[gf^I^JMI^IM^IMIM[^ THE LIBRARIES COLUMBIA UNIVERSITY ^^ HEALTH SCIENCES LIBRARY J i g|ui][ runiruilffugfr5TJIrm3f?ugf^ Digitized by the Internet Arciiive in 2010 with funding from Open Knowledge Commons http://www.archive.org/details/studyofexperimenOOwogl STUDIES IN CANCER AND ALLIED SUBJECTS THE STUDY OF EXPERIMENTAL CANCER A REVIEW STUDIES IN CANCER AND ALLIED SUBJECTS Conducted under the George Crocker Special Research Fund at Columbia University VOL.1. T H E S T UD Y O F E XPERI M E N T AL CANCER. A Review. By William H. Woglom, M.D. Illustrated with many plates. In two bindings, Quarto, boards, or 8vo, cloth, pp. xi -\- 288. Price, $5.00 net. VOL. II. PATHOLOGY. Illustrated with many plates and charts. Quarto, boards, pp. vi + 267. Price, $5.00 net. VOL. III. FROM THE DEPARTMENTS OF ZOOLOGY, SURGERY, CLINICAL PATHOLOGY, AND BIO- LOGICAL CHEMISTRY. Illustrated with many plates. Quarto, boards, pp. ix + joTDtians, who were ac- quainted with cancer as early as 1500 B.C., although, according to Joachim,^ they included the most various swellings as part and parcel of the disease. But the lesion which we now recognize as cancer was probably not unknown to them, for they described an ulcerating disease and treated it with an arsenic salve. ^, It was certainly familiar, at any rate, to the Egyptians of a somewhat later period, for manu- scripts prepared about 800 B.C. contain descriptions of cancer of the breast„^ which leave no doubt that the disease was recognized at that time ; even before the dates just quoted, however, cancer was being extirpated in India, and a salve applied to the wound in the hope of preventing recurrence of the disorder.^ With cancer of the breast Hippocrates was fairly well acquainted and he recognized the occurrence of malignant disease in certain of the internal organs as well. He believed it to be due, in common with all other pathological conditions, to a deficiency or an excess of bile, blood, or mucus. Before the day of the Roman physician Celsus, the term carcinoma included the most bizarre collection of swellings. Celsus,*" however, distinguished cancer from carcinoma, including under the former head- ing many lesions which are now recognized as simply inflammatory. 1 Die Lehre von der Krebskrankheit, etc., Wolff, Jena, 1907, Teil I, 3. 2 Arndt, cited by Wolff, 3. ^ Oefele, cited by Wolff, 3. 4 Wolfler, cited by Wolff, 3. ^ Wolff, 7. I 2 BRIEF HISTORICAL REVIEW He not only separated several of the benign neoplasms (among them ganglion and lipoma) from the maHgnant growths, but described the enlargement of the axillar}- l}Tnph nodes accompanying cancer of the breast, and seemed to be familiar with carcinoma of the Hver and spleen. Nor was he unaware of the serious nature of cancer, for he expressed the opinion that it could be cured only in the first, or indurative stage, counseled ablation of the less maHgnant growths, devised operative measures for removal of carcinoma of the Hp, and extirpated cancer of the breast. Galen, ^ who in the second century- elaborated the theories of Hippoc- rates, recognized with his predecessor four cardinal fluids upon whose proportion to one another depended the state of the body. These fluids were blood, mucus, yellow bile (derived from the liver), and black bile fsecreted by the spleen) ; to the collection in undue amount of the material last mentioned he ascribed the origin of malignant growths. If the bile were sharp and irritating, there arose an ulcerating cancer ; if mild, one of the non-ulcerating t}'pe. The spread of the disease was well known to Galen, but. ignorant of the existence of the lymphatics, he beheved that it took place by way of the veins. His therapeutic endeavors were directed particularly toward attaining a diet which should not contain any of the substances productive of black bile, while purging played no unimportant part in the treatment. Con- sidering both procedures of the highest utility, he nevertheless did not entirely neglect surgical inter^-ention, which he conceived, however, to be of secondar}' value only, reasoning that it was illogical to treat a constitutional disease by measures so purely local as surgery. Leonides - was the first investigator wdth courage requisite to dis- card the advice of Hippocrates, never to operate on an ulcerating cancer. He practised with the knife and cautery the most thorough and energetic treatment of carcinoma of the breast, and was the first to appreciate the importance of retraction of the nipple as a diagnos- tic sign. From the time of Leonides, about i8o B.C., until the Renaissance, the art of the physician progressed only to a small degree, if, indeed, it advanced at all ; but -^dth the awakened interest in art, Hterature, and science which made this such an eventful period, there was aroused a 1 Wolff, lO. 2 Wolff, 14. BRIEF HISTORICAL REVIEW 3 renewed attention to medicine. Vesalius ^ exposed several of the errors of Galen and his predecessors, while his pupil Fallopius ^ made important contributions toward the diagnosis of cancer ; and although he preferred to treat the disease with arsenic, his contemporary, Ambroise Pare,^ was fully ahve to the necessity of total ablation whenever its performance was possible. Throughout more than a thousand years Galen's hypothesis of the four humors had ruled supreme in the domain of medicine, without an attempt having been made to depose it. With Paracelsus,"* however, there appeared signs of an imwilHngness to accept it in its entirety, and in the next century the final overthrow was ac- complished through the simultaneous operation of several factors, the most important of which were the demonstration of the circula- tion of the blood, and the revelation of the red blood cells and lymph channels. With these discoveries, and that of the cellular structure of cork by Hooke, hypothesis had begun to give place to observation, and follow- ing the improvement of the microscope came the description of the nucleus by Brown. '^ However, the cells of Hooke and his immediate successors were but chambers in the plant tissue, while it is doubtful from Brown's description whether he recognized the importance of his observation. It remained for Schleiden ^ to appreciate the significance of the cell as a unit in the organism of the plant, and for Schwann^ to apply this conception to the animal tissues. This new work en- hanced greatly the significance of Miiller's ^ earlier announcement of the cellular structure of certain growths, and encouraged him to reexamine tumors in which no cells had at first been found, as a result of which he ^ was enabled, with lenses of higher power, to- perceive a cellular structure also in these. But even after the cell had been accepted as the tissue unit, specula- iWolfE, 32. 2 Wolff, 34. 3 Wolff, 42. nvolff, 51. ^ Observations on the Organs and Mode of Fecundation in Orchideoe, etc., London, 1831. ^Arch.f. Anat., Physiol., etc., (Miiller), 1838, 137. ^ Mikroskop. Untersiichimgen iiher die U ebereinstimmimg in der Struktiir u. dem Wachstum der Thiere u. Pflanzen, Berlin, 1839. ^ Arch. f. Anat., Physiol., etc., (Miiller), 1836, p. ccxviii. ® Ueher denfeineren Bau und die Formen der krankhaften Geschwillste, Berlin, 1838. 4 BRIEF HISTORICAL RZ\1EW tion regarding its actual source continued rife, and the relation of the intercellular substance to cell genesis became the subject of much debate. Believed by Schwann to be destined for transformation into new cells, its insignihcance in this process was first appreciated by Remak.- A series of publications begun in 1S41 culminated in a description - of cell division in which it was denied that the inter- cellular substance was able to produce new cells, and declared to be very probable that all animal cells were created through the progres- sive diidsion of preexisting cells. According to Virchow"s conception.'^ it was just as impossible for a non-cellular material to elaborate a cell as it was for the decomposition products of animal or vegetable matter to give origin to an infusorian. Omnis cellida e cellula. as animals were derived only from animals, or plants from plants. The cells of a tissue ' which was to become the seat of a tumor began to swdl and divide, possibly as a result of irritation. If di^-ision went on rapidly, and if the members of successive generations became progressively smaller, the tissue arrived finally at the granulation stage in which, like granulation tissue, it was indifferent in appear- ance, the concUtion being analogous to that obtaining in the embryo in the early days foUo^dng fertilization. Upon this state there super- A'ened one of differentiation, leading to the ffnal stage in which the tumor assumed its finished form. The growth of a mali^gnant tum^or took place, not by increase of its elements, but by the fusion of accessory foci evolved in its neighborhood, a method of dissemination which proved that the primar}' nodule exercised a certain stimulation upon the surrounding parts through the agency of a secreted fluid. The process was ex- actly the same in metastatic tumors, but whether only a fluid excre- tion were concerned in this case was a question difficult of decision. It was not at all improbable that in some instances ceUs were actuall}' transported, to act as the infective material; and although ^ Untersuchungen iiber die Eiitwkklung der Wirbelthiere. Berlin, 1855. ^Arck.f. Anat., Physiol., etc., Qliiller), 1852, 47. ^ Arch. f. path. Anat., etc., (Virchow), 1855, via, 3. Die Cellidarpathologie, 4te Aufl., Berlin, 1871, 24. * Die krankhaften Gesckwiilste, Berlin, 1863, 89. BRIEF HISTORICAL REVIEW 5 this was not a common occurrence, still, as the cells within the tumor produced the injurious fluid it was not imreasonable to imagine them able to carry it to other locaHties. It was certain, however, that metastases were not a result of proHferation of the transported cells, but rather the product of healthy elements in the neighborhood, incited to mahgnant growth by the excretions of cancer cells deposited there. The vahdity of Virchow's h}pothesis was seriously affected by the later work of Thiersch,^ who advanced proof that the cells of epitheho- mata developed only from epithehum — never from connective tissue. And even though the examination of serial sections failed to demon- strate in every case an actual connection between the tumor and the epithehum, it was not unreasonable to postulate a pre^'ious association with this tissue, since it might have been present in the form of cell islands which had become estranged from their surroundings during the development of the embryo. In Thiersch's opinion, metastases in the l^miph nodes were to be ascribed with much more reason to the actual proHferation of transported tumor cells than to growth of the normal elements of the node under stress of stimulation by the h}^o- thetical growth-exciting fluid. Furthermore, Waldeyer,- amphfying the investigations of Thiersch, demonstrated that all carcinomata were of epithehal origin and denied that a transformation of connective tissue elements into cancer ceUs ever occurred. He agreed with Thiersch that metastases were the product of proHferation on the part of transported cells and attributed a large share in the production of secondary deposits to the ameboid motion which, in conjunction with Carmalt,^ he had demonstrated in the cancer cell. To this point Lambert and Hanes ^ have again called attention • within the past few years. Thus was overthrown the first half of Virchow's h}'pothesis. The remainder, which sought partially to explain by irritation the launch- ing of ceUs upon a career of lawless growth, is stiU intact, and accepted • ^ Der Epithelialkrehs , namentlich der Haiit, Leipzig, 1865, 58. ^ Arch. f. path. Anat., etc., (Virchow), 1867, xli, 470. ^Arch.f. path. Anat., etc., (Virchow), 1872, Iv, 481. ^ Jour. American Med. Assoc, 1911, Ivi, 791. Arch.f. path. Anat., etc., (Virchow), 1912, ccix, 12. 6 BRIEF HISTORICAL REVIEW by many investigators of the present day as an explanation, in part at least, of the inception of mahgnant proliferation. The occurrence of cancer of the mouth in smokers, of carcinoma of the stomach upon the ground of a gastric ulcer, of cancer of the gall-bladder in connection with cholelithiasis, and many other instances of an apparent relation between chronic irritation and mahgnant growth have been cited time and again in favor of this conception. To mention specifically but a few observers. Neve ^ has directed attention to the natives of Kashmir who wear underneath the clothing a basket of glowing charcoal, and in whom epithelioma is very common upon the anterior abdominal wall at the site where the basket is carried, while carcinoma of the mouth is, according to the experience of Chalmers and Perr}^^ very frequent among the women of India who chew the betel-nut and retain it in the mouth during sleep. Nor are instances wanting among the lower animals. Bashford ^ has cited the observations of Captain Brodie-Mills upon the great frequency of squamous cell carcinoma at the root of the right horn among the cattle used in India for draught purposes — the right horn being used by the natives for the attach- ment of agricultural implements or wagons. PUcque"^ noticed that carcinoma in horses was commonly found where the bit irritated the corners of the mouth, and subcutaneous fibromata ahnost invariably at the point of pressure by coUars or girths, while in dogs the most posterior mamm^, those most often engorged, were the ones more frequently affected by cancer. In cats, the upper hp, which was the one more hkely to be wounded, was the site of election for mahgnant disease. Thiersch ^ ascribed the development of epithehoma to a disturbance in the equihbrium between epithelium and connective tissue brought about by senile atrophy of the latter, while the pioHferative or bioplastic energy of the epithehum remained at the same time unim- paired. Mahgnant growth thus occurred not by reason of an in- crease in the ofi'ensive power of epithelium, but through a decrease 1 British Med. Jour., 1910, ii, 589. 2 Cited by Bashford, Third Sci. Report, Imperial Cancer Research Fund, London, 190S, 19. ^ Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 22. * Revue de Chir., 1889, ix, 521. ^ Der Epithelialkrebs, namentlich der Haiit, Leipzig, 1865, 78. BRIEF HISTORICAL REVIEW 7 in the defenses of the connective tissue. Chronic irritation, inflamma- tion, or trauma might under certain conditions act as the final stimulus and thus be an indirect cause of mahgnant growth. Waldeyer,^ on the contrary, urged an opposite conception : That with increasing age epithelial structures lost their vigor and became surrounded by a replacement fibrosis, the result of an increased activity on the side of the connective tissue. The isolation of epithelial ele- ments might be attended by any one of three possible results ; most frequently such cells underwent necrosis, at other times there super- vened cyst formation, while in the third place they might assume malignant growth. An important prelude to this type of prolifera- tion was the appearance of changes in the connective tissue extraor- dinarily Hke those accompanying inflammation. Hauser,- although admitting that epithelium was the most vigorous of all the tissues, maintained that Thiersch's hypothesis did not explain the enormous increase in growth energy so frequently encoun- tered in cancers of the cylindrical cell type. This phenomenon was but insufficiently explained by the assumption of a decrease in the physiological defenses, and there was required the addition of a posi- tive quantity to the opposite side of the equation, which might well be an augmentation in the assimilative power of the epithelium coupled, perhaps, with an increased avidity for food-stuffs. Furthermore, if the h3q3othesis were correct, one would anticipate the more frequent occurrence of multiple cancers, whereas the disease was almost without exception local in origin and limited to a single organ, if not to one part of that organ. Finally, connective tissue was often not so passive as the hypothesis assumed, for in many scirrhous carcinomata growth of the stroma proceeded with the greatest vigor and occasionally out- stripped even the epithehal proliferation. v. Hansemann ^ criticized Thiersch's hypothesis on the ground that it explained neither the relatively uncommon occurrence of carcinoma, nor its presence in the young, and extended the latter objection to include the fact that carcinoma in early life was exceptionally malig- nant. Were the conception correct, the case should be reversed. ^ Arch. f. path. Anal., etc., (Virchow), 1867, xli, 470; 1872, Iv, 67. ^ Das Cylindercpithel-carcinom des Magens wid des Dickdarms, Jena, 1890, 135. ^ Die mikroskopische Diagnose der hosartigen Geschwiilste, Berlin, 1902, 215. 8 BRIEF HISTORICAL REVIEW Although Cohnheim ^ was not the first to suggest a connection be- tween mahgnant growths and embryonic rests, it is with his name that the hypothesis is generally associated. Having described a congenital myo-sarcoma of the kidney, he suggested that the tumor might have originated from germinal muscle cells which had been snared off at a time when the foundations of the urinary organs were being laid down. He assumed that in early embryonal hfe more cells than requisite were produced and that the unutiHzed elements, with all their inherent power of proHferation, were cut oft" at a very early period, probably corre- sponding to the interval between differentiation of the germinal layers and completion of the foundations for the various organs. Cohnheim sought support for his suggestions in the undift'erentiated embryonal appearance of tumor cells as well as in the congenital or early postnatal occurrence of neoplasms, although admitting that it would go hard with his hypothesis were it forced to depend upon the latter alone for substantiation. Still, the conception did not in the least require that the tumor itself should be congenital, but merely the foundation for it. What were the circumstances able to initiate growth in embryonic rests he made no definite effort to estabhsh, although offering the suggestion that one of them might be repeated arterial congestion or even inflammatory hyperemia, so that the idea of traumatic etiology inight be true wdthin limitations. It was impos- sible, however, to apprehend the influence which excited or released the power of indefinite proHferation, for it had not yet been learned in what manner growth was normally inhibited. It might be that normal tissues were able to exert a certain control over the germinal displacements and thus keep them from proHf crating, while trauma might so weaken this physiological restraint in the neighborhood of a cell rest as to permit the inception of growth. Cohnheim believed that his explanation accounted for the variety of tissues so often found in tumors, as well as for the occurrence of certain neoplasms in specific localities. It was a clinical observation of many years' standing that epithehal growths were common at the various orifices of the body, and, indeed, this fact had been made use of by Virchow to sustain the hy- pothesis of mechanical insult. But if injury were the factor determin- ^ Arch. f. path. Anat., etc., (Virchow), 1875, Ixv, 64. Vorlesungen iiber dig. Path., Bd. i, Berlin, 1877, 634. BRIEF HISTORICAL REVIEW 9 ing continuous proliferation it was hard to understand why tumors should be so rarely met with on the hands and feet, which of all parts of the body were most eminently exposed to trauma. An elucidation of the frequent occurrence of epithelial growths at certain sites must, therefore, be sought elsewhere, and most reasonably in the fact that at the apertures of the body there was a comphcation in the embryonal structure at one or another stage of development. R. Hertwig ^ thought that a hypothesis which would refer the origin of neoplasms to embryonal rests was irreconcilable with the fact that tumors, and particularly mahgnant tumors, were more frequent in the higher age periods. It must be assumed that the body was everywhere provided with nutritive materials which were only await- ing utilization. How was it, then, that for decades the colonies of embryonal cells did not make use of this material? Either they were unable to do so (for lack of the very power which Cohnheim's concep- tion emphasized and upon which it was built), or they were by some means excluded from sharing the universal food-stuff — a supposi- tion which was inconceivable. Furthermore, Hertwig did not beHeve that a comparison could be instituted between embryonal and tumor cells on a common ground of absence of differentiation. Those of the embryo were undifferentiated, it was true, but they possessed an increasing tendency to differentiate, or, in other words, to exchange cytotypic for organotypic growth, while the slight disposition originally present in the elements of a tumor became progressively weaker and weaker. The problem of malignant growth, according to Hertwig, was a double one, since it must first be explained how, in an organism which had reached its Hmit of size, the cells normally lost their autonomous growth and subjected themselves to their surroundings, and secondly it must be ascertained what changes had to occur in cells in order that they might escape from the control exercised over them by the organism. Cohnheim's hypothesis has been criticized by Bashford ^ in a dis- cussion of cancer in the cheek and abdominal wall which, almost 1 Festschrift Ernst Haeckel, Jena, 1904, 347. ^ Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 24, and Introduction, p. ix. 10 BRIEF HISTORICAL REVIEW unknown in Europeans, occurs ^^•ith great frequency among the natives of India follo-^ing chronic irritation produced by the use of the betel-nut or Kangri basket. '". . . it becomes necessary to postulate further a uniform and abundant distribution of " embryonic rests ' over the body, or to assume a different distribution of "embryonic rests" in Europeans and native races, coinciding with the points they respec- tivelv select for the indulgence of various practices invohdng the appli- cation of peculiar irritations."" And further: '' Suffice it to sa}^ that if all forms of cancer are to be explained by such speculations and the interA'ention of congenital 'germs" is to hold good for all cases, then it must be assumed that such germs are as uniformly distributed, e.g. over the surface of the body of all vertebrates as the skin is itseh. and thus the explanation becomes no explanation at aU." While, in the opinion of Ribbert.