CASTRATION EXERTS NO INFLUENCE UPON THE GROWTH OF TRANSPLANTED OR SPONTA-

NEOUS TUMORS IN MICE AND RATS.

BY JOSI~ S. H I L A R I O , M.D.

(From Columbia University, George Crocker Special Research Fund, New York.)

(Received for publication, May 18, 1915.)

Graf 1 was the first to seek some relationship between the internal secretion of the sexual organs and the development of transplanted tumors. The sexual glands were removed from 58 male and 60 female mice, which were afterwards inoculated with a tumor having an inoculation percentage of from 80 to IOO, 40 male and 31 female unoperated mice serving as controls. He concluded that castration exerts no essential influence either upon the percentage of takes or upon the rate of growth. A few years later, AlmagiA 2 described five experi- ments, comprising 30 castrated and 50 normal mice inoculated with a tumor characterized by slow growth and a comparatively low inoculation percentage. Among the 50 normal animals, 42 had tumors and IO had not, while in the cas- trated group only 6 were positive, 24 having no tumor. I t was observed, fur- thermore, that the injection of testieular extract into tumor-bearing mice exag- gerated the development of their growths, and that castrates which had proved refractory to inoculation could be restored to a condition of receptivity by the introduction of this material. He concluded, therefore, that the development of a tumor is stimulated, either directly or indirectly, by the sexual glands, and inhibited by their absence.

Rohdenburg, Bullock, and Johnson, 3 like Graf, expressed the opinion that castration exerts no effect, or, at the most, a very slight one, upon either the inoculation percentage or the rate of growth.

An inference exactly contrary to that of Almagi~ has recently been recorded by Sweet, Corson-White, and Saxon, 4 whose material consisted of 73 castrated mice and 14 castrated rats, compared with 73 normal mice and I5 normal rats. These authors used a tumor with but moderate powers of adaptation and pro- liferation, since such a growth would be more sensitive to any slight degree of immunity present than would one with a very high inoculation percentage.

1 Graf, R., Centralbl. f. Path., 19o9, xx, 783. 2 Almagi/t, M., Bull. d. r. Accad. d.i reed. di Roma, 1912, xxi, lO2. a Rohdenburg, G. L., Bullock, F. D., and Johnson, P. J., Stttdies in Cancer

and Allied Subjects, New York, 1913 , iii, 87. 4 Sweet, J. E., Corson-White, E. P., and Saxon, G. J., Jour. Biol. Chem.,

1913, xv, 181.

l b 8

Josd S. Hilario. 159

They asserted that receptivity for such a tumor is increased by castration, and that the proliferative power of the neoplasm is augmented.

Since each of the three possibil i t ies--(a) that castration in- creases tumor growth, (b) that castration retards tumor growth, (c) that castration exerts no effect upon tumor growth has been advocated, it is necessary that the experiment be repeated on a larger number of animals and with tumors of various types.

The following experiments, amounting to nine in number, in- clude in the final reckoning 256 male animals, of which 88 were castrated mice, 35 were castrated rats, lO 3 were normal mice, and 3 ° were normal rats. The animals in every experiment except VI and IX were of approximately equal weight and age, and all those in any one experiment were obtained from the same dealer. Castra- tion was performed under ether anesthesia, through an incision in the median line of the scrotum, care being taken not to crush the glands in the forceps so that complete removal of all testicular tissue might be assured. The wound was closed by a continuous catgut suture.

From ten to twenty-six days after castration the animals were inoculated in the right axilla, seven series with carcinoma, and two with sarcoma. The castrated animals and their controls were inoc- ulated at one sitting from the same tumor and with grafts of simi- lar size, and in every instance the castrates and their controls were kept in adjoining boxes and on the same diet. Details regarding tumor dose, weight of animals, etc., are to be found under Text- figs. I to IO, in which the tumors are reproduced in silhouette, their actual size being readily calculable from the IO cm. scale at- tached. The first charting was made ten days after inoculation, subsequent ones at weekly intervals; a J- sign indicates the death of an animal. The results were read at a time when the largest number of animals could be included. More exact information on this point will be found in the detailed discussion of the charts.

