TO MAMMARY CANCER IN FEMALE MICE

SUSCEPTIBILITY TO FOLLICULAR HORMONE AND DISPOSITION TO MAMMARY CANCER IN FEMALE MICE P. J. VAN GULIK AND R. KORTEWEG (From the Laboratory of the Antoni van Leeuwenhoekhuis, Netherlands Institute for Cancer Research, Amsterdam) There are numerous indications that the follicular hormone plays an important part in the origin of cancer in the mammary gland of the mouse. Among spayed females of a highcancer strain mammary tumors occur much less frequently than among normal controls (Loeb; Murray; Korteweg, and others) ; they have been produced in varying percentages of castrated males of a highcancer strain by ovarian grafts (7 per cent, Murray, 1; 27 per cent, Korteweg, 2), and have also arisen in male mice following injections of estrin (Lacassagne, 3; Burrows, 4; Bonser, 5). The estrous cycle would seem to be not quite identical in all strains of mice. One of the writers (6) has observed that with advancing age the cycles diminish more rapidly in frequency and intensity in Little s lowcancer strain, C57 black, than in the highcancer dilute brown strain of Murray and Little. Lacassagne (7) also found a more regular and more distinct cycle in animals of a highcancer strain (R3) than in those of a lowcancer strain (XVII). Bonser (8), however, found no differences of importance in this respect between a highcancer (Bagg albino) and a lowcancer (Little s black agouti) strain. Rrunschwig and Bissel (9), who studied the estrous cycle in normal mice, found a greater number of cycles per timeunit and a longer duration of each separate cycle in females of the lowcancer strain, C57 black, than in those of the highcancer dilute brown strain. Their series of mice, however, numbered only 8. While the observations of these various investigators show some divergences, the differences in the estrous cycle in mice of highcancer and lowcancer strains do not seem on the whole to be very great. We do not feel justified, however, in concluding without further investigation, as did Bonser and Brunschwig and Bissel, that no causal relationship exists between a high follicular hormone production and a disposition to mammary cancer. A normal cycle is unquestionably necessary for the preservation of the species. If in mice of certain strains the genital organs are relatively immune to follicular hormone, it seems probable that these animals may normalize their cycle by the production of a larger quantity of hormone. The presence of a normal cycle consequently gives no indication as to the amount of hormone produced. We are concerned here with two questions. Is there a difference in susceptibility to follicular hormone among different strains of mice? If the answer to this question is in the affirmative, is it possible to establish a relation between such a difference and a difference in disposition to cancer of the mammary gland? 56

FOLLICULAR HORMONE AND MAMMARY CANCER IN FEMALE MICE 57 MATERIALS AND METHODS Three strains of mice were used: the dilute brown strain of Murray and Little (dba), with a high incidence of spontaneous cancer of the mammary gland (1); Little s strain C57 black (CS?), with a low incidence of cancer of the mammary gland (one out of each two or three hundred females); strain 2 Leeuwenhoekhuis (2). The first two of these strains are well known. About the third strain, built up in our Amsterdam laboratory, one communication has appeared (14). It was developed by brothertosister mating of albino mice (which were not in any way related to mice of any of the familiar strains) for twenty generations, followed by cagebreeding, so that a rather high standard of purity has been attained. The animals of this strain are strongly disposed.to lung cancer. Mammary cancer, however, seldom occurs, having been found in only.3 of 272 adult virgin females. Besides mice of these pure strains, a number of F, hybrids were used. These are classified as follows: dilute brown females X C57 black males, designated as dbf,; C57 black females >(: dilute brown males, as BdF,; 2 females >(: dilute brown males, as OdF,. The living conditions of the mice are described elsewhere ( 11 ). The female mice were anesthetized and castrated