Dr. Andervont (National Cancer Institute): Probably the best point you made, Dr. Muhlbock, is that here is a virus which is not the whole story in the production of malignant cells. There are other factors that must be present before it can function. And this, of course, is provocative.

Dr. W. U. Gardner (Yale University): On this occasion I think that we could appropriately mention the late Dr. Leo Loeb who about 40 years ago first proposed a simple formula for mammary cancer in mice-genes plus hormones give cancer. This is doubly appropriate today because Dr. Muhlbock has stated that the transplantation of pituitary glands abets mammary tumorigenesis, even in mice without identifiable mammary-tumor agent. Dr. Loeb and his associates first showed, about 21 years ago, that, in mice showing a low tumor incidence, the incidence of mammary tumors could be greatly increased by implanting one, two, or three pituitary glands, even when they could not find the transplants at necropsy. Dr. Muhlbock's observations have revealed the importance of proper hormonal conditions in mammary carcinogenesis in mice and thus added evidence to the formula.

Our early work indicated the importance of a maternally transmitted influence on the appearance of multiple localized hyperplastic nodules in the mammary glands of female mice or of mice under adequate hormone treatment. These nodules appeared in large numbers in the mammary glands of older multiparous mice of the C3H and DBA strains which showed a high incidence of spontaneous mammary tumors. They increased in number with age and decreased with the postlactation interval. The glands of the high tumor mice of strain A showed fewer nodules but the ducts more frequently gave origin to tumors. Nodules were lacking in comparable mice of the low tumor N strain. We believed these nodules to be induced by the Bittner agent after this agent was defined biologically. But mammary nodules are of two types: (1) those that represent localized areas that do not regress as rapidly as adjacent areas (retention type), and (2) those that represent localized proliferation of otherwise quiescent mammary tissue. The retention type disappears during the postlactation period. Qualitatively the two types of nodules may be difficult to distinguish between in gross preparations. We assumed that the agent promoted growth of the localized hyperplastic nodules in a suitable hormonal environment and that other influences determined their neoplastic transformation.

Hypophysectomized mice quickly show regression of these nodules for the greater part. A few, not over 5 percent, persist and these probably were already adenocarcinomas at the time of hypophysectomy. Mice given testosterone propionate will also show very few of the local hyperplastic mammary nodules, even when estrogen is given, and will acquire few tumors. After nodules have formed, the injection of testosterone will not prevent mammary cancer or the persistence of some nodules, though it does reduce their number.

Genetic influences are also of importance in mammary carcinogenesis in mice. Hybrid offspring of a low mammary tumor strain (C57) and a high mammary tumor strain (CBA) give F1 hybrid mice, which tend to show an incidence of mammary adenocarcinomas that parallels the maternal CBA strain and mammary tumors of a squamous type occur at a lower incidence among the reciprocal hybrids. The adenocarcinomas contain agents, the others do not. The backcrosses of the F1 hybrids to the parental strains indicated that the C57 genotype, even in the presence of the Bittner milk influence, resulted in few mammary tumors. This was also revealed by other hybrid crosses. In all these experiments, the agent was transmitted through the milk by nursing rather than injected or fed after concentration from milk or some other source.

We must recall that mammary tumors have been induced under a variety of conditions in mice: (1) known agent + hormone + adequate genotype, (2) adequate genotypes and hormones (pituitary and gonadal) - known agent, and (3) hormones and carcinogenic hydrocarbon - known agent. All must act over comparatively prolonged periods before mammary cancers appear. It seems to me that we must consider the possibility either (1) that we are dealing with more than one type of disease showing similar symptoms, or (2) that all the known mammary carcinogenic influences act through a common causal agent, possibly with the formation of a causal agent or "iso-carcinogen" in the mammary tissue itself.

Dr. Beard: It would seem appropriate to comment on the influence of the host in the occurrence of the malignant process associated with virus infection. To me it would be impossible to have any understanding of the contribution of the host, qualitative or quantitative, to viral response without some idea of the mass factors concerned with the material used for the transmission of the diseases.

Twenty years ago we were greatly impressed with the work of Dr. Ray Bryan which showed that the ID50 dose of the Shope papilloma virus was about 80 million particles in domestic rabbits. Such a relationship has been observed with all the virus tumors thus far studied on a quantitative basis. In myeloblastosis, the ID50, measured in a highly susceptible strain of chickens, is about 24 million virus particles and in erythroblastosis the comparable infectious unit is 24 thousand particles. It is thus quite evident that, although the virus determines the cellular characteristics of host response, it is largely the chicken or the rabbit which determines whether or not disease will occur. Many strains of chickens hardly react at all to either avian virus, but some respond with high incidence.

In relation to mouse mammary carcinoma, there has been much study of the relative influences of hormones, virus, and genetic constitution. The truth is that what has been needed, though yet lacking, in the study of mouse mammary carcinoma, as well as other virus tumors and tumor viruses, has been the correlation of the various concepts as to the variety and the degree of influence of the factors which can contribute to the processes of induction and maintenance of malignancy.

I do not have the slightest doubt that mice may develop carcinoma of the breast without the participation of virus, and this might very well be influenced to a certain extent by kind and quantity of hormone. In contrast, it was disappointing to hear the serious consideration given to the possible influence of the virus on the production of hormone. While such a concept seems entirely remote, there would be no difficulty in understanding that, on the other hand, the hormone might well influence the preparation of breast tissues which would make possible reaction of the cells to the virus.

In principle, viruses are much like bacteria, with the difference that bacteria live outside cells and viruses exert their influence and are synthesized intracellularly. Everyone knows that bacteria cannot grow without proper media and some of these organisms, like tubercle bacilli, have very special requirements. If this is the case with bacteria, how could one suppose that it would not be even more necessary to have a responsive medium inside the cell for viruses, whose requirements are so much more specific than those of bacteria? As far as I am concerned, the matter with respect to the mouse mammary carcinoma is relatively simple. There exists a specific viral agent which has the capacity to induce specific malignant reaction in the cells of the mammary gland. Without the proper conditions for the growth and the propagation of the virus inside the cells, no response is observed. Given the opportunity for proper conditioning of the physiological state, however, the cell develops the capacity to respond to the virus, and carcinoma arises.