Growths of various kinds were employed to test the presence or absence Of resistance. Thus, among the mouse carcinomata, Crocker Fund Tumors Nos. I I, 15, and 46 are adenocarcinomata with an inoculation percentage of from 5 ° to 75, which grow

160 Castration in Mice and Rats.

slowly and occasionally retrogress, while Tumor 18o, a carcinoma of the solid type, grows in every mouse with extreme rapidity and never recedes. The Ehrlich mouse sarcoma, with an inoculation percentage of from 5 ° to IOO, grows quickly and sometimes re- gresses. As for rat tumors, the Flexner-Jobling adenocarcinoma grows with moderate speed in from 50 to IOO per cent of inoculated rats, receding rather frequently; the Jensen sarcoma grows more rapidly, gives an inoculation percentage of from 5o to IOO and often retrogresses.

With this preliminary description of experimental material, the text-figures may be introduced.

Experiment I . - - (Text-f ig . I.) The result was read at the fourth charting, in order to include Nos. 6, 24, 26, 31, and 32, which died before the fifth. The control group contains 17 positives (85 per cent) and 3 negatives, the castrates 8 positives (67 per cent) and 4 negatives. The rate of growth is the same in both series. In the control group 2 tumors (7 and 18) receded, and in the castrates none.

Experiment//.--(Text-fig. 2.) Here the computation was made at the third charting to include Nos. 2, 9, I3, and 32, No. 15 being counted among the posi- tives since it developed a tumor before the next charting. Hence there are I5 positives (75 per cent) and 5 negatives in the control, and 13 positives (62 per cent) and 8 negatives among the castrated mice. There were IO cases of invo- lution among the control animals (Nos. 3, 6, 7, IO, II, I2, 17, I8, 19, 20) and 2 (Nos. 29, 36) in the castrated. Growth was more vigorous in the castrates, but this circumstance cannot be regarded as significant because this is the only experiment in which such a phenomenon was observed, and, secondly, because the rate of growth in the castrated animals is characteristic of Tumor 46 under normal conditions. It would, therefore, be more accurate to say that prolifera- tion was diminished in the control groups, by some factor unconnected with the experiment. The high mortality suggests an infection, more severe, perhaps, in the controls, since it is known that tumors do not grow so well in sick animals.

Jos~ S. Hilario. 161

CONTROL C A S T R A T £ D . . 24 3, ~ o m Io . . x N m l

' - • • e 4 2 t . • 0 4 0

z . • • e e

3 - . , e e

4 •

S •

6 •

7 - • • e e

I •

9 •

I 0 •

s s . • • • e

2 3 • • • t I

• t e e . . . • e t

• e Q e 2s . . ' e Q

• o o t u . • • t t

L P / - - • • Q

• • • • 2 e • • • • •

• ' e g 2 e . . . . .

• ,, • 0 3 0 . . . . .

31 . . . .

3 2 . . . .

I I . • g • •

12 . . ~ • •

13 • • ~ • •

I S • • •

16 •

17 . • • • •

18 - - - - - -

19 . -

2 0 i , . . . . . , , , |

t o CM.

T~xToFIo. I. Experiment I. I z 0 . Nos. I-2o, normal male controls (average 2 0

weight I4.8 gm.). Nos. 21-32, castrated Dec. Is, I914 (average weight I5.4 gm.). All mice inoculated Dec. ~.8, I914, in the right axilla with o.oi gm. by the needle method.

162 Castration in Mice and Rats.

C O N T R O L t 0 t7 2 4 3 1 : 9 D A Y S

t • • " , e Q

2 ., , • 1"

3 . • • • •

4 - - . ~ • ,~

S , . , 4 •

| . • • • •

C A S T R A T E D lf0 17 24 ~ 38DAYS

2 2 . ~ A q l

z 3 - . , • 0

z 4 . . , • 0

ss • t • i l

Z6 -- . • 0 •

z7 • . • • •

7 -- • t A t 2 8 • • * • •

• • q 9 • 2 9 • * 9 • •

9 • • • 1" 3 0 • • • * e

1 0 . • % 1 % 31 - - . .