on either side through an incision in the midline of the back. The ovaries, fallopian tubes, and the upper part of the uterine horns were liberally removed. Following this, vaginal smears were made daily for at least two weeks in order to determine whether castration had been successful. At the conclusion of each experiment the mice were killed and autopsied. The follicular hormones used for these experiments were preparations of menformon and oestron, manufactured by Organon Ltd. at Oss, Holland, and kindly placed at our disposal by the Pharmacotherapeutical Laboratory of the University of Amsterdam. In the first experiment the hormone was administered with the food; in the other experiments it was injected subcutaneously. A fine needle, 5 cm. in length, was used for the injections. This was inserted under the skin as far as possible to prevent leakage of the hormone and the skinwound was pinched together for a few minutes after withdrawal of the needle. The vaginal smears were made with the aid of a platinum loop daily at 9 a.m. and on Wednesday, Thursday, and Friday at 4 p.m. as well. The smears were left to dry in the air and were stained by the Giemsa method. A good stain is important and requires much care, as the leukocytes are otherwise not sufficiently visible, and an erroneous interpretation may result. The estimation of the stage of the cycle on the stained vaginal smears was made by two persons working, independently of each other, in accordance with the observations of de Jongh and Laqueur (12). These authors distinguished eight stages, represented by the letters a, b, c and d (negative smears), e, f, g and h (positive smears). As the stages e and f are under consideration in the experiments to be described, a short description of these is appropriate. In stage e (positive) the epithelium consists of more than 5 per cent of nucleated cells; not more

58 P. J. VAN GULIK AND R. KORTEWEG Strain 3 x 1u 4 x 1u c 57....... 43% f 12.1 86 % f 9.76 dba........ 4% f 5.2 79.3% Et 4.45 I I 5 x 1u 93 % f 2.1 73.2% f 1.61 Strain 3 x 1u 4 x 1u c 57........ 43% f 12.1 86 f 9.76 dba........ 53.77 f 5.19 5 x 1u 93 % f2.1 67.3 % f 1.68 than 5 per cent of the entire cell material may consist of leukocytes (proestrus). In stage f (more strongly positive) more than 5 per cent of the epithelium is made up of cornified cells; 5 per cent at most of the cell material may consist of leukocytes (estrus). Since with adequate staining it happens only rarely that no leukocytes at all are to be seen, a smear is called " positive " if these number less than 5 per cent of the total cell material. EXPERIMENTS EXPERIMENT I (March lboct. 28, 1933): Sixtyeight C57 and 19 dba spayed female mice, from eight to fifteen months of age (most of them from ten to thirteen months), were given menformon in the diet. For this purpose an oily menformon solution of fixed concentration was mixed with skimmed milk. During the first six weeks the mice received 1 units on two successive days each week. Since, however, this dose appeared to be too low, it was increased and for eleven weeks each animal received 81 units per week, divided among three days. The percentages of positive vaginal smears (f and > f; estrus) were as follows: C57, 74 per cent & 1.7; dba, 13 per cent 1.88. EXPERIMENT I1 (Oct. 2Dec. 15, 1934): Eleven C57 and 15 dba spayed females, eight to thirteen months of age, were injected subcutaneously with an aqueous solution of menformon, containing 1 unit per.1 C.C. The injections were commenced on Mondays and if, for instance, it was desired to give 5 units per cycle, two injections of 1 unit each were given on Monday and Tuesday and a final injection of 1 unit on Wednesday. Injections were made as follows: one week 3 X 1U; one week 4 X 1U; nine weeks 5 x 1U. The percentages of mice reacting with positive smears (e and > e) are given in Table I. The percentages which reached stage f and higher (cornification, estrus) are shown in Table 11. From a comparison of these tables, it appears that the C57 mice showing a positive reaction invariably attained the stage of cornification. Among the dba mice. however,

