority of the results reported. We also found that the prediction model correlating cancer pathway bioactivity scores with in vivo carcinogenic effects in rodents was not robust. A change of classification of a single chemical in the test set was capable of changing the overall study conclusion about the statistical significance of the correlation. Furthermore, in the subset of pesticide compounds used in model validation, the accuracy of prediction was no better than chance for about three quarters of the chemicals (those with fewer than 7 positive outcomes in HTS assays representing the 11 histopathological endpoints used in model development), suggesting that the prediction model was not adequate to predict cancer hazard for most of these chemicals. Although the utility of the model for humans is also unclear because a number of the rodent responses modeled (e.g., mouse liver tumors, rat thyroid tumors, rat testicular tumors, etc.) are not considered biologically relevant to human responses, the data examined imply the need for further research with HTS assays and improved models, which might help to predict classifications of in vivo carcinogenic responses in rodents for the pesticide considered, and thus reduce the need for testing in laboratory animals.

Challenges in using the ToxRefDB as a resource for toxicity prediction modeling.

Developing and evaluating toxicity prediction models requires selection and use of datasets of known positive and negative agents for the endpoint(s) of interest. EPA's Toxicity Reference Database (ToxRefDB) is a publicly available dataset containing detailed study and effect information on more than 400 chemicals, and it has been used by EPA researchers to develop toxicity prediction models. During an initial evaluation of reproductive toxicity, however, limitations were uncovered in applying data from ToxRefDB that involved interpretation of toxicity effects and designation of toxicity endpoints, core attributes of the database that are critical to its use. These limitations for reproductive toxicity were found to be related, at least in part, to challenges faced in (1) evaluating the source of the original study data (EPA Data Evaluation Records (DERs)) for input into ToxRefDB and (2) interpretation of the biological significance of responses. These limitations of the ToxRefDB have important implications for the wider use of the database as it currently exists. Our results point to a need for improvements to the existing ToxRefDB and/or for researchers to independently evaluate, assign and verify positive or negative designations to data from ToxRefDB before use in development or validation of prediction models or testing frameworks.

An enhanced tiered toxicity testing framework with triggers for assessing hazards and risks of commodity chemicals.

This paper presents an enhanced integrated testing framework based on tiered testing and endpoint-specific decision triggers envisioned for application to commodity chemical safety assessments. The framework has two tiers in which exposure information can be integrated with hazard data at each Tier. Tier 1 tests are used to screen chemicals for major toxic effects (i.e., acute toxicity potential, target organs of repeat dose toxicity, genotoxicity potential, neurotoxicity potential, reproductive toxicity potential, immunotoxicity potential, and developmental toxicity potential), and to direct planning for more complex and targeted testing in Tier 2. The proposed decision triggers coupled with information on use and potential for exposure allow for scientifically-based decisions to be made about further testing in Tier 2, indicating which specific endpoints and tests warrant further evaluation, and which do not. The testing framework addresses risks to humans during all stages of development and provides data relevant to assessing hazards to sensitive subpopulations, such as infants and children. The REACH program in Europe and TSCA in the United States have led to an increased focus on development of hazard and risk information for chemicals used in industrial processes and consumer products. This framework and its toxicity decision triggers will allow for scientifically justified evaluation of chemicals that is comprehensive in terms of hazard screening, focuses resources on the specific complex tests that are most important for hazard characterization, and minimizes the use of animals.

Tiered toxicity testing: evaluation of toxicity-based decision triggers for human health hazard characterization.

A set of biologically-based toxicity testing decision triggers was developed and analyzed within a tiered testing and decision-making framework for evaluating potential human health hazards and risks associated with chemical exposures. The proposed three-tiered toxicity testing approach starts from a base set of toxicity studies (acute toxicity, in vitro genetic toxicity, in vitro cytogenetics, repeat dose/subchronic toxicity, developmental toxicity, reproductive toxicity) and then uses the toxicity triggers to identify which specific additional tests are needed to adequately characterize a substance's hazard potential. The toxicity triggers were initially evaluated using published information for eight chemicals, representing diverse classes. A retrospective validation study was then conducted using seven chemicals which had completed the USEPA's Voluntary Children's Chemical Evaluation Program (VCCEP). The toxicity triggers were shown to identify appropriate higher tier tests and to be reasonably predictive of the results expected in higher tiered tests. Employing these toxicity triggers within a tiered testing framework could lead to a reduction in the number of laboratory animals without diminishing the degree of scientific certainty necessary for hazard evaluations. The toxicity triggers appear to be suitable for identifying which specific endpoints and tests warrant further evaluation, and which do not, and for documenting the scientific basis for such decisions.

Humans appear no more sensitive than laboratory animals to the inhibition of red blood cell cholinesterase by dichlorvos.

Inhibition of red blood cell (RBC) cholinesterase is a consistent and sensitive indicator of exposure to dichlorvos (DDVP). Absent human data, default 10-fold adjustment factors for potential interspecies and intraspecies sensitivity differences would be used in developing a reference dose from the no observed effect levels for this endpoint obtained in toxicological assessments of laboratory animals. However, many studies of the cholinesterase-inhibiting effects associated with DDVP exposure have been conducted in humans, including healthy male volunteers, other healthy subpopulations, and diverse clinical subpopulations. Indeed, ample human data exist to permit a data-based assessment of potential interspecies sensitivity differences in RBC cholinesterase inhibition associated with DDVP exposure. In aggregate, these data demonstrate that the DDVP doses producing inhibition in humans are virtually identical to those eliciting the same levels of inhibition in laboratory rats, mice, monkeys, and dogs. Thus, healthy humans appear to be no more sensitive than laboratory animals to DDVP's effects on RBC cholinesterase, and an interspecies uncertainty factor of 1 is appropriate and scientifically warranted for use in DDVP risk assessments.