^ many carcinomata almost cer- tainlv took their origin from embryonal cell rests, it was evident that such structures were not invariably able to assume the malignant tvpe of growth, since many of them lay indefinitely quiescent. .\lthough they possessed, in common with other epithehal cell groups, the normal tendency to penetrate underhing connective tissue, this latter must, under ordinary conditions, exert some sort of inhibitory influence preventing invasion on the part of the epithehum. and not until this opposition had been removed could the potential energies of the epithelium be unfolded. The release of the normal invasive faculty was brought about through the presence in the connective tissue of simple hyperemia, or of inflammatorv' lesions such as round cell infiltration, or proliferation of the fixed elements with a con- comitant production of blood vessels, an analogous situation occur- ring in the development of glandular structures in the embryo, where downgrowth of epithelium was always preceded by alterations in the subjacent tissue. Xevertheless. in the fully developed organism at least, cellular changes in the connective tissue did not invariably ehcit epithehal invasion, for inflammation, even though it continued over an extended period of time, was not always followed by the development of carcinoma. Hence, the inflammatory lesions which "^ Arch. f. path. Anat., etc., rvircliovrj. 1894, cxxx\-, 433. Geschu'idstlehre, Bonn, 1904, 552. Das Karzinom des Menschen, Bonn, 1911, 482. BRIEF HISTORICAL REVIEW II had been found to precede carcinoma must be distinguished by some special characteristic, and the whole process could best be explained on the assumption that they were the product of certain agents elab- orated by the epithelium itself — that the epithelial cells prepared the soil into which they were later to penetrate. Such a chain of events, however, might result merely in the occurrence of heterotopia, and to explain actual malignancy it was necessary to postulate in addition that the epitheHal cells suffered a loss in differentiation,^ through which they became enabled to proliferate more luxuriantly. In the case of embryonal rests, malignant growth was the more readily initi- ated because the cells had not yet undergone complete differentiation, and were therefore more ready to start upon a career of indefinite multipHcation. Ribbert's work was the anatomical expression of a biological law formulated by Weigert,^ who had suggested that as tumors, either benign or malignant, developed without the interference of external stimuH or after the action of those which, under other circumstances, did not give rise to tumors, stimulus could be considered merely as an accessory cause. Such an etiological factor could only arouse a latent disposition toward the growth of a tumor or accelerate the proliferation of one already present. To explain this disposition, it might be imagined that there was a misrelation between the proHf-. erative energies of some region of the organism and its neighboring tissues, which had occurred either through increased growth activity on the part of certain cells or through decreased resistance of those tissues determining the bounds of physiological growth. The for- mer might take place in the way that Cohnheim had postulated, or else single cell groups might retain their growth energy after the embryonal period had passed. A decreased resistance, on the other hand, might be congenital for one or another locahty or develop in later life by the more rapid aging of certain tissues — a conception which had been advanced by Thiersch and with which Weigert expressed his agreement. For neither non-maUgnant nor malignant proliferation, however, was it necessary to assume the application ^ Ribbert distinguished very sharply between such a condition, and one of specific cancerous degeneration. 2 Ueher Entziindiing, Eulenburg's Real-Encyclopddie der gesammten Heilkunde, erste Aufl., 1880, iv, 644; zweite, 1886, vi, 325. 12 BRIEF HISTORICAL REVIEW from without of a direct stimulus to cell growth, for in both cases there were concerned only those proclivities toward continuous multipli- cation which had been inherent in the cells since the embryonic period. Hauser/ in a series of articles in which, to support his contention, reference was made to his own previous observations,^ denied strenu- ously the correctness of the view held by Ribbert regarding the develop- ment of carcinoma. This disorder, in his opinion, could be produced only through a fundamental change in the biological properties of the implicated cells, and with the assumption of such a change certain morphological data were in entire accord. The loss of physiological function, the increase in the size of the cell and its nucleus, the higher chromatin content of the latter, the presence of abnormal mitotic figures, and finally, the enormous capacity for multiplication — all indicated an entirely new type of cell. The process was, in short, a specific cancerous degeneration, and a weakening of the defenses set up by the connective tissue against epithelium could play but a sub- ordinate role in the process. While Ribbert had always contended that a new growth increased simply by the multiplication of its own cells — ans sich heraus — Hauser, on the contrary, maintained that it spread by a wave of cancerous degeneration which successively attacked normal structures at its margin. In this connection the views of Adami ^ are of interest. He de- scribed an early multiple tumor of the adrenal cortex " . . . in which occasional cells in the immediate neighborhood of the small mass of new growth, while still retaining their relationship to the columns of the zona radiata, by their enlarged and deep-staining nuclei stood out as of the can£erous_t^pe.'' He quoted, furthermore, certain observa- tions of Horst Oertel upon the apparently direct cancerous transfor- mation of liver cells in a case of multiple hepatic carcinoma, and concluded with the statement : "For my part I regard it as wholly dem- onstrated that it does occur ; and it must be taken into account in the development of any adequate theory of neoplasia." ^Arch.f. path. Aiiat., etc., (Virchow), 1894, cxxxviii, 482. Beitr. zur path. Anat., etc., (Ziegler), 1897, xxii, 587. ^ Das chronische M agengeschwiir , etc., Leipzig, 1883. Das Cylinder epithel-carcinom, etc., Jena, 1890. ^ The Principles of Pathology, Philadelphia and New York, 1910, Vol. i, 836. BRIEF HISTORICAL REVIEW 1 3 This view has been adopted in part also by Verse/ a pupil of Mar- chand, in so far as concerned carcinoma ta of the gastro-intestinal tract. An account of the argument between Ribbert and Hauser may be found in an article by the latter author,- whose paper is of additional value because it contains a review of all the cancer literature from 1891 to 1897. An enthusiastic disciple of Ribbert was Borrmann;^ who refuted Hauser's interpretation of the spread of carcinoma and quoted Kokubo,^ Petersen,^ Pforringer,^ and himself ^ in support of Ribbert's behef that cancer spread only through the proliferation of its own cells. He ascribed the origin of nearly all tumors to the growth of embrv'onal rests, and fully supported Ribbert's view that an inflammatory reac- tion always preceded the abnormal growth of epithelium. While Borst ^ did not accord such unqualiiied approval to Ribbert's hjpo thesis, he nevertheless admitted that its propounder had done no inconsiderable service in calling attention to the many mistakes which had been made in explaining the transformation of normal into abnormal epithehum and by arousing pathologists to the too exclusive attention which had been devoted to the behavior of the epithelium. His own observations led him to beheve that an iiiflammatory infil- tration of the corium occurred almost without exception — a true h^-perplasia of its connective^ tissue, however, not so frequently. While both processes were certainly in part secondary to the car- cinoma, it could not be denied that inflammation in the subepitheKal tissue might frequently afford the special sort of stimulation necessary for the development of such a growth. The initiatory invasion, how- ever, was usually on the part of the epithelium ; and although he was at variance with Ribbert on one point, he nevertheless agreed entirely ^ Uber die Entstehiing, den Ban u. das Wachstum der Poly pen, Adenome, n. Karsinome des Magen-Darmkanals {Arb. a. d. Path. Inst, zii Leipzig, 1908, Bd. i, Heft 5). - Cenlralbl. f. allg. Path., etc., 1898, ix, 221. ^ Zeitschrift f. Krebsforsch., 1904, ii, i. * Festschrift fiir Orth, 1903. ' ^ Beilr. zur klin. Chir., (v. Bruns), 1902, xxxii, 543. ^ Beitr. zur. klin. Chir., (v. Bruns), 1904, xli, 687. '' Das Wachstum und die V erbreitungswege des Magencarcinoms, Jena, 1901. (Erster Supplementband, Grenzgebiete der Medizin und Chiriirgie.) ^ Die Lehre von den Geschwiilsten, Wiesbaden, 1902, ii, 686. 14 BRIEF HISTORICAL REVIEW with Ribbert's conception of the spread of carcinoma, concluding that the continuous cancerous transformation of normal glands in the vicinity of a fully developed carcinoma of a mucous membrane did not occur. Janeway ^ expressed the conviction that the great majority of cutaneous cancers were a direct outgrowth from normal epithelial cells and that their development was associated with insignificant changes, or no changes at all, in the connective tissue. V. Hansemann,^ extending the observations of Klebs,^ investigated very thoroughly the types of mitoses found in tumor cells and de- scribed three varieties, distinguished by the amount of chromatin contained — hypochromatic, normal, and hyperchromatic. He ex- plained the increased growth energy of the cancer cell by postulating a gain in the power of independent proKferation with a concomitant loss of differentiation ; this condition he called anaplasia. Extensively keratinized cells in a carcinoma of the skin would be less anaplastic, because departing less from the normal type, than would those of a similar tumor that were entirely free from keratinization, while, similarly, an intestinal carcinoma that still preserved some sem- blance of glandular arrangement or that contained cyhndrical or goblet cells, would be less anaplastic than one composed of polymorphous elements. The anaplasia increased directly as the degree to which the cells were biologically removed from the tissue in which they had orig- inated. It should not be lost sight of, however, that anaplasia was a relative, not an absolute, term — that cells were not of themselves anaplastic, but only in comparison with their mother tissue. Nor should anaplasia be likened to embryonalism from which, on the contrary, it was to be sharply distinguished, for embryonal elements were still undifferentiated, while those in a state of anaplasia had been differentiated at one time but had lost that property. Cells in be- coming anaplastic did not necessarily retrace their way over the road by which they had departed from embryonahsm, but rather, charac- ^ Zeitschriftf. Krebsforsch., 1909-1910, viii, 403. ^ Arch. f. path. Anat., etc., (Virchow), 1890, cxix, 299. Studien ilber die Spezificitdt, den AUruismus iind die Anaplasie der Zellen, Berlin, 1893. Berl. klin. Woch., 1909, xliv, 1850. ^ Die allg. Paihologie, Jena, 1889, Theil 2, 524. Deut. med. Woch., 1890, xvi, 518. BRIEF HISTORICAL REVIEW 1 5 teristics previously held in abeyance now became able to assert them- selves. One of the criteria of anaplasia was an unequa l division of the chromosomes during mitosis ; and if this were followed by an asym- metrical division of the cell it must be assumed, for the following reason, that elements of different biological character would result. In the developing ovum each unequal segmentation was followed by a series of equal divisions whose purpose it was to enlarge and perpetu- ate characteristic cell groups. The stages in development at which these unequal divisions occurred v. Hansemann called generation stages. Every generation stage was accompanied or shortly followed by a change in the direction of growth and by a degree of differentia- tion previously absent as, for example, in the invagination of the vege- tative cells of the blastula to form the gastrula, which was the first instance of the process under discussion. To these generation stages each asymmetrical division of a cell was comparable,- and it was accord- ingly accompanied by a change in the energy and the direction of growth. V. Hansemann did not wish to be understood as considering ana- plasia the prime cause of malignant proliferation, since according to his conception a growth stimulus was required in addition. When such a stimulus acted upon a normal cell the result was hyperplasia — upon an anaplastic cell, malignant growth. Those mitoses were asymmetrical in which the allotment of chro- mosomes to the daughter stars was unequal, but as the actual number of chromosomes going to each diaster could not be counted except in rare instances, it was necessary to be content with an approximate estimation. Irregular mitoses v. Hansemann believed characteristic of the carcinomata, as he had never succeeded in finding them in sarcomata, or in hyperplastic, regenerating, or normal tissue. Asymmetrical mitoses had, however, been discovered previously by Podwyssozki ^ in the cells of regenerating liver tissue, and soon after the appearance of v. .Hansemann's first article further instances of their presence in cells other than those of carcinomata were reported. Ribbert,^ although able to substantiate the occurrence of asym- ^ Beilr. ziir path., Anat., etc., (Ziegler), 1886, i, 301. ^ Deut. med. Woch., 1891, xvii, 1183. l6 BRIEF HISTORICAL REVIEW metrical mitoses in carcinoma, could not agree entirely with the con- clusions of V. Hansemann regarding their signiiicance, for he had been unable to distinguish between this type of karyokinesis and other patho- logical division figures. In later articles Ribbert ^ denied the existence of anaplasia in the sense of v. Hansemann and strongly emphasized his behef that the characteristics of cells were so firmly ILxed by hered- ity that single elements or their ancestors within the same organism could never suffer such a radical change as unequal and undiffer- entiating segmentation. Stroebe - examined about a dozen carcinomata and found asym- metrical mitoses present in greater or smaller number without excep- tion, but was successful in demonstrating them also in sarcomata, in benign tumors, in heahng wounds of healthy tissue, and in the germinal centers of lymph nodes. He suggested, therefore, that they were not pecuhar to any special tissue, either pathological or normal, but were to be found in any location where there was a large number of mitotic figures. Beneke ^ beHeved that the mitotic anomalies were secondar)^ to a disturbance of specific functional cell power. In the cells of mahg- nant new growths a loss of function was accompanied by a gain in growth power and this condition, kataplasia, constituted the only difference between the cancer cell and the normal element. In the history of developing cells there could be no reversion and thus ana- plasia, or a return from later to earlier stages through the agency of pathological mitoses, could not possibly exist. V. Hansemann's critics had attacked him by disputing both the diagnostic value of irregular mitoses and the conception that anaplasia could give rise to mahgnant growth. His answer was that they had not read, or at least had not understood, the explanations of his hy- pothesis. To the first criticism he ^ replied that the impossibility of diagnosing carcinomata only by the presence of pathological mitoses had been recognized and emphasized in his papers ; but because such 1 Dent. med. Woch., 1896, xxii, 471. Bihliotheca Medica, 1897, Abt. C, Heft 9, 32. ^ Beitr. zur path. Anat., etc., (Ziegler), 1892, xi, i; 1893, xiv, 154. ^Arch.f. path. Anat., etc., (Virchow), 1900, clxi, 100; 1901, clxiii, 174. * Die mikroskopische Diagnose der hosartigen Geschwiilste, Berlin, 1902, 97, 191. BRIEF HISTORICAL REVIEW 1 7 structures were not specific for carcinoma it did not necessarily follow that the entire h>pothesis should be discarded. To take such a course would be "to pour out the child with the bath." In answer to the second objection he repeated that the h^-pothesis of anaplasia did not concern itself with setiology. By far the most perplexing attribute of the mahgnant cell has been its capacity for unlimited multiplication. As the study of the cell pro- gressed and cytologists became famiUar with the events running their course in the fertihzed egg, the fact seemed significant that segmenta- tion did not begin until another cell had entered and fused with the ovum after which, however, this element became endowed with the most remarkable proKferative power. An analog}^ between fertiliza- tion and the growth of the cancer cell seemed reasonable and did not escape the observant mind of Virchow/ who suggested that be- cause tumor juices appeared to act on certain elements as a fertihz- ing agent the behavior of these tissues was, to this extent, entirely comparable to that of the ovum. Leucocytes had often been described between or even in cancer cells and Klebs - suggested that they might, by conjugating with the elements of epithelium, excite mahgnant growth. Farmer, Moore, and Walker ^ described the occurrence in mahgnant growths of heterot^^ical mitoses similar to the ones found in game- togenic cells and dift'ering from those of somatic cells in possessing but one-half the number of c hromosom es which, moreover, assumed the forms_of rings, Joops, etc. They felt justified, therefore, in corre- lating the appearance of these gametoid neoplasms with the result of a stimulus which had altered the normal somatic course of cell develop- ment into that characteristic of reproductive (not embryonic) tissue. V. Hansemann^ denied the presence of heterotypical_mitoses in malignant growths and referred the reduction of chromosomes de- lineated by these authors to asymmetrical mitoses and the degenera- tion of occasional chromosomes. 1 Die krankhaften Geschwiihte, Berlin, 1863, i, 87. - Die allg. Pathologic, Jena, 1889, Theil 2, 524. ^ Proc. Roy. Soc, 1903, Ixxii, 499. Proc. Roy. Soc, Series B, 1906, Ixxvii, 336. *Biolog. Centralbl., 1904, xxiv, 189; 1905, xxv, 151. 15 BRIEF HISTORICAL REVIEW While Bashford and ^Murray ^ at first con&rmed the findings of Farmer, Moore, and Walker, dissociating themselves, however, from the conclusions which had been drawn, more extended observation convinced them - that the mitoses of cancer were somatic rather than heterot}'pical and that the heterot\^e appearance had been conferred by variations in the development of the achromatic figure, the peculiar forms of the chromosomes, and their mode of attachment to the spindle. Walker and Whittingham ^ have recently reaflfirmed the observation of Farmer, IVIoore, and Walker, adding the statement that the char- acteristic shapes of the chromosomes belonging to the first meiotic (heterotypical) division had been encountered in very considerable numbers among the cells of mahgnant growths. Critzmann.