Relative vs. Additive Risk for Predicting Human Cancer Risk from Animal Data
David W. Gaylor, National Center for Toxicological Research, FDA, Jefferson, AR 72079
Cancer risk assessments commonly assume equal excess (absolute) risk per unit dose across species with appropriate interspecies dose scaling. It is equally appropriate to assume equal relative risk across species. Then, the human cancer risk is estimated to be the excess site-specific animal risk multiplied by the ratio of the background site-specific human cancer risk to the background site-specific animal cancer risk. In the absence of information regarding potential tissue sites for human cancers, a maximum value of 2% for the lifetime incidence is used for common human tumors, other than skin cancer or lung cancer associated with smoking. Hence, for cancer risk estimates based on common animal tumors, e.g., liver tumors in male mice with a greater than 2% incidence in control animals, human cancer risk estimates would be reduced. For rare animal tumors with a historical background rate in controls of less than 2%, estimates of potential human cancer incidence would be increased.