Although regression analyses between the acute oral or intraperitoneal LD50 values for both rats (Rattus norvegicus) and mice (Mus musculus) and the 96‐h LC50 values for fathead minnows (Pimephales promelas) or rainbow trout (Salmo gairdneri) for a broad spectrum of environmental chemicals were statistically significant, the large variability in the regressions limits their practical utility for interspecies extrapolations. To gain a better understanding of this variability, analyses were performed to reveal the structural features characteristic of those chemicals that deviate substantially from the interspecies regressions. The criterion selected to indicate excessive variability was a greater‐than‐fourfold difference between the observed rodent LD50 and that estimated from the regression against the fish data. Discriminant function analyses showed aldehydes, esters and organophosphorus insecticides to be the primary outliers in the fish‐to‐rodent comparisons. Rodents were generally less sensitive to intoxication with carbonyl‐containing compounds but more sensitive to organophosphate anticholinesterase poisoning than were fish. This species‐selectivity for both classes of chemicals correlates with species‐related differences in metabolic clearance and/or pharmacodynamic differences in target organ sensitivities. Since these same idiosyncracies in metabolic disposition and receptor sensitivity would be suspected to be operative for structurally related chemical compounds, interspecies extrapolations of toxicity information for both carbonyl compounds and for anticholinesterase agents should take into consideration these important factors, which confer certain degrees of species‐selectivity for these chemical classes.
- Interspecies comparative toxicity
- Interspecies extrapolation
- Structure‐activity relationships