~.

From the Cancer Research Laboratory, the University, Liverpool.

An experimental study of athreptie immunity in carcinoma.

By

A. J. 6elarie, .~I. D.

Since the discovery that a malignant growth could be successfully cultivated by means of transplantation from one animal to another, the experimental study of carcinoma has been possible. The fact that an animal could be inoculated with carcinoma and the further discovery that immunity to inoculations could be produced have suggested the possibility that the processes oecuring are analogous to those which are known in bacteriology.

So far as our present knowledge goes, we are justified in the belief that the growth of the tumor in the inoculated animal is due not to anything of the nature of recognisable bacteria or other organisms inoculated with the turnout, but solely to the growth and multiplication of the turnout cells themselves, which are approximately on the same developmental level as the ceils of the host. The inoculation of carcinoma is therefore not quite similar to the introduction of microorganisms into the body. In the lat ter ease the individual components of the material injected are of an entirely different type to those found in the host. Experiments have also shown that the immunity produced to carcinoma does not conform to the laws that hold good when dealing with bacterial diseases. In the first place, natural immunity to carcinoma may be met within individuals or it may affect a whole raee. Secondly, experimental imnmnity is not specific. Active immunity to a part icular tumour may be produced not only by inoculation of that tumour, but also by a tumour of a different strain and by various organs of animals of the same species;

1) The experiments detailed in this paper were carried out in the L. C. R. with the exception of those described on pages 16 to 19.

A. J. Gelar ie , An experimental study of athreptic immunity in carcinoma. 121

E h r l i c h has found immunity to carcinoma can be produced by sarcoma and vice versa. This he has called ,,Panimmunity".

How are these facts to be explained? Why is it that some animals are susceptible to carcinoma, whilst others are quite immune, and further how is it that animals can be artificially rendered immune to transplanted carcinoma not only by a tumour of the same strain, but also by normal tissues o.f the same species? Is the immunity due to the presence of antibodies? Assuming that the enormous cell proliferation of cancer, whether by altering the metabolism of the host or in some other way, causes an animal with a tmnour to produce antibodies, the antibodies cannot be said to be specific, because as already mentioned, immunity can be produced with non specific substances such as blood plasma, embryonic tissue, spleen etc., if these are taken from the same species of animal. This question will~ however, be discussed later. So far, the lormation of antibodies has not been proved 1) either by complement deviation or precipitation, and the fact that an animal which is naturally immune to transplanted carcinoma, or has been made so artificially, may develop, a spontaneous tumour, shows that the experimental facts require further explanation.

E h r l i c h put forward a theory in explanation of the facts observed in immunity to carcinoma. His theory was that if an animal is susceptible to carcinoma it means that the implanted cancer cells have found a sui- table medium in the organism in which to grow and multiply. In other words, the turnout cell is supposed to possess u great avidity for a specific food material within the host. If the implanted cells find this particular material then they live and the transplantation is succesful. On the other hand if the host does not furnish the specific food, the cancer cells die and the graft is unsuccessful. This E h r l i c h has called ,,Atrepsia".

The following experiment will serve as an illustration of athreptic immunity. If mouse tumour is transplanted into a rat it grows, according to E h r l i c h and A p o l a n t , for eig'ht to ten days, after which time it i)ecomes absorbed. If, however, before the eight days have elapsed it is transplanted from the rat to a mouse it grows vigorously. These facts are explained in this way: the transplanted cells grow for several days in the rat, because they have carried with them a certain amount of the specific food. When this becomes used up, the tumour cells being unable to procure more from the new host (the rat), they die of starvation. If, however, before this happens they are retransplanr into a mous% they

1) cp. J. O. Wakel in B a r r a t t , Complement deviation in mouse carcinoma. Proc. roy. soc. 1911. Ser. B. Vol. LXXXIV. p. 277. - - Ueber Komptement- ablenkung bet Menschencarcinom. Zeitschr. f. Krebsforsohung. 1912. Bd. 11. H. 2.

12"2 A . J . Gelarie~ An experimental study of athreptic immunity in carcinoma.

go on living, being now able to obtain in the organism of the mouse that specific material, which is necessary for their growth and development.

This type of immunity forms the subject of the present paper. The experiments and observations made were carried out on mice implanted with E h r l i c h ' s carcinoma Stature 11 and with a carcinoma received from the Cancer Research Laboratory in Glasgow. lcIistologically, these two turnouts do not show any characteristic differences. They are both of the glandular type, the Glasgow tumor exhibiting slightly more connective tissue stroma than E h r ] i c h ' s tumour, and further the epithelial cells of the parenchyma of the Glasgow turnout seem to me to be a trifle smaller thaa those of E h r l i c h ' s . Biologically it differs from E h r l i c h ' s tumour Stature 11 in that it grows very rapidly in a given time, being three to four times larger than E h r l i c h ' s tmnour. It, however, gives about the same percentages of successes in grafting as E h r l i c h ' s tumour. Experiments have also been done on pregnant mice, and observations made on heredity.

E x p e r i m e n t I. 12 nfice were implanted at the same time on the right side with E h r l i c h ' s tumonr and on the left with Glasgow turnout. As controls 8 mice were implanted on both sides with E h r l i c h ' s tumour and 6 were transplanted on both sides with Glasgow tmnour. The following data were obtained:

12 mice inoculated on the right with E h r l i c h ' s tmnour and on the left side with Glasgow turnout.

Percentage result.

