Microbial and Phenotypic Definition of Rats and Mice: Proceedings of the 1998 US/Japan Conference (1999)
Chapter: CIEA/NCRR/NIH Genetic and Microbiological Monitoring of Mouse and Rat Resources: Directions for the Future
responsiveness to various pharmaceutical products. A set of similar studies with outbred animals may show a difference in response to treatment at some dose level. For example, if a chronic toxicity study is performed in a contract facility by gavage, and a second study for dose ranging is done in the in-house facility by dietary administration using dietary restriction (because it is much less expensive to plan the carcinogenicity study by dietary administration), and the studies result in different outcomes regarding the toxicity of the product, one will not know whether those differences are related to the modalities of the ingestion of the drug, the husbandry of the stock being treated, outbred differences from different sources, dietary restriction, or some other factor.
Need for Integrated Findings
When we interpret carcinogenicity studies, we no longer can simply state, ''It was positive in this site, and it was negative in all of the other sites.'' As a part of the ICH guidance for pharmaceuticals (with which the US Environmental Protection Agency [EPA] agrees), we now perform integrated assessments of carcinogenicity—weight of the evidence. It is necessary to consider all of the data collected on the product and try to make a determination of whether or not those findings are important for human risk. In so doing, it is often necessary to rely on historic data to interpret the findings. When looking at historic data, it is important to note its relevance to the particular strain of animal and, unfortunately, to the animal stock.
Managing Changes Over Time
For some facilities that have been testing for 20 years, changes over time can clearly be seen in the historic response rate for spontaneous tumor incidence. Other facilities have no historic data but instead, rely on published data. How relevant is that published data to that in-house-contained and bred outbred or inbred strain? The answer to such a question is critical for the assessment and determination of the carcinogenic risk for humans. It is imperative for us to understand and manage test data accurately.
One approach to managing test results is to try to control as many variables as possible. That method is the best solution—to try to control the dietary supplies, the strain of animals, the stock of the animals, the dosing regimens, and the like so that one can say that the corresponding data are reliable for interpretation. If a second person repeats the experiment or study, the person can be reasonably confident of obtaining the same result. However, controlling all of the variables is probably not feasible in the global development arena. How can you ensure that all animals have the same diet, no matter where the test facilities are located? I believe we can at least begin in this direction by controlling stocks of animals.
Nevertheless, we must be very careful not to select a single stock of a single strain of animal and proceed with that alone. It is known that some animal strains metabolize and respond to pharmaceuticals more like humans than do other strains. Having those strains for testing to find the most relevant response is important so that resulting data from a well-controlled strain is relevant to humans. If we pick only one stock and standardize it, we will loose a great deal because even though we will have a reproducible result with that strain, we may not want to use that strain for much of our testing.
