A category approach to predicting the repeated-dose hepatotoxicity of allyl esters

► A category approach was tested to predicting the hepatotoxic effects of allyl esters. ► A category was built using a new toxicity database and mechanistic information. ► Untested analogs were categorized by predicting a key event linked to the toxicity. ► Hepatotoxicity of the untested ones was estimated based on available toxicity data. We tested a category approach to predict the hepatotoxic effects of repeated doses of allyl esters using a new database for repeated-dose toxicity. Based on information on hepatotoxic mechanism of allyl acetate, the category was defined as allyl esters that are hydrolyzed to allyl alcohol. Allyl alcohol is readily oxidized to acrolein in the liver, causing hepatotoxicity. Seventeen marketed allyl esters were obtained and grouped into category by identifying or predicting allyl alcohol formation. Allyl esters with a saturated straight alkyl carboxylic acid moiety (allyl acetate, hexanoate and heptanoate as tested species, and allyl butyrate, pentanoate, octanoate, nonanoate and decanoate as untested species) are likely similar in rate of ester hydrolysis, thereby defining subcategory 1. NOAEL and LOAEL for the hepatotoxic effects were estimated at 0.12 and 0.25mmol/kg/d for the untested species, based on those of allyl acetate. The remaining nine allyl esters with other alkyl or aromatic carboxylic acid moieties were placed in subcategory 2: their hepatotoxicity levels were not predictable due to an unclear match between their degree of structural complexity and rate of hydrolysis. Our results demonstrate the usefulness of the category approach for predicting the hepatotoxicity of untested allyl esters with saturated straight alkyl chains.