' 3 2 . . • 1"

1 2 . . , . . 3 3 . ~ : 4 1"

t 3 . • - t • • t

t 4 . • • • 3 S - - - • , 1 .

I S . . . . 3 6 -

16 • 1" 3 7

17 . . 3 8

18 • 3 9

19 . . . . . 4 0

2 0 , 4 !

4 2

21 4 3 ) , . , , , . . . , ~

WCJL

TEx¢-FIo. 2. Experiment IL -~1~" Nos. I-2I, normal male controls (average

weight z7.5 gra.). Nos.,~2,43, castrated Dec. I, I914 (average weight I7.3 gm.). All mice inoculated Dec. I5, I914, in the right axilla with o.ox gin. by the needle method.

Jos5 S. Hilario.

COIITROL

- • • * O l O

2 - - . " • 8 8

3 - . • • I • •

4 - - • • @ g •

S . • . • • •

I- . • 4 • •

7 . * • I • •

I - • I • g

I - • • • I • @

!1 - - . • • ,s • • e • •

1 2 - - - - * * * •

1 4 - - • * • Q

l S - - , , o

! ~ o • • a t

1 7 I

1 1 - -

t 9 - -

2 0 I

1 6 3

CASTRATrO t0 I"I 2 4 31 3 8 4S S t-DAYS

2 3 - * * • • • •

I S - - - - • • • • •

2 6 . • • • • • •

2 7 . • ~ • • • •

~ - - - - . • • •

• • . * • • •

3 0 - • • • • • •

3 1 - - - - - - • •

36

m _ m I i

~ cM.

15 TExT-FIG. 3. Exper iment III . ~ . Nos. 1-22, n o r m a l m a l e controls (average

weight 17.4 gm.) . Nos. 23-38 , castrated Dec. IO, 1914 (average weight 17.5 grn.). All mice inoculated Dec. 28, 1914, in the r ight axilla with o.oi gin. by the needle method.

Experiment III.--(Text-fig. 3.) The calculation in this case was made at the seventh charting. 16 controls (73 per cent) were positive and 6 negative, while in the castrates II (73 per cent) were positive and 4 negative. There were 3 cases of involution (Nos. 9, 16, I7) among the control mice, and 4 among the castrated (Nos. 27, 3o, 32, 34). The rate of growth was slightly decreased in the castrated animals.

CONTROL CASTRATL'I) M 17 2 3 ~ JO T/ 2 3 ORS

I - e O zJ • o •

2 ' e • ?.Z • • •

3 " + 4 U " • Q

4 • • • ~ 4 " t •

S • • • 2 S • • •

6 • • • t $ , 0 4

7 • • • 2 7 • • •

O • + 4 2 8 ' o •

9 e • 29 ' • ql

gO • 4 3 0 • • t

I I • • • 31 , • •

12 o • 3 2 . . •

13 • ,~ • 3 3 • • •

14 - o 9 4 . • •

iS • " • 3S • • •

18 . , • 3 8 - - •

17 . . • 3 7 - • ,

i 8 • t 3 8 • • •

19 - - - - - 3 9 "

I 0 - - t 4 0 ' " t I . . . . A ..... +.,~

MeN.

i8o TExT-FIG. 4. Experiment IV. - ~ . Nos. i-2o, normal male controls (average

weight 16.4 gin.). Nos. 2I-4O, castrated Nov. 28, 191- 4 (average weight 15.7 gin.). All mice inoculated Dec. 8, 1914, in the right axilla with o.o1 gm. by the needle method.

Experiment IV.--(Text-fig. 4.) At the third charting, 17 controls (94 per cent) were positive and I was negative. Of the castrates 19 ( ioo per cent) were positive and none were negative. The rate of growth was the same in both groups, and there was no instance of regression.