6 mice showed turnouts on both sides 50 pCt., (i ,, Glasgow tumour only 50 ,, 0 .. ,. E h r l i c h ' s .. 0

C o n t r o l E x p e r i m e n t w i t h E h r l i c h ' s t u m o u r .

S mice inoculated on both sides with E h r l i c h ' s tumour.

8 ,, showed tumours on both sides 100 pe t .

C o n t r o l E x p e r i m c n t w i t h G l a s g o w t u m o u r .

6 mice inoculated on both sides with Glasgow tumour.

6 ,, showed tmnours on both sides 100 pCt.

t)f the 12 mice inoculated with both tmnours, in 6 both grew, but E h r l i c h ' s tunmur was rudimentary as compared with the Glasgow tnmonr and also smaller than the tunmurs i~ the control mice in which E h r l i c h tumour had been implanted on both sides. In 6 only the Glasgow tmnour grew, but in none was E h r l i c h ' s found aloue. The

A. J. Gelarie~ An experimental study of athreptic immunity in carcinoma. 123

avidity of the more virulent turnout to the speeifie food prevented the

growing of the less virulent tumour.

E x p e r i m e n t II. The next experiment consisted in the total ex- stirpation of Glasgow tumour from mice in which both tumours were growing, thus leaving E h r l i e h ' s turnout alone present. This still grew slowly as compared with the control experiments with E h r l i e h ' s tumour. The removal of the tunmur was carried out with every aseptic precaution. An incision was made round the turnout and the skin covering the turnout was removed with the latter. The wound was washed out with a 5 pCt. solution of chloride of zinc to prevent recurrence and only these mice in which the wound had completely healed without any recurrence of that growth, were used for experiment. The mice lived 30 to 48 days after the operation.

In 6 animals with two tumours the Glasgow tumour was removed,

E h r l i e h ' s turnout being left; In all 6 animals E h r l i e h ' s tumour became very large ten days later. Here again we see, that after the removal of the more virulent turnout,

the less' virulent grows more rapidily and shows considerable increase in size.

E x p e r i m e n t III. The Glasgow tmnour was implanted on six mice on which E h r l i e h ' s tumour was already growing. In 4 of those the Glasgow tmnour grew (66,6 pCt.).

E x p e r i m e n t IV. 8 mice on which Glasgow turnout was already growing, were inoculated with E h r l i e h ' s turnout. At the end of fore'teen days no E h r l i e h ' s turnouts were of reeognisable size and a second im- plantation of E h r l i e h ' s tumour was made on all the mice. At the end of twenty three days there was still no sign of any growth of E h r l i e h ' s turnout and all the mice were once more regrafted. On this occasion 2 mice did develop E h r l i c h ' s tumour. These results may he smnmarised as follows :

Percentage result.

8 mice with Glasgow turnout were grafted with E h r l i c h ' s tumour.

In 0 ,, did E h r l i e h ' s turnout grow . . . . . . . . . . 0 pCt., ,, 0 ,, . . . . . . ,, on 2ndgraf t ing 14 days later 0 ,, ,, 2 . . . . . , ,, ,, ,, 3rd ,, 23 ,, ,, 25 ,,

TILe results of experiments III and IV are striking: Transplantation of a more virulent turnout upon mice ah'eady bearing a less virulent tmnour is successful in 66,6 pCt. of the eases, whilst if a less virulent tumour is grafted on mice already bearing a more virulent tumor, there is only

194 A..I. Gelar i% An experimental study of athreptic immunity in carcinoma.

25 pCt. of successes even after three separate implantations; no successful graftings at all oeeured after either first or second occasions. Experiment IV will be discussed further in connection with experiment VIII. Here I would only like to emphasise the necessity of inoculating at least three times in the course of ten weeks before regarding a mouse as immune. It very often happens that mice which do not develop a tumour after the second grafting, nevertheless do so after the third, l have observed three mice, two of which were inoculated four times in the course of three months without success and yet on the fifth occasion a turnout developed which, however, grew very slowly. A third mouse l had under observation for eight months. During that period it was grafted seven times, but at no time was a tnnmur seen. 14 days after the 8th. grafting a small tumour was seen. This remarkable resistance will be referred to again.

E x p e r i m e n t V. 10 mice were grafted on one side with E h r l i e h ' s turnout and three days later a second implantation was made on the opposite side, also with E h r l i e h ' s tumonr. The.results are given below:

Percentage result.

10 mice grafted twice on opposite sides with E h r l i e l F s turnout, three days interval between the two grafts.

4 ,, showed tumours on both sides . . . . . . 40 pCt. 3 ,, .. ,, only on side of first grafting . 30 ,, 1 ,, . . . . second ,. 10 ..

2 ,, remained resistant to subsequent inoculations.

The control animals grafted with the stone turnout at the same time and in the usual way, i. e. on one side only, yeilded 75 pCt. of successes on the occasion of first grafting of the experimental mice, whilst on the second occasion the graft was successful in 96 pCt. of the control mice. In the experimental mice SO pCt. grew tmnours but in only a half of those were the turnouts present on both sides. In the other half, tumours were present on only one side, of these 3 or 30 pCt. represented grafts made on the first date and 1 or 10pCt. showed grafts made on second date. namely three days after the first.

E x p e r i m e n t VI was similar to the last experiment except that the second inoculation was made 12 days instead of 3 days after the first.

The details of experiment are as follows: Percentage result.

