This series of analyses provided several interesting results. First, while the use of two or more assays for mutation increases the correlation with the TD50 beyond that obtained using the Ames test alone, the increase was not statistically significant. The use of mutation and acute toxicity data combined did however yield a significantly higher correlation (0.76 = r = 0.85, depending on the chemicals selected) than was obtained with the use of mutation or acute toxicity data alone. When the analysis was restricted to carcinogens affecting specific target organs (lung or liver), correlation coefficients in the neighborhood of r = 0.9 were obtained. Using all of the RTECS assays, the correlation of the composite relative potency index with the minimum TD50 across sites was r = 0.80, 0.87 or 0.79, depending on whether data for rats, mice, or the most sensitive species was used. Although this last index included any data on tumorigenicity available in RTECS, Travis et al. (1990a) noted that exclusion of the tumor data from the index did not appreciably alter the results obtained.
Recently, Goodman & Wilson (1992) calculated the correlation between the TD50 and LD50 for 217 chemicals that they classified as being either genotoxic or nongenotoxic. The correlation coefficient for genotoxic chemicals was approximately r = 0.4 regardless of whether rats or mice were used, whereas the correlation coefficient for nongenotoxic chemicals was approximately r = 0.7.
McGregor (1992) calculated the correlation between the TD50 and LD 50 for different classes of carcinogens considered by the International Agency for Research on Cancer. The highest correlations were observed in Group 1 (known human carcinogens) with r = 0.72 for mice and r = 0.91 for rats, based on samples of size 9 and 8 respectively.
5. Low Dose Risk Assessment
5.1 Correlation Between Upper Bounds On the Low Dose Slope and MTD
Krewski et al. (1989) noted that the values of q1* derived from the linearized multi-stage model fitted to 263 data sets were highly correlated on a logarithmic scale with the MDTs in those experiments. As with the TD50, this association between q1* and the MDT occurs as a result