CONTROL C A S T R A T [ D ,o . ~4 3, w s ,o . , ~ o , , s

, . • o Q u . , * O 2 . , o t - • e O

~ . , • 0 z6 e e O

4 . ' e e 2? ,, 4 Q

s . o O 2 o . , O t

s • , O t l 2o ,, • O

7 • L ¢ 4 ~o . • • O

e . • e g 31 - . . •

s . • o ( I ~z • • • •

t o . . • • 33 - • e t

t l . " i l 0 3 4 • • • 8

t 2 - - • • • 3 S • • • •

~3 . , • • 3 6 - - • •

t 4 • " • • 3 7 - - # •

t S - - • • • 3 8 - - -

t 6 - • • • 3 9 . . . .

t 7 - - • • • 4 0 - - • .

1 o . • O 1" 4,

1 9 • . • • 4 2

2 0 - . - * t 4 3

2 1 • • ,

4 4 2 2 , . .

2 3 | l l o l l l l l t l

ES TExT-FIG. 5. Experiment V . ~ . Nos. 1-23, normal male controls (average

weight I7.5 gm.). Nos. 24-44, castrated Dec. 9, I914 (average weight I5.5 gm.). All mice inoculated Dec. 24, I914, in the right axilla with o.oI gm. by ~he needle method.

Experiment V.--(Text-f ig . 5.) The reading in this case was made at the third charting, to include Nos. 2, x8, 28, and 33. In the controls there were 22 positives (96 per cent) and I negative, and among the castrates I7 positives (8x per cent) and 4 negatives. The growth rate was similar for the two groups, and no tumors regressed.

]66 Castration in Mice and Rats.

CONTROL CASTRAT[D 10 1 7 1 4 3 1 D A Y S fO t 7 2 4 3 1 D A Y S

1 . . • • t l . , e I

2 . . • t t 2 . • • •

3 - - . o • t 3 . . , ,

4 . . . . t 4 . . . +

S * 1" 1 | . . . .

6 - 4" t 6 . • , •

7 ' - - - - - - t 7 - - • • "

I I - - _ 4" 1 9 • • . 4"

S . . . . 1 9 . . . .

I 0 - - - - - - t . 2 O •

2 1 . . . .

I i1 |111,111

FRC TExT-FIG. 6. Experiment VI. I--~-" Nos. I-IO, normal male controls (average

weight 85.0 gm.). Nos. 11-21, castrated Dec. 14, 1914 (average weight 51.7 gm.). All rats inoculated Jan. 9, 1915, in the right axilla with o.o2 gin. by the needle method.

Experiment VI.--(Text-fig. 6.) This experiment offers another example of retarded growth in sick animals, except that, in contrast to Experiment II, both groups are about equally affected. To include this series, it is necessary to read the result at the second charting, where 5 of the controls (55 per cent) were positive and 4 negative. Among the castrates, 9 (82 per cent) were posi- tive and 2 negative. The rate of growth was approximately the same in both groups. Two minute tumors regressed in the controls (Nos. 4, 7) and 6 in the castrates (Nos. 15, 16, 17, 19, 20, 21).

Jos~ 8. Hilario. 167

CONTROL C ASTRATE,D I 0 t 7 2 4 31 38 4 S S 2 D M S t 0 I? ?.4 35 3 8 4S S29AYS

I - • , • • • • S - " : ~ o ° | •

Z , - • • • • • T - - - - - , " • •

3 - • • • • • • II . . . . . ° "

4 - • . , qu • • S - - . . - - - • " •

S - • • • q' • • 1 0 - .

t t .

I i , * , r . . . . I ¢N

FRC TExT-Fro. 7. ExperlmentVII. I - ~ " Nos. I-5, normal male controls (average

weight 5o.o gin.). Nos. 6-11, castrated Dec. 3, 1914 (average weight ~.3 ~ - ) . All rats inoculated Dec. 14, 1914, in the right axilla with 0.02 gin. by the needle method.

,Experiment VH.--(Text-fig. 7.) At the seventh charting, 5 controls (ioo per cent) were positive and none negative, while 4 (67 per cent) of the castrates were positive and 2 negative. The rate of growth was somewhat diminished in the castrated rats, only 2 small tumors having regressed (Nos. to and xI).

