As our knowledge of the specific genetic factors involved in the development of certain types of neoplasms increases, the characterization of subsets of specific neoplasms as “complex traits” is also enhanced. In some cases, such as retinoblastoma, where the penetrance of the gene exceeds 95%, it is unlikely that many cases can be attributed to a complex genetic trait, although other genetic factors such as imprinting may play a significant role in the expression of a few such neoplasms. On the other hand, with some of the more common neoplasms in which there is a known genetic predisposition, an understanding of the genetic foundations for the disease has become clarified. One such example is that of breast cancer, where (see above) two genes, the BRCA1 and BRCA2, have been characterized as causally related, when mutated, to a significant number of inherited cases. But if one examines the genetics of mammary and ovarian cancer in general, one sees a spectrum of diseases ranging from Cowden’s disease to postmenopausal breast cancer, in which it is likely that all women affected with this dominantly inherited predis- position to multiple neoplasms exhibit either frank malignancy of mammary tissue or mammary dysplasia (cf. Wolman and Dawson, 1991). Figure 5.11 shows a pie chart indicating the fre- quency of several inherited syndromes in which mammary cancer is predominant, these fading off into a larger group of polygenic diseases, with the majority of breast cancer resulting from unknown, presumably environmental causes and thus termed sporadic mammary cancer (Lynch et al., 1990). However, even in individuals bearing a mutated BRCA1 gene, the risk of develop- ing mammary cancer is not absolute and is partly age-dependent, being greater prior to age 50 and decreasing somewhat by age 65 or 70 (King et al., 1993). Interestingly, these workers de- scribe families in which some members develop mammary cancer and others ovarian cancer, but the mutation in the BRCA1 gene is identical. A variety of environmental factors can increase or decrease the risk of development of mammary cancer even in patients exhibiting genetic predis- position. These include diet, age at first pregnancy, parity, abortions, or artificially imposed menopause (Henderson, 1993; Andrieu et al., 1993). The age of onset of the disease is also re- lated to the breast cancer incidence of immediate relatives, such that the risk is much greater among women whose mother or sibling was diagnosed with breast cancer prior to age 40, but the risk remains elevated even for those whose mothers were diagnosed with the disease at the age of 70 years or older (Colditz et al., 1993). An equally important example of genetic predisposi- tion to a common human neoplasm is that seen in polygenic colon cancer. While the two major Mendelian genetic causes of colon cancer are shown, there are several others, as can be noted in
Figure 5.11 Schematic representation of the genetic heterogeneity seen in breast cancer in humans. (Af- ter Lynch et al., 1990, with permission of authors and publisher.)
Table 5.1, that contribute very small numbers to the chart. During the last decade there have been at least two studies suggesting that polyps and colorectal cancers occur primarily in genetically susceptible individuals in the human population, although the studies could not eliminate non- heritable causes of the disease (Cannon-Albright et al., 1988; Houlston et al., 1992). These stud- ies suggested that perhaps as many as 80% of patients with this disease exhibit some sort of genetic predisposition in which one or a few genes may be the primary factors but multiple genes are probably involved in the ultimate expression of the disease. As with breast cancer, development of the disease at an earlier age increases the risk to siblings and close relatives (Fuchs et al., 1994; St. John et al., 1993).
On the other hand, although there is little if any evidence that human leukemia is inherited in a Mendelian fashion, there is good evidence to indicate that a number of patients with leuke- mia may have had a significant predisposition for the disease. The groups at greatest risk are listed in Table 5.9. One of the most interesting aspects of this listing is that, when one member of identical twins develops leukemia at a relatively early age, the risk is about 20% that the unaf- fected twin will also succumb to the disease; but this means that 80% do not appear to be at risk.
At least one extensive study (Gunz et al., 1975) indicated that polygenic mechanisms may be involved in the incidence of certain cases of chronic lymphocytic as well as acute leukemia but not of chronic myelogenous leukemia.
Of the examples given in Table 5.9, the majority are associated with radiation effects or genomic abnormalities. Several investigations of radiation-induced damage and DNA repair in individuals with and without a family history of cancer have indicated that abnormalities tend to occur, especially in patients with cancer or who are cancer-prone (Parshad et al., 1983; Pero et al., 1989; Kovacs and Langemann, 1991). Hsu and associates (1991) used a standardized test with the mutagen bleomycin to assay the mutagen sensitivity of the general population. These investigators demonstrated that moderately sensitive to hypersensitive individuals make up al- most 23% of the total population. Recently Lipkowitz et al. (1992) demonstrated an increased frequency of abnormal rearrangements of immune receptor loci in peripheral blood lymphocytes of agricultural workers. This may relate to an observed increased risk for lymphoid malignancy in this occupational group. As noted earlier, patients who are heterozygous for the genetic dis- ease ataxia-telangiectasia (see above), composing 0.7% to almost 8% of the population (Mur- nane and Kapp, 1993), exhibit a slight but significant increase in radiosensitivity to ionizing radiation. Since the genetics of DNA repair in response to chemical and physical insults is quite complex (Chapter 3), multiple genes are likely to be involved in the genesis of many complex trait neoplasms.