ACUTE TOXICITY ESTIMATION OF MULTICOMPONENT PLANT PROTECTION PRODUCTS USING CALCULATIONS, IN SILICO AND IN VIVO METHODS. PERSPECTIVES FOR UPDATING APPROACHES TO CLASSIFICATION AND RISK ASSESSMENT

Abstract

Aim of study is to assess different alternative approaches to acute toxicity assessment of PPP, including calculation based on the assumption of additivity and in silico methods. Acute toxicity of eight PPP, containing from 2 to 5 active ingredients (AI) pesticides and number of co-formulants was assessed. In vivo studies of acute toxicity were conducted according to OECD 425. To derive in silico prediction of LD50 of active ingredients EPA Toxicity Evaluation Software Tool (T.E.S.T.) was used. Calculation of Acute Toxicity Estimate for mixture was done using GHS additivity formula, taking into account in vivo LD50 of only active ingredients of PPP, of all ingredients of formulation and LD50 of active ingredients predicted by T.E.S.T. On the basis of in vivo results, calculations and in silico predictions PPP were classified according to GHS and Ukrainian Hygienic classification. Additivity coefficients were calculated to assess extent of interaction. Results of mentioned in vivo studies, in silico modelling and calculations are presented. Mean deviation of predicted by T.E.S.T LD50 values from experimental was 62,95%. Mean deviation of calculated ATE from in vivo LD50 for mixture was 74,5% (based on in vivo LD50 data for AI only), 21,1% (based on based on in vivo LD50 data for all ingredients) and 56,4% for ATE calculated using T.E.S.T predictions. Correlation coefficients for mentioned calculations were 0,69;0,84 and 0,60 respectively. Differences of calculated and tested values of acute toxicity estimates for eight multicomponent PPP did not lead to their misclassification in up to 75% of cases according to GHS when based on in vivo data. Differences in calculated values of acute toxicity estimates based on in silico predicted results lead to misclassification of the half of the formulations, however it may be lower if account to variability of experimental results and small number of mixtures tested here. Underestimation of the hazard according to GHS classification happened only in 12,5% of the mixtures studied here. Further studies will include assessment in the similar way as presented here of larger sample of multicomponent plant protection products and other mixtures representing wider range of acute toxicity categories and development of the list of mixture type-specific LD50 values for active ingredients (e.g. depending on solvents) and their application in the PPPs classification and risk assessment.

Keywords

• acute toxicity
• ld50
• plant protection products
• structure-activity relationship
• classification
• mixtures.