A Biological Basis for the Linear Non-Threshold Dose-Response Relationship for Low-Level Carcinogen Exposure

Abstract

There is no escaping the need to quantitatively assess the risks from low-level carcinogen exposure in order to judge how much regulatory effort is appropriate to reduce cancer hazards to acceptable levels. However, it is clearly recognized that the levels of exposure and the associated cancer risks from contamination of the environment by carcinogens are almost invariably far below the level which are directly measurable either by animal experiments or epidemiological studies in exposed human populations. We are therefore forced to use mathematical extrapolation models to define the relationship between dose and effect at levels which can never be ascertained directly. There are a number of mathematical extrapolation models which fit dose-response data for tumor induction and yet which predict risks differing by order of magnitude at very low levels of exposure.1 None of these models has a sound basis in biological fact since we do not understand the pathogenesis of neoplasia at the cellular level. Perhaps the most plausible is the linear non-threshold dose-response model which was first adopted for use in the assessment of cancer risks from ionizing radiation2 and then more recently from chemical carcinogens.3 Its plausibility rests largely on the association between tumorigenicity and mutagenicity as common manifestations of genotoxicity and the linearity at low dose levels of the dose-response relationships for mutagenicity by ionizing radiation4 and chemicals.5