The consolidation of formal genetic, epidemiologic, cytogenetic and molecular genetic data has provided a complementary model that invokes the unmasking of recessive predisposing mutations by aberrant mitotic events. This model proposes the fixation of monoallelic genetic damage and, subsequently, its elucidation through mitotic malsegregation or recombination. The net result of these sequential events is a single somatic
tumor progenitor that is homozygously defective at the relevant
tumor locus. These events have been reasonably well delineated in the cases of
retinoblastoma and
Wilms tumor and limited studies indicate the applicability of the model to other solid
tumors as well. The relevance of these approaches to clinical medicine promises to be substantial. The public health value of accurate empirical designation of high- and low-risk groups for specific
cancer development may be large. At present, the involvement of various environmental agents in these processes is unknown. However, should the interplay between
genetic predisposition and genotoxic damage be defined, the implication for work force assignment may be of great interest. The application of these analyses to premorbid diagnosis has already been successful in
familial retinoblastoma cases. Hopefully, these latter approaches will provide the conceptual basis for their extension to other human
cancers. Genotypic information that does not rely on phenotypic expression may provide a useful adjunct to conventional histopathologic diagnoses, particularly in diseases such as the
soft tissue sarcomas that are often equivocal with the latter analyses. Finally, the identification of molecular alterations in
tumors associated with particular genetic syndromes may lead to more rapid and efficient isolation of linked
genetic markers; and these approaches will certainly be useful in estimating the number and genetic location of similar mutations with an etiologic role in the more common
cancers. The elucidation of interactions of these genetic loci with
environmental carcinogens as well as the dominantly acting class of oncogenes holds great promise in uncovering the molecular basis of
cancer. Once these single or bipartite mechanisms are identified, the limitations of our understanding of features of the disease caused by genes with partial penetrance or expressivity, or by compound genetic loci, promise to become a reality at long last.