FOLLICULAR HORMONE AND MAMMARY CANCER IN FEMALE MICE 59 TABLE 111 : Experiment ZZ: Percentages of Positive Reactions (e and > I i) on Various Days First day Dose (day of Second day Third day Fourth day Fifth day injection) 3 x 1u c57... 4 x 1u c57... 5 x 1u c 57..... 1.2% 4.% 73.7% 11.3% 49.1% 43.% 85.5% 52.6y 9.% 62.% 26SY.8% were many which got no farther than stage e (proestrus). The difference between the two strains is especially clear after a dosage of 3U. Table I11 shows the percentages of mice with a positive reaction (e and higher) on the various days, reckoning the day of injection as the first day. TABLE IV: Experiment ZZZ: Positive Smears (e and > e) after Various Doses of Menformon in Olive Oil 1+2+2u 2+2+2u c 57.... dba... Bd Fi.... db F1..... 1% f 5.39 35% f 6.87 1% f 4.57 35% f 6.77 75% f 5.25 66% f 5.79 6% f 5.32 5% f 5.32 65% f 7.44 95% f 3.84 6% zk 6.8 55% f 7.41 I TABLE V: Experiment ZZZ: Percentages of Mice Attaining Cornijication (f and > f) after Various Doses of Menformon in Olive Oil Strain I ~ + I + I U I 1 + 2 + 2 ~ I 2 + 2 + 2 ~ ~~~ ~ c 51..... dba... Bd FI.... db F1..... 1% f 5.39 1% f 5.19 5% f 3.29 3% f 6.87 75% f 5.25 ' 33% f 6.8 37% f 5.65 35% f 5.15 657 f 7.44 257 f 7.42 25% f 6.47 257 f 6.47 EXPERIMENT I11 (Jan. 25Feb. 22, 1937): Seventeen C57, 15 dba, 21 BdF,, and 2 dbf, spayed female mice, five to six months old, received three injections of a solution of menformon in oliveoil, in twentyfour hours (Monday at 9 a.m. and 4 p.m. and Tuesday at 9 a.m.), as follows: one week 1 + 1 + lu, i.e. 3U; two weeks 1 + 2 + 2U, i.e. 5U; one week 2 + 2 + 2U, i.e. 6U. The percentages reacting with positive smears (e and > e) are given in Table IV. From this table it would appear that the dba mice reacted better than the blacks. That this is not so, however, appears from Table V, in which the percentages attaining stage f and higher (cornification) are given.

5 1 P. J. VAN GULIK AND R. KORTEWEG TABLE VI : Expedment ZZZ: Percentage of Positive Reactions (c and > e) on Various Days Dose First day (day of injection) Second day Third day Fourth day Fifth day 1+1+1u c57... dba............. Bd Fi........... db Fi........... 26.6% 5.% 1.% 11.6% 5.% 25.% 1+2+2u c57... dba...........,. Bd Fi........... db Fi........... 3.7, 56.67 27.% 32.5% 69.8% 28.1% 3.% 3.7 47.1 9$ 6.6% lo.o%, 1.% 2 + 2 + 21J C57... dba............. Bd Fi......, db Fi........... 17.7% 93.3% 4.% 4.% 48.37 26.3% 25.% 2.% 35.3% 1.% 5.% Table VI shows the percentages with positive smears (e and > e) on the various days after injection of different doses of menformon. EXPERIMENT IV (Nov. 28, 1938Feb. 21, 1939): Five C57, 14 dba, 15 BdFI, and 14 dbf, spayed females, eight to nine months old, received subcutaneous injections of menformon in oliveoil, in different concentrations, over twentyfour hours, viz, at 9 a.m. and 4 p.m. on Monday, and at 9 a.m. on Tuesday, In all, injections were given for seven weeks, namely: two weeks 3 X 1/3U; one week 3 X 2/3U; three weeks 3 X 1U and one week 3 X 2U. TABLE VI I : Experiment Z V: Total Positive Reactions (e and > e) afkr Injections of Menformon in Olive Oil c 57............, 73.5% f 5.12 dba...................... 45.97 f 3.37 Rd Fi...........,. 73.3% f 2.9 db FI.......................... 6.2% f 3.3 Table VII shows the percentages in which a positive reaction (e and > e) was obtained, taking all injections together. Table VIII gives the percentages of positive reactions (e and > e) after different doses of menformon. Strain 3 x 4u C 57..... 4% f 1.1 dba.... 14% f 4.45 Bd Fi... 15% f 4.38 db Fi.... 187 f 4.92 3 x tu 3 x 1u 2 x 3u 8% f 12.1 8% f 8.76 1% 64% f 8.8 45% f 6.28 93% f 4.51 73% f 7.95 82i% f 4.72 9.37 f 4.58 75% f 7.75 69% f 5.59 1% The percentages of mice attaining f and > f (cornification) reactions after different injections are shown in Table IX, while Table X gives the percentage

