The results from a survey of the expert information that is publicly accessible on the use of the dog as test species during the regulatory evaluation of agricultural chemicals and pesticides are reported. Methods that are being used or considered in order to reduce the number of dogs used for this purpose are described. Regulatory evaluation aims at establishing threshold values for safe human exposure; it is based on no-observed-adverse-effect levels (NOELs) determined in animal studies. Current regulations require testing in two species, a rodent species (usually rat or mouse), and a non-rodent species (usually the dog). Subchronic (90-day) and chronic (12-month) repeated-dose feeding studies must be routinely conducted in dogs. This report first focuses on the results from a retrospective study analysing data on 216 pesticides kept on record by the Bundesinstitut für Risikobewertung, BfR (German Federal Institute for Risk Assessment), the competent regulatory authority in Germany. The study was sponsored and coordinated by SET, the German Foundation for the Promotion of Research on Replacement and Complementary Methods to Reduce Animal Testing (Stiftung zur Förderung der Erforschung von Ersatz-und Ergänzungsmethoden zur Einschränkung von Tierversuchen, Mainz) and conducted by the BfR. Since the data submitted for registration of a product is the property of the manufacturer, the study could only proceed with the collaboration of the German Association of Manufacturers of Agricultural Chemicals (Industrieverband Agrar, IVA). To ensure confidentiality, designated codes were used instead of the compounds' proper names when the study was published. The results support two major conclusions. The use of the dog for the testing of pesticides is indeed necessary because the dog has proved to be the most sensitive species for about 15% of the compounds examined. However, chronic studies are only of limited value since they only provide essential information that cannot be obtained in sub-chronic studies in about 5% of cases. These conclusions are supported by several retrospective analyses using data on pharmaceutical drugs carried out in the context of the International Conference on Harmonisation of Technical Requirements for the Registration of Pharmaceuticals for Human Use (ICH). Over 90% of drugs elicited no toxic symptoms in 12-month studies in dogs in addition to those that had been recorded previously in studies conducted for 90 or 180 days in dogs and rats. Another approach comparing the results from pre-clinical animal studies with clinical studies noted that animal studies predicted about 70% of the effects observed in volunteers, and in about 94% of cases the effects occurred in animal studies lasting not more than one month. Furthermore, the report summarises the current methods under consideration that could refine or reduce the use of dogs in toxicity testing: industrial data sharing and harmonisation of guidelines, in vitro methods, human studies, computational prediction models, and integrated testing approaches. The integrated Agricultural Chemicals Safety Assessment (ACSA) testing scheme, which is currently being developed in an international project initiated by the International Life Sciences Institute (ILSI, USA), is of particular relevance, since an ambitious attempt is being made to design a new comprehensive test framework incorporating modern scientific approaches and covering most aspects of current regulatory testing requirements. The ACSA project has access to the pesticide database of the US EPA's Office of Pesticide Programs (OPP). Preliminary results have confirmed the two major conclusions from the joint SET/BfR study conducted in Germany. Taking these results into account, it is recommended that the regulatory requirement for 12-month studies to be routinely carried out in dogs should be abandoned. While 90-day studies should be conducted in both rats and dogs, chronic studies should only take place in rats. If the dog is more sensitive than the rat in the 90-day study, an additional safety factor to the NOEL value obtained in the chronic rat study should be applied in order to set the threshold for safe human exposure, instead of conducting a 12-month study in dogs. This safety factor may be calculated from chronic NOEL data available in several pesticide databases. Chronic tests using dogs would then only be required if the test compound belongs to a new class of chemicals that has never been tested before. Thus, the report concludes that, according to current scientific knowledge, the routine 12-month studies in dogs are no longer required for agricultural chemicals and pesticides, and international regulations should be changed accordingly. Active international support of such measures is welcomed, from both an economical and an animal welfare perspective.