9 mice grafted twice on opposite sides with E h r l i e h ' s tnmour, a period of twelve days intervening

between the 2 grafts. 2 .. showed tmnours on both sides . . . . . . 22,2 pCt.,

A. J. Gelar le~ An experimental study of athreptic immunity in carcinoma. 125

Percentage result.

3 mice showed tumours only on side first grafted 33,3 pCt. 1 animal showed tumour only on side second grafted 11,1. ,, 3 ,, remained ' res is tant to subsequent inoculations.

The control animals grafted at the time of first implantation gave 75 pCt. of successes and those done at the time of the second gave 67 pCt. In the mic'e in this experiment~ 66,6 pCt. in all showed turnouts. Of these only one third or 22,2pCt. had turnouts on both sides, whilst in the remaining two thirds having tumours on one side~ only 33,3pCt. were tumours representing grafts made on the first occasion and only 11,1 pCt. represented grafts made 12 days later.

From the last two experiments (V and VI) we see that if a second implantation is made 3 days after the first, 40 pCt. of the animals show tumours on both sides, whilst if the second transplantation is not made until 12 days after the first, only 22,2 pCt. show tumours on both sides, the percentage of mice having tumonrs on one side only being practical ly the same in both experiments and the proportion of successes from first and secor/d implantations also remains the same.

E x p e r i m e n t VII. 8 mice which had been successfully grafted twenty three days before, were regrafted. At the end of 17 days none of these mice had developed a second tltmour. An attempt was therefore made to produce an additional tumour on the opposite side. This also was unsuccessful.

These results may be summarised as follows:

Percentage result. 8 animals bearing E h r l i c h ' s tumour of 23 days growth

inoculated on opposite side. 0 animals showed tumours on opposite side at the end

of 17 days . . . . . . . . . . . . . . . . 0pCt. , Alle mice grafted again. 0 animals showed tumours at the end of 20 days 0 ,,

The various attempts to regraft had no noticeable effect on the growth of the original tumour which increased at the same rate as tumours in control mice1).

In this experiment we see that by stil l further delaying the second implantation the percentage of mice which can be implanted succesfully is nil. According to E h r l i c h ' s theory the facts brought out in the last

1) By control mice are meant (unless otherwise stated) normal mice grafted on one side with portions of the turnout employed for experiment; the control mice were implanted at the same time as the experimental mice.

1~6 A . J . G o l a r i e , An experimental study of athreptic immunity in carcinoma.

3 experiments can be accounted for by assuming that as soon as a tumonr is implantod~ it begins to attach to itself by means of nutriceptors the specific food material necessary to its growth. In this way at the end of 23 days there is none of this food material still remaining available for the needs of a second implantation.

E x p e r i m e n t VIII. Mice with large turnouts broken and par t ly degenerated in the centre which had been growing for 68 to 74 days were now grafted for a second time on the opposite side. This second grafting was succesful in two cases or 25pCt. Thus:

Percentage result. 8 animals bearing E h r l i c h ' s tumour of 68 to 74 days

growth inoculated on opposite side. 2 animals showed tumours on both sides . . . . . 25 pCt.

The remaining 6 animals died without a second turnout. This result at first sight seems to contradict that of experim6nt VII.

A similar result was got in experiment IV; after the tumour had become very large a second implantation was successful in 25 pCt. of the m~ce used.

Various explanations of these results suggest themselves. On Eh r l i ch ' s theory, they might be exph~ined on the assumption that the contin/led demand of the growing tumour for the specific m~tterial causes the host to produce more of this substance or even an excess, in this way possibly rendering the animal even more susceptible to subsequent inoculations with mouse tmnour. Or is it that a large tumour has reached the limit of growth and can assimilate no more food material and thus there is some available for the use of the second implant? It might be thought that the continued growth of the tumour would produce a substance possibly of an enzyme nature inimical to the growth of the turnout either directly or by inter- fering with the production of the specific food material. The lat ter can be excluded at least in some cases, as in at ]east 25pCt. of the cases a second implantat ion can be successfully made. Still another possibility suggests itself. Could the enornmus cell proliferation of the big turnout not have altered the metabolism of the animal in such a way, that it caused a hypersusceptibili ty of an anaphylaetic nature?

With a view to obtaining some information as to the value of these theories, further experiments were made:

E x p e r i m e n t IX. Mice with large tumours broken on the surface and of 60 to 70 days growth were employed. The tumours were removed with the superjaeent skin and the wound washed out with a 5 pCt. solution of zinc chloride to prevent recurrence. Many mice did not stand this operation and died either immediately or 10 to 14 days later. Only mice

A. J. G e l a r i o , An experimental study of athreptic immunity in carcinoma. 127

in which the wound healed 10 to ]4 days after the operation without suppuration and which lived a long time (120 to 200 days,) afterwards were used in these experiments. The regraft was made on the opposite side of the animal so that a local recurrence of the first transplantation could be excluded. Details of experiments are the following:

Percentage result.

10 animals in which large tumours had been success- fully removed were employed.

5 ,~ were grafted 14 to 18 days later. 3 ,, developed tumours . . . . . . . . . 60 pCt., 2 ,, in which tumours did not develop were

regrafted 28 days later. 0 ,, developed turnouts . . . . . . . . . 0 , 2 ,, again regrafted 13 days later. 1 ,, developed a tumour . . . . . . . . 50 ,,

The 5th. animal died without a tumour.