^ having noticed a relation between twin births and cancer in certain families, suggested that both anomalies resulted from the development of two ova, twin births following the simultaneous fertilization of two ova, cancer the inclusion of one of these within the other. A cancer was thus an abortive fetus retained ■^'ithin a developed one — the brother of him who bore it. De Morgan ' concluded that in certain persons there was a local tendency toward tumor formation, impressed upon the tissues perhaps during embr\'onic life, but lying dormant until finally awakened by irritation. The slumber cell hypothesis became more wideh- known, however, through its adoption by Grawitz,^ according to whose concep- tion some of the embryonal connective tissue elements lost their cellular nature and were no longer demonstrable T\dth any of the known stains. They remained slumbering in this state, although still taking part in metaboHsm, until, awakened by stimulus or other cause, they were finally enabled to prohferate. Beard '' modified Cohnheim's hypothesis in conformity with his own ^ Proc. Roy. Soc, 1904, Ixxiii, 67. ^ Proc. Roy. Soc, Series B, 1906, Ixxvii, 226. * Jour. Path, and Bad., 191 1, x\a, 185. ^ Le Cancer. These de Paris, 1S94. Cited by Wolit, 443. ^ Lancet, 1871, ii, 155. ^ Arch. f. path. Anat., etc., CVirchow), i89i,cxxvii, 96. Berl. klin. Woch., 1892, xxix, 109. ^ The Enzyme Treatment of Cancer, London, 1911. BRIEF HISTORICAL REVIEW 1 9 views on the course of normal development as an alternation of asexual and sexual generations. He regarded cancer as an attempt of displaced or aberrant germ cells to reproduce the asexual phase of development (the chorion in mammals and its homologue elsewhere), while nor- mally the degeneration of the asexual generation was ascribed to the activity of ferments, especially tr>7)sin, produced by the growing embryo. On this series of assumptions he based his advocacy of the now discarded trypsin treatment of cancer. CIL\PTER II ATTEMPTS TO PRODUCE TUMORS The possibility of inciting cancer in tissues previously normal has attracted and maintained the interest of a large number of experi- menters. The recorded attempts are numerous and if. as is probable, there have been still others which, because they miscarried, have not been thought worthy of pubhcation. the number of times that the experiment has been made must be very great. The ingenuity of man has been sorely taxed to tmd ways of conferring the power of mahgnant growth upon normal cells and the keenest intellects of two continents have grappled with the question year after year. eternaUy hoping, vet alwavs baffled. The greater number of trials have been molded to ht one or the other of the two most v^ddely knov^m h}-potheses of cancer genesis — the h}-pothesis of cell irritation and that of cell dis- placement. Following these two h^-pothetical clues, investigators of the question of mahgnant growth have irritated ceUs or misplaced them in everv conceivable way. and stiU the riddle remains unanswered to us as it did to our forefathers. \Miat relations, if any. obtain between irritation and cell growth can hardly be re\dewed here, but those interested in them may find a fuh discussion of the question in the writings of Mrchow. AVeigert. Ribbert. and others. Xor can one consider at this point the second hypothesis which, fathered by the great Cohnheim, has become in later years the protege of Ribbert and his school, A number of pathologists have sought to initiate mahgnant growth bv introducing the cells of one animal into another. Xone of them, however, has been able to produce a true neoplasm, and the utter hopelessness of the attempt has been emphasized in recent years by the fact that no malignant new growth has ever been reported among the thousands of mice and rats inoculated with normal tissues for the purpose of conferring immunity toward the transplantation of cancer. ATTEMPTS TO PRODUCE TUMORS 21 Among the first to attack in this way the problem of the origin of cancer was Zahn,^ who transplanted the tissues of adult rabbits into other rabbits without a single successful result. When, however, he substituted embryonal cartilage, the grafts proliferated, the amount of growth being proportional to the vascularization of the tissues surrounding them. The transplantation of fetal bone was equally successful. Leopold,^ who engaged in similar experiments, also found that while the tissues of young or adult animals were quickly absorbed after transplantation, fetal cartilage would remain ahve and grow, occasionally reaching a size which exceeded by two or three hundred times that of the original graft. In this way there was estabhshed a true enchondroma which persisted for a considerable period. The younger the embryos, the more vigorous was the growth of their cartilage. Kaufmann^ sewed in portions of epithehum of the combs and wattles of fowl, and created in this way cysts inclosed by a giant cell tissue which grew very vigorously for weeks and exhibited no signs of regression even after months had passed. Martin^ injected into the jugular vein of a rabbit oil of sweet almonds which had been rendered irritating by the addition of i to 2 per cent of croton oil. He discovered some time later the existence of epithehal growths in the lung recalhng exactly in their structure the type of epithehoma that Malassez had named epithelioma muquotde. Hanau,-^ in the course of protracted attempts to produce atypical epithelial growth, painted various preparations of tar upon the scrotum of white rats and upon the vulva, nipple, and mamma (both secreting and non-secreting) of bitches. All of the experiments, however, were entirely barren of result even when irritation was kept up for months. Considerable interest was aroused for a time through the descrip- tion of an experiment by Lack,^ in which he left the scrapings from a 1 Congr. internal, des Sci. med. de Geneve, 1877, 5™^ Session; Compt. rend, el Memoires. Geneve, 1878, 658. Arch.f. path. Anal., etc., (Virchow), 1884, xcv, 369. ^Arch.f. path. Anat., etc., (Virchow), 1881, Ixxxv, 283. ^ Arch. f. path. Anat., etc., (Virchow), 1884, xcvii, 236. * Cited b)-^ Ledoux-Lebard, Arch. gen. de Med., 1885, civ, 423. ^ Silzungsber. d. Gesellsch. d. Aerzte in Zurich vom 9 Mdrz, 1889. Fortschrilte der Med., 1889, vii, 338. ^Journal Path, and Bad., 1900, vi, 154. 22 ATTEMPTS TO PRODUCE TUMORS rabbit's ovary in the peritoneal cavity until the death of the animal fourteen months later. At the autopsy, masses of new growth exhibit- ing the features typical of ovarian cancer were found in the abdomen and thorax. Shattock/ however, pointed out that as there was in this animal a uterine tumor which resembled exactly, in the gross, a columnar cell carcinoma which he himself had observed in the uterus of a rabbit, it was at least possible that the growths ascribed by Lack to his experimental measures might have been metastases from a primary carcinoma of the uterus. That such tumors in the rabbit are not un- common may be inferred from the fact that Boycott^ has recently reported four cases of epithelial neoplasm in the uterus of this animal, and Leitch,^ as well as Marie and Aubertin,^ spontaneous carcinomata in the same organ. Fraenkel ^ carefully repeated Lack's experiment fourteen times, but always with an unsuccessful issue. He further described attempts to produce a chorion epithelioma by engrafting rabbit placenta into the rabbit from which it had been removed, into other gravid rabbits, into puerperal rabbits, and into one male. But not one of these trials was followed by the outcome that he was trying to procure, nor could he detect any evidence of growth in embryonal ova after their inocula- tion. Birch-Hirschfeld and Garten^ injected very young emulsified em- bryos into the livers of goats, rabbits, fowl, salamanders, and frogs, without being able to induce the formation of permanent tumors, although nodules of cartilage did grow for a time in the Kver and lungs before they were finally absorbed. The embryonal cells underwent a certain amount of differentiation in spite of their unusual environment, and growth in the liver seemed to be favored when that organ was periodically stimulated by gentle heat. Brosch''' beheved that he had succeeded in obtaining atypical epithe- lial growth as the outcome of the following experiment : An area of 1 Trans. Path. Soc. London, 1900, li, 56. ^ Proc. Roy. Soc. Med., 1911, iv, Path. Section, 225. ^ Proc. Roy. Soc. Med., 191 1, v, Path. Section, i. ■* Bull, de V Assoc. Franq. pour V Etude du Cancer, 191 1, iv, 253. ^ Centralbl. f. allg. Path., etc., 1903, xiv, 666. ^ Beitr. zur path. Anat., etc., (Ziegler), 1899, xxvi, 132. '^ Arch. f. path. Anat., etc., (Virchow), 1900, ckii, 72. ATTEMPTS TO PRODUCE TUMORS 23 skin on an animal's back was crushed between the blades of a forceps and a few days later a solution of paraffin in xylol was rubbed into the ulcer which remained after the removal of the scab. Further appHca- tions were not undertaken until after the infiltration of the wound had receded. The atypical proHferation was observed from six to ten weeks after the begirming of the experiment. Stahr ^ noticed that rats fed for a long time on oats would develop papihary epithehomata of the tongue, and he traced these tumors to the irritation set up by the lodgment of small vegetable fibers. The same author- produced areas of chronic irritation in rabbits, rats, and mice by repeatedly scraping a mucous membrane and painting the eroded part with x>'lol-parafiin, soot, or tar; still, in spite of the chronic lesion thus induced, there had been no evidences of atypical epithelial prohferation. Kelhng^ denied that the cells of mahgnant growths originated in the body of the individual in whom the tumor developed, and beheved rather that these elements were cells of organisms lower in the ph^do- genetic scale which had obtained entrance and begun to proHferate. When he submitted this hypothesis to experimental proof he was unable to demonstrate its vaKdity to his satisfaction until the health of the animals had been dehberately undermined before inoculation, or until the injections were made in the neighborhood of wounds. By implanting emulsions of flies, gnats, snails, etc., he had been successful in producing a fibro-sarcoma, an adeno-carcinoma, and a mixed cell sarcoma, and, furthermore, in recognizing in some cases the cells of the organism introduced. In the following year he ^ described the evolu- tion of other mahgnant tumors in five out of seven dogs, through the inoculation of fowl embryos emulsified in sahne solution. None of Kelhng's observations have received any widespread acceptance. Ribbert ° obtained small but typical papillomatous growths on the inner surface of the rabbit's hp by repeatedly scraping certain points, again denuding them as often as the epithelium was regenerated, and 1 Centralhl. f. allg. Path., etc., 1903, xiv, 4. ^ Munch, med. Woch., 1907, liv, 1178. ^ Wien. med. Woch., 1903, liii, 1431. Miinch. med. Woch., 1903, 1, 923. ^ Miinch. med. Woch., 1904, li, 1047, 1909. ^ Geschwtilsllehre, Bonn, 1904, 352. 24 ATTEMPTS TO PRODUCE TUMORS finally allowing them to heal. It appeared that a state of chronic inflammation played an important role in the causation of this and other types of fibro-epithelial tumors, for, as Ribbert remarked, when irritation and inflammation had ceased tumors of this class not infre- quently retrogressed. Loeb^ injected or transplanted portions of embryos into adult animals with the general result that most of the tissues not only grew, but even underwent a certain amount of differentiation. Thus epithe- lial cells arranged themselves in glandular form and could be seen to contain granules denoting cell activity. As in the experiments of Zahn and Leopold, the most pronounced growth was observed in the case of cartilage. The implantation of adult tissues afforded analogous results except that growth on the part of cartilage was not so marked. With tissues other than cartilage embryonal cells did not proliferate more actively than those from animals fully grown, and, indeed, in some cases the adult elements seemed to grow even more vigorously than the others. The data thus acquired did not, therefore, justify the widely accepted idea that embryonal tissues were greatly superior to adult in their power to proliferate after transplantation. In all of the experiments the displaced cells finally ceased to grow and even- tually underwent absorption.^ The same author ^ succeeded in produc- ing, in the uterus of the guinea-pig, nodules of decidual tissue possessing the power of temporary growth. Incisions were made into the pregnant uterus, sometimes in various directions, while at other times part of one horn was spKt longitudinally and the mucous membrane everted. When the uterus was thus incised on the fifth or sixth day of pregnancy there developed in the wounds a number of nodules composed of typi- cal decidual tissue. Immediate contact with the ovum was evidently not necessary, for serial sections of a number of these " deciduomata " were searched in vain for embryos. In experiments prosecuted during earlier or later stages of pregnancy decidual nodules were not evolved, although small ones could still be produced as late as the tenth day, and in one case followed an operation undertaken during the first two ^ Jour. American Med. Assoc, 1903, xl, 974. 2 It is not quite clear whether these experiments were done by Loeb himself, or whether the author is recapitulating the work of others. ^ Centralbl. f. dig. Path., etc., 1907, xviii, 563. ATTEMPTS TO PRODUCE TUMORS 25 days. From three to four weeks after operation the nodules were partly or entirely necrotic. Commenting on the relation between these nodules of decidua and the malignant tumors, Loeb ^ said that during the early period of its existence the corpus luteum secreted a chemical body which united with the mucosa of the uterus and sensitized it. When incisions were made into the uterus the freeing of its inner surface from tension acted as an external stimulus which caused the sensitized tissue to react with the production of tumor-like formations, where under ordinary conditions the uterus would have shown only the usual processes of wound healing. For the initiation of malignant growth, therefore, one factor alone did not always suffice. There were necessary a chemical sensitizing substance and an external exciting cause — facts which must be kept in mind in attempting to interpret the origin of cancer. The experiments of Wilms ^ concerned the implantation of five-to- seven-day-old embryo chicks into young fowl, a procedure which was followed by considerable growth and differentiation. In trying to overcome the reaction against the newly introduced cells, by making several successive inoculations at intervals of eight days, Wilms gained the impression that more energetic development took place. There seemed to be a certain disposition favorable to growth in one of the birds, for all of the grafts had developed into palpable nodules at a time when those in five other birds were still indistinguishable. As in the experience of other investigators, so there occurred here a certain tem- porary growth and differentiation of the embryonal tissue in spite of its new surroundings, succeeded by final regression and entire absorption. Careful and extensive experiments of the same type were undertaken by Nichols ^ to determine whether normal tissue could acquire after transplantation the power of unHmited growth and give rise to metasta- ses. Rabbits and guinea-pigs, sixty-two in all, were inoculated with the cells of various organs but always with an unsuccessful issue, at least in so far as unlimited , proHferation and the production of metastases were concerned, although occasionally the implanted structures were possessed of enough proliferative energy to give origin to growths re- 1 New York Med. Jour., 1909, xc, 145. 2 Verhandl. d. deiitschen path. Gesellsch., 1904, 8 ^ Tagung, 79. ^ Jour. Med. Research, 1904-1905, N.S., viii, 221. 26 ATTEMPTS TO PRODUCE TUMORS sembling dermoid cysts or teratomata. The materials used included testis, ovary, kidney cortex, Hver epithelium (adult and fetal), adrenal gland, adult or fetal epidermis, adult uterine epitheUum from pregnant and non-pregnant animals, fetus (entire) , fetal cartilage, and placenta. Among all these tissues only adult and fetal epidermis, uterine epithehum, entire fetus, cartilage, and placenta were capable of proliferation ; these produced tumors which varied in size from a small nodule to one exceeding by ten or twenty times the amount intro- duced, but not one of which was at all comparable to a true mahgnant growth. In an endeavor to create tumors from normal cells already possess- ing, or supposed to possess, high proliferative power, Levin ^ implanted or injected rabbits with cells from the ovary, with embryonic cells or those from healing wounds in the adult hver, with epithehum, and with pigmented cells from the iris. In no case, however, could he convince himself that any of the elements so transplanted had taken on mahgnant growth. Reinke^ recorded an example of at3q3ical epithelial growth which he had procured by the injection of 4 % ether into the eye of the adult salamander. The proHferating epithehum of the crystalline lens thus obtained was transplanted intraperitoneally into other sala- manders, where it continued its growth, lost its similarity to lens epithe- lium, and finally came to resemble carcinoma. Reinke had found, also, that the mitosis educed by treatment with ether could be inhibited. He rubbed down salamander lens with ether, which he poured off and allowed to evaporate, afterward making a sahne emulsion of the residue so gained. This he inoculated into the eyes of salamanders that had undergone an injection of 4 % ether eight days previously and found subsequently no mitoses, or only a few at the most. The conclusion was accordingly drawn that there was present in the tissues a ma- terial which inhibited mitotic division and which could be extracted with ether. The continued action of ether upon normal cells dissolved this substance or altered it in such a way as to permit the occurrence of mitotic division or atypical growth ; but if the material were supphed 1 Jour. Med. Research, 1901, N.S., i, 145. 2 Deutsche tnedismal Zeitung, 1907, xxviii, 579. Munch, med. Woch., 1907, liv, 2381. ATTEMPTS TO PRODUCE TUMORS 27 in sufficient amount, it could be taken up anew by the cells and again exert its inhibitory action upon division. Askanazy/ stimulated by these experiments, mixed emulsions of rat embryo with 4 % ether before inoculation, and obtained a series of large rapidly growing teratoid tumors such as he had never seen following the inoculation of untreated embryos, even in pregnant animals. The fat stain, Scharlach R, which Fischer believed capable of exciting epithelial growth exerted, on the contrary, an inhibitory effect upon the proliferation of embryonic tissue. Freund- succeeded in producing teratoid tumors in 74% of rats that had been inoculated intraperitoneally with an emulsion of rat embryo. The age of the injected rats was immaterial, but the results were more uniformly successful in females. Previous treatment of the embryo tissues with ether water or with solutions of indol with or without ether had no effect upon their subsequent proliferation, nor was growth any more vigorous after autoplastic than after homoplastic transplantation. Finally, tumors followed the intraperitoneal in- oculation of animals in which subcutaneous injection had been un- successful, a fact proving that resistance had not been conferred by the single antecedent unsuccessful treatment. V. Hippel ^ observed the evolution of a teratoma from the head of a twelve-day-old rabbit embryp which had been injected as a saline emulsion into the eye of an adult rabbit. Six weeks after the begin- ning of the experiment, the tumor had reached a length of i centimeter and a thickness of 0.8 centimeter and was composed of elements of the outer and middle germinal layers, but contained none of the tissues of the hypoblast. Embryos were transplanted into adult animals by v. Tiesen- hausen ^ also. In his earlier experiments he made use of mammalian material, but in later investigations he employed the chick, because of the readiness with which embryos of known age could be procured from artificially incubated eggs. About one hundred and fifty inocu- ^ Wien. med. Woch., 1909, lix, 2518, 2578. CentralU.f. allg. Path., etc., 1909, xx, 1039. ^ Beitr. zur path. Anat., etc., (Ziegler), 1911, li, 490. ' Verhandl. d. deutschen path. Gesellsch., 1906, lo*^^ Tagung, 56. ^ Arch. f. path. Anat., etc., (Virchow), 1909, cxcv, 154. 28 ATTEMPTS TO PRODUCE TUMORS lations were made into young and old fowls at different sites with chicks varying in age from one to eight days. Five-day embryos gave the best results, whereas inoculation with those less than two days old was attended by a uniformly negative outcome. The sites chosen for implantation included the anterior and posterior chambers of the eye, the brain, the pectoral muscles, the anterior abdominal wall, the great omentum, the peritoneal cavity, the comb, and the wattles. In no case was progressive growth obtained, even in the birds from which the chick embryos had been derived, nor were any metastatic deposits observed. Growth went on actively for about eight weeks, but grad- ually ceased after from six to twelve months. Fichera ^ inoculated rats subcutaneously with a bouillon suspension of rat embryos, causing nodules which grew progressively until, at the expiration of two or three months, they were at least twenty times the original size. But by the end of six months their volume had greatly diminished, and they could not be considered in any sense as malignant new growths. When several successive inoculations had been made grafts became necrotic as early as the tenth day. The question whether pregnant animals might not perhaps offer a more favorable soil for the growth of embryonic tissue has not escaped investigation. Askanazy - and his pupil Jentzer ^ thought that preg- nancy in the host exerted a distinctly adjuvant effect upon the growth of experimental teratomata in white rats, and Fere ^ noticed increased growth in brooding hens ; but Shattock, Sehgmann, and Dudgeon,^ on the contrary, did not find that the proKferation of ingrafted fetal cartilage was stimulated in rabbits that repeatedly bore young. Rous ^ cited certain investigations of Fichera ^ which seemed to es- tabHsh the fact that embryonic grafts grew better in the rats from which they had been removed than they did either in non-pregnant or in other pregnant rats. The observations of Rous, which were ^ Arch, de Med. exp. et d^Anat. path., 1909, xxi, 617. 