FOLLICULAR HORMONE AND MAMMARY CANCER IN FEMALE MICE 51 1 Strain C 57..... dba....,. Bd Fi.... dh FI.... 3 x IU 3 x :u 3 x 1u 3 x 2u 4% f 1.1 soy, f 12.1 8% f 8.76 1% 14y f 4.45 57'z f 8.76 36% f 5.5 71% f 8.8 26% f 5.39 73% f 7.95 82% f 4.72 937 f 4.58?% f 3.23 72% f 8.9 647 f 5.5 87% f 6.16 TABLE X : Experiment I V: Percentages of Positive Reactions on Various Days Dose First day (day of injection) Second day Third day Fourth day Fifth day 3 x tu c57... Bd Fi........... db F1........... 3 x au c57... Bd FI..,........ db Fi........... 3 x 1u c 57............ ' Bd Fi....... db Fi..,........ 2.% 21.5,$ 14.3% 28.5% 14.4% 16.6% 4.7, 7.2% 17.% 1.9% 8.% 5o.oyo 73.3% 78.5% 78.5% 4.57 69.5% 66.5% 7.1% 2.3% 6.5% 3 x 2u c 57... dba..,..,....... Bd Fi........... db Fi........... 4.7, 71.5% 6.% SO.Ooj, 1.% 71.5% 93.3% 85.7% of the mice showing positive smears (e and > e) on the various days after injection. EXPERIMENT V (Feb. 6July 2, 1939): In this experiment only one weekly injection of oestron was given, namely at 11 a.m. on Tuesdays. Twentytwo C57, 29 dba, 5 2, 25 BdF,, 28 dbf, and 49 OdF, spayed females were used. As not a sufficient number of C57 and dba mice was available at the initiation of the experiment, it was begun with 4 of the former and 3 of the latter, the numbers being gradually increased to 22 and 29 respectively. The age was two to three months for the C57 mice and three to four months for the others. Injections were given during twelve weeks; one week 1U; one week 2U; one week 2l/iU; one week 3U; one week 4y~U; two weeks 6U, and five weeks 9U. The percentages showing positive (e and > e) smears, taking all injections together, is given in Table XI. Tables XII, XIII, and XIV show the reactions arising after injection of a low dose (lu, 2U and 2MU), of a medium

512 P. J. VAN GULIK AND R. KORTEWEC dose (3U and 4Y'U), and of a high dose (6U and 9U) of oestron. The stage e and > e reactions are given in Table XI1 and stage f and higher (cornification) in Table XIII, while Table XIV shows the percentages with positive smears (e and > e) on the various days after injection. SUMMARY OF EXPERIMENTS The susceptibility to follicular hormone (menformon and oestron), of 437 spayed female mice of different genetic composition was ascertained in five experiments carried out in 193339. For these experiments there were used 123 C57 black, 92 dilute brown (MurrayLittle), 5 2 Leeuwenhoekhuis, 61 BdF,, 62 dbf,, and 49 OdF, animals. The hormone was administered in four different ways: per os, mixed with milk; by subcutaneous injection in an aqueous solution, given over several days; by subcutaneous injection in oliveoil, in divided doses over 24 hours; and, in the last experiment, in a single injection in oliveoil. TABLE XI : Expen'mcnl V: Tot1 Positive Readions (e and > c) afkr Injection of Oestron C 57..... 73.4% f3.1 dba..... 37.3% *2.96 2..... 48.6% i2.2 Rd FI..... 53.% i 2.9 db FI..... 46.8% f 2.36 Od FI..... 49.6% f2.9 In Experiment I, in which menformon mixed with milk was administered per os, the possibilities for error are great. As the mice were kept four in a cage, there was never any certainty that each received an equal quantity of the hormone. In the other experiments, in which the menformon was injected subcutaneously, leakage from the injection channel was sometimes observed in spite of the precautions taken. This technical difficulty may account for the fact that the results occasionally did not come up to the expectations. The response to an injection is moreover influenced to some extent by the previous injection. If during a certain week a large dose is administered, with consequent stimulation of the mice, as it were, the reaction to a small dose in the following week will be better than if no stimulating injection had preceded. This may explain the differences in the reaction of a mouse to the same dose on different occasions. Our experiments offer further proof of the fact, which was indeed already known, that better results are obtained with several injections of follicular hormone than when the same quantity is administered in one injection. The mice generally reacted better in the spring than during the winter. EXPERIMENTAL RESULTS 1. The genital organs of females of the C57 black strain were found to be more susceptible to follicular hormone than the genital organs of females of the dilute brown strain. Roughly estimated the amount of the hormone necessary to cause a given reaction in the blacksis only one third of that required to produce the same reaction in the dilute brown mice.

