Etiology of Wilms' tumor.

Present knowledge of the pathogenesis of Wilms' tumor can be used to plan further epidemiologic investigations in a logical manner. Table 3 summarizes the epidemiologic evidence presently available that implicates various exposures. The proportion of cases of Wilms' tumor that arise from germline mutations is unknown. Future case-control studies could contribute to providing such information by systematically collecting and preserving constitutional genetic material from cases and their parents and neoplastic genetic material from their tumors. The studies that have demonstrated germline mutations so far have involved small numbers of subjects, being the equivalent of "case reports" (36, 37, 45-47, 50, 53). The techniques of molecular biology will be necessary to distinguish germline mutations or equivalent events that have been transmitted by parents from those that have arisen de novo. (Since it has already been shown that phenotypically unaffected parents can transmit relevant mutations to offspring who become affected (45, 46), the presence or absence of phenotypic abnormalities in the parents cannot be used to infer whether germinal mutations have arisen de novo.) Such studies will be laborious and expensive. The best strategy to minimize cost and labor would be restriction on age; the cases diagnosed relatively early in life (before the age of 2-3 years) are more likely to have arisen from germline mutations than those diagnosed later (34). The existence of de novo germline mutations relevant to the development of Wilms' tumor should prompt epidemiologists to search for causal exposures before conception. Preliminary evidence suggests that males have a higher germinal mutation rate (48). On the basis of the epidemiologic evidence to date, Olshan et al. have suggested the following: "paternal preconceptional exposures may play a more important role in the etiology of Wilm's tumor than maternal factors"(8,p.943). Pathologic and molecular biologic studies have shown that Wilm's tumor arises from nephrogenic rests that represent renal stem cells, the development of which has been arrested before normal differentiation occurs (2,44). Exposures that occur during intrauterine life are likely to be relevant to this process; these can be studied with epidemiologic methods, and clues have emerged (7,86,89). It is important to realize, however, that the fetus may be exposed in utero to substances to which the mother was exposed prior to conception, if these substances are excreted slowly or not at all, such as the persistent organochlorine pesticides (96,98). Study of the determinants of the growth behavior of nephrogenic rests (persistence, regression, generalized growth, or malignant transformation (2)) may be possible only in an animal model of Wilm's tumor, if one can be found in which nephroblastoma arises from nephrogenic rests. Ethylnitrosurea, an alkylating agent, causes renal tumors identical to Wilm's tumor in the opossum when given early in postnatal life (103) and in the rabbit when given transplacentally (104). Nephroblastoma in animals is one of the very few types of tumors that is inducible by chemicals via the transplacental route. This mode of induction should be regarded as as an etiologic possibility in humans as well as animals (105). As Hard, however, has pointed out, "nodular renal blastema [now known as nephrogenic rests] or nephroblastomatosis...have [sic] never been described in conjunction with nephroblastoma of animals" (105,p.184). the renal stem cells that give rise to Wilm's tumor normally disappear 4-6weeks prior to birth (19). If they persist after birth, as nephrogenic rests or otherwise, postnatal exposures could contribute to the development of Wilm's tumor. Cases diagnosed later, especially those presenting with lower stage disease, are more likely to have arisen from mutations occurring after birth. Potential carcinogens in children's diet and the home environment are worthy of close study.