2 Verhatidl. d. deiitschen path. Gesellsch., 1907, 11 ^ Tagung, 82. ^ Rev. med. de la siiisse Romande, 1908, xxviii, 329. * Compt. rend. Soc. Biol., 1899, li, 824; 1900, Hi, 737. ^ Proc. Roy. Soc. Med., 1909-1910, iii, Path. Section, 132. ^ Proc. Soc. Exp. Biol, and Med., 1909-1910, vii, 71. Jour. Exp. Med., 1911, xiii, 248. ^ Policlinico, Sez. pratica, 1909, xvi, 692. Etiologia del cancro, Rome, 1909. ATTEMPTS TO PRODUCE TUMORS 29 undertaken independently of Fichera's, demonstrated that embryonic mouse tissue obtained at operation and implanted in the mother would grow well if she no longer carried young ; and although growth was no more rapid than in favorable non-pregnant aliens, yet it per- sisted for a longer period and resulted in a greater variety of tissues. On the other hand, when a mouse still bearing embryos was implanted with embryonic tissues from her own uterus the grafts failed either to grow or to differentiate, although they did become vascularized. Still, they did not die, and sometimes proliferation was inaugurated after the conclusion of pregnancy. Freund ^ concluded that pregnant rats offered a somewhat more suitable soil for the growth of embryo emulsions than did non-preg- nant females, although the difference amounted to but 2 %. Rous - investigated in mice the behavior of embryonal cells mixed with tumor, hoping that as some transplantable carcinomata were able to originate sarcomatous transformation in their stroma, so close asso- ciation with tumor cells might cause embryonic tissue to assume ma- lignant properties. He found that growth of both components took place and that there was often an intimate histological association of the two ; but for the occurrence of these conditions a balance of avidity was necessary, otherwise the embryonic cells were necessarily soon overgrown by those of the rnore rapidly proliferating tumor. It seemed, therefore, that the enormous proliferative capacity exhibited by embryonic tissue in lUero depended at least as much upon the ex- cellent nutritive arrangement as upon inherent cell energy. In a mixed graft that had only partially succeeded, tumor and embryo tended to grow or to fail together, while in a number of quantitative experiments it was found that tumor and embryo proliferated better alone than when mixed. Gougerot and Laroche ^ obtained a keloid in a tubercular guinea- pig at the point where an inoculation of tuberculin had been made. The ridge of growth was twenty-five milHmeters long and five milH- meters wide, appeared -to involve the needle track, and when excised 1 Beitr. zur path. Anat., etc., (Ziegler), 191 1, li, 490. ^ Proc. Soc. Exp. Biol, afid Med., 1909-1910, \-\i, 73. Jour. Exp. Med., 1911, xiii, 239. ^ Compt. rend. Soc. Biol., 1908, Ixv, 342. 30 ATTEMPTS TO PRODUCE TUMORS and examined, was found to consist of collagenous fibers inclosing a few cells. This case was the sole success in a long series of experi- ments upon fourteen guinea-pigs. The development of sarcoma in a rat that had been exposed repeatedly to X-rays was reported by Clunet.^ Fragments of the growth were transplanted into eight rats, in one of which there was found a nodule the size of a pea about two months after inoculation. Recession, however, set in immediately, and two months later the tumor had disappeared entirely. Considerable discussion has ensued upon Fischer's description ^ of a series of experiments in which he had succeeded in producing lesions imitating very closely the histological structure of carcinoma. After having failed to produce a suitable irritation by means of agar impreg- nated with calcium salts he tried a saturated solution of the stain Scharlach R in olive oil, injecting it into the ears of rabbits just beneath the epithelium and under considerable pressure. After a few days an augmented number of mitoses, of which a few were atypical, was detected in the germinal layer of the epithelium, hair follicles, and sebaceous glands, and there was an increase in the thickness of the whole epithelial layer with an attendant over-production of keratin. In later stages the epithelial proHferation was still more marked, hair follicles and sebaceous glands had either disappeared or become strands of squamous epithelium, and these structures then began to invade the underlying parts. When the intrusions reached the particles of oil lying in the connective tissue the epitheHum grew around the globules in an irregular way, producing a lesion which in appearance was comparable only to a squamous cell carcinoma in man. Such areas could not, however, be called true epitheUomata because they were not destructive and because, furthermore, growth did not continue after the oil had been absorbed. It was found that the two fat stains Sudan III and indophenol, chemically unrelated though they were, exerted a similar action. A series of experiments on the epithe- lium of the stomach, intestine, and mamma, had been inaugurated, but at the time of writing the subepithelial inflammation resulting from 1 Recherches exp. siir les Tumeurs malignes, Paris, 1910, 297. 2 Verhandl. d. deutschen path. Gesellsch., 1906, lo*^^ Tagung, 20, 22. Munch, med. Woch., 1906, liii, 2041. ATTEMPTS TO PRODUCE TLTIORS 3 1 the inoculation of solutions of Scharlach R produced no trace of epi- thelial growth in these locations. Other solutions of the dye were in- vestigated, among them a saturated ethereal solution which was injected into four of the mammary glands of a rabbit. In all four, numerous lobules were transformed into islets and strands of squamous epithe- lium of exactly the same shape as a gland lobule and connected with the duct. It seemed most probable that the parenchyma had been destroyed and then replaced through regeneration (often of duct epi- thelium), and that during the process there had resulted the substitu- tion of squamous epithehum for gland lobules. No such lesions were to be found in the mammae that had not been injected. To explain the action of Scharlach R Fischer suggested that there might be a specific substance — an attraxin — contained in it, able to exercise a posi- tive chemotactic influence on epitheHal cells. The growth of the epi- thehum about the oil globules could, at any rate, hardly be explained by the weU-known disposition of epithehum to grow along a free surface, for it did not take place about collections of other oils and fats ; and, moreover, if Scharlach R exerted only an ordinary stimulus upon the epithehum it was hard to see why the effects above set forth did not follow its external apphcation, even when this was continued for months. The demonstration of a chemotactic action on the part of the dye seemed to Fischer, therefore, the most significant feature of the experiment, and pursuing his h;ypothesis further he interpreted the hmitless growth of malignant tumors as the product of a hypotheti- cal attraxin occurring naturally and exerting its influence upon em- bryonal cell rests. McCoimell,^ in repeating these experiments with rabbits. Belgian hares, guinea-pigs, and white rats, found that Belgian hares were on the whole the most suitable animal in which to reproduce the lesions described by Fischer. Jores- pubHshed a general confirmation of Fischer's results in which, however, he laid somewhat more stress upon the role which was played by the germinal epithehum of the hair folhcles. Indeed, when moderate pressure had been employed the outcome of the injection was mainly a prohferation of this layer, and the h3'pertrophy of the surface epi- 1 Jour. American Med. Assoc, 1907, xlix, 1498. ^ Miinch. med. Woch., 1907, liv, 879. 32 ATTEMPTS TO PRODUCE TUMORS thelium occurred only when the oil had been injected under greater pressure. The sebaceous glands, at first unaffected, were gradually lost to \-iew amid the encroaching hair foUicles. If the epithehal growth exceeded a mere thickening of the hair foUicles a lesion of cancerous appearance was the result ; still, such a condition was difficult of attainment and Jores had seen it only once. He had never been able to reproduce it in scar tissue even after repeated injections, pos- sibly because of the absence of hair follicles. The process as a whole seemed to be the sum of two components — action b}^ the dye upon the upper part of the hair follicles, and proliferation of the epithehum which lay in direct contact with the oil globules. The latter phenom- enon one might be tempted at first to refer to the tendenc}^ of epi- thelium to grow over any surface. But particles of Scharlach oil were more readily inclosed than other foreign bodies, a fact which, in con- junction with the characteristic action of the stain upon the hair foUicles, led the author to assign to the dye an actual influence upon squamous epithelium, although it might not be chemotactic. It was very unlikely, too, that Scharlach R exerted any real stimulus to growth, and the proliferation attending its injection might be considered with more justification as a replacement process exceeding, in the manner described by Weigert, the bounds of physiological regeneration. Hehnholz^ confirmed and extended the experiments begun by Fischer, and proved that the eff'ects of Sudan III and Scharlach R were exercised not only upon various t^pes of epithehum but also upon those connective tissues which produce cartUage. He did not adopt, however, the conclusions which Fischer had drawn, and expressed the behef. not that the dyes possessed a specific attraxin. but rather that they contained some substance which, by interaction with con- nective tissue, produced a soU appropriate for epithehal develop- ment. In contrast to the foregoing workers Snow - could not substantiate Fischer's observ-ations. although it should be noted that his injections were not made under the high pressure upon which Fischer had insisted. ^ Johns Hopkins Hasp. BidL, 1907, xvnii, 365, 369. ^ Jour. Infect. Diseases, 1907, iv, 3S5. ATTEMPTS TO PRODUCE TUMORS 33 Stahr/ on the other hand, was able to corroborate Fischer, but only by following strictly the technic which that author had described ; wherefore he emphasized the necessity of injecting the oil under very high pressure. He found that atypical epithelial growth would follow even a single injection of a saturated solution of Sudan III in oHve oil. While Stahr's results were thus confirmatory, he disagreed with Fischer's theory of a specific chemotactic action and was inclined to refer the lesions to the co-result of a stimulus plus a complex of condi- tions, not the least among the latter being the anatomical structure of the tissue into which the dye had been injected. Wyss- had concluded from observations on X-ray carcinomata that such tumors were consequent upon altered nutrition, and this upon a narrowing or obliteration of the subepithelial blood vessels supplying a certain group of cells. Areas deprived of nourishment in this way underwent proliferation and finally became parasitic on the neighboring tissues. Wyss believed that the proliferation described by Fischer was not due to chemotactic influences but to a withdrawal of normal nutrition through the obhteration of blood vessels, and emphasized the fact that the most striking results were seen after injection had been made at high pressure. Levin ^ investigated in white rats the action of a saturated solution of Scharlach R in paraffin or oil when injected subcutaneously or into the mamma. Only in the connective tissue elements could any proHf- eration be demonstrated, an observation well in accord with a previous finding by the same author,^ that the rat was prone to vigorous reac- tions on the part of its connective tissues. Seyberth '" confirmed the observation of Rehn, that tumors of the bladder were not infrequent among those who worked in anihne dye factories, and recorded five such cases, three non-malignant although showing excessive epithehal prohferation, and two mahgnant. Of the latter, one was an adeno-carcinoma and the other a carcinoma. These growths, the author beheved, were all without doubt the prod- ^ Miinch. med. Woch., 1907, liv, 11 78. ^ Miinch. med. Woch., 1907, liv, 1576. ^ Jour. Exp. Med., 1908, x, 815. ^ Med. Record, 1907, Ixxii, 974. ^ Miinch. med. Woch., 1907, liv, 1573. 34 ATTEMPTS TO PRODUCE TUMORS uct of one cause, namely, chronic stimulation of the bladder wall by urine containing aniline bodies, and Fischer's experiments had thrown a significant light upon the relations between aniline dyes and epi- thelial proliferation. Stoeber^ experimented with certain components of Scharlach R and Sudan III, all of which were found to elicit epithehal prohferation. Some of these, however, which were toxic, and others which caused necrosis, were of course unsuitable for further investigation. Con- trary to Fischer's behef the chief source of the new epithehum was not the surface layer but,, as Jores had contended, the germinal epithelium of the hair follicles. The same writer ^ recorded the next year an ex- periment in which oily solutions of Scharlach R, amidoazo toluol, and a-naphthylamine had been inoculated into the foot of an old man with ununited fracture of the leg, before amputation. When this was performed fourteen days later it was found that Scharlach R and amidoazotoluol had induced epithelial changes identical with those evolved in the rabbit's ear. They were not so extensive, however, partly because hair follicles and sebaceous glands were only sparsely distributed in the inoculated region. The overlying epithelium was thickened, and the down-growing projections appeared to have been derived from thickened hair folKcles and the ducts of sweat-glands. No sebaceous glands were recognized in the sections. In conjunction with Wacker,^ Stoeber sought to discover what effect would follow the injection of various products found in the body under physiological or pathological conditions, and especially of such as resulted from the splitting and decomposition of albumins. It was apparent that only organic substances of a basic nature and soluble in fats, had any definite action upon epithehum. Employing Fischer's technic, but using small amounts and making two injections with two or three days intervening, the authors obtained epithelial proliferation with 2 and 5 % solutions of pyridin in olive oil, and with 5 % indol or skatol in rabbit fat. The inoculation of indol solution followed by a small amount of skatol produced a tumor which attained the size of a hazel-nut within fifteen days. The histology of the lesions provoked by pyridin were less characteristic than in the case of indol and skatol. ^ Miinch. med. Woch., 1909, Ivi, 129. ^ Milnch. med. Woch., 1910, Ivii, 739. ^ Munch, med. Woch., 1910, Ivii, 947. ATTEMPTS TO PRODUCE TUMORS 35 The picture produced by the two last mentioned was similar to that of human squamous cell carcinoma, so similar, indeed, that it would have been dithcult for even the most skilled diagnostician to distinguish one from the other. ]Meyer ^ injected Scharlach oil into the renal arteries, under the cap- sule, and into the parenchyma of the kidney, in a series of rabbits and dogs, but without being able to produce any effect upon the renal epithelium. In certain other experiments, in which inoculations had been made under the skin, he found that when a vein or an artery had been ligated the action of the stain was hastened, and that under these conditions even oil alone, or oil and paraffin, would inaugurate epithe- lial proHferation. The growth of epitheHum in the rabbit's ear was, therefore, according to this author, due to chronic inflammation asso- ciated with other circulatory disturbances. This view was challenged by Greischer,^ who, working with Schmincke, induced anemia in one ear of each of a series of rabbits by ligating the carotid, arterial hyperemia in others by cutting the cervical sympathetic, or finally, venous hyperemia by the hgation of veins at the base of the ear. The lesions accompanying inoculation of the operated ear were always the same as those obtaining in the normal one of the opposite side, and epithelial proliferation could never be dis- covered in the prepared ear when it had not occurred in the normal one. Disturbances of circulation, therefore, played a role only in so far as they shortened or protracted the length of time during which the injected material could act upon the epithelium. White ^ introduced oleic acid, either pure or with the addition of carbon dioxide or methyl oxalate, into the ears of rabbits, the mammary glands of guinea-pigs, and the backs of mice. Notwithstanding the fact that his findings corresponded so exactly with those of Fischer that, as he himself said, the illustrations pubhshed by that author might have been prepared from his own specimens, White could not find the slightest evidence of any chemotactic influence. Thus, there was no intrusion by the epithelium, and hair folHcles outside the inflam- 1 Beitr. zur path. Anal., etc., (Ziegler), 1909, xlvi, 437. '^Inaugural Disserlation, Miinchen, 1911. Miinch. med. Woch., 1911, Iviii, 1608. * Jour. Path, and Bad., 1910, xiv, 450. 36 ATTEMPTS TO PRODUCE TUMORS matory zone remained entirely unaltered ; and although it was true that epithelial cells lined the abscesses, this was merely an expres- sion of the normal tendency of these elements to grow over a free surface and could not be attributed to any specific property of the fatty acids employed. The injected materials appeared to act merely as chronic irritants. Although White's results were negative with oleic acid in liquid paraffin and with a mixture of oleic acid and cholesterin, a positive action was claimed for these substances by Wacker and Schmincke.^ Benthin^ injected the ears of animals, usually rabbits, with paraffin, oHve oil, agar, agar with calcium carbonate and calcium phosphate, Sudan III, Scharlach R, indophenol, amidoazobenzol, paratoluidin, amidoazotoluol, a-naphthylamine, indol, skatol, paraffin with soot, gum arable alone and with Sudan III, oleic acid, and glycerine. No growth was obtained after olive oil, agar, or gum arabic alone, and none following agar with calcium carbonate or calcium phosphate, while the injection of parafiin and soot gave rise only to epithelial cysts. Oleic acid, glycerine, paratoluidin, and amidoazobenzol were unsuitable, for, as Stoeber had already pointed out, they caused too much necrosis. The most marked growth accompanied the use of Sudan III and Schar- lach R, and Fischer's statements regarding the action of these dyes were verified by Ben thin, who, nevertheless, did not believe that the far-reach- ing conclusions drawn by that author could be justified. The growth of epithehal cells around the oil droplets seemed to occur only when there was contact between the two, and deeply lying masses of oil, even though of considerable size, did not become inclosed by epithelium. The author ascribed the epithelial proliferation to a complex of causes among which inflammation, tissue tension, perhaps circulatory distur- bances, and, most important of all, the presence of certain chemicals, were jointly responsible. It was noticed that wounds following opera- tive removal of the injected areas healed with thick keloid-like scars as though the fat stains had the property of stimulating growth. Wacker and Schmincke,^ reviewing the various materials used by themselves and others to educe epithelial proHferation, pointed out ^ Miinch. med. Woch., 1911, Iviii, 1607. ^ Zeitschrift f. Krebsforsch., 1910-1911, x, 227. ^ Miinch. med. Woch., 1911, Iviii, 1607. ATTEMPTS TO PRODUCE TUMORS 37 that all which had afforded a positive result had this factor in common, that they possessed the solubilities of Hpoids — produced, that is, a physical change in the Hpoid membrane of the epithelial cell. Still, a few materials were unable to inaugurate proliferation, possessed of this property though they were. That the stimulating action of the various agents was not primarily a chemical process, in the sense of a splitting up of the lipoid membrane surrounding the cells, the authors inferred from the following experiment. Bergel ^ having already dem- onstrated the presence of a fat-splitting ferment in lymphocytes. Wacker and his associate injected the ears of rabbits with emulsions of rabbit spleen, lymph nodes, or pleural lymphocytic effusions, without, however, being able to find any evidence of the epithehal growth which should have taken place were lipoidolysis a factor in its initiation. McConnell - tested the question whether Sudan III dissolved in oil would exert any influence upon the cells of an epithehal tumor. Hu- man carcinoma was suspended in sahne solution and injected into SLxteen white rats, of which eight were subsequently inoculated at the same site with three minims of Sudan oil. No growth took place either in the inoculated animals or in the controls. Werner^ investigated the effect of Scharlach R upon growing mouse tumors and found that concentrated solutions in oil excited growth, while alcoholic solutions caused the tumor at first to shrink and finally to slough away. The excitation of growth caused by the oily solution was not the result of chemotactic influences, but of a true growth stimu- lus. Powdered Scharlach R was added to tumor emulsions before in- oculation by Albrecht and Hecht,^ to see whether the vigor of growth would be augmented. In contrast to Werner's findings regarding tumors already established, the dye neither incited the cells to increased proliferation nor did it inhibit their growth. A review of certain aspects of the more recent attempts to estabHsh mahgnant growth has been pubhshed by Herxheimer and Reinke.^ ^ Miinch. med. Woch., 1909, Ivi, 64. Miinch. tJted. Woch., 1910, Ivii, 1683. ^ Jour. Med. Research, 1908, N.S., xiii, 381. ^ Munch, med. Woch., 1908, Iv, 2267. * Cenfralbl. f. allg. Path., etc., 1909, xx, 1039. ^ Ergebnisse der dig. Path., etc. (Lubarsch and Ostertag), Abt. ii, 1909, xiii, 416. 