Quantitative Evaluation of DNA-Binding Data In Vivo for Low-Dose Extrapolations

Abstract

Introdu.ctionThe risk oftumor formation from exposure to a chemical carcinogen is dependent on the exposure, the potency of the carcinogen, and the imiividual host reaction. Rumans are exposed to chemical carcinogens at dose levels which are orders of magnitude below the levels used in"animal studies on carcinogenicity. The latter experiments provide significant data only at high-dose levels which lead to tumor incidences on the order ofpercent (Fig.1). For humans, tolerable exposures producing not more than one additional tumor in one million lives should be defined. The extrapolation range therefore covers four to five orders of magnitude. Instead of using purely mathematical models for the extrapolation, it woulä. be desirable to have a biologically relevant indicator which could be investigated in the dose range to be bridged.A large group of chemical carcinogens is known to bind covalently to DNA in the target cell. Under appropriate conditions, this primary DNA lesion can be expressed as a mutation finally leading to cancer. The primary interaction of the · carcinogen with DNA can be investigated with appropriate techniques (radiolabeled test compound, phosphorylation with 32 P, or immunological methods) at low dose levels which would not give rise to a detectable increase in tumor yield with a limited number of animals treated. It is therefore possible to investigate the shape of the dose-response curve in the region of interest. MethodsWith radiolabeled test compound,. the limit of detection is dependent on the specific activity, on the binding potency ( covalent binding index, CBI; Lutz 1979), on the amount of DNA analyzed, and on the radioactivity in a vial considered significant. U nder optimal conditions met, for instance, with tritiated aflatoxin B 1 of 8 Ci/mmol and a CBI of 10 000, a single dose of 1 ng/kg rat was the o bserved limit of detection for liver DNA binding (1 0 dpm net radioactivity in 6 mg DNA).

Dissimilarity in aflatoxin dose-response relationships between DNA adduct formation and development of preneoplastic foci in rat liver

“…These effects are accentuated at higher, more acutely toxic, doses. Lutz has suggested that, although DNA binding may be proportional to dose, cytotoxi-city-induced cell division induced disproportionately at higher doses may enhance tumor development [36,37]. Eaton and Gallagher also suggest that tumorigenic response does not necessarily follow linearity at all doses because of the cytotoxic effects of AFB at higher doses [38].…”

Opinion of the Scientific Committee on a request from EFSA related to A Harmonised Approach for Risk Assessment of Substances Which are both Genotoxic and Carcinogenic

“…Studies on covalent binding of substances that are both genotoxic and carcinogenic to DNA show a linear dose-response relationship in the lowdose range, with no indication of a threshold (Neumann, 1980;Dunn, 1983;Lutz, 1987;Beland et al, 1988). This might be thought to suggest a linear decrease of genotoxicity, and eventually of cancer risk, at low doses and implies that exposure to even a single molecule of a genotoxic substance could produce DNA damage and thereby some degree of risk.…”

Endogenous genotoxic agents and processes as a basis of spontaneous carcinogenesis

“…At very low dose Ievels, where all processes follow first-order kinetics, the additional darnage is expected to be proportional to the dose (Lutz, 1987). In view of the analysis discussed above it is doubtful whether this proportionality also holds between DNA darnage and tumor incidence.…”