5 animals grafted 24 to 30 days after operation. 5 ,, developed tumours . . . . . . . . . 100 ,

In this experiment 10 mice with large tumours of the type described in Experiment VIII and IX were employed, the turnouts being removed. 14 to 1~ days later 5 of these animals were regrafted. 3 of these were successful. 2~ days after the last grafting the 2 mice which appeared immune were again grafted, but without any result. 13 days la ter im- plantation was again attempted and in one of the mice a tumour developed. 24 to 30 days after the operation the 5 remaining mice were regrafted and all 5 developed tumours.

E x p e r i m e n t X was of a similar nature but was carried further. Large turnouts of the type already described (in Experiment VIII and IX) of 58 to 63 days growth were removed from 6 animals. 20 days after the operation all 6 mice were regrafted on the opposite side and in all of these turnouts grew. These tumours were in turn removed 23 days after the last grafting and 21 days after this operation the mice were again grafted on the back. Turnouts once more developed in al l 6 animals and these were in turn removed 18 to 21 days after the implantation. 18 days after the removal of the third series of tumours all 6 animals were again regrafted on the neck. Turnouts appeared. These tumours were also removed 14 days after implantat ion and 22 days later a fresh implantation was made on the chest. Once more each mouse developed a tumour. Thus:

128 A . J . Gelar ie~ An experimental study of athroptic immunity in carcinoma.

Percentage result.

t; animals with large tumours were selected for ex- periment.

The tunmurs were successfully removed. 6 ,, grafted 20 days later ; 6 ,, developed ~umours . . . . . . . . . 6 tumours 23 days later removed and regrafted

21 days later: 6 developed tumours . . . . . . . . . 100 ,. (~ ,, tumom's 18 to 21 days later removed and

regrafted 18 days later; 6 ,, developed tumours . . . . . . . . . 100 6 tunmurs 14 days later removed and regrafted

22 days later; ~ ,, developed tmnours . . . . . . . . . 100

100 pCt.,

The results of the last two experiments (IX and X) would seem to indicate theft mice which have had a tumom; instead of being immune are hypersensitive to subsequent inoculations. In the 2nd part of experiment IX the percentage of successful graftings is 100, which is nmch higher than that obtained when normal animals are grafted for the first time. It is possible that this hypersensibility is of an anaphylaetic nature and that the ,,taking" of a second implant is analogous to the anaphylaetic fit which occurs when the reacting dose is given to a sensibilised animal? In this case the first implantation would correspond to the sensibilising dose. Experiment X, however, seems to exclude the possibili ty of this theory being tenable, as nothing corresponding to an antianaphylactie reaction was obtained. On the contrary the mice continued to respond to each successive inoculation. These results, however, as pointed out in the discussion on experiments IV and VIII are quite in accord with E h r l i c h ' s theory of athreptic immunity.

The next step was to see whether mice with well developed but not degenerated tmnours of 30 to 40 days growth would behave in the same way, as the mice with large tumours and degenerated in the centre of 60 to 70 days duration.

E x p e r i m e n t X[. The small tmnours were removed in the same maturer as in the previous experiments and as before only those mice were used in which the wound healed by first intention and which did not show any sign of recurrence. Tile results may be tabulated thus:

A. J. Gelarie~ An experimental study of athreptic immunity in carcinoma. 129

Percentage result.

10 tumours of 30 to 4=0 days duration completely removed. 4 of above mice were grafted on 23rd or 29th day after

operation; 0 mice developed tmnours . . . . . . . . . . OpCt., 4 of above mice were reg~'afted on 47 th day after operation; 1 mouse developed a turnout . . . . . . . . . ~ ,, 3 negative mice were grafted on 62 nd day after operation; 1 mouse developed a turnout . . . . . . . . . 33,3 ,, 3 mice grafted on 32nd or 44:th day after operation; 3 ,, showed turnouts . . . . . . . . . . . 100 ,, 3 remaining mice grafted on 50th or 75th day after

operation; 3 mice developed tumours . . . . . . . . . . 100 ,,

In Experiment IX we saw that mice with large turnouts of 60 to 70 days duration were successfully reinoeulated in 60 pCt. of the eases~ 14 to 18 days after the removal of the original tumour at a still later period the percentage of successes became 100. On the other hand in the experiments just recorded (Experiment XI) in which mice with smaller tmnom's of only 30 to 40 days growth were used, no positive results were obtained 23 to 29 days after operation. But if the first attempt to inoculate a second time was deferred until 32 days or more after the removal of the original turnout, the second tumour grew in 100 pCt. of the mice and if we include the 4 minus from the 1st part of the ex- periments in which more than one attempt to grow a second tumour was made, we get 8O pCt. of successes.

Considering Experiments IX, X and XI, together the percentage of successes when grafting a 2nd mmour after the total exstirpation of the original tumour is 88,5pCt., whilst in the control ~) animals of these experiments the average percentage of successful grafting is 54.

The results of these experiments seem to exclude the formation of antibodies unless these are exceedingly evanescent. If antibodies against cancer cells were formed by the organism, a subsequent inoculation ought to be withouk result. The objection might be raised that in these experi- ments, the formation of antibodies is interfered with, the operation of removing the original tunmur being so severe that the orgmfism is weakened and its nutrition impaired to such an extent that antibodies cannot be formed. In answer to such objections, I am able to say that the mice

1) See footnote p. 125. Zeitsehrift ffir Y, rebsforsehung. ]4, Bd. 1. Heft.

