A Retrospective Appraisal of the Ability of Animal Tests To Predict Reproductive and Developmental Toxicity in Humans

Abstract

The reproductive and developmental toxicology areas have undergone numerous changes in the 30 years since the thalidomide tragedy. It would be comforting if such changes have resulted in a greatly decreased likelihood that human conceptuses will develop malformations. However, there is little evidence that such testing has better enabled us to identify human reproductive or developmental toxicants. Although there may be justification for performing animal testing on drugs and pesticides, a simple acute toxicity test may be just as predictive as currently employed reproductive and developmental toxicity tests for other synthetic chemicals. Support for this premise is provided by comparing available acute toxicity information on many compounds, identified as (potential) human developmental or reproductive toxicants, with their respective likely human threshold doses. Acute LD50values of other potential human reproductive or developmental toxicants, not documented as being a threat to humans, also were compared with their time-weighted averages or their lowest teratogenic dose in the same species. The data generated support the premise that most reproductive or developmental toxicants are a potential threat to humans only if exposure levels approach those that are lethal in the rat or mouse. In contrast, unequivocal evidence of developmental toxicity for drugs, such as thalidomide and isotretinoin, to which the human embryo is highly sensitive, were not apparent in the screens used now for routine developmental toxicology testing. Thus, requirements that all chemicals be tested for reproductive and developmental toxicity should await the development of testing methodology capable of better assessing human risk.