A second conclusion, again under conditions of a permanent increase in mutation rate, is that at the new equilibrium between mutation and selection (which will be achieved several tens—if not hundreds—of generations later, depending on the amount of increase in mutation rate and selection coefficients), the MC will become 1.0. In other words, for a sustained increase of x% in mutation rate, there will be an x% increase in the frequency of chronic diseases at the new equilibrium. This conclusion holds for several different combinations of assumed parameter values (selection coefficients, thresholds, numbers of loci, environmental variances, interactions among genes) and consequently can be considered robust.
Finally, if the population sustains radiation exposure in one generation only, the increase in MC will be transient and small, followed by a progressive decline to zero. The result will be a transient small increase in disease frequency followed by a decline toward the baseline frequency in subsequent generations.
This committee uses the 2% value in its calculations as the best MC estimate for the first postradiation generation, which was also the case for the ICRP (1999) Task Group and UNSCEAR (2001).
Mouse data on rates of induced mutations (incorporated in the DD estimate) provide the basis for genetic risk prediction in humans. In predicting the risk as a product of P, 1/DD, and MC (i.e., Equation (4-3) noted in the section on mutation component), an important assumption is implicit: mutations will be induced in those genes at which spontaneous mutations in humans cause disease (i.e., the quantity P), the average rate of induced mutations in mice is applicable to induced human germline mutations, and such induced mutations will be compatible with viability and hence recoverable in the offspring of irradiated individuals. However, thus far, no radiation-induced genetic diseases have been found in the offspring of those who have sustained radiation exposures (e.g., Byrne and others 1998; Meistrich and Byrne 2002; MGSC 2002).
Advances in human molecular genetics and radiation genetics during the last decade support the view that there are several fundamental differences (in mechanisms, nature, etc.) between spontaneous mutations that cause disease and radiation-induced mutations studied in experimental systems such as the mouse. More specifically, they suggest that a major proportion of human genes of relevance from the disease point of view may not yield “recoverable” induced mutations. Stated differently, the rate at which induced disease-causing mutations are seen in human live births following parental radiation exposures may be much lower than that of induced mutations in mice.
Since there is no alternative to the use of mouse data on radiation-induced mutations for risk predictions in humans, methods have to be devised to bridge the gap between induced mutation rates in mice and the risk of genetic disease in humans. One such method has been developed recently and is based on the incorporation of a correction factor, termed the potential recoverability correction factor (PRCF), in the risk equation (Sankaranarayanan and Chakraborty 2000a). As a consequence, the risk now becomes a product of four quantities instead of the original three:
where P, 1/DD, and MC are as defined earlier and PRCF is the disease-class-specific potential recoverability correction factor. Since PRCF is less than one, the estimate of predicted risk will be smaller when PRCF is incorporated than when it is not.
The differences between spontaneous disease-causing mutations in humans and radiation-induced mutations studied in experimental systems, which constitute the basis for the development of the PRCF concept, are discussed in detail by Sankaranarayanan (1999) and Sankaranarayanan and Chakraborty (2000b) and summarized in Annex 4D.
To assess PRCF, it was necessary first to define criteria on the basis of information available from molecular studies of radiation-induced mutations, to apply these to human genes of interest on a gene-by-gene basis, and to examine which among them can be considered candidates for potentially recoverable induced mutations. The operative words are the italicized ones, since there is as yet no evidence for a radiation-induced germ cell mutation in humans, our understanding of the structural and functional genomics of the genome is incomplete, and the criteria will undoubtedly change with advances in knowledge.
Among the attributes considered in defining the criteria are gene size, location, normal function, known mutational mechanisms, spectrum of spontaneous mutations, “gene-richness” or “gene-poorness” of the region, whether intragenic (including whole-gene) deletions and multigene deletions are known, and whether disruption of the gene or genomic region by rearrangements is associated with a mutant phenotype. Under the assumption that a deletion is induced in a genomic region containing the gene of interest, the question asked was, Given the structural and functional attributes of the gene or genomic region, can this deletion be considered potentially recoverable?