38 ATTEMPTS TO PRODUCE TUMORS The effect of ether upon the cells of the salamander, discovered by Reinke, corresponded entirely with experiments in plants where it had been found that hibernation could be overcome with ether, and premature blossoming brought about. It was further parallel with hemolysis, which was due to solution of the lipoids of red blood cells, and Reinke had therefore suggested that blastosis, or stimulation to mitotic cell division, might perhaps be interpreted as the final outcome of a melting, saponification, or solution of hpoids in the cells and their nuclei, taking place in the same way as in hemolysis. The sug- gestion was strengthened by Askanazy's observation of the increased growth power of embryo rat tissue after exposure to ether. Somewhat later, and independently of Reinke, J. Loeb had reached similar con- clusions, deciding that it was the fat-dissolving properties of the substances with which he was working that enabled him to achieve artificial parthenogenesis. Furthermore, Scharlach oil and the afhhated substances that had been utihzed by Fischer and his successors were also solvents of Hpoids, a circumstance that could be correlated with Overton's theory of a hpoid ectoplasm. Although any serious injury to the cell was followed by cytolysis, as could be demonstrated most easily through the appHcation of heat, if the action of this agent were not so intense as to lead to destruction the protoplasm became more active, metaboHsm was increased, ferments were set working, and latent vital processes were awakened. Ether caused a change in the condition of those important elements, the lipoids, whose role in cell division could be readily harmonized with Weigert's conception of proKferation, according to which tissue injury was interpolated between interference with the cell and subsequent growth. Herxheimer and Reinke could see no reason, therefore, to abandon Weigert's view that there could be no direct stimulation of cell growth, and it seemed justifiable to ac- cept with him the existence of an external functional stimulus, but to refer growth to a preceding injury. CHAPTER III EARLIER OBSERVATIONS ON THE TRANSMISSIBILITY OF CANCER Long before it was realized that the question of the transmissibility of cancer could be approached through experiment upon the lower animals, its solution had been attempted by means of the much less exact method of cUnical observation ; and although numerous in- stances of transfer of the disease have been recorded, most of them so seriously exceed the limits of creduHty that the occurrence of cancer in the two patients may be attributed to chance. Butlin^ has laid down with admirable clearness certain conditions which reasonably may be demanded before it is agreed that any given case illustrates the transmission of malignant disease from one person to another. " . . . all cases of reported contagion of cancer in which the disease is not of the same variety must be unhesitatingly rejected ; . . . cases must not be accepted where there is no evidence that the affected parts of the two individuals were from time to time in con- tact; ... it is extremely improbable that contagion should take place from a covered cancer, such, for instance, as a cancer buried in the breast ; or that the disease could be implanted in a part the covering of which has not been broken." . A few cases of the reputed transmission of cancer may now be examined to see in how far they conform to Butlin's requirements. TRANSFER FROM ONE PERSON TO ANOTHER Accidental Zacutus - reported the following case, perhaps the first to appear in the literature. "A poor woman who had suffered for many years from an ulcerated cancer of the breast, slept at night on the same - ^ British Med. Jour., 1907, ii, 255. - Praxis MedicaAdmiranda, p. 3i,obs. 124 (Appendix to Opera, tome ii, Lugduni, 1649). 39 40 E.IRLIER OBSERVATIONS ON THE TRANSMISSIBILITY OF CANCER couch with her three sons. All three became affected with a like disease. Five years after the death of the mother, two of the sons died ; but the third, being more robust, recovered with difficulty after having submitted to excision of the cancer at the hands of a surgeon." Tulpius ^ expressed surprise that there should exist any doubt as to the contagiousness of ulcerating cancer, and related a case that had come within his own experience. "Adriana Lambert, a woman advanced in years, was afflicted with a cancer of the breast that had advanced to such a putrid condition as to infect by its exhalations her body-servant who attended upon her, living in close contact with the mistress. Some evil spark lighted such a conflagration as to destroy the maid no less than her mistress, for such a foul, irregular ulcer invaded the breast and armpit of each that I am undecided which of these two was tortured with the greater savagery." Juncker - believed that cancer was contagious, but not virulently so unless it were implanted in situations suitable for its reception, and especially in a place where solution of continuity of the tissues had been effected. That there existed in Juncker's day a popular belief in the con- tagiousness of cancer is shown by an episode which Guelliot ^ has reported. Cancer patients having been refused admission at the Hotel Dieu in Reims, Jean Godinot, a canon celebrated for his charity, offered the sum of twenty-five thousand livres for the construction of a hospital for such patients, who until then had been able to subsist only through what they could gain from compassionate worshippers by exposing their sores at the church door. A house was bought in the center of the town and opened as the Hopital Saint-Louis, and here cancer patients were cared for, but on condition that they were not to leave the building. The neighbors, however, complained that, as cancer was contagious, the health of the public would be endangered if such an estabhshment were tolerated in a populous quarter, and the administration was finally compelled to transport its charges to a lazaret formerly used for idctims of the pest. It was not until 1841 ^ Observationes Medicae, Amstelredami, 1672, editio nova, 292. ^ Conspectus Chirurgiae tarn Medicae, Halle, 1721, 327, ^ Gaz. des Hopit., 1892, Ixv, 1210. EARLIER OBSERVATIONS ON THE TRANSMISSIBILITY OF CANCER 4 1 that cancer cases were admitted to a general hospital, and even then they were kept in a separate ward. Viel-Hautmesnil ^ reported the three following cases as apparent instances of the transmission of cancer. Mr. Smith, of St. Thomas's Hospital in London, tasted the contents of a small cyst in a mammary cancer which he had just removed. The acrid taste remained per- manently in his mouth, and he became subject to attacks of vomiting which ultimately determined his death. A doctor was said to have died of cancer which had apparently been communicated to him by his wife. The third example was a case reported by Tulpius, con- cerning a surgeon whose wife suffered from mammary cancer. Ac- customed to apply his mouth and tongue to the affected part in order to relieve the intense pain, he soon afterward died of a terrible cancer which destroyed the interior of his mouth. The foregoing cases of transference are interesting, of course, only in so far as they show the age of the question now under discussion, for before the time when the microscope had come into general use the diagnosis of cancer was often so highly problematical that reports of transmission of the disease may be dismissed with scant curtesy. Still, even after the importance of histological diagnosis had been es- tabhshed, examples of contagion continued to be described, most of them, however, on a par with the earlier instances through the neglect of the observers to advance adequate microscopic justification for their claims. Budd - has related the case of a strong and healthy girl of nineteen who regularly washed dressings and linen soaked with the discharge from an advanced case of uterine and vaginal cancer. Six months after the death of the patient the young woman was admitted to the hospital with a large axillary cancer, which eventually proved fatal. Budd noted, further, that five surgeons of the same institution had died of cancer within a few years, and expressed the opinion that such a mortality could hardly be conceived except on the supposition that the disease had been communicated, at least to some of them, during manipulations on patients sufifering from cancer. Whitehead ^ had under his care a man with an extensive epithe- 1 Consid. gen. medico-chir . sur le Cancer; These de Paris, 1807, 24. ^ Lancet, 1887, ii, 1091. ^ Lancet, 1887, ii, 1040. 42 EARLIER OBSERVATIONS ON THE TRANSMISSIBILITY OF CANCER lioma of the lower lip, attributed by the patient to his habit of drinking daily out of a small vessel used by his father, who had died during the previous year of the same disease. Adam ^ described the case of a medical man who married a woman subsequently found to be the victim of mammary and uterine cancer, from which she soon afterward died. Within a short time the widower had developed cancer of the liver and stomach, but before his death he married again, and a year or two afterward' it was discovered that his widow was the subject of cancer of the breast. Smith ^ reported a woman with sloughing carcinoma of the mamma, whose husband was attacked by cancer of the stomach, and Behla ^ collected fourteen examples of contagion which had occurred among patients under his own observation. The first in the series was a man of seventy-six who died of carcinoma of the kidney three years after his wife had died of cancer of the liver. The other thirteen instances were similar. To these cases of cancer a deux, Behla added one of cancer a trois which had been communicated to him by Elsler. Herr B, who died of carcinoma of the rectum, was nursed by his son-in-law who for six months administered daily nutrient enemata. A short time after the death of the patient the son-in-law developed a carcinoma of the lip, and during his illness a cancer of the breast was discovered in his wife. The following examples of " cancer houses " have been drawn frorn Behla's monograph.^ In the village of Breitenbach, with a population of four hundred and fifty, four persons died of cancer in 1906, among them a man with a can- cer of the lip who lived in the house where a man had died of carcinoma of the esophagus in 1905. A citizen of Rehfeld died of cancer of the stomach, and the man to whom the house was sold selected for his living room the one in which the previous owner had died, erecting his bed in the place where that of his predecessor had formerly stood. Although he had been in perfect health before moving into this house, the new tenant died after about a. year from carcinoma of the stomach. ^Lancet, 1887, ii, 766. 2 Med. Record, 1895, xlviii, 383. ^ Deut. med. Woch., 1901, xxvii, 427. ^ Die Bestdtigung d. kunstlichen Zuchtung des Krebserregers, Berlin, 1910, ly. EARLIER OBSERVATIONS ON THE TRANSMISSIBILITY OF CANCER 43 Newsholme ^ has criticized the usual pubhshed accounts of cancer houses on the score that the writers always forgot that they were deahng with small figures ; they forget, for instance, that if a number of persons were set to tossing coins, it would happen eventually that one of them would toss a thousand heads without a break. One who remembered the occurrence of such coincidences as this would attach little importance to numbers of cases in particular houses. The same idea has been expressed also by Adami,- who said : ^'. . .by the law of chance, just as one individual in a thousand may be of gigantic proportions, so one house in a thousand may show a great excess of cases of cancer — or of twin births — over the or- dinary run of houses." Experimental Although the attempt has been made to decide, by means of ex- perimental inoculation, the question regarding the transmission of tumors from one person to another, the number of these experi- ments is so small that no deduction can with any safety be drawn from them. On October 17, 1808, at the Hopital de St. Louis, and in the pres- ence of several physicians and students, Alibert ^ allowed himself to be injected with ichorous material from a cancer of the breast. The experiment was performed at the same time upon M. Fayet, a medi- cal student, and the next morning upon MM. Lenoble and Durand. Except for an inflammatory reaction the experiment was without sequelcB. A week later Alibert inoculated himself a second time, and his colleague, M. Biett. He himself escaped with a result similar to that which followed the first trial, but M. Biett developed a somewhat more severe infection, which involved the axillary and cervical lymph nodes. Senn ^ transplanted into himself a fragment of lymph node in which carcinomatous invasion had been demonstrated by the mi- croscope. Although a nodule the size of a pea appeared at the implan- tation site and remained stationary for two weeks, it vanished soon ^ Trans. Epidemiological Sac. London, 1906-1907, N.S., xxvi, 73. ^ Principles of Pathology, Philadelphia and New York, igio, Vol. i, 839. ' Description des Maladies de la Peaii, Paris, 1825, 118. * Jour. American Med. Assoc, 1901, xxxvii, 811. 44 E.4RLIER OBSERVATIONS ON THE TRANSMISSIBILITY OF CANCER afterward, and seven weeks later a red linear scar was the only indica- tion of the experiment. Lanz ^ inoculated his gardener on the back of the hand with finely minced common warts (verruca vulgaris) removed from another persons- arranging the series in the shape of the letter "J." The first sign of growth did not appear until fully one and a half months had elapsed, when two or three very small nodules were recognizable. Two months after implantation there were eight warts forming a "J " and each about the size of a pinhead, while a month later the number had in- creased to twelve. Lanz said, further, that as a result of rubbing down the warts of a patient, he himself had acquired several of these tumors. TRANSPLANTATION OF TUMORS INTO THEIR BEARERS Accidental Although it may be true that tumors can be transplanted by con- tact under certain conditions, not a few of the instances that have been reported to illustrate this type of implantation may have been cases either of multiple tumors or of retrograde metastasis ; and the amount of care which is necessary to eliminate error is well shown in an article by Petersen.^ In a case of inoperable carcinoma of the uterus, the skin in the neighborhood of the vulva contained a number of small nodules and a few indurated ulcers which seemed clearly enough to be the outcome of auto-inoculation. Examination of serial sections, however, proved definitely that they were the result of retrograde lymphatic metastasis. Full as the Hterature is of examples of reputed transfer by contact, there are but few which satisfy Ewing's demand ^ that "... the transferred tumor shall exhibit a structure similar to that of the orig- inal, but different from that spontaneously arising in the invaded tissue." A useful set of standards for the guidance of those engaged in the critical study of these cases has been formulated by ButHn ^ as follows : "The disease must be of the same variety in the primary carcinoma 1 Dent. med. Woch., 1899, xxv, 313. 2 ^^.^/^ y Dermat, u. Syph., 1904, Ixx, 313. 3 Arch. Internal Med., 1908, i, 177. ^ British Med. Jour., 1907, ii, 256. EARLIER OBSERVATIONS ON THE TRANSMISSIBILITY OF CANCER 45 and in the reputed contact-cancer. The identity of the disease must be proved by microscopical examination. The primary disease must have been exposed at the time at which the contact is known to have taken place ; and there should be such evidence of contact of the pri- mary carcinoma with the seat of the reputed contact-cancer as would satisfy a jury." The necessity for microscopical control was illustrated by an anec- dote. A specimen that had been exhibited as an example of contact transfer having been finally cut and examined, it was found that an ulcer on the cheek immediately opposite a carcinoma of the gum was a simple inflammatory lesion. Klebs ^ described three instances of squamous cell epithelioma of the stomach, two of which were, in his opinion, undoubtedly due to infection of the mucous membrane by particles implanted on it, while the third had probably been so produced. The primary tumors involved respectively the esophagus, the face, mouth, and throat, and the tongue. Kaufmann ^ recorded the case of a woman with an ulcerating epi- theHoma on the dorsal surface of the right hand, who was in the habit of rubbing her eye with the back of the afifected member. About three years after the discovery of the tumor a small ulcerating growth developed on the inner side of the right lower lid. Both ulcers had all the microscopic characteristics of a cancroid. Kraske ^ had seen two patients with primary cylindrical cell car- cinoma situated high up in the rectum, who had developed similar growths close to the anus and separated by a considerable distance of healthy tissue from the higher tumor. The lowermost were in the region normally covered by squamous epithehum, a circumstance which seemed to exclude the possibility of their having been primary, V. Bergmann ^ discussed a patient with an ulcerating epithe- lioma on the lower lip and a tumor of the same type on the upper at a precisely corresponding point. The former growth had appeared three months before the man came under observation and the latter six or seven weeks later, but not until after the lower tumor had ul- ^ Handbuch d. path. AnaL, Berlin, 1868, Bd. i, erste Abt., 190. "^ Arch. f. path. Anat., etc., (Virchow), 1879, Ixxv, 317. ' Centralbl.f. Chir., 1884, xi, 801. '^ Berl. klin. Woch., 1887, xxiv, 891. 46 EARLIER OBSERVATIONS ON THE TRANSMISSIBILITY OF CANCER cerated. In answer to an inquiry from Butlin^ the author replied that the growths had been subjected to microscopical examination and found to be epitheliomata. A woman under the observation of Hamburger ^ was admitted to hospital with an epithelioma of the left labium minus about the size of a hen's egg, first noticed two years before. On admission there was discovered at a corresponding point on the right labium minus a small, raised, ulcerating tumor which the history stated was of only two months' duration. Microscopical examination proved that both growths were squamous cell epitheliomata. Thorn ^ found a carcinoma of the cervix uteri coupled with a car- cinomatous ulcer in the left side of the vagina, but without involve- ment of the fornix. Perimetritic adhesions had drawn the uterus to the right, tilting an exceptionally long cervix to the left so that the cer- vical lesion had been kept in constant contact with the left vaginal wall. Both tumors presented the same histological structure. In a second case, a carcinoma involved both right labia and a shallow carcinoma- tous ulcer occupied a corresponding point on the left side, although the tissues intervening had remained intact. No statement was made regarding the structure of the tumors in the latter case. Butlin ^ collected two cases in which an epithelioma had arisen on one labium at the exact place of contact with a similar tumor on the other side, and one where a labial tumor opposite to an epithelioma was not malignant. He further reported three examples of apparent transfer from one side of the larynx to the other, and a fourth instance in which an ulcer of one vocal cord immediately opposite to an epi- thelioma on the other was demonstrated by microscopical examination to be merely an inflammatory lesion. Among various examples re- garding other parts of the body, there was described an ulcerated epithelioma of the lower Hp in a patient who had the habit of pressing his forefinger against the ulcer and then rubbing it on his nose. Twenty months after the excision of the tumor the patient presented a large ulcerated patch on the nose. This lesion, first noticed three months after the operation on the lip, was proved by the microscope to be of identical structure with the first tumor. ^ British Med. Jour., 1907, ii, 259. 2 Hospilahtidende, 1892, x, 81. ^ Ceniralbl. f. Gyndkol., 1894, xviii, 228. '^British Med. Jour., 1907, ii, 257. EARLIER OBSERVATIONS ON TKE TRANSMISSIBILITY OF CANCER 47 Secondary growths of the Fallopian tube arising by implantation of cells from malignant tumors in the abdominal organs have been described by many authors, one of the most recent being Wakasugi.^ The nodules are found on the outer aspect of the tube, or, as occurs not infrequently in the case of ovarian carcinoma, in the mucous membrane. It has often been observed that particles of tumor, eluding the vigilance of the surgeon at operation, have been spread through the wound to demonstrate later in the most unhappy way the possi- bility of ingrafting a tumor into the indi\ddual in whom it has orig- inated. Thus Quincke - recorded a case of peritoneal carcinosis in which, as the result of abdominal puncture for ascites, there was found in the subcutaneous tissue at autopsy a carcinoma identical in structure with the peritoneal nodules, although the intervening tissues were free from growth. The secondary tumor had been clinically recog- nizable as early as the tenth day after puncture. Thorn ^ saw a patient who had submitted two years pre^'iously to vaginal hysterectomy, during which lateral incisions had been made to overcome the narrowness of the vaginal orifice. Renewed hemor- rhage necessitating subsequent examination, a small recurrence was discovered in the scar, and upon microscopical examination it was shown that this nodule exactly resembled the original carcinoma of the cervix. A second and similar instance of implantation was de- scribed by the same author. Lack ^ discussed two cases in which the rupture of a carcinomatous lymph node during operation was followed by recurrence in the scar, a third where recurrence supervened at the site of a tracheotomy wound after excision of a sarcoma of the lower jaw, and two examples of apparent implantation in tracheal wounds during operation for cancer of the larynx. Richardson '^ reported a patient wdth carcinoma of the breast in whom an exploratory puncture had been performed four weeks previ- ^ Beitr. zur. path. Anat., etc. (Ziegler), 1910, xlvii, 483. ^ DeuL Arch. f. klin. Med., 1875, ^vi, 134. ^ Centralbl. f. Gyndkol., 1894, xviii, 228. ^Lancet, 1896, i, 1638. ^ Boston Med. and Surg. Jour., 1898, cxxix, 414. 