J30 k. J, Gelarie~ An experimental study of athreptie immunity in carcinoma.

used in Exp. IX, X and XI, were only some from many others which have been operated on in the same manner. Such animals have lived from 28 to 221 days after the operation and have been in exceptionally good eondition. Some of them have e~'en had litters, to which reference will be made later.

It will be noticed in ~he above experiments t.ha~ some m~ee were ultimately successfully implanted with a tmnour after having been previously inoculated several times without result. To this fact I atlaeh the greates* importance as bearing" on the biological behaviour of carcinoma. Further reference to this eircmnstanee will be made later~ in eonnection with experi- ments about to be described.

E x p e r i m e n t XII. A htrther series or experiments was undertaken, to assertain whether after the total exstirpation of the external turnout metastases would be fotmd in internal organs, especially in the lungs, Tilt idea behind tiffs series of experiments is to give those cancer cells which may have been transported by the blood stream to internal organs an opportmfity of proliferating.

By renmving the external turnout the special f,)od stuff may be set free and in that way provide with the proper nutritive material the few cancer cells which may be lodging in internal m'gaus. 20 mice were employed for these experiments and only those were used in which the wound healed in ten to fourteen days without suppuration and which did not show any reeurrenee after the removal of the tmnor. The results obtained are exhibited below:

No of mice from which tumours were removed

8

i

12 J

Age of turnout

20 to 38 days

45 to 6"2 ,.

No of days after renmva[ of tmnours,

at which death oeeured

5 mice 32 io 48 days '~ 60 1 130 ,.

4 ,, 20 to 33 $ ,. 46 to 71 ,,

Metastases.

Absent.

} AbsenL

Turnouts of '20 to 38 days duration were completely removed from 8 mice, 5 died from tlle ; /2nd to -t,q th day. The lungs did not show any metastases on microscopical examination. 2 died at the end of the 60 th and one at the end of the 130th day after operation and still no metastases were found. Tmnours of 45 to 62 days duration were comple- tely removed from 12 mice. 4 of these died at the end of 20 or 33 days after operation - - no me~asta.ses were found. S died at the end of 46 to 71 days and again no metastases were seen.

3.. J. Gelarie~ An experimental study of athreptie immunity in carcinoma. 131

Special attention was paid to the lungs in examining for metastases in this experiment. Serial sections from different parts of each lung were examined. Not in one single instance could metastases be detected with certainty. Groups of 3 to 5 cells were seen, which looked very much like carcinoma cells, but I had serious doubts if they real ly were such, especially after Dr. B a r r a t t called my attention to similar groups of 3 to 5 cells found in the lung of rabbit 's which had died from fibrinaemia and also, in the lungs of apparent ly normal animals. I am therefore of opinion, that those authors are mistaken who have described single cells or small groups of cells as metastases in the lungs of mice suffering from trans- planted carcinoma. It appears to me that cancer cells can only be identi- fied with certainty when such cells are found definitely arranged in columns.

It is impossible to diagnose a group of 3 to 5 cells with absolute certainty. Metastases in the form of columns of cells were not found. If isolated cells had been present in the lungs of the mice at the time the original tmnours were removed, they ought during the time that elapsed before the death of the animals to have formed reeognisable tumours.

E x p e r i m e n t s on p r e g n a n t mice .

Some of the previous experiments have shown that the presence of a tumour prevents the growth of a second transplantation made subsequently, and it was pointed out that the only satisfactory explanation of this phe- nomenon is to be found in the hypothesis that it is dependent upon athreptic influences. With a view to assertaining whether pregnancy would likewise inhibit the growth of a tumour~ the following experiments were undertaken: As a rule pregnant animals are resistant to transplantations, and if a pregnant mouse already posseses a tumour, it either remains stationary or grows very slowly, so that the tumour is only about one third the size of tumours of control 1) animals.

In order to observe the effect of interrupting pregnancy in mice with turnouts, it was decided to remove the uterus and foetuses in such mice by abdominal section.

~Iice which had been grafted at the same time and with the same tumour were used as controls.

Mice bear the abdominal operation very well and are quite l ively afterwards.

Observations were made on 7 mice. During the first week after the operation the tumours of three animals very slightly increased in size, whilst in the remaining -I animals they remained stationary.

In the 2nd. week the tumours of all 7 mice began to grow very

1) See footnote page 125.

9*

132 A . J . G e l a r i % An experimental study of athreptic immunity in carcinoma.

rapidi ly and at the end of tile 3rd. or ~lth. week after operation reached a very large size, being nearly twice as large as the tumours of the control mice.

Do athreptie influences account for this result'? It seems as if they might. As long as the foetuses were growing, the turnouts remain statio- na<v or ~'row very slowly because the struggle betweeff the nutrieeptors of the turnout and those of the foetuses for the specific food has ended in a victory for the foetuses, which grow and develop mor~ rapidily, having a greater avidity ( E h r l i e h ' s ereptive power) than the tumor for the specific food material of the host. When the foetuses are removed, the supply of tlLis special food material is relat ively greater and the tunloltr increases h~ size.

The following experiments are very interesting: They were carried out on pregnant mice otherwise perfectly normal i. e. in mice havhlg no turnout. 5 to 20 days after the urea'us and foetuses had been removed, the mice were grafted. Observations were made on 40 mice. '2 grafted 5 and 12 days after the operation died at the end of from 5 to 7 days after grafting and so were of no avail.

