A UNIFIED APPROACH TO RISK CHARACTERIZATION

Abstract

Listen

Translate article

A fundamental issue of toxicology is to determine levels of human exposure to toxic substances, commonly referred to as acceptable daily intakes (ADIs) or reference doses (RfDs), that are presumed to cause no or negligible health risks. The goal is to determine relatively safe levels for human exposure to toxic substances using scientifically defensible methods and data, including a discussion of the imprecision of the process. Ideally, an understanding of the biochemical mechanisms of toxicity would allow the construction of biologically based dose-response models that are applicable across species in the range of human exposure levels. In the absence of such information, low-dose linear extrapolation to zero risk, without a threshold dose, generally is used by U.S. regulatory agencies to estimate cancer risk. For noncancer effects, the no-observed-adverseeffect level (NOAEL) is divided by safety (uncertainty) factors to arrive at an ADI (RfD) that is presumed to be below a threshold dose and to pose no risk, or at worst, negligible risk. The dichotomy of approaches for cancer and noncancer risk assessment based on the threshold concept can be questioned. Animals and humans with no known exposure to a particular toxic substance sometimes exhibit an noncancer adverse health effect such as kidney failure or a defect at birth. Hence, other exogenous and/or endogenous exposures produce that adverse health effect, particularly in susceptible individuals. Thus, a threshold dose, if one exists, is already surpassed in a population. If a toxic substance contributes by the same pathway to the background dose, additional adverse effects are expected with no threshold for the added exposure. For cancer, threshold doses may exist for nongenotoxic (secondary mechanisms) modes of action. Examples of secondary mechanisms are the production of bladder cancer at high doses that produce