850 THE CARCINOGENIC STIMULUS II 



papilloma filtrates. It was thought at first that in the wild rabbit the lesion re- 

 mains benign, but it has been demonstrated (Syverton et al., 1950) that in some 

 wild rabbits carcinomas develop from the initial benign lesion. Progression to a 

 carcinoma occurs more readily in the domestic rabbit; from the metastatic cancer 

 a virus can no longer be demonstrated (Friedewald and Kidd, 1939; Syverton etal., 

 1950). Either the virus is "masked" in the carcinoma, or it has served only to 

 initiate a precancerous lesion, the resultant carcinoma being virus-free. If the 

 carcinoma is actually without virus, then the filterable agent cannot be considered 

 necessary for the continuity of the neoplastic process. 



The work on the Shope papilloma provides a basis for speculation concerning 

 undiscovered tumor viruses. In the case of a neoplasm of known viral causation 

 the carcinoma did not yield the virus. Perhaps in other neoplasms the apparent 

 absence of a virus might then be explained on the basis of "masking", either by 

 antibody, or by intimate combination with cellular constituents to the extent that 

 separation by usual methods is not feasible. 



{b) Mouse mammary cancer 



That cancer viruses might easily escape detection is emphasized by the experi- 

 ments which revealed the mouse mammary tumor agent (Bittner, 1942). Mammary 

 cancer of mice was thought to be influenced in its development primarily by hor- 

 monal factors. Early experiments indicated, however, that the incidence in Fi 

 hybrids between two high and low cancer strains (A and C57) was very much 

 higher when the mother was of the high rather than the low mammary cancer 

 strain (Korteweg, 1934; Staff, Jackson Memorial Laboratory, 1933; Murray and 

 Little, 1936). Bittner (1936b) demonstrated by foster-nursing experiments that the 

 "maternal influence" (tumor agent) is transmitted in the milk during suckling. 

 High tumor strain A mice did not develop mainmary cancer if foster-nursed 

 immediately after birth by a low tumor C57 female. Ingestion of only relatively 

 small amounts of milk from a high tumor (A) female made likely the development 

 of mammary cancer. 



Agent-free A mice give rise to progeny which are agent-free and consequently 

 low-cancer. If the agent is reintroduced, either by foster-nursing or injection of 

 tumor extracts, then the line again becomes "high tumor". The agent can be 

 transferred to females by males of certain inbred mouse strains (Bittner, 1952). 



Neither foster-nursed high tumor A mice, nor low cancer C57BL mice, develop 

 mammary cancer if given estrogenic hormone. It appeared, then, that the 

 hormonal factor, estrogen, acts only in conjunction with the mammary tvimor 

 agent or virus. The virus is the essential carcinogen, according to this concept, but 

 only in hosts of appropriate and specific genetic constitution. The importance of 

 the propagation of the virus has been demonstrated by Heston etal., (1956) ; passage 

 through hosts of certain genetic constitution results in disappearance of the virus. 



To recapitulate, it was postulated that hormonal, viral and genetic factors are 

 interrelated in the genesis of spontaneous mammary carcinogenesis in mice of 

 highly inbred strains, and that in the absence of any one the incidence of mam- 

 mary cancer sinks to a negligible level. If the caloric intake is restricted, so that 

 only two-thirds of the adult weight is attained, restricted mice do not breed, nor 