In 7 mice grafted 12 to 20 days after operation the tumour was markedly affected, having reached only one third the size of tile tumours of the control animals, whereas 2 grafted 10 and 15 days after the ope- ration remained immune to 2 subsequent inoculations. The tumours of the renmining three animals grafted 17 and 20 days after operation grew at about the same rate as in the control nfiee.

The control animals gave about the same percentage of successful grafting as the mice used in this experiment.

Are these phenomena analogous to those recorded in experiments IX, X and XI, in which the suhsequent inoculation, made 18 to 20 days after the total exstirpation of the tumoun gave less successful results than when made 50 to 70 days after operation? I prove or to disprove this point, as I was at an early date after tile removal of the reals might not be lost. The fact that it

have not been able either to very anxious to graft the miee

foetuses, in order that the ani- is not always easy to diagnose

pre~.naney with certainty and fm'ther that the mice only became pregnant at certain seasons, put difficulties i~1 the way of carrying out this work on a large scale.

However, we see how immunity, in this ease part ial immunity (as shown by slow growth of the tmnour during pregnancy, or by inoculation after the interruption of pregnancy) may be closely related to hypersensi- bil i ty (as shown by large size of the lumour after the interruption of pregnancy) - - both immunity and hypersensibility, as seen, in this special ease, can be paral led by examples from physiology, were we see that

A. J. G e l a r i e , An experimental study of athreptio immunity in carcinoma. 133

living matter can be influnced in two opposed directions by chemical sub- stances, by alterations in temperature, and by electricity1).

In one direction we have stimulation of the growth and the function of the cells and in the other direction diminution and ult imately para- lysis -~ of growth and function. For example nux vomiea, in small quanti- ties increases .function but in large doses causes paralysis. Chloroform and many other substances act in the same way. This apparent contra- diction may be explained thus - - the increased vitality induced by the small dose demands an increased amount of oxygen and when this becomes insufficient, the result is suffocation.

The rapid growth of the tmnour after the interruption of pregnancy may be satisfactorily explained in an analogous way. I n the two experiments on pregnant mice we get apparently contradictory results. On the one hand we have tumours when already present at the time of the inter- ruption of pregnancy growing with abnormal rapidi ty after the operation. On the other hand, we have pregnancy interrupted in mice without turnouts and tumours subsequently implanted with a result that tl~ey grow ab- normally slowly, being only one third the size of tumours implanted at: the same time in normal mice. The interruption of pregnancy in the first ease would seem to give results analogous to those obtained in experiment III and IV, where after the removal of a well de~'eloped virulent tumour a less virulent one already present grew very rapidily. On the other hand," the results in the second experiment on pregnant mice seem of a different character and appear to require further explanation.

Taken by itself, the results in this last experiment might be accounted for ill either of the following ways; - - on the one hand the pregnancy may ha~-e used up al l the available specific food material, whilst on tlle other hand it may have left behind an excess of this special material which inhibits or entirely paralyses those cells which produce this part i - cular substance.

The key of the explanation~ however, nray lie in the fact that the turnout which is already present at the time of the interruption of pregnancy may be assumed to have its nutriceptors already anchored to the receptors of this specific material and so be able to take full advantage of this specific food when that is set at liberty. On the other hand where the

1) cp. Nernst~ Ghttingec Nachr. Mathem.-physik. Klasse. 1S99. S. 104. - - Nernst und B a r r a t t , Ueber die elektrisehe Nervenreizung durch Weehselstr6me. Zeitschr. f. Elektroehemie. 1904. Bd. 35. S. 664.

2) J. O. W a k e l i n B a r r a t t , The theory of nerve exoitation. Brit. reed. journ. 1912. Sept. 28th. - - The lethal concentration of aeide and bases in respect of Paramaecium aurelia. Proc. roy. soe. 1904. Voh LXXIV. p. 100.

134 A . J . G e l a r i e , An experimental study of athrel~tio immunity in carcinoma.

tumour is implanted after the interruption of preg~mncy, the overproduction of the specific nutritive material left behind, has caused an alteration of the metabolism of the host and in that way has caused the turnout cells to al ter their nutriceptors for another kind of specifc food. This new specific food, which the host may or may not produce in sufficient quantily, has caused the tumour cells to grow slowly or not to grow at all.

In this connection it seems to me that in- the removal of uterus and foetuses the ovaries in all probabil i ty are also removed or their blood supply is interfered with to such an extent that they are rendered functionless. It may be that these organs have something to do with the formation of that ,,spedfic" food material. Reliable data concerning the removal of ovaries in the human subject would therefore be of considerable interest.

With reference to what has been said concerning the possible alterations in the production of the specific food material and what has been stated in regard to modifications of the nutriceptors, the following experiments may be of interest:

9 mice which had been grafted three to four times with E h r l i c h ' s tumour without result, were then grafted with Glasgow tumour. This was also without result~ and a 2nd inoculation with Glasgow turnout 21 days ]ater was likewise negative. On grafting again with E h r l i c h ' s tmnour, 5 out of the 9 previously immune mice developed E h r l i c h ' s tumour, which however grew very slowly scarcely reaching one third of the size o~ turnouts of control mice which had been grafted at the same time with portions of the same tumour.

Here we see, that repeated inoculations with at first a less virulent turnout and ult imately with a more virulent tumour does not necessarily render the animal imnmne to a subsequent inoculation even with a less virulent turnout.