~... FOLLICULAR HORMONE AND MAMMARY CANCER IN FEMALE MICE 513 Strain One low dose One medium dose One high dose (6 9U) c 57...... dba....... 2...... Bd FI..... dbf1... Od FI..... 57.2% f 4.99 4.% f 2.56 2.y f 3.77 26.4% f 4.4 28.1% f 5.39. 16:% f 3.5 4.% f 4.65 32.% f 4.38 44.4% f 7.54 48.% f 4.72 8S.lyO f 3.5 5.% f 3.63 66.1% f 2.42 64.4% f 2.49 54.5% f 2.72 61.37 f 2.69 TABLE XIII: Experiment V: Percentage of hlice Attaining Cornijication (f and > f) after a Single Low, Medium or High Dose of Oestron Strain c 57....... dba........ 2....... Bd FI...... db FI...... Od FI...... Low dose (1 21U) 53.2% f 4.85 4.% f 2.56 6.% f 2.22 13.2% f 3.57 21.8% f 5.2 1.% f 3.36 Medium dose High dose (3 43U) (6 9U) 61.47 f 4.37 19.2% f 3.2 12.% f 3.2 26.4% f 2.35 24.% f 4.4 32.6% f 2.69 33.3% f 7.41 26.1 Yo f 2.49 28.Y f 4.31 35.2% f 3.36 TABLE XIV: Experiment V: Percentages of Mice with Positive Reactions (e or > e) on Various Days Dose First day (day of injection) Second day Third day 7ourth day Fifth day One low dose (1 24U) c 57................... dba... 2......... Bd Fi........ 57.% 4.7 2.% 26.3% 19.5% 16.% 1.6Y 2.4% c 57........ dba.,.................. Bd F1.................. db Fi.................. Od Fi.......... 4.7 32.% 44.4% 48.% 4.% 2.% One high dose (6 9U) c57... db F1............. Od FI............. 83.87 42.3 % 61.5y 66.7 48.7% 68.% 3.2% 1.7% 2.2% 1.7% 8.% 2%

514 P. J. VAN GULIK AND R. KORTEWEG 2. Reactions after the administration of follicular hormone began later in the black than in the dilute brown strain, but lasted longer in the former. 3. Of the mice giving a positive reaction the blacks showed the highest percentage attaining cornification. 4. As regards the susceptibility of the genital organs to follicular hormone, the strain 2 Leeuwenhoekhuis was found to be between the black and dilute brown strains. 5. The susceptibility of the reciprocal F, hybrids was not greatly divergent; in general it lay between the susceptibility of specimens of the parent strains. DISCUSSION The first question put by us in the introduction, whether in different strains of mice there is a difference in susceptibility to follicular hormone, has been answered affirmatively, and we may now pass to the discussion of the second question: Is there any relation between these differences in susceptibility to follicular hormone and the differences in disposition to cancer of the mammary gland in various strains of mice? In our experiments the genital organs of the females of the highcancer dilute brown strain appeared to be much less susceptible to follicular hormone than those of the lowcancer strain C57 black. In the 2 strain, whose susceptibility to follicular hormone lies between that of the other two strains, cancer of the mammary gland is rarely found. The disposition to mammary cancer is probably higher, however, in this strain than in the C57 black strain. Many of the 2 mice die of lung cancer at an early age, which may tend to decrease the frequency of cancer of the mammary gland. Bonser, who administered estrin to spayed females of different strains, observed that the reactions were slightly weaker in animals of a highcancer strain (Little s line of Bagg albinos) than in animals of a lowcancer strain (Little s black agouti line) and that the reactions in the former lasted for a shorter time. She considered these differences of little importance. From her Table 111, however, it appears that, though no distinct differences were observed after very high doses of estrin, the response to smaller doses, spread over several days, showed decided differences. As in our case, so also in hers, the differences appeared most clearly not when a very large, quite unphysiological dose of follicular hormone was administered, but when a more physiological dose was given. The observation made by Bonser on a much smaller number of mice would seem therefore to bear out our