48 EARLIER OBSERVATIONS OX THE TR.\XSMISSIBILITY OF CAXCER oush'. At operation a small nodule was found in the pectoralis minor, and the entire mass was therefore thoroughly removed. ''The examination of the tumor and the pectoralis major showed a clear space between the two. A nodule was found among the fibers of the pectorahs major corresponding in position and size to that in the pectoralis minor, and in the same general line ^^dth the two just de- scribed there was a stellate maHgnant mass in the axillary fat, not connected mth the enlarged lymph nodes. " The infection of the pectorahs major, the pectorahs minor, and the axilla, in a straight line, shows, I think, a direct contamination of pre- viously healthy parts by the exploring punch, for the nodules were of about the same size and age, and in the muscles, at least, they were the only ones found." Discussing the transplantation of tumors into their bearers, Richardson added that he had seen the "... stitch- holes of a closed abdominal wound, after nephrectomy for sarcoma of the kidney, burst out with luxuriant masses of recurrent disease." Experimental The problem whether or not a tumor can be transplanted success- fully into its bearer has even been subjected to experimental investi- gation, although, as E^^'ing ^ has said: "Xo experimental e\ddence is needed to show that a mahgnant tumor may often be grafted froni one part of the patient's body to another, since the several recognized modes of metastasis daily demonstrate this process." Hahn - seems to have been the first to approach this question from the experimental standpoint. In a case of inoperable recurrence of a marmnary carcinoma, three portions of infiltrated skin were trans- planted into the sound integument overlying the normal breast on the opposite side. About eleven weeks after the operation all three grafts had developed into nodules which were demonstrated by the microscope to be carcinomata. Mrchow ^ objected, however, that this experiment did not demonstrate the transplantation of a tumor into its bearer, because, as the entire skin had been transferred, 1 Arch. Internal Med., 1908, i, 176. 2 Berl. klin. Woch., 1887, xxiv, 892. BerL klin. Woch., 1888, xxv, 413. See also Frank, Dent. med. Woch., 1891, r\ii, 933. ^ Berl. klin. Woch., 1887, xxiv, 892. EARLIER OBSER\'ATIOXS ON THE TRANSMISSIBILITY OP CANCER 49 the tumor had grown in its new location only in so far as it was situ- ated in the implanted integument. The case, therefore, was entirely analogous to the ingrafting of a portion of skin containing hair, which, even though successful, could not be held to have demonstrated the transplantabihty of hair. Senn ^ transferred a fragment of epithelioma to the subcutaneous tissue of the patient's leg, but nothing remained of the graft after the expiration of four weeks. Cornil - reported two cases of the inoculation of a tumor into its bearer, by a surgeon whose name was withheld. In the first instance a graft from a spindle cell sarcoma of the mamma, inserted into the normal breast on the other side, had reached the size of an almond at the end of two months. This nodule, when removed and subjected to microscopical examination, proved to be a spindle cell sarcoma similar to the primary' tumor. The patient died somewhat later of an intercurrent disease, but no secondary tumors were demonstrable at autopsy in spite of the most painstaking search. The second case was similar except that the implanted growth was an adeno-carcinoma. The secondary nodule, while it had all the clinical characteristics of a new growth, was not removed, on account of the patient's reluctance to submit to another operation, and microscopical proof of its nature could not, therefore, be advanced. Thorn ^ wrote that in six cases of inoperable carcinoma of the uterus he had made about twenty implantations in the respective patients, either by se^\'ing small pieces of growth into the mucous membrane of the vagina or by rubbing fragments into small vaginal wounds. Although one of these grafts seemed to proHferate for three weeks it disappeared soon afterward, and the outcome in all the other experi- ments had been negative. To the subject of implantation cancer as a whole Milner ^ has con- tributed an exhaustive critical review which includes two hundred references. 1 Surgical Bacteriology, Philadelphia, 1889, 261. ^ Semaine med., 1891, xi, 259. Bull, de I'Acad. de Med., 1891, xxv, 906. ^ Centralbl. f. Gyndkol., 1894, xviii, 228. ^Arch.f. klin. Chir., (v. Langenbeck), 1904, Lxxiv, 669, 1009. 50 EARLIER OBSERVATIONS ON THE TRANSMISSIBILITY OF CANCER ATTEMPTS TO TRANSFER HUMAN TUMORS TO ANIMALS Since the day when Peyrilhe ^ made the first recorded experiment, attempts to transfer cancer from man to the lower animals have been almost continuously in progress, and although a positive result has been reported more than once, the consensus of opinion has been for many years that such claims cannot be seriously entertained. Peyrilhe's account of his investigation appears in the English trans- lation of the Dissertation in the following words : "I will relate here an experiment which I myself made with the cancerous virus. I procured about two drachms of it from a cancerous breast, and introduced it by means of a syringe, into a small wound made in the back of a .dog. I covered the wound with a plaster and bandage, and in three days removed the dressing ; the retraction of the skin afforded an ulcer, which already afforded a very disagreeable smell : it was of a dark violet color, and the parts all around it were emphy- sematous. I covered it again with the same plaster, and in forty- eight hours opened it again for the second time. The effects were then more violent. The whole skin from, the head to the tail was completely emphysematous. A little ichorous blackish matter flowed from the wound. The eyes of the animal were vivid, and he seemed to have a great thirst : in this state the poor creature was perpetually howling. At length my maid, disgusted by the stench of the ulcer, and softened by the cries of the animal, put an end to his life, and thus prevented my observing the ultimate effects of this disease." Dupuytren ^ fed animals with cancerous material, introduced it into the abdomen, injected cancer juice into the peritoneal cavity and the veins, and inoculated the pus of an ulcerated cancer, but without succeeding in transmitting the disease. Langenbeck ^ injected into the femoral vein of a dog the juice from a medullary carcinoma of the humerus, mixed with the dog's own serum, and two months afterward, when the animal was autopsied, 1 Disscrtatio Academica de Cancro, Antverpiae, 1775. A Dissertation on Cancerotis Disease, London, 1775. 2 Cited by Viel-Hautmesnil, Considerations generales medico-chirurgicales sur le Cancer, Paris, 1807, 23. ^ Schtnidt's Jahrbticher, 1840, xxv, 99. EARLIER OBSERVATIONS ON THE TRANSMISSIBILITY OF CANCER 51 several small, round, bluish nodules were found in the lungs. When submitted to the microscope, they appeared without doubt to be carcinomata. Virchow/ however, who saw the drawings of these tumors, said that the structure was more similar to that of spontan- eous cancer in the dog than to that of carcinoma as it occurred in the human subject. Lebert's treatise on cancer^ contained the description of an experi- ment performed in conjunction with Follin. Part of a mammary cancer was emulsified in water and about sixty or eighty grams of the resulting fluid, in which the presence of cancer cells had been confirmed by the microscope, was injected into the jugular vein of a dog. The animal died fifteen days afterward, and the autopsy dis- closed a number of nodules in the walls of the heart, beside some of smaller size in the liver, all of which contained cancer cells with round or elliptical nuclei possessing one or more nucleoli, while numerous free nuclei were also present in the preparations. Lebert did not believe, however', that this single experiment was of any great signi- ficance and declined to draw definite conclusions therefrom, realiz- ing that the dog might have been the subject of spontaneous cancer when the injection was first undertaken. Many investigators beside Dupuytren had reported their failure to transplant human tumors iruto animals, and with the idea that the miscarriage of their efforts might have been due to the length of time intervening between removal and implantation, Billroth ^ planned a series of experiments in which this period should be reduced to a minimum. He inoculated three dogs subcutaneously and six intra- venously, but with negative results in every case despite his care- ful technic. For the subcutaneous inoculations two carcinomata and a giant cell sarcoma were used, while for the intravenous three carcinomata, a struma, a lymphoma, and a giant cell sarcoma were chosen. Alberts ^ was the first to appreciate that the lesions so often de- 1 Die krankhaften Geschwiilste, Berlin, 1863, i, 87 (footnote). ^ Traite pratique des Maladies cancereuses et des Affections curables confondues avec le Cancer, Paris, 1851, 136. ^Wien. med. Woch., 1867, xvii, 1137, 1153. '^ Das Carcinom in historischer 11. experimentell-pathologischcr Bezichung, Jena, 1887, 183. 52 e.JlRlier observations ox the til\xsmissibility oe cancer scribed as transplanted carcinomata might have been septic emboli, and to conduct experiments in such a way as to exclude sepsis. Tumors that had been aseptically removed by operation were inoculated into dogs, but in no case did a successful result ensue. Klebs ^ examined fragments of human carcinoma removed from the peritoneal cavities of white rats at varying intervals after introduc- tion, and found that in the great majority of cases the epitheHal con- stituents of the graft had vanished by the third day. Duplay and Cazin - transplanted various t^'pes of tumors from man into different locaKties in animals of several species, but the out- come was always unsuccessful, even when they tried to produce by trauma a soil suitable for implantation. Nor did they meet mth any more favorable results after the introduction of fragments that had been kept for a period outside the body to allow the life cycle of a h}^othetical parasite to be completed, in accordance with the sugges- tion of Metchnikof f.^ The authors decided that mahgnant tumors did not appear to be transmissible from one species of animal to an- other. This conclusion may be accepted as representative of most authori- ties, and even where inoculations have been made into species so closely related to man as the anthropoid apes, the attempted transfer has failed, as has been shown by the work of Roux and Metchnikoff ,^ and of Jobling.^ The reputedly successful results have all been submitted to careful scrutiny and found wanting in one respect or another, v. Hansemann,^ for example, ha^^ing pointed out that typical metastases have never been described, although their presence would be a most important proof that implantation had been accompUshed. Ribbert ^ has said that transplantation could be performed only ^ Deut. med. Woch., 1890, xvi, 710. ^ Semaine med., 1892, xii, 61. Compt. reiid. de I'Acad. des Sc, 1892, cxiv, 325. Semaine med., 1893, xiii, 329. ' Ann. de I'lnst. Past., 1892, \d, 158. * Bull, de I'Acad. de Med., 1903, 1, loi. ^ See Flexner, Med. Record, 1909, Ixxv, 783. Monographs on Medical and Allied Subjects, Rockefeller Institute, New York, 1910, No. I, 120. ^ Berl. kUn. Woch., 1905, xlii, 314. ^ Verhandl. d. deutschen path. Gesellsch., 1904, S^e Tagung, 104. EARLIER OBSERVATIONS ON THE TRANSMISSIBILITY OF CANCER 53 between individuals of the same or of a very closely related species, and has extended this rule to include tumors, since no satisfactory evidence of their transfer from man to the lower animals has yet been adduced. TUMOR TRANSPLANTATION WITHIN THE SAME SPECIES (jHanaujJ described the first successful transfer of carcinoma within the same species. Involving the vulva of an old rat there was dis- covered an ulcerated cancer which had metastasized in the inguinal lymph nodes on both sides and in those of the axilla on the right. Microscopical examination showed that it was a keratinized squamous cell carcinoma. A fragment from one of the still unulcerated nodes was sewn under the skin of the scrotum in two old rats. One of these animals died forty- seven days after inoculation, presenting a few small nodules on the quasi-mesentery of the right vas deferens. The omen- tum-like structure which, in the rat, accompanies the vasa spermatica interna, was studded with tumors, while the great omentum had been converted into a nodular mass. The small omentum was also in- volved, and behind the stomach and between this organ and the spleen, were still other growths. Microscopical examination of two of the nodules in the omentum showed that they were carcinomata of ex- actly the same character as the tumor which had been inoculated. En- couraged by these findings, Hanau examined the second rat on the fifty- ninth day after implantation and discovered a round ulcer affecting the prepuce, and a firm movable tumor in the right half of the scrotum about one-half the size of the testicle. On the sixty-first day this rat was sacrificed. No trace of epithelial proliferation was found in the ulcer, but on the right gubernaculum hunteri there was a flat, white nodule about 2.5 millimeters in diameter, while between the testicle and the tail of the epididymis lay the larger growth which had been felt during life. A complete histological description of Hanau's material has been published by Jenny.^ Pfeiffer,^ in an article on the pathogenic protozoa, said incident- 1 Forlschritte der Med., 1889, vii, 321. Arch. f. klin. Chir., (v. Langenbeck), 1889, xxxix, 678. "^ Arch. f. klin. Chir., (v. Langenbeck), 1895, li, 269. * Centralbl. f. Bakt. etc., erste Abt., Orig., 1890, viii, 802. 54 EARLIER OBSERVATIONS ON THE TRANSMISSIBILITY OF CANCER ally and without giving any details regarding the experiment, that he had succeeded two years previously in transferring a melanotic carcinoma of the mouse to other mice. V. Eiselsberg ^ discovered a tumor in an adult rat invoMng the right shoulder and about the size of a hen's egg, hard, nodular, and movable on the underlying parts. The examination of fragments removed for diagnosis proved that the growth was a spindle cell sar- coma. Some days later portions of the growth were excised and sewn into mesenteric folds in two half-grown rats. The tumor rat having died under the anesthetic, an autopsy was immediately under- taken, during the course of which it was found that the neoplasm was connected by dense adhesions to the periosteum of the scapula, the bone itself, however, ha\'ing been spared. Nothing was said of the presence of metastases. In the second month follomng inoculation the two rats were examined under narcosis. In one there was nothing abnormal, but in the abdomen of the second there was discovered a tumor the size of a nut. Five months after inoculation this rat died during the night, and examination twelve hours later disclosed a firm nodular tumor in the mesentery, about the size of a hen's egg. .No other pathological condition was encountered. The microscopic structure of the daughter tumor was identical with that of the spon- taneous growth except that a larger number of spindle cells were present. A fragment was immediately transplanted into another rat but ^\ithout result — doubtless, said v. Eiselsberg, because the tumor cells were already dead. A tumor about as large as a hazel-nut was found by Morau - in the axilla of a white female mouse of unknown age. The nodule was re- duced to an emulsion, which was inoculated subcutaneously into other mice, and among these a number of tumors were found three months later. From one of the growths of this, the first generation, cultivation was continued for about three years, seventeen transfers being made in all. IMorau drew the following conclusions from his series of experi- ments : CyUndrical epitheliomata can be transferred to other mice by ^ V/ien. klin. Woch. 1890, iii, 927. 2 Compt. rend. Soc. Biol., 1891, xliii, 289. Compt. rend, de I'Acad. des Sc, 1893, cxvii, 62. Arch, de Med. exp. et d'Anat. path., 1894, vi, ()Tj. EARLIER OBSERVATIONS ON THE TRANSMISSIBILITY OF CANCER $5 inoculation ; heredity plays a considerable role in the development and evolution of these tumors ; traumatism hastens and favors their generalization ; pregnancy accelerates their growth ; the tumors possess a variable toxicity, which may destroy the host's life ; they seem to lose in virulence as they develop in new animals ; when they are not ulcerated, they do not contain microbes. Firket ^ published a preliminary note describing the transfer of a spindle cell sarcoma from rat to rat. A fragment from one of the many abdominal tumors in the animal primarily affected was trans- planted into the peritoneal cavity of another rat, which died about six weeks later with generalized growths in the abdomen. Grafts from this generation were transplanted into three rats, in all of which growth took place, and with one of the resulting tumors two other rats were ingrafted, one of them living long enough to develop small nodules. The tumor retained its original sarcomatous structure throughout the experiment. VeHch - discovered in the femur of a white rat a subperiosteal sarcoma, a portion of which, transplanted subcutaneously into an- other rat, produced a nodule as large as a walnut after a week's growth. This increased so rapidly in size that in three weeks it had attained a length and breadth of five centimeters, and a thickness of three. At the end of six weeks, when the rat died, the tumor was about one- third the length of its body. Before the death of the animal pieces had been removed from the tumor and implanted in three rats, in all of which nodules were appreciable a few days after inoculation. Al- together this sarcoma was carried through eight generations, but the growth energy became progressively more feeble until finally the tumor died out. VeHch recorded the fact that the rats used after the eighth generation were from another source and considered the possibiHty that a strange breed might not have offered favorable conditions for growth, dismissing it, however, because a decrease in proliferative energy had set in before the new strain was introduced into the ex- periment. The daughter tumors were all spindle cell sarcomata and similar in every respect to the primary growth. To none of these communications was any especial attention devoted, ^Bull. de I' Acad. Royale de Med. de Belgique, 1892, vi, 1147. 2 Wien. med. Blatter, 1898, xxi, 711, 729. 56 EARLIER OBSERVATIONS ON THE TRANSMISSIBILITY OF CANCER for their significance was not grasped at the time of their appearance ; and it was only after the earlier articles of Jensen, Loeb, and Borrel had been pubhshed that experimental pathologists, appreciating the possibiHties which they had until then neglected, commenced a con- certed attack upon the biological side of the cancer problem. A communication from the pen of Jensen^ which initiated this era of acti\ity, has became a cj^s_sic in the literature of cancer research as the masterly paper of Koch had become classic in the annals of tuber- culosis — for its evidences of patient, accurate, and exhaustive in- vestigation. Jensen - had described previously the results attend- ing the transplantation of tumors from mouse to mouse, but the fruits of his labor first became accessible to the general pathological public upon the appearance of an article in the German literature. The facts recorded in the earher papers were, in short, that he had been able to transmit a mouse cancer through ei ght generati ons, that nqjmicro- organisms had been found in it, and that experiments deahng with the resistance of the cancer cell to various agents had been undertaken. In the German paper Jensen described in great detail the tumor and the results attendant upon its transplantation. The primar}^ growth was about the size of a hazel-nut, situated under the skin of the back, and microscopical examination disclosed the fact that it was a t}'pical carcinoma. Daughter tumors were of entirely similar architecture, a relatively sparse stroma inclosing a large number of alveoli contain- ing cells of various contours. The cell bodies were large and fairly homogeneous, the nuclei large and provided mth nucleoli and a promi- nent chromatin network. As a rule the mitotic figures, which were by no means rare, were of the normal type, although atypical forms were occasionally to be found. A part of the nodule was emulsified in physiological saHne solution and a small amount of this suspension was inoculated subcutaneously into five mice, three of which developed tumors that were afterward transplanted into other series of mice. The experiments had extended over a period of two and a half years, the tumor had been carried through nineteen generations, and would grow in about half of the mice inoculated. During all the time that ^ Centralbl. f. Bakt., etc., erste Abt., Orig., 1903, xxxiv, 28, 122. ^Hospitalstideirde, 1902, x, 489. HospUalsiidende, 1903, xi, 549, 5S1. EARLIER OBSERVATIONS ON THE TRANSMISSIBILITY OF CANCER 57 the neoplasm had been under cultivation metastases in internal organs had never been discovered, and the regional lymph nodes, while usu- ally swollen, were only once found to contain anything like a secondary deposit. Unfortunately, however, the specimen had been mislaid be- fore it was ready for examination. The details of Jensen's experiments, in common with those of the investigations which have been prosecuted by the two succeeding authors, will be discussed in the following chapters under their ap- propriate headings. The period that saw the pubHcation of Jensen 's first article witnessed also the accounts of Loeb's early work.