Two possibilities for the mechanism of action of the mammary-tumor agent can be considered: (1) The agent acts via an influence on the hormone production or the hormone balance. (2) The agent influences the susceptibility or reactivity of the mammary gland to the hormonal influence.

To investigate the first possibility, the effect of the mammary-tumor agent on the estrous cycle was determined. Forty-seven animals of the agent-free hybrid F1, known to be susceptible to the agent, received intraperitoneal inoculations at the age of 4 weeks of extracts of mammary tumors of the C3H strain. Three months later daily vaginal smears were taken for a period of nearly 2 months. Forty-nine untreated animals of the same hybrid served as controls.

The length of the estrous cycle was determined as shown in text-figure 1. During the experimental period, the controls showed 493 complete cycles and the treated animals 446 cycles. The duration of the cycles varied in both groups from 3 to 16 days, as calculated from estrus to estrus. In both groups the large majority of the cycles had a duration of 4 to 6 days. There was no difference between the animals with the mammary-tumor agent and the controls in the average duration of these short cycles.

The second group of cycles (duration from 9 to 16 days) formed 14 percent of the total number of cycles in the treated animals as against 13 percent in the controls. The prolongation observed was always due to a lengthening of the diestrous periods; this group of cycles constitutes the so-called spontaneous pseudopregnancies, in which functional corpora lutea are present, which indicate prolactin secretion by the pituitary. We must thus conclude that the mammary-tumor agent has no significant effect on the incidence of these pseudopregnancies, i.e., on prolactin secretion.

Length of estrous cycle in days

Text-figure 1. Length of estrous cycle in days in (C57BL X DBAf)F1 mice.

We next determined the duration of the various stages of the estrous cycle in the animals with the mammary-tumor agent and the controls (text-fig. 2). Here too we failed to observe any difference between the 2 groups.

Length of various stages of the estrous cycle

HOURS

Text-figure 2. Length of various stages of the estrous cycle in (C57BL X DBAf)F1 mice.

In summary, we can conclude that this experiment failed to produce any indication that the mammary-tumor agent affects either pituitary function in respect to gonadotropin or prolactin production, or directly or indirectly the ovarian hormonal function.

The following evidence will, however, show that the phenomenon of spontaneous pseudopregnancy has a direct bearing on the question of the ultimate effect of the mammary-tumor agent, if only in a rather negative sense. The incidence of these pseudopregnancies appears to vary greatly between strains, as is shown in table 5.

Table 5. Spontaneous Pseudopregnancy In Various Inbred Strains Of Mice And Fi Hybrids*

Strain

Number of cycles studied

Pseudopregnancies (percent)

IF

105

44

C57BL

115

23

DBAf

116

27

O20

186

14

C3Hf

332

12

CBA

282

11

A

329

4

(♀ O20 x ♂ DBAf)F,

312

12

(♀ C57BL x ♂ DBAf)F,

939

14

*Four animals per cage.

The incidences thus vary between 4 and 44 percent. Of special interest is the occurrence of only a low percentage in the mammary-tumor-agent-containing strain A. This strain is exceptional in another respect: It is the only agent-containing strain in which we can observe a low incidence of mammary carcinoma in virgin animals and a much higher incidence in animals that have been made pregnant or pseudopregnant only a relatively small number of times. This phenomenon can occur only in a strain with a normally low number of spontaneous pseudopregnancies, as otherwise induced pseudopregnancies could not be expected to have such a marked influence.

If the agent does not influence the hormone production, it would follow, per exclusionem, that its effect has to be sought in an influence on the susceptibility or reactivity of the mammary gland. Experimentally this is difficult to prove. No effect of the agent could be found in young animals in tests for the reactivity of the mammary gland to estrone (table 6). It is probable that the agent acts only after a certain latency period. Thus far the only visible effect of the agent is the development of hyperplastic nodules in the mammary gland.

It is conceivable that the host response can be influenced by factors which act on the propagation and transmission of the agent.

There are strains of mice that propagate the mammary-tumor agent and transmit it from generation to generation continuously, and there are strains that cannot propagate the agent continuously and thus can be considered as naturally agent-free. A proper genetic constitution is thus the premise for the propagation and transmission of the agent. Little is known of how this mechanism acts.

The question can be raised whether the capacity to propagate the mammary-tumor agent and transmit it from generation to generation continuously is correlated with susceptibility to the development of mammary cancer after hormonal treatment only. For this reason, we compared the mammary-tumor incidences after standardized hormonal treatment in two agent-free sublines of strains that naturally possess the agent (C3Hf and WLLf), one strain (IF) which can propagate the agent, without, however, being infected naturally, and three strains (O20, CBA, and C57BL) which cannot propagate the agent continuously from generation to generation and thus are naturally agent-free.

Table 6. Threshold Dose Of Estrone For Growth Of Mammary Gland In Castrated Mice With And Without The Agent Estrone (G.)

Strain

0.01

0. 05

0. 1

0.2

0. 3

♀ DBA (with)

_

( + )

+

♀ DBAf (without)

-

( + )

+

♂ DBA (with)

-

( + )

+

♂ DBAf (without)

-

(-)

+

♀ A (with)

-

+

+

♀ Af (without)

-

+

+

♀ C57BL (with)

-

( + )

+

♀ C57BL (without)

-

( + )

+

As hormonal treatment we chose the most effective method known: the isografting of pituitaries in otherwise intact female animals. The transplants produce prolactin continuously and thus induce a hormonal constellation which favors mammary-gland carcinogenesis. The results are shown in table 7.

Table 7. Mammary-Tumor Frequency After Subcutaneous Iso-Grafts Of 20 Pituitaries In Inbred Strains Of Mice Without The Agent

Strain

Number

Percent

Average tumor age (days)

Average age at death (days)

C3Hf

18

89

419

386

WLLf

50

18

622

583

IF

19

0

-

477

O20

40

67

531

548

CBA

14

50

527

576

C57BL

32

12

454

583

The conclusion must be that a correlation between the susceptibility to the development of mammary cancer after hormonal stimulation and the propagation and/or transmission of the agent does not exist.

The limited data available suggest, however, that there does exist a correlation between the susceptibility to develop mammary carcinoma after hormonal treatment and the susceptibility to develop these tumors after administration of the agent. Strains with a high incidence of mammary carcinoma in the presence of the agent generally show also a high incidence of these tumors in the absence of this agent, after suitable hormonal treatment.

In strains which possess a high susceptibility to mammary carcinoma after either hormonal treatment, i.e., implantation of pituitaries, or after infection with the agent, it is difficult to show clearly the interaction after combined treatment. The C57BL strain, however, is suitable for the demonstration of this interaction. Hormonal treatment alone or the mammary-tumor agent alone induces but a few tumors in otherwise untreated female animals of this strain. Combined treatment induces a high incidence of mammary tumors, as is shown in table 8.

Table 8. Mammary Carcinoma In C57bl Female Mice After Various Treatments

Treatment

Number

With tumors

Without tumors

Percent

Average tumor age

(days)

Average age at death (days)

Controls (virgins)

Fostered on (♀C3H X ♂ WLLf)

193

0

-

621

F1 (virgins)

61

11

507

602

20 Pituitaries, subcutaneous

32

12

454

583

Fostered on (♀ C3H X ♂ WLLf) F1 and with 20 pituitaries, subcu-

taneous

25

56

460

491

From the evidence presented, the host response to the mammary-tumor agent, i.e., the development of mammary cancer, is determined primarily by hormonal factors. Thus the host response is dependent not only on factors directly or indirectly acting on the mammary-tumor agent as such, but also on those factors acting on the endocrine system.