results. The susceptibility of the genital organs to follicular hormone is characteristic for different strains of mice and consequently is determined genetically. It further appears that of the five strains of mice experimented on, the two with a high disposition to cancer of the mammary gland are the very two in which the genital organs of the females evince little susceptibility to follicular hormone, so that these females probably produce a relatively great quantity of this hormone. In other words, there seems to exist a positive correlation between the quantity of follicular hormone produced and the disposition to cancer of the mammary gland, This suggests that this relation is at the same time a causal one. Only a very careful investigation on the relative

FOLLICULAR HORMONE AND MAMMARY CANCER IN FEMALE MICE 515 susceptibility to follicular hormone in a much greater number of strains will finally solve this problem. The two species of F, hybrids evince an almost equal intermediary susceptibility to.follicular hormone, which suggests that the susceptibility to this hormone bears no relation to the maternal extrachromosomal factor of Loeb and Lathrop, Little, and Korteweg, but is chromosomally determined. If the mammary glands of mice of highcancer and lowcancer strains possessed an equal susceptibility to the cancerogenic action of follicular hormone, the surplus of this hormone produced by specimens of the highcancer strain might exercise a deleterious influence on their mammary glands. In this way a causal relation between degree of susceptibility to follicular hormone and disposition towards cancer of the mammary gland in the female mouse would be conceivable. The correctness of this hypothesis, which was advanced by one of us as early as 1934 on the strength of the results of a preliminary research (13), seems to become a little more probable in view of the results of our later experiments. Even if this hypothesis should prove correct, however, only a part of the facts observed can be explained by it. SUMMARY AND CONCLUSIONS The susceptibility of 437 spayed female mice of different genetic composition to follicular hormone (menformon and oestron), administered in different ways, was determined. Great differences in susceptibility, probably genetically determined, appeared to exist between the different strains. From the data collected by Bonser and by us it appears that the genital organs in our two highcancer strains were less susceptible to follicular hormone than those in our three lowcancer strains. The probability is suggested that the females whose genital organs are least susceptible produce the greatest quantity of follicular hormone under normal conditions. As a result of our studies on mice the possibility is suggested that there may exist a causal relation between the degree of susceptibility of the genital organs to follicular hormone and the degree of disposition to cancer of the mammary gland. LITERATURE 1. MURRAY, W. S.: J. Cancer Research 12: 18, 1928. 2. KORTEWEC, R.: Acta, Unio internat. contra cancrum 2: 137, 1937. 3. LACASSACNE, A.: Compt. rend. SOC. de biol. 114: 427, 1933. 4. BURROWS, H.: Rapport de la 4"'" Conference de la Leeuwenhoekvereeniging, Amsterdam, June 1935, p. 82. 5. BONSER, G. M.: Twelfth Annual Report, British Empire Cancer Campaign, 1935, p. 95. 6. KORTEWEG, R.: Nederl. tijdschr. v. geneesk. 77: 438, 1933. 7. LACASSACNE, A.: Compt. rend. SOC. de biol. 115: 937, 1934. 8. BONSER, G. M.: J. Path. & Bact. 41: 33, 1935. 9. BRUNSCHWIG, A,, AND BISSEL, A. D.: Arch. Surg. 33: 515, 1936. 1. KORTEWEG, R.: Genetica IS: 337, 1936. 11. KORTEWEG, R. : Genetica 18 : 35, 1936. 12. DE JONGH, S. E.. ASD LAQUEUR, E.: Abderhaldens Handb. d. biol. Arbeitsmethoden, Abt. V, Teil 3B: 16391666, 1938. 13. KORTEWEG, R.: Nederl. tijdschr. v. geneesk. 79: 1463, 1935. 14. KORTEWEG, R.: Nederl. tijdschr. v. geneesk. 83: 5989, 1939.