^ Loeb discovered a cystic sarcoma of the thyroid in a white rat and succeeded in transmitting it to other rats by inoculation, but not by causing its ingestion. Dur- ing the fifteen months that the growth had been under propagation the histological structure had been preserved unaltered. Neither the primary tumor nor those belonging to subsequent generations had produced metastases, although in one instance tumor cells had been found penetrating a blood vessel. In a succeeding paper Loeb - reported that the neoplasm in ques- tion had been carried through about forty generations in the course of twenty months, but had finally become so infected that further propagation had been rendered impossible. Other rat tumors had, however, been successfully transplanted. While the experiments of Jensen and Loeb were in progress, Bor- rel ^ was engaged upon the same problem. Although transplantation succeeded in only about lo % of the mice inoculated, daughter tumors were obtained of such rapid growth that after the expiration of forty days they weighed more than the mice themselves. Metastases were discovered at autopsy in the blood vessels of the lungs as well as in the lymph nodes, always preserving the structure of the main growth. A critical review of the work of Jensen's antecedents will be found in a summary by Sailer^ of the literature relating to the inocula- bihty of carcinoma. '^ Jour. Med. Research, 1901, N.S., i, 28. Arch. f. path. Anal., etc., (Virchow), 1902, clxvii, 175. ^ Jour. Med. Research., 1902, N.S., iii, 44. ^ Ann. de VLnst. Past., 1903, xvii, 112. ^American Jour, of the Med. Sciences, 1900, cxx, 190. CHAPTER IV THE TRANSPLANTED TUMOR THE STROMA REACTION As soon as it had been demonstrated that mahgnant growths were transmissible, it was necessary to know whether the cells of the daughter tumors were direct descendants of those that had been introduced in the graft, or whether they arose from the tissues of the new host. This was, of course, tantamount to asking : Is the transfer of a tumor an instance of tr ansplantatio n or of infection ? The vital importance of settling this question did not escape Jensen.^ He followed carefully the fate of implanted fragments, examining them at daily intervals after transplantation, and found that, while many of the parenchymal cells perished soon after their introduction into the new animal, there was still a fairly large number that remained alive, and that in larger grafts the cells able to retain their vitality were those situated at the periphery. The stroma of the newly in- grafted tumor became hyaline and nearly all the connective tissue cells in the ceritral parts disappeared, while at the same time many recently formed blood vessels and fibroblasts appeared about the fragment and in its margin. The final fate of the stroma was not absolutely clear to Jensen, but it seemed as though it were gradually penetrated from without by blood vessels and fibroblasts and that it eventually suffered absorption, although there remained the possibility that part of it might continue to live. The surviving cells of the parenchyma were always sharply demarcated from those surrounding them, and there was no evideiice of the creation ^f tumor cells from any of the eLlements of the host. Without doubt, therefore, the process was a real trans- planteiion. ^ Centralbl. f. Bakt., etc., erste Abt., Orig., 1903, xxxiv, 127. 58 "^ ^r^r. .-^J^. S^Afi^ /^) Q^ 'd^r^. "^^"t "^ ^^(3 @ ® V, X.r-^'^iL^M^*^^^": / -V" d;^^ 4 ■ \'% %n- -At, 3 Jensen's mouse carcinoma: fully developed tumor, from which grafts were used in the investigation of the early stages shown in the three following plates. f-/- n « -4. Gj > ;j _aj ^ TS X (J Cj OJ '^ o ^^ ■'■■rrsr-K 5^' ^ ^ ■2 X £ 5 o THE TRANSPLANTED TUMOR 59 Loeb^ was unable to decid e defin itely between transplantation and infection, possibly because his initial observations were conducted with a sarcoma — a tumor type in which the analysis of early growth is fraught with the greatest difficulty. In a later article he^ expressed the view that the peripheral cells of the graft remained aHve and developed, mingling with the surrounding elements of the connective tissue. It was very probable that tumor cells themselves were transplanted in addition to the tumor-producing agency. V. Leyden,^ while admitting that inoculated tumors resulted from the proHferating ceUs of the transplanted graft, could understand this fact only by assum ing the presence, of an intrac ellular paras ite. This supposition, he thought, did not oppose the conceptions which had been developed by the pathologist, and was, moreover, the only one capable of explaining the phenomena associated with cancer in man. Jensen's findings were confirmed and amplified by Bashford and Murray^ in conjunction with Cramer,^ through the investigation of several tumors with parenchyma and stroma of different types. The results, which were similar in all the growths studied, were described most fully for Jensen's tumor, which had been submitted to a more ex- tensive examination than the others. The authors found that the first evidence of reaction to the graft was a r apid aggregat ion of polymor- phonuclear leucocytes in the surrounding tissue s of the host, beginning about two hours after the introduction of the tumor fragment and of short duration only. The leucocytes collected about any necrotic material present and even penetrated between the tumor cells. About fifteen hours after the introduction of the graft there occurred prolifera- tive changes in the surrounding areolar connective tissue of the host, the cells of which became shorter and thicker, while their nuclei divid ed by amitosis. The proliferating cells migrated into the cleft which origi- nally separated the connective tissue of the host from the newly intro- duced fragment, and appHed themselves to the surface of the graft. The stroma had already- become hyahne after twenty- four hours, and commencing degeneration could often be detected in its cells. Such changes were even more distinct after thirty-six hours, and by this time ^ Jour. Med. Research, 1901, N.S., i, 37. "^ Jour. Med. Research, 1902, N.S., iii, 52. ^ Berl. klin. Woch., 1905, xlii, 348. * Proc. Roy. Soc, 1904, Ixxiii, 70. ^ Sci. Reports, Imperial Cancer Research Fund, London, 1905, No. 2, Part ii, 24. 6o THE TRAXSPLAXTED TUMOR the nuclei had become small, irregular, and deeply staining, while fatty degeneration of the cytoplasm had set in. The capiUaries of the graft w-ith their blood corpuscles were still recognizable, although the endo- thehum showed degenerative changes. Outside of the graft fibrin fila- ments had appeared in the exudate and amitotic di\dsion of the con- nective tissue cells was still in progress, while spindle-shaped wandering cells had begun to penetrate the transplanted tumor. Three days after inoculation the cleft between the tumor and the host's tissues was almost obhterated, and wandering cells could be distinguished in the interstices of the graft, where they were di\"iding by mitosis. ^Mitotic di^'ision was in e\idence, also, in the fibroblasts of the host nearest the tumor. In the transplanted stroma the collagenous fibrils had fused into homogeneous glassy bundles and the connective tissue cells ex- hibited, beside chromatolysis. unmistakable fatty degeneration of their protoplasm. Xo development of new blood vessels had yet occurred, and the whole mass of transplanted tissue remained ^\'ithout vascular supply. Four days after transplantation, however, an ingrowth of capiUaries had taken place and vascularization was in su£h rapid progress that aU stages of new capillary formation might be found in the same specimen. Fibroblasts frorn^ the_host were streaming into the tumor from every side, some of them appearing to exercise a phagocytic function and finally to degenerate, while others sur^dved to form the new stroma. The old connective tissue stiU remained, but was in the last stages of degeneration. From four days onward there was a continual recession of the stroma originally introduced, and an orderly progress of vascularization until, by the eleventh day. fibrils of new-formed coUagen could be found in the new stroma which, at the eighth, had been very cellular. Thus the cancer cells manifested the faculty of continuous growth and a power to make the tissues of the new host subserA-ient to their needs. The proHferative power, however, attached only to the parenchymal cells, and was not acquired by any of the elements with which they had come in contact in successive hosts. This stroma reaction was shown further to be a specific one. That is to say, the character taken on by the newly formed scaft'olding was determined not by the reacting tissues themselves, but by some^uiij ki iown influe nce exerted upon them by the parenchymal cells, which o 7* " St 3; y" o /iT ,> ? ci- / 1 O Small mincing machine devised by Haaland for emulsifying firm tumors. Natural size. JJI |N LUERaPAP.IS |TJ . S.C.B.g 15! \jJ \l'-glass syringe of 0.5 cubic centimeter capacity and graduated to o.oi cubic centimeter, employed to inoculate accurate doses of tumor emulsion. Four-fifths natural size. Hollow platinum-iridium needle for the inoculation of intact tumor fragments. Natural size. THE TRANSPLANTED TUMOR 69 In Ehrlich's institute/ on the other hand, the emulsion method has been found preferable, and the technic there employed was de- scribed as follows. The mouse bearing the tumor for inoculation is decapitated with scissors, placed in alcohol-subhmate, and, having been rinsed off in alcohol, is laid out on a board freshly cleansed with subHmate solution. The tumor is removed with sterile instruments, chiefly by blunt dissection, minced with scissors and forceps, and then broken down in a mortar into an emulsion without the addition of any extraneous fluid. Inoculation is done with capillary pipettes carrying a stopper of cotton wadding at the mouth end, the pipette being entered at a point over the lower abdomen which has been shorn and washed with alcohol, and the emulsion deposited in the axilla. Absolute asepsis must accompany all the steps of the operation. For the more ready preparation of tumor emulsions, Haaland^ devised a smaU mincing machine constructed on the principle of the ordinary meat grinder. In Bashford's laboratory^ both needle and syringe have been em- ployed. Of the two methods, that in which small fragments of tumor were introduced intact by means of a hollow needle was preferred, for the reason that it caused less damage to the tumor cells, a fact which explained the superiority of the results achieved. Although ac- curate doses could be administered by means of this method under certain conditions, the EngHsh workers, reflecting on the dis- similar results obtained in various laboratories, felt the necessity for a method which would permit of more uniform dosage without serious damage to the cells. The following procedure was accordingly adopted. After removal the tumor was reduced to an emulsion, which in the case of firm tumors was most conveniently effected with Haaland's mincing machine, although softer growths could be emulsified by repeatedly clipping them with sharp scissors. The Pasteur pipette, which did not permit of accurate dosage, was replaced by a small all-glass syringe of 0.5 cubic centimeter capacity, cali- brated to ox»i cubic centimeter. A hypodermic needle of sHghtly wider bore than that used in a serum syringe was fitted to the nozzle, 1 Arb. a. d. Konigl. Inst. f. Exp. Tlierap., 1906, Heft i, 79. 2 Bed. klin. Woch., 1907, xliv, 714. ^ Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 267. 70 THE TRANSPLANTED TUMOR and the injections were made in the axillary region or along the flank by introducing the needle into the groin, pushing it toward the axilla, and injecting the emulsion during withdrawal. Between injections the needle was wiped carefully on a pledget of sterile cotton-wool moistened with absolute alcohol. It was unnecessary, except in special cases, to epilate the site of inoculation, although it might with advan- tage be moistened with alcohol. With a hollow needle,^ the dose could be varied between 0.005 ^^^ 0.03 gram. Bashford, Murray, Haaland, and Bowen ^ recorded that the doses used by them in many thousands of experiments varied between 0.005 and 0.03 gram where a tumor fragment was inoculated by means of a needle, and between 0.025 ^^^ o-^S cubic centimeter where an emulsion was injected with a syringe. While practically all sponta- neous mammary tumors of the mouse could be induced to grow by inoculating large numbers of young animals with small intact frag- ments, the number yielding a successful result was much smaller when 0.05 to o.i cubic centimeter of tumor emulsion had been in- oculated. The foregoing authors, therefore, as well as Murray^ and Gierke,^ were of the opinion that variations in success were to be explained by technical differences, pointing out that the conversion of a tumor into an emulsion involved considerable mechanical injury to the paren- chyma, and, furthermore, that the inevitable absorption of part of a large dose of emulsion might readily be responsible for the evolution of sufficient acquired immunity to preve nt the estabHshment of the tumor. According to Bashford ^ the optimum conditions of transplantation for different strains could only be reached after trial and error more or less prolonged. Most tumors were best transferred by the frag- 1 These platinum-iridium needles are provided with plungers which fit closely enough to allow the tumor fragment to be aspirated by negative pressure upon their withdrawal. The needles are sterilized between inoculations by being flamed off with alcohol. (W. H. W.) ^ Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 269. ' Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 105. * Beitrdge zur path. Anat., etc., (Ziegler), 1908, xliii, 332. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 119. ^Fourth Sci. Report, Imperial Cancer Research Fund, London, 1911, 137. THE TRANSPLANTED TUMOR 7 1 ment method, and some could be propagated continuously only by this means. Others, again, gave consistently better results when larger doses of emulsion were inoculated. The intervals between successive transplantations were of great importance, for while some tumors could be cultivated only by rapid passage, in the case of others this was unsuitable. The influence of short intervals was particularly noticeable in the frequent experience that rapid passage might lead to a temporary exhaustion of the tumor cells, from which they could recover only after a prolonged sojourn in one animal. It was not always the most rapidly growing tumors that could be transplanted at the shortest intervals, and a permanent improvement in the rate of growth and the inoculation percentage might merely indicate that the optimum interval for grafting had been selected, either deliber- ately or accidentally. Other means of implantation have been employed, although less extensively. Thus Hertwig and Poll,^ as well as Stahr,^ imbedded fragments in small incised wounds which were sutured afterward with silk. In Gaylord's laboratory the method preferred was, as described by Clowes,^ emulsification of the tumor in a mortar with the addition of salt solution and removal of the connective tissue residue by means of fine rakes. Relative Importance of Soil and Graft It was found, however, that in spite of the most careful technic, transplantatj^on was not always successful, and that in some^animals the tuniorj:ells, unable to proliferate, underwent final abso rption. Thus Jensen * had attempted the transplantation of several tumors before he succeeded in getting one to develop, and Borrel ^ found that even those growths which could be successfully transferred might yield but about io^^of_daughter tumors. Ehrlich^ wrote that only a small proportion of spontaneous tumors were propagable, and that of one hundred and eight such growths 1 Ahhandl. d. Konigl. Preiiss. Akad. d. Wisseiischaften, 1907, 6. ^ Centralbl. f. allg. Path., etc., 1909, xx, 869. ^ British Med. Jour., 1906, ii, 1549. * Centralbl. f. Bakt., etc., erste Abt., Orig., 1903, xxxiv, 29. ^ Ann. de I'lnst. Past., 1903, xvii, 112. ^ Zeitschrift f. Krebsforsch., 1907, v, 61. 72 THE TRANSPLANTED TUMOR only about 8 % had been cultivated continuously. The hemorrhagic cyst-adenomata seemed to resist all efforts at transplantation. Bashford, Murray, and Cramer/ however, succeeded in trans- planting hemorrhagic tumors, and ascribed their success to the fact that they had implanted small fragments into a large number of mice instead of inoculating large doses into a few animals. But even though the most suitable dose be chosen the outcome is still uncertain, for it is determined not alone by the power of the malignant cell to proliferate in a strange host. Another condition for continuous growth is that the new animal shall offer a proper soil. Loeb^ found that, as a general rule, when two tumor fragments were transplanted into a rat, either both of them grew or else neither one did, and this, he thought, indicated differences in the soil rather than in the grafts introduced. The relative importance of these two factors was investigated also by Bashford, Murray, and Cramer,^ who inoculated the members of one group of mice with a single graft each and those of a second series each with five fragments of tumor. If idiosyncrasy of individual mice alone determined whether a tumor would grow or not, the per- centage of success per animal should be the same in both series, with multiple tumors in the successful cases ; but the series with m ultiple in oculation s gave a h igher percent age of success than those with . single implantations. The authors accordingly attributed a minor importance to the influence of the soil in determining success or failure, and a much greater to the introduced cells. This decision received additional support from the constancy with which different sporadic tumors gave a uniformly low percentage of success or even completely negative results, while others .yielded a much higher outcome in most cases. The authors were driven to the same conclu- sion by the subsequent behavior of the different descendants of one and the same tumor, when one series grew quickly and another slowly, with high and low percentages of success, especially since this result might be reversed in subsequent series. Even more remarkable was 1 Proc. Roy. Soc, Series B, 1907, Ixxix, 170. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 321. ^ Jour. Med. Research, 1902, N.S., iii, 58. ^ Sci. Reports, Imperial Cancer Research Fund, London, 1905, No. 2, Part ii, 51. THE TRANSPLANTED TUMOR 73 the disparity sometimes seen between t wo tumor s arising alongside each other and derived from separate groups of cells at a single in- oculation, one gr o wing quickly and the oth er slow ly. These same' characters were again presented by the daughter tumors which oc- curred when the two growths were transplanted into new mice. Importance of Uniform Dosage The importance of accurate dosage was first suggested by Loeb/ emphasized later by Clowes and Baeslack,^ and Gaylord and Clowes,^ and has been upheld consistently by Bashford and his colleagues. Fig. 2. — Diagram illustrating the effect on the amount of tissue produced of increase in the number of cells introduced or surviving after transplantation (effective initial dose), under conditions where the arithmetical factor alone is assumed to be of moment. Thus: ^ "The size the tumours of any one strain attain in a given time is in part determined by the proportion of the introduced cells which adapt themselves to the new conditions : i.e. by the size of the effective initial dose. - That this is the case can .be demonstrated by experiments in which the initial dose of tumour material varies. ^Jonr. Med. Research, 1902, N.S., iii, 59. "^ Med. News, 1905, Ixxxvii, 970. ^Surgery, Gynecology, and Obstetrics, 1906, ii, 634. ^ Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 273 et S3q. 74 THE TR.A.XSPLAXTED TUMOR '■'The simplest phenomenon is encountered in the experiments with tumours which grow \\-ith a rapidity proportional to the initial dose introduced. A transplantable spindle cell sarcoma of the rat, for which we are indebted to Professor Jensen, illustrates the subject now under discussion in a diagrammatic manner. . . . The size attained by the tumours in the animals inoculated \Yith. the larger dose.