Evaluation of the carcinogenicity of chemical substances
In 1978, the Health Council of the Netherlands recommended a method for evaluating the carcinogenic (cancer-inducing) properties of substances, as well as for estimating the cancer risk associated with exposure to substances with such properties. That method uses a classification system that divides carcinogens into two categories on the basis of their mechanism of action: genotoxic and non-genotoxic. Genotoxic carcinogens are assumed to pose a cancer risk at any concentration, in other words: with these carcinogens there is no such thing as a safe level of exposure. However, non-genotoxic carcinogens are assumed to have a threshold value, a level of exposure at and beneath which cancer will not be induced. With a genotoxic carcinogen, linear extrapolation is the recommended method for estimating the risks associated with a given level of exposure. Where a non-genotoxic carcinogen is involved, the appropriate procedure is to estimate the threshold value. In the present report, a Health Council committee has addressed the issue of whether this approach can still be justified given the state of the art.
The Committee provides a summary of the state of the art. The principal developments cited by the Committee in this connection are a more precise understanding of how a healthy cell gives rise to a tumour and the availability of mathematical models to describe this process. This new information about the way in which cancer develops lends support to the system, advocated in 1978, that classifies carcinogenic substances into two groups on the basis of their mechanism of action. This classification system has also been recognized in other parts of the world. As regards recent developments in the field of mathematical modelling, the Committee feels that models are not (yet) suitable for use in risk assessment since the data they require is simply not available for the majority of substances. The method proposed in 1978 is suitable, however, since it is easy to use and requires little data on the substance in question. In addition, it is consistent with the new developments outlined above and is based on a widely accepted hypothesis about the effect of carcinogens at the molecular level on the incidence of cancer observed in a population. Partly because this method uses linear extrapolation for genotoxic carcinogens, the Committee considers that it is relatively safe. In other words, when using this method there is a high probability that the risk will be overestimated. The Committee considers this safety to be important, given the lack of data sufficiently detailed to be of use in risk assessment. More specifically, little detailed information is available concerning the intervening steps between the first harmful molecular changes in a healthy cell and the incidence of (one or more types of) cancer in people or experimental animals.
The Committee identifies the uncertainties that pervade evaluation of the carcinogenicity of substances and risk assessment. These uncertainties stem from a variety of factors, including (unavoidable) chance variation and inadequate understanding. Of the two, the latter is the more important. This concerns the limited understanding of how cancer arises and of the part played by carcinogenic substances in this process. So any conclusions regarding a substance’s carcinogenic properties or any estimates of cancer risk necessarily involve certain assumptions. According to the Committee, while great strides have been made in cancer research, the nature of this progress means that it does not facilitate any improvements in the risk assessment method. The overwhelming majority of uncertainties associated with risk assessment in the late seventies have not gone away, nor does the Committee expect them to be resolved in the near future.
Given the new data about how cancer arises and the part played by substances in this process, and in view of the above-mentioned uncertainties, the Committee considers it advisable that the 1978 approach be retained. However, it feels that it would be sensible to reserve the option of departing from this approach in special cases. The Committee found a few instances where this was justified: data concerning the mechanisms of action of some genotoxic carcinogens prompted the Committee to favour the derivation of a threshold rather than the use of linear extrapolation in these cases. Accordingly, the Committee recommends that the classification of carcinogens into two groups for the purposes of risk assessment be retained, as should the method of risk assessment itself. The Committee further recommends exceptions be made in specific cases, where the data indicates this to be an appropriate course of action.