A comparative assessment of the CLP calculation method and in vivo testing for the classification of plant protection products.

In Europe, animal testing for the purpose of regulatory plant protection product (PPP) assessment should be undertaken only as a last resort. Nevertheless, there is a need to improve the acceptance of alternative methods, which has been slow due to a lack of data regarding the predictivity of in vivo effects. The CLP calculation method is an alternative method based on the concentration addition of all adverse substances in a mixture. It is often applied as a conservative approach for the estimation of toxicodynamic interactions. However, PPPs consist of pesticides and co-formulants, which in combination can also exhibit altered toxicokinetic properties. Our analysis revealed that oral and inhalation toxicity was underestimated for approximately 45% of the in vivo classified products by the CLP calculation method as compared to in vivo testing. With regard to skin and eye irritation, the CLP calculation method underestimated the irritating potential in 22% and 6% of PPPs, respectively. Based on specific concentration limits, skin sensitisation was underestimated in 34% of PPPs. Similar false negative rates have been reported for PPP in vitro testing. Hence, we suggest the development of an integrated assessment strategy, weighing all available information and considering relevant parameters influencing predictivity and uncertainty.

Comparative cytotoxicity of plant protection products and their active ingredients.

Toxicological testing of plant protection products (PPPs) is a legal requirement in the EU. The whole PPP formulation is tested for acute endpoints in vivo during approval procedure of PPPs. However, alternative methods such as the CLP calculation method (CM) are employed increasingly. In the first part of this study we analysed PPPs for the correlation of GHS classifications resulting mainly from in vivo LD50-values with classifications obtained from calculated LD50-values using the CM. Accordingly, the CM predicted 80% of the PPPs correctly. However 31% of classified products were not identified revealing a considerable inaccuracy of this method. Based on these results ten PPPs and corresponding ASs were further tested in a cytotoxicity assay employing 3T3 and hFF cells (one PPP and corresponding AS were tested in HepaRG cells). The study outcome revealed that the cytotoxicity data did not reliably reflect differences in toxicity between ASs and PPPs. Especially organic solvent based formulations demonstrated a higher cytotoxicity than water based formulations independently to their toxicity in vivo. Overall, the cytotoxicity test did not provide a more robust method than the CM. However, the database consisting of ten PPPs was small and therefore no robust conclusions can be drawn.