Cell Proliferation and Killing: New Twists in Multistage Cancer Risk Extrapolation. K. T. Bogen, Health and Ecological Assessment L-453, Lawrence Livermore National Laboratory, University of California, CA 94550-9900
Dose-related increases in cell proliferation and cell killing play potentially important roles in predicting low-dose cancer risk, according to "cell-kinetic" multistage cancer theories. According to these theories, a malignant state arises from accumulations of (e.g., 2 or 3) critical somatic mutations forming a progression of normally irreversible stages, cells at each of which may be subject to either clonal expansion or clonal extinction or both. Dose-induced net proliferation of intermediate ("premalignant") cells may be induced by an otherwise nontoxic (e.g., hormone- or growth-factor-mimetic) mitogen, or as focal regenerative hyperplasia in response to dose-induced cell killing in surrounding tissue, or both. In the absence of such net proliferation, corresponding low-level cancer risk for a "purely genotoxic" compound is expected to approximately linearly proportional to dose. In contrast, to the extent that net (e.g., regenerative) proliferation induced by a "purely cytotoxic" compound is expected to have a threshold-like dose-response, corresponding cancer risk is expected to be virtually zero at sufficiently low dose levels. Many genotoxic compounds are also cytotoxic, in which case dose-response at very low levels of increased risk is expected to be an approximately linear function of dose, corresponding to the genotoxic component of increased risk. Thus, for genotoxic compounds (that may or may not be cytotoxic), linear dose-response extrapolation from cancer-bioassay data is often assumed to yield a plausible upper bound on increased risk, e.g., for regulatory purposes. However, for many genotoxic chemicals, it is known that cytotoxic resistance may arise in chronically exposed cells and also that this may not be associated with corresponding resistance to mutation. Multistage models incorporating plausible assumptions regarding the occurrence of such cytotoxic resistance to genotoxic chemicals predict that linear extrapolation from bioassay data may substantially underestimate increased risk at low doses. This possibility is illustrated using cancer-bioassay data on rats chronically exposed to aflatoxin B1 and to nitroso-1,2,3,6-tetrahydropyridine.
This work was performed under the auspices of the U.S. Department of Energy at Lawrence Livermore National Laboratory under Contract W-7405-ENG-48.