It is true that mice can be made immune a r t i f i c ia l ly*)and retro- gression and complete disappearance of t ransplanted mouse tmnour may be produced either by injection of the tulnour pulp itself or of the pulped organs of animals of the same species as well as of pulped nmuse foetus.

I have already alluded to the fact that imnmnity and hypersuscepti- bil i ty are closely reluted to each other. It seems to n:e that the artificial production of immunity is in reality due to an artificial alteration of the metabolism of the animal and fnrther that the metabolism can be altered in two directions, i. e. the animal can either be rendered immune or hypersensitive, in other words, the very substances which produce

1) cp. J. O. Wake l in Barra t t~ The experimental production of immunity to carcinoma in animals. Liverpool Medico-chirurgical Journal. January 1913.

A. J. G e l a r i e , An experimental study of athreptie immunity in carcinoma. 185

resistance to tumour growth, may also produce susceptibility. Coming back to the alteration of the nutriceptors of the turnout, we might assume, that by repeated inoculation with a Glasgow turnout, the meta- bolism of the mouse has been so altered, that the organism now produces substances which are suitable for the growth of E h r l i c h ' s tumour, or it might be that the altered metabolism has caused the transplanted tumour cells to develop nntriceptors for another kind of substances. The slow growtti of the tumours could then be accounted for as evidence of a gradual adoptation to a new environment.

Further we see that immunity to one part icular strain of tumour does not necessarily render the animal immune to another strain of turnout.

16 mice which were resistant to E h r l i c h ~ s carcinoma on repeated (3 to 5 times) implantations were grafted with ~ (~las~ow tumour, g mice developed a tumour on first grafting, 2 on second and the remaining 10 remained immune to 2 subsequent graftings, that is to say 37,5 pCt. of mice which were immune to one strain of carcinoma were susceptible to another, more virulent strain. 8 of the above 10 mice (2 died) on which a Glasgow tumour did not grow after repeated graftings, sub- sequently also remained resistant to repeated (twice) inoculations witli E h r l i c h ' s tumour.

E x p e r i m e n t s on H e r e d i t y .

E x p e r i m e n t I. S young mice were chosen which formed the pro- geny of mothers in which the turnouts (Ehr l i ch~s ) had been completely removed after each of four successive graftings.

Final ly the fifth implantation also resulted in the development of a tumour. The male parent mouse was immune to three inoculations with E h r l i c h ' s tunmur. The 8 young were grafted with E h r l i e h ' s tumour, 2 died 6 to S days after transplantation, so that a palpable turnout could not be in these cases felt, but microscopical examination showed that the tmnour was growing; the other 6 mice developed good tumours.

E x p e r i m e n t If. 16 young mice were used from 4 mothers which were immune to E h r l i c h ' s carcinoma, implantat ion having been attempted 4 to 5 times unsuccessfully. The fathers of these mice had tunmurs. Out of the 16 mice, 2 developed tumours on the first grafting, 3 on the second, while the remaining 11 young remained resistant to repeated (3 times) inoculations with the same tumour strain as that growing on the father. The fourth inoculation was made with Glasgow tumour and three mice developed tuumurs. The remaining 8 died 5 to 12 days after the inoculation with Glasgow turnout, and it was impossible to say whether ti~e tumour was growing or not.

Tllese facts seem to me to be of the greatest interest. In the second

136 A.J. Gelarie, An experimental study of athreptic immunity in carcinoma.

experiment an inmmne mother passes on at least partial imnmnity to 6~,,q_ pet. of" its offspring, whilst in the first experiment the immune father is apparently devoid of this power. Again in the first experiment the susceptible mother passes on susceptibility to all or 100 pet. of the off- spring, whilst as in the second experiment when the father is susceptible only 31,2pCt. of the off,spring are susceptible0.

Again, 6.q,,~ pCt. of the offspring which were immue to E h r l i e h ' s turnout were susceptible to Glasgow tumonr. I do not venture to put forward a theory in explanation of this fact. The turnouts vary both bio- logically and morphologically. The susceptibility to Glasgow tnmour may be due to its higher virulence: tt~e connective t issne strolna which, as l have ah'eady mentioned, is slightly more abmxdat than in E h r l i c h ' s tmnonr, may account for its greater virulence and avidity to the nutritive material or the host. In this connection it is interesting to remember that R i b b e r t and others attach great importance to the part the stroma plays in the genesis of carcinoma.

Returning to the question of altering the metabolism of the mfimal in an earlier part of the paper, some experiments have been mentioned which give som(~ indication that the metabolism may be afl'ected in two opposite w a y s - i. e. the animal may be rendered immune or hyper- sensitive to carcinoma. [ shmlld like to add some fnrther experiments bearing on lhe same point:

A mouse, which had been grafted once with E h r l i c h ' s tmuour without result was grafted 24 days later with a turnout from another mouse injected several weeks before with pulped mouse tnmour which had been exposed to the temperature of solid CO~ for 10 mimltes. 10 days later, the Ih'st implantation began to grow vigorously whilst the second illoculation made with turnout from a treated mouse, did not grow at all.

