Toxicity data collected in a laboratory setting are the primary source of potency information used for regulatory, modeling, or risk assessment purposes. However, the relative reproducibility of such toxicity data is rarely discussed. This study investigated the reproducibility of growth impairment data for the freshwater ciliate Tetrahymena pyriformis exposed to a structurally diverse group of chemicals of varying hydrophobicity within different modes of toxic action, either non-covalent narcosis or covalent electro(nucleo)philicity. The proportions of chemicals representing each mode of toxic action, or mechanism of action within each mode, were not chosen to emulate the occurrence of manufactured chemicals or chemicals within the TETRATOX database. Chemicals for which prior toxicity data existed were re-tested and reproducibility was evaluated. The toxic potency values of the selected chemicals were largely reproducible after re-testing of the toxic potency, as 98% of the chemicals had re-test toxicity values within one log unit of the original potency value. To further scrutinize the reproducibility of toxicity values, differences between values were investigated by mode of toxic action. A stringent criterion for reproducibility was enforced, which dictated that the re-tested toxicity value must be encompassed by the fiducial interval (FI) of the original toxicity value and vice versa for the chemical to be considered reproducible. Toxicity values of 28 of the 50 re-tested chemicals conformed to the criterion set for reproducible values. Of the nonreproducible chemicals, seven were narcotics: four nonpolar or neutral narcotics and three other narcotics (e.g. polar narcotics). However, four of these seven narcotics did have toxicity values encompassed by one FI, but not the other FI. The remaining chemicals that did not have reproducible potency measurements were electro(nucleo)philic in nature. Certain toxicophores were highly represented among these chemicals. These included quinone derivatives, electron releasing amino and hydroxyl moieties, and electron withdrawing nitro substituents, often in tandem with strong leaving groups (i.e. halogens), and unsaturated alcohols. Lack of reproducibility was common among the chemicals that elicited toxicity after either abiotic or biotic transformation. There was no clear trend between hydrophobicity and lack of reproducibility. While data are limited, these results suggest that toxic potency values of chemicals acting via the electro(nucleo)philic mode of toxic action could be more susceptible to non-reproducibility. Ramifications of such lack of reproducibility could manifest in predictive toxicology models and their use in regulatory and risk assessment endeavors.
Available at: http://works.bepress.com/terry_schultz/67/