Correlation of data from the strain A mouse bioassay with long-term bioassays

Fifty-nine chemicals that had completed National Cancer Institute rat and mouse 2-year carcinogenicity tests were tested in the strain A mouse pulmonary tumor assay. Without knowledge of chemical identity, 53 chemicals were tested in strain A mice in one laboratory and 30 were tested in a second independent laboratory. Strain A tests on 24 of these chemicals were conducted in both laboratories. The strain A results were generally not predictive of the 2-year rat and mouse carcinogenicity test results. Furthermore, there was poor agreement of strain A results between the two laboratories. Although a variety of explanations may be invoked to explain the lack of concordance between the strain A tests and the 2-year rat and mouse tests, no one factor is sufficient to rationalize the poor concordance between strain A and 2-year carcinogenicity bioassay results.