168 Castration i,~ Mice and Rats .

CONTROL CASTRATED to 17 24 ~ m s n ~ u ~ m s

• ' q / t . - " t t

Z • • * . ~ t 3 • : ~ q

3 * : qk qk t 4 * " * Q

4 - " • • 1 6 " ~" ~" 1~

| . . , • | 6 * t ~ 4 % •

6 • ". *' ¶ t7 " ~ ** *'

7 o • • • . t l l " • ' "

8 * • " * 1 9 * " : "

9 . ~ Z O , "

to T zt T

~i T ,,2 T , s "f

IO CN.

FRC T~xT-F:G. 8. Experiment VI I I . I-~'U" Nos. I -H, normal male controls (aver-

age weight 44.5 gin.). Nos. 12-23, castrated Dec. 4, 1914 (average weight 50.0 gin.). Al l rats inocnlated Dec. I8, I914, in the right axilla with o.o2 ~ . by the needle method.

Experiment VIII.--(Text-fig. 8.) Among the controls 8 (ioo per cent) were positive at the fourth charting, and none negative, while among the castrated rats 9 (ioo per cent) were positive and none negative. The rate of growth was. similar in both groups and none of the tumors regressed.

Jos~ S. Hilario. 169

CONTROL a 'rRAT[D t 0 17 2 4 M 5 10

t . 4 ~ 9 •

z . O ~ t o ,

t7 24 M ~

, 9

, 0 3 . 0 0 . . e O 4 - O 0 I t . • •

S . • • t $ . • .

6 . • • t 4 . . *

"/ * - - - I S * • -

I I . . - - 16 • - - - -

7 - - - - - -

| I | I I I I J | I |

10 f411.

JRS T~xT-Fm. 9. Experiment IX. ~-~B-" Nos. z-8, normal male controls (average

weight 39.o gin.). Nos. 9-17, castrated Dec. 15, z914 (average weight 72.7 gin.). All rats inoculated Dec. 2'9, I914, in the right axilla with o.o2 gin. by the needle method.

ExperiTnent IX.--(Text-fig. 9.) At the third charting, 6 controls (75 per cent) were positive and 2 negative, while of the castrates 6 (67 per cent) were positive and 3 negative. There were more large tumors in the control than in the castrated rats. Two small tumors (Nos. 7, 8) receded in the controls and 3 (Nos. 13, I5, 16) in the castrates.

170 Castration in Mice and Rats.

Summing up, it appears that with the exception of Experiments VI and VII, where it was 27 and 33 per cent, there was a difference of less than 2o per cent between the castrated animals and their controls, a disparity which does not exceed the rather wide margin which must be allowed for experimental error in the biological in- vestigation of cancer.

When the figures are set in two parallel columns (Table I) with their differences between, it is seen that in two experiments ( I I I and VI I I ) the inoculation percentages were identical, in two ( IV

TABLE I.

I 8 67

Experiment Controls. Difference. Castrated. No. i . . . . .

8 5 I I . . . . . . [ 75 I 3 62

I I I . . . . . . 73 I o } 73 I I V . . . . . . ! 94 6 I o o [

V . . . . . . 96 I5 ] 8 i [

Ex p~rioment Controls. Difference.

vI . . . . . . I ss I 27 I V I I . . . . . . ] IO0 I 33 I

v m . . . . . . I ~ o o / o I Ix . . . . . . I 7s I s !

Castrated.

82 67

IOO

67

and VI) this percentage was larger in the castrated group, and that in five of the experiments (I, II, V, VII, and IX) it was larger in the controls. Although this might appear to indicate that castra- tion decreases the receptivity for a transplanted tumor, it must be recollected that in two experiments (II and VI) the animals were probably sick, and that in one (VI I ) the number is very small, and it must be noted, also, that it is in two of these three experiments (VI and VI I ) that the greatest differences between the control and the castrated animals are found (27 and 33 per cent, respectively). In the other six experiments the difference lies between o and I8 per cent, averaging but 8 per cent.

