The understanding of the obligatory steps in the carcinogenic process and the knowledge of the sequence of events leading to cancer have become increasingly clear over the past decade. The stringent distinction in regulatory toxicology between genotoxic and non-genotoxic carcinogens, with distinctly different approaches to cancer risk assessment, are no longer justified. In 1998 the German MAK Commission took these advances into account by introducing an extended classification scheme. The projection, by simple linear extrapolation, of the tumour response in animals to high doses onto the very low doses anticipated in man or actually determined in the occupational environment can no longer be considered as reflecting the current state of knowledge. The calculation of a lifetime cancer risk by assuming a rectilinear relationship in humans between exposure and a cancer incidence from zero to 25 %, as proposed by the “T25“ method currently discussed in the EU, looks precise but is almost certainly wrong and therefore not acceptable. Instead of estimating the risk by calculating the T25, for workplaces it is proposed that the margin of exposure (MoE) be determined. This expresses the magnitude of difference between the level of human exposure and the human benchmark dose for a 5 % risk level (HBMD05). The decision as to which MoE would be considered acceptable depends on substance-specific and socio-economic factors. The proposal is substantiated by a number of carcinogenic substances and their current technical exposure limits (TRK), which are established in Germany for carcinogens. The MoEs so far calculated are in the range from 1.4 to 265 and should serve as an appropriate scientifically-ground basis for setting the priority of management procedures.
