248 PHYSIOLOGIC GENETICS 



inheritance and the relationship of the genetic and nongenetic factors involved in their 

 etiology. 



Methods have been worked out for analysis of heritability of threshold characters 

 by Robertson and Lerner, 1062 Dempster and Lerner, 245 and others, and data on tumors 

 as all-or-none characters with continuous underlying genotype could be collected and 

 subjected to such analyses. Furthermore, in certain tumors where a recognizable 

 precancerous lesion occurs, data on the three classes — tumor, precancerous lesion, and 

 nontumor — could be analyzed for number of genes involved much as Green 482 and 

 McLaren and Michie 825 have analyzed the inheritance of number of vertebrae in the 

 mouse. 



The usual approach, however, has been by transposing the phenotypic expression 

 of the character from alternative expression to continuous variation. One way of 

 doing this is by measuring susceptibility in terms of latent period of the neoplasm as 

 Anderson 23 did in order to ascertain heritability of bovine ocular carcinoma. The 

 more susceptible the animal, the earlier the tumor appears. Another way is to measure 

 susceptibility by counting the multiple tumors that occur. Multiple tumors occur 

 particularly when the tumors are induced. This manner of measuring response 

 quantitatively was particularly advantageous in estimating the number of genes 

 involved in the inheritance of pulmonary tumors in mice. 557 In this work response 

 was also measured on the basis of latent period 557, 559 and the results from both types 

 of measurement were roughly parallel. 



More progress has been made in linkage studies of neoplasms than in any other 

 group of threshold characters. Without the cancer genes being individually identified, 

 however, it has not been possible except in certain situations to distinguish between 

 true linkage and a pleiotropic effect of the gene tested. 



Probably the greatest advantage that neoplasms have over many of the other 

 threshold characters is the number of nongenetic factors in this underlying continuous 

 variation that can be identified. These may be any of a host of physical and chemical 

 carcinogens, endocrine factors, nutritional factors, and viruses. Much of our know- 

 ledge of the physiologic paths of genie action leading to neoplasia is revealed through 

 the analysis of the interrelationship between these endogenous and exogenous non- 

 genetic factors and the genes. 



It is with the methodology employed in these facets of the genetic approach to the 

 etiology of cancer that this paper will be concerned. The genetics of transplanted 

 tumors is considered by others elsewhere in this volume. 



INBRED STRAINS FOR CANCER RESEARCH 



As early as 1909, Dr. C. C. Little foresaw that in order adequately to analyze the 

 inheritance of neoplasms when the presence or the absence of the tumor was not a full 

 indication of the genotype, it would be necessary to develop inbred strains in which the 

 genotype was uniform. That was the year in which he started inbreeding strain DBA 



