Probing Carcinogenicity Limits

When the Environmental Protection Agency determines exposure limits for genotoxic carcinogens, it typically extrapolates a dose-response line several orders of magnitude below the data cutoff from high-dose toxicity studies. Now, a group led by George S. Bailey of Oregon State University has examined 40,800 rainbow trout to determine the incidence of cancer from exposure to extremely low doses of a polyaromatic hydrocarbon (Chem. Res. Toxicol., DOI: 10.1021/tx9000754). The results provide an experimental estimate for how conservative EPA limits might be for carcinogen exposure. The research team gave sets of trout feed containing 0 to 225 ppm dibenzo[a,l]pyrene (DBP) for four weeks, then returned the fish to a regular diet for nine months before examining them for tumors. The experiments extend the dosages tested by about three orders of magnitude below previous studies and show that DBP dose-response is not linear. When the data were extrapolated to determine how much of the carcinogen will induce one extra cancer case per million animals, the researchers found that the dose of DBP is 500- to 1,500-fold higher than EPA's linear extrapolation from high-dose data, indicating that EPA's method is underestimating the dose and overestimating cancer risk.