Genes, Ethnicity, and Risk Assessment

GENES, ETHNICITY, AND RISK ASSESSMENT P. UNRAU and K. DOERFFER* Humans are clonal organisms, all the cells in the body being derived from the initial genetic fusion of sperm and egg. The characteristics which keep tissues and cells stable during our lives are controlled by the DNA in somatic cell nuclei. When, owing to inherited genetic factors or to environmentally induced mutational events, genetic variants arise in our clonal cell lineages, they may either be selected against, be null, or lead with some chance to cancer. Cancer risks are thus the major somatic risks of concern. They will reflect both our original genetic constitution and the effects of the environment on that constitution expressed in our somatic cells. Risk assessments are designed to calculate the real or imagined effect of an environmental risk source upon a population which is thought to be at risk [I]. After such a calculation is performed, the results are discussed as if they had meaning, and new regulations and changes are introduced [2-4]. Usually, such calculations are performed very precisely. However, the degree of precision of the calculation usually exceeds the understanding of the biological factors which may be involved . Substantiating the accuracy of the calculation with hard biological and genetic knowledge has proven to be very difficult. Thus, we wind up in the position of Gold, et al., who, using a simple index to rank possible carcinogenic hazards in the environment, have seen the resulting calculations driven to absurdity in the name of risk assessment and prevention [5]. Epidemiological methods used to calculate genetic and somatic (cancer ) risks require certain simplifying assumptions. Genetic and somatic risks have both a genetic and an environmental component, and the sources and magnitude of the genetic component are usually ignored The authors gratefully acknowledge support from CANDU Owner's Group for a contract to K. Doerffer. They also gratefully acknowledge the interesting discussions with John Perehinec, John Unrau, Tony Waker, and Richard Osborne that have helped clarify their approach to the origins of genetic risks. *Chalk River Laboratories, Chalk River, Ontario, Canada KOJ IJO. 0031-5982/95/3901-0934$01.00 Perspectives in Biology and Medicine, 39, 2 ¦ Winter 1996 239 in risk assessment. This is because the analysis of genetic risk sources is still difficult and expensive; the genetic analysis can only be avoided by using simplifying assumptions. In order to be able to distinguish inherent risk sources from environmental ones, so that the actual effects of the latter can be measured, we have looked at some of the simplifying assumptions to see how much they affect our picture of genetic and cancer risks. Presently, risk assessment can only be performed with the underlying assumption that all members of a population are essentially alike. This means that the genetic characteristics of a population are viewed as being uniform, a perspective that Mayr calls the "essentialist" view of organisms [6]. While the essentialist view permits groups of organisms to be aggregated for probabilistic calculations, Mayr points out that all sexually reproducing organisms are in fact genetically unique (leaving aside identical twins). He proposes that assessments of individuals in sexually reproducing species must take into account their genetic uniqueness. If genetic components turn out to be important in risk assessment, the present essentialist view of humans and human populations will have to be replaced by an individualist view. At the logical extreme, the prediction of individual risk can only be based on individual characteristics , and this has implications even for how we talk about risks and populations. If a population is a collection of genetically unique individuals , then assigning the same, average risk to all members ofthe population does not seem to be valid. Populations do not have or show risks, only individuals do. Risks are usually expressed in fractional terms. For example, the risk of having a particular genetic disorder is given as 1/30,000, or the risk associated with a particular dose ofionizing radiation is given as 1 cancer in 100,000 people at risk. But Mendel pointed out that for practical purposes units of heredity were indivisible, and they were independently inherited [7]. We would say now they are binary in nature. So there...