Thought-Starter: The concept of human exposure-based toxicity testing for agrochemicals

摘要

Agrochemicals are thoroughly evaluated to ensure human safety. Despite efforts to introduce in vitro methods, animal studies identify toxicological hazards and set reference doses for human health risk assessments. Very high dose levels are employed in animals, with no relevance to realistic human exposure scenarios. A relevant, exposure-based and 3Rs-focused approach would be to limit high dose levels to a fixed multiple of human exposure; this assessment assumed a limit of 1000-times the predicted human exposure level. A dataset was established to evaluate how this approach would have impacted the dose setting for toxicity tests of existing agrochemicals. Chronic and acute dietary exposure scenarios were considered for 161 agrochemical reviews published by EFSA (of these 140 had ADIs and 100 ARfDs). Values were obtained for ADI, ARfD, critical NOAEL and predicted human exposure levels. The results showed that for 67% of ADIs and 59% of ARfDs there was no toxicity at 1000-times the predicted maximum human exposure level. In the remaining cases, 1000X human exposure exceeded the critical NOAEL, but the majority of the exposure levels were still lower than the critical LOAEL. For the limited number of chemicals where exposure levels were higher than the critical LOAEL, a case-by-case analysis investigating the expected nature of effects was conducted. In conclusion, the majority of the agrochemicals reviewed in this study were tested in animals at high doses not relevant to human exposure. These high doses often result in unnecessary, non-specific toxicity, and understanding the scientific relevance of these effects at doses with no relationship to actual human exposures leads to additional animal testing. These data highlight the potential refinement that human exposure-based testing could provide to the safety assessment of agrochemicals from a 3Rs and scientific perspective.