When the single experiments are gathered into two groups, one including mice and the other rats, it is shown that among 88 cas- trated mice 68 (77.3 per cent) had tumors and 2o (22. 7 per cent) had none, while among Io 3 normal controls 87 (84.5 per cent) had growths and I6 (I5. 5 per cent) had not. Of 35 castrated rats 28 (8o per cent) developed tumors and 7 (2o per cent) proved refractory, the group of 3 ° normal controls containing 24 (8o per cent) with, and 6 (2o per cent) without tumors. The total inocu-

Josd S. Hilario. 171

lation percentage in the castrated rats and their controls is there- fore identical, and, in the case of the mice, is nearly so, since the 7 per cent decrease is so slight that it may be disregarded. Indi- vidual disparities in the various experiments thus vanish when all the groups are added together, and it is quite possible that the slight difference in favor of the controls might similarly disappear if a still larger number of experiments were available.

Nor is there any distinct evidence that, once the graft has ob- tained a foothold, the conditions governing its further development are either more or less favorable in the castrated series than in the controls. In five experiments (I, IV, V, VI, VI I I ) the growth rate is identical in the two groups; in one ( I I ) it is better in the castrates, possibly by reason of a source of error already pointed out; in three (III , VII, IX) it is more rapid in the controls. In the control series seventeen tumors receded wholly or partially, and the same number in the castrated animals.

The fact that the growth rate of transplanted tumors in normal control mice and rats is occasionally a little in excess of that in castrates, coupled with the observation that the inoculation per- centage is a trifle higher in normal animals, might seem to indicate some slight decrease in receptivity after castration. In both cases, however, the difference is so small that it may justly be referred to an outside factor such as minor nutritional disturbances or some mild and unavoidable infection at the time of operation. Cer- tainly if castration exerts a specific effect upon tumor growth, its results should be more evident.

Almagi/t described a higher degree of cachexia among the con- trols than among the castrates, a statement which the experiments now under consideration do not support, since there was no sign of such a condition in either of the two groups. This corresponds with the experience of the majori ty of observers, that cachexia does not appear until the tumor has ulcerated, and that it is then to be attributed to infection.

While it is well known from the work of Beatson, Abbe, and others, that oophorectomy in young women may check temporarily the growth of a carcinoma of the breast and even induce the dis- appearance of .small metastatic nodules in the skin, yet it is equally

172 Castration in Mice and Rats.

, i~ -_6_ : : • Q

, ~ . 6 1 , . , , , , b O i & , ~ o l o , , - * ' l 4'

, / ~ t d J 6 6 0

, ~60 d O O • , ~ 6 ~ I w W ~ T

• ~ 6 0 9 qP

~ l e • 9

~3

(O ¢N,

,, ~ l o e O l l TExT-Fro. zo. Experiment X. Mice bearing spontaneous mammary adeno-

carcinomata, and castrated at the periods shown by the arrows. The tumors were charted weekly. The removal of the ovaries exerted no influence upon the growth of the neoplasms.

Experiment X.--(Text-f ig . Io.) From these II mice, old females with spontaneous carcinoma of the mamma, the ovaries were removed at periods designated by the arrows, the tumors being charted at weekly intervals thee'e- after. No. I was castrated in order to observe the effect upon the inguinal growth; in the meantime, recurrence took place in the axilla from a tumor re- moved about 4 weeks previously. The decrease in the dimensions of the growth in No. 6 was brought about by a hemorrhage. The other mice require no com- ment; it is perfectly evident that the removal of the ovaries had no effect on" the activity of the tumor in any of the cases.

Jos~ S. Hilario. 173

well known that none of these cases are permanently benefitted. The period of influence is generally not more than six months. That no effect followed removal of the ovaries in this experiment may be due to the fact that the animals used had passed the age when these organs have any marked influence upon the blood sup- ply of the breast.

CONCLUSIONS.

Castration neither increases nor decreases the inoculation per- centage of transplantable carcinomata and sarcomata of the mouse and rat, nor does it either stimulate or retard their proliferation; it exerts not the slightest effect upon the growth of spontaneous carcinomata of the mouse.

The writer wishes to express his obligation to Professor Francis C. Wood, Director of the George Crocker Special Research Fund, for the privilege of working in the laboratory of the Fund, and to Drs. William H. Woglom and Frederick Prime, Jr., for many help- ful suggestions during the prosecution of these experiments.