7 mice were injected snbcutaneouslv with an emulsion of equal parts el Glasgow and E h r l i c h ' s turnout. As a control 5 mice were similarly injected with an enmlsiou containing E h r l i c h ' s tunmur only and 3 with an emulsion of Glasgow tumom' only. All the inoculations were made at the same time and with the same tumours. Of the 7 mice which were injected with the emulsion of both tmnours, 3 mice (4:3,s pCt.) developed tumours and these grew very slowly indeed. Of the 5 controls injected with E h r l i c h ' s turnout 2 (4(1 pCt.) g'ave positive results; whilst of the controls with Glasgow tmnour one out uf three (,33,3 pCt.) developed a

1) I am very sorry that 1 omitted to make a note of the sex of tile offspring, which might be of importance. Similar observations on the human subject would be of interest and might possibly lead to the discovery of an underlying principle in heredity.

A. J. G e l a r i e , An experimental study of athreptic immunity in carcinoma. 137

tumour. All the tumours in the control mice were large and well deve- loped. This difference in the size of the tumours in the control and experimental mice is interesting and contrary to the facts observed in Experiments III and 1V, where E h r l i e h ' s and Glasgow tumour were implanted on opposite sides of the same mouse.

Does the explanation lie in the fact that 1low the struggle between the nutrieeptors of the different turnout strains is taking place and at close quarters has caused the animal to develop a par t ia l resistance to both tumours; or have the nutrieeptors of each tumour been altered so that now they demand a different nutritive material which the host only produces in small quantities? The results in this experiment cannot have been affected by the disintegration, as the material injected into the control and experimental mice, had been treated in exactly the same way1).

This experiment taken in conjunction with those previously mentioned seems to show that under different conditions, substances wh%h at one time produce resistance to turnout growth may at another time produce increased susceptibility. The following experiments may also be of interest:

5 mice with small tumours were injected intraperi toneaiiy with pulped mouse foetus: placenta and amniotie fluid previously heated for half an hour to 56 o C. The tumours on the mice very rapidly increased in size, being twice as large as tumours in mice which had been grafted at the same time and with the same turnout, but hat not subsequently had an intraperitoneal injection of pulped foetuses.

Similar results were first recorded by F l e x n e r and J o b l i n g (Proe. of soe. f. exp. biol. and med. 1907): According to these observers, rats injected intraperi toneally with heated tumour emulsions were made more susceptible to turnout transplantation and if repeated injections were made at intervals of ten days, the susceptibili ty was fm'ther increased. Even animals in which turnouts remained stationary or became absorbed, could be made susceptible t-o a subsequen~ inoculation.

When we review the facts brought oat by the preeeeding experiments, it seems as if the formation of antibodies to implanted mouse carcinoma is unlikely, if indeed it cannot bee excluded altogether.

We do not know, whether the infective process in carcinoma depends upon a special virus or a parasite. ,~s far as our knowledge goes, arti- ficial infection is produced by a transplantation of tissue. - - The cause of the formation of cancer i. e. the transformation of the normal cell into

1) The disintegration was carried out by an apparatus designed by J. O. Wakel in B a r r a t t and consisted of 2 metal syringes separated by a finely per- forated metat disc. The material is pressed from one syringe to the other and in this way is squeezed through the perforated plate.

138 A . J . Gelar ie~ An experimental study of athreptic immunity in carcinoma.

a cancerous one does not seem entirely dependent upon one single etio- logical factor, but upon many, chemical or mechanical, thermal or even actinic influences, such as X rays~), Finsen light, while at times a para- site 2) or its toxiu may be tl~e exciting cause.

E h r l i e h ' s theory of atl~repsia seems at the present time to form the only satisfactory explanation of the phenomena observed in all the various and puzzling aspects of the cancer problem and appears to be valid not only for transplanted carcinoma, but even for spontaneous tumours. That there is a certain predisposition in the organism cannot be denied. This is i l lustrated by the fact that certain organs are highly susceptible to carcinoma, whilst others are never or at any rate comparatively rarely attacked. The distribxltion by the circulatory system in man does not give a satisfactory explanation of why carcinoma cells should form meta- stases onIy or more frequently in certain organs than in others. Chemical substances must play an important part in this connection. As shown by F r e u n d , K a m i n e r and N e u b e r g the blood of an organism affected with carcinoma does not destroy cat'cinoma cells. The spleen and the heart seldom become affected, yet if cancer cells are present in the blood of the animal, role would think thev must pass through those organs also. Other organs also appear to be unsuitable for the growth of carcinoma: for example cartilage. Carcinoma of the larynx never breaks through the cartilago. Again. the spread of carcinoma from one organ to neighbouring organs 1nay show peculiarities. Carcilmma o[ the oesophagus for example very seldom affects the aorta, whilst the lungs as a rule are infected, and metastases of the glands of the axilla or the retroperitoneal glands as a rule leave the arteries untouched.

From the above, it seems clear that chemical substances play a very important part i~l the development of new growth, and all the experiments recorded are in favour of the theory of athreptic immmlity, i. e. that a tumour originates only from such cells as possesses receptors with a greater avidity than the receptors of the body cells for specific ehemical substances.

Before concluding this paper I have much pleasure in expressing my great indebtness to the Director of this Laboratory, Dr. J. O. W a k e l i n B a r r a t t , for his kind ~ttention and constant assistance, iikew~se t.o Dr. M a r y M. G a r d n e r for the services she has rendered in writing out this paper.

1) ep. J. O. W a k e l i n Barra t t~ The action ofSharlaeh g upon X rayed skin. The Lancet. February 157 1913.

2) B o r r e l , Annales de l'lnstitute Pasteur. 1905. No. 3 et 1910. No. 10.