^ at the end of ten days, and their subsequent progress, present a striking contrast to those of the tumours arising from the small dose. They are nearly twenty times as large, and the animals succumb more quickly. It is important to note that the initial percentage of successes is the same in both series, viz. loo per cent. ... ■'"WTien experiments with these doses are made with other tumours, and especially ^^dth transplantable mouse carcinomata. the same result is rarely obtained. . . . Half the mice ... of this experi- ment"' ("inoculation sviih a squamous ceU carcinoma) "were inocu- lated A^dth 0.025 c.c. of tumour emulsion, and the other half . . . with 0.15 c.c. of the same material. An initial proliferation took place in ah. but although the doses were as i : 6, the sizes of the tumours arising from the larger doses are only in a few instances greater than those originating from the small doses, and even then little more than t^^dce as large. . . . '"Similar experiments i^dth many tumours show that the contrast between the results of inoculation of large and small doses are often. the reverse of those which we have been considering so far. . . . This is the most usual result when mice are inoculated vdih large and smaU doses. Sm aU dos es grojwjprogressively and well, whereas large^doses, even when followed by a mor^j) ronounced j3roliferation, give tumours which remain stationary or disap pear spontan eouslv. The natural resistance of the mice cannot be invoked to explain the anomaly. . . . '•'The different behaviour of the tumours arising from larger and smaU doses, and the temporary diminution in size or disappearance after initial prohferation. must be referred to the effects of the absorp- tion of more or less tumour material, inducing an adequate specific resistance of the animals in the one case and not in the other. The nature of the difference between tumours which support transplanta- 1 The doses employed v,-ere 0.2 c.c. and o.oi c.c. of an emulsion. (W. H, W.J THE TRANSPLANTED TUMOR 75 tion by large doses, and those which do not, is by no means clear. It is not merely a difference in capacity for independent life of the cells. . . . The strength of the specific resistance following absorption of the same quantities of different tumours seems to vary from one strain to another, as does also the susceptibility of the tumour cells to such altered resistance, and it is obvious that differences of this kind will manifest themselves also by varying susceptibiHty to dosage. "This process, which may be described as a concomitant active immunization, is in part responsible for the low percentage of success attending the primary transplantations of sporadic tumours. The ratio of absorbed tissue in them is higher, and the initial prohferation is less, in conformity with the failure of adaptabihty. This effect is exaggerated when primary transplantation is carried out with large doses, and is diminished when minute grafts are introduced. Where mice are naturally resistant to the tumour inoculated, con- comitant immunization greatly enhances it." In a paper by Bashford, Murray, and Haaland,^ there were described two experiments with the same tumor strain, in one of which pro- gressively growing tumors were obtained from 0.05 cubic centimeter of tumor emulsion, but not from o.i cubic centimeter. In the second, large doses of emulsion (oT5~~cubic centimeter) gave a higher per- centage of more quickly growing tumors than did simultaneous inocu- lation of similar mice with a small dose (0.025 cubic centimeter). The explanation of this different behavior at different times was, the authors thought, to be sought in alternations of the biological quaHties of the tunior cells and corresponding alternations in their vulnerability to unfavorable environment. Equal emphasis was given in this paper to the importance of accu- rate dosage when tumors or normal tissues were used to produce the resistant state. An amount of material much less than the smallest dose mentioned in the preceding paragraphs may give rise to a tumor. Uhlenhuth, Haendel, and Steffenhagen ^ described an experiment in which growth took place after the Jensen rat sarcoma had been rubbed into a scarified area in the skin, and another in which this growth was transferred 1 Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 364. ^ Arb. a. d. Kaiserl. Gesundheilsamte, 191 1, xxxvi, 477. 76 THE TRANSPLANTED TUMOR through the medium of a pin which had been plunged into it and after- ward inserted under the skin of a healthy rat and immediately with- drawn. These results led them to try the transmission of tumors through the bites of bed bugs, stinging flies (Stomoxys), and leeches, but of six trials with bed bugs and fhes and an equal number with leeches, not one was successful. Importance of Uniform Soil The desirability of using mice of the same age and race was pointed out almost coincidently by Bashford, Murray, and Bowen,^ and by Clowes and Baeslack.^ The former authors, having observed differ- ences in the suitabiHty of animals of different colors, even among the ordinary English tame mice, were convinced of the necessity of using the same race throughout any one experiment. The wild mouse, they thought, would probably offer more uniform conditions than the tame mouse, but a sufficient stock of uniform age was difficult to obtain and supervise. Besides having the mice of the same race it was important that they should be of the same age, preferably from five to seven weeks old. When the indicated precautions were observed, the individual variations in the general suitability of dif- ferent mice of the same race and age were negUgible, if implantation were performed at the same site, and provided sufficiently large num- bers of animals were used. Inoculation Site In Bashford's laboratory,^ the subcutaneous tissue of the back was at first elected, attempts to perform collateral series of intraperitoneal inoculations having been abandoned owing to the frequency with which growth within the peritoneum occurred secondarily from tissue implanted in the abdominal muscles. The axilla was finally^ chosen, however, as the most suitable location for the following reason: ''On 1 Proc. Roy. Soc, Series B, 1906, Ixxviii, 196. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 285. 2 Jour. Exp. Med., 1906, viii, 484. 3 Proc. Roy. Soc, Series B, 1906, Ixxviii, 197. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 286. * Proc. Roy. Soc, Series B, 1967, Ixxix, 175. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 327. THE TRANSPLANTED TUMOR 77 a series of 1 1 passages, it was found that the axilla was invariably a more suitable site than the dorsal subcutaneous tissue : 59 mice out of 286 developed tumours both back and front, 18 mice had tumours only on the back, and 59 only on the front. . . . The most nat- ural explanation of the difference would seem to be that the con- nective tissue reaction . . . without which the grafts cannot grow, is more readily supplied by the connective tissue of the mammary region." EhrHch^ also preferred the axilla, and routine inoculations were always made in that location. Implantation has been undertaken in still other locations, as, for example, successfully in certain of the intra-abdominal organs by Goldmann,^ and with indifferent success in the testicle by Flexner and JobKng.^ Interval after which Growth becomes Apparent When implantation has been successful a tumor appears at the inoculation site after a certain period, which was set by Jensen^ at about fourteen days, or in exceptional cases six to eight days. Borrel ^ found small nodules after from twelve to twenty days, Clowes and Baeslack^ after two to four weeks, and Michaelis^ stated that one could tell by about the third week whether an inoculation had been successful or not, although occasionally the nodule might not be palpable until more than two months after implantation. Still, the appearance of the tumor may occasionally be delayed for as long as five and a half months,^ eight months as described by Bridre,^ or even between eight and ten months as recorded by Stahr.^° ^ Arh. a. d. Konigl. Inst. f. Exp. Therap., 1906, Heft i, 79. ^ Arb. a. d. Konigl. Inst. f. Exp. Therap., 1906, Heft i, 79 (footnote). ^Monographs on Medical and Allied' Subjects, Rockefeller Institute, New York, 1910, No. I, 35. * Centralbl. f. Bakt., etc., erste Abt., Orig., 1903, xxxiv, 125. ^ Ann. de VInst. Past., 1903, xvii, 113. 8 Med. News, 1905, Ixxxvii, 968. ' Med. Klin., 1905, i, 204. ^ Oral communication regarding a sarcoma of the rat, by Dr. J. A. Murray of the Im- perial Cancer Research Fund, London. ^ Ann. de VInst. Past., 1907, xxi, 762. 1" Centralbl.f. allg. Path., etc., 1909, xx, 874. 78 THE TEANSPLANTED TUMOR It is essential to distinguish early tumors from inflammatory swell- ings, as has been pointed out by Clowes and Baeslack/ Bashford, Murray, and Cramer,^ and Michaelis.^ In the experiments of Clowes and Baeslack, inflammatory swelKngs frequently occurred soon after inoculation, but differed from malignant tumors either by undergoing absorption or by ulcerating out in the course of two or three weeks. A true tumor, on the other hand, developed somewhat later and was generally firmer than an inflammatory mass. Inoculation of Stationary or Receding Tumors Loeb ^ removed half of a stationary tumor from a rat and re-im- planted it in the same animal, after which both the new graft and the remaining piece began to grow at a rapid rate. In this case it was sug- gested that the liberation of the tumor from the tension of its capsule might have permitted the resumption of proliferation. A fragment taken from a growth actually diminishing in size grew so weE after inoculation into another rat that a number of tumors were after- ward transplanted from it. The original tumor, however, continued to recede until only a few small nodules were left. White and Loeb^ later entered more fully into this question and found that, in general, the tumors resulting from the transplantation of stationary or receding growths did not have a proHferative activity equal to those descended from vigorously growing tumors. In the opinion of Bashford, Murray, and Bowen,^ diminished trans- plantabihty was due to a real alteration in the parenchymal cells, an inabihty to establish themselves in new hosts, coinciding with the spontaneous cessation of proliferation in an animal in which growth had already become established. 1 Med. News, 1905, Ixxxvii, 968. 2 Sci. Reports, Imperial Cancer Research Fund, London, 1905, No. 2, Part ii, 63 (footnote). 3 Med. Klin., 1905, i, 204. * Jour. Med. Research, 1901, N.S., i, 34. Arch.f. path. Anat., etc., (Virchow), 1902, clxvii, 175. Jour. Med. Research, 1902, N.S., iii, 47. 5 Centralhl.f. Baki., etc., erste Abt., Orig., 1910, Ivi, 488. ^ Proc. Roy. Soc, Series B, 1906, Ixxviii, 208. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 299. THE TRANSPLANTED TUMOR 79 Transplantation of Metastases The lung metastases of four spontaneous tumors were transplanted by Murray/ with success in two cases. Uhlenhuth and Weidanz- also attempted the transplantation of pulmonary metastases, but their mice died of sepsis before the comple- tion of the experiment. A somewhat more encouraging issue attended their endeavors to cultivate lymph node metastases, and one tumor was obtained among four inoculations. Unfortunately, however, the mice into which this growth was implanted died of mouse typhus before the experiment was finished. Gay^ was successful in transplanting metastases of the Flexner- Jobhng rat carcinoma, and ventured the opinion that there occurred after several generations an increase in the virulence of the resulting tumors which was exemplified by a greater proliferative activity and the production of wider spread metastases of a more pronounced epi- thelial type. Inoculation of Tumor Mixtures Mixtures of tumors have been transplanted in a few instances. Apolant and Ehrlich"* recorded a^ series of inoculations with a mixture of three alveolar carcinomata and one malignant adenoma, in which the first generations showed no departure from the structure of an ordinary transplantable alveolar carcinoma. Sarcoma development occurred in this strain between the twelfth and fourteenth generations. Haaland ^ mixed Jensen's carcinoma with a sarcoma from Ehrhch's laboratory, and obtained the carcinomatous strain pure again in Danish mice and the sarcomatous in Berlin mice. The Danish mice were susceptible to carcinoma and refractory to sarcoma, while the Berhn strain was highly susceptible to sarcoma and resistant to Jen- sen's carcinoma. Ehrlich^ and Apolant'' described the results following the trans- * Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 106. ^ Arb. a. d. Kaiserl. Gesiuidheitsamte, 1909, xxx, 441. ^ Proc. Soc. Exp. Biol, and Med., 1908-1909, vi, 74. ^ Berl. klin. Woch., 1906, xliii, 38. ^ Bcrl. klin. Woch., 1907, xliv, 716. ^ Zeilschrifl f. Krebsforsch., 1907, v, 67. '' Zcitschrift f. Krebsforsch., 1907-1908, vi, 251. 8o THE TR.1XSPLAXTED TUMOR plantation of mixtures of carcinoma and sarcoma, carcinoma and chondroma, and chondroma and sarcoma. ]\Iixed tumors in which both components were intermingled, and which were exactly analogous to the carcinoma sarcomatodes of v. Hansemann, were readily ob- tained by the inoculation of a mixture of sarcoma and carcinoma. It was much more difficult to produce a mixed tumor with combinations of chondroma and carcinoma, or sarcoma, because the elements varied so much in \'itahty and in prohferative energ}' that pure chondromata, carcinomata, or sarcomata, were usually the outcome of the experiment. But even when both components grew there was never the intimate comminghng which would allow one to claim the production of a dis- tinct new variety of tumor, for both remained isolated and each one produced its o"«ti t}^ical form of neoplasm. It had been found that the inoculation yield was much less aff ected by lo w degrees of tempe ra- ture than was the proliferative energ^^ and yet when sarcoma that had been expo sed to — lo"^ C. for periods even as long as nineteen days was inoculated together mth chondroma, it was impossible to produce a mixed tumor with an equal distribution of the two com- ponents. There occurred merely two independent tumor tissues grow- ing side by side ^ never an amalgamation into a distinct new tApe. The importance of these experiments for general pathology lay in the demonstration that a mixed tumor could result or persist only when the biological conditions of growth for the two components were approximately equal. Resistance Offered by the Cancer Cell to Various Agents Jensen^ discovered that no_g rowth follo wed the in oculat ion of c rushed cells , and that the resistance which the cancer cell was able to oppose toward various injurious agents AA'as perhaps somewhat lower than that displayed by the cells of normal tissue. \\'liile his tumor was still capable of successful transplantation after an exposure of about eighteen days to a temperature of i° to 3° C. and of about twelve days to room temperature, it remained ahve hardly twenty- four hours if kept at body heat. When warme d for five minutes at 47° C, or cooled for a few minutes at — 20 ° C, the tumor was kill ed. 1 Centralbl.f. Bakt., etc., erste Abt., Orig., 1903, ssxiv, 129, 131. THE TRANSPLANTED TUMOR 8 1 Intense light (Finsen) , in so far as it was able to penetrate the tissiie^ was fatal, and partial drying, or a five^ minute expos ure to a - j %^ solution of car bolic aci d, also rdDbed the cells of their power to pro- Uferate. Loeb^ investigated the maximum temperature to which a rat sar- coma could be exposed without losing its power to grow. He found that fragments kept at 43^ to 44° C. for forty minutes re mained viab le, as did one heated for twenty-five minutes at 43° C. and then for fifteen minutes at 45° C. On the contrary, pieces exposed for thirty minutes to a temperature of 45° C. as well as those exposed for a similar length of time to still higher degrees (up to 50° C), did not grow in any single instance. Fragments of tumor kept on ice for five days were successfully transplanted, and it was thus clear that the tumor^pro- du cing facto r did not materially lose^n power after having been kept for this length of time at a temperatur e of fr om 2° to 4° C. That intact cells were necessary for growth was demonstrated by the neg- atiy e resul t which attended the in]'e£tiQiLQ|._filt ered emuls ions. Michaelis^ achieved a suc cessful outcom e from the transplantation of Jensen's tumor after it had lain in an ice-b ox for five days, and even after it had been expos ed to liq uid air for h alf an hou r. C hloroform water, however, rob bed the c ells of their po wer to_ grow. Moore and Walker,^ several years later, observed that a half-hour's exposure to liquid_air at a temperature of about — 195° C. was not always fatal for cancer cel ls, wldle Gaylord^ found that they could withstand this tempera ture for e ighty minu tes. In the experiments of Clowes,^ all tumor cells seemed to be destroyed at 45° C. Toward organi c disin fectants, hov/ever, their resistance was very^Jiigh. Treatment for an hour with mercuric .chloride at a concentration of 1-3500 was insu fficien t to prevent the development of a small proportion of slowly growing tumors, and it was not until a strength of 1-2000 was reached that compl£te_destruction of the cells was effected. They were killed also by m ercuric io dide in a solution ^ Jour. Med. Research, 1902, N.S., iii, 62. Arch.f. path. Anat., etc., (Virchow), 1903, clxxii, 345. ^ Med. Klin., 1905, i, 204. ^Lancet, 1908, i, 226. * Jour. Inf. Diseases, 1908, v, 443. ^British Med. Jour., 1906, ii, 1549. 82 THE TR-JlXSPLAXTED TUMOR of from I-2000 to 1-2500 and ammonium fluoride in a concentration of i-iooo. but potassium cyanide in an -^— solution was unable to 100 effect the death of either the Jensen or the Brooklyn tumor, although bacteria are e ntirely destro yed by an — — . solution. ■ '" ' 20a Ehrlich ^ considered that the maxima of temperature giyen by Jensen and Loeb represented the maxima for his o^vn carcinomata and sarco- mata. But the few minutes' exposure to —20^ mentioned by Jensen as haying sufficed to destroy his tumor was, in Ehrhch's experience, not enough, for he had often seen growth power preserved after ex- posure for forty-eight hours to a temperature of from 25° to 30° be- low zero C and in one case he had been able to transplant a carcinoma kept for two years at 8' to 10^ below zero. In the case of a chondroma he had been able to discover microscopic evidence of temporary growth in material heated for an hour at 50" C. an exposure which would have destroyed the cells of a sarcoma or a carcinoma, while a success- ful result had attended the transplantation of a chondroma that had been kept th ree d ays at the tem perature of hquid a ir. In these ex- periments, however, as the vitaUty of the cells had been very seriously impaired, tumors larger than a pea were never obtained, and these gradually disappeared instead of gro\\ang progressively. Bridre - found that mice developed growths only exceptionally after the inoculation of , finely ground tu mor, and ne\^ after the injection of an emulsion in^s ahne solu tion (filtered or unfiltered), or of tumor dried or heated above 50° C. In short, it was necessary to inoculate int act ce lls. From the work of Haaland,^ it appeared that the cells of a sar- coma were better able to resist heat than those of a carcinoma, for he was able to purify a mixed tumor of its carcinomatous elements by exposure to 44' C. for thirty-five minutes or longer. Lewin '^ was able to substantiate for a carcinoma of the rat all the statements which had been apphed to mouse carcinomata, and for an ^ Zeitschrift f. Krebsforsch., 1907, v, 65. 2 Ann. de I'lnst. Past., 1907, xxi, 767. 3 Berl. klin. Woch., 1906, xliii, 40. ^ Zeitschriftf. Krebsforsch., 1907-1908, vi, 304. THE TRANSPLANTED TUMOR 83 artificial mixed tumor the observation of Haaland just cited. He succeeded also in repressing the development of keratin by exposure to 46° C. for a quarter of an hour. The experiments of Uhlenhuth and Weidanz ^ yielded results quite analogous to those of other observers. Tumor cells were killed by an exposure to 56° C. for twenty-five minutes, but not by being subjected for the same length of time to 45° C. ; nor was a several days' sojourn in an ice and salt mixture, at a temperature far be- low the freezing point, fatal to them. Loeb and White ^ kept mouse tumor at 44° C. for varying lengths of time and plotted the results of inoculation in a series of curves depicting the growth energy, the latent period, and the number of receding tumors. The curves corresponded rather closely, and showed that up to a certain temperature the result was roughly proportional to the degree of heat appHed. In all the curves there was a critical point between forty-five and fifty-five minutes, and at the end of the latter period fundamental alterations in the cell set in. The three properties above mentioned probably depended, therefore, upon one and the same factor of the cell protoplasm. The cancer cell, however, may be prevented from growing by damage more subtle than any of the comparatively gross injuries so far de- scribed, for Haaland ^ showed that the a^tion^Ljadium was fa^al^ to itsjife, although the anatomic al structure remained u naltered. COMPARATIVE GROWTH RATE OF THE MALIGNANT CELL While much has been written in the past about the enormous pro- liferative capacity of the cancer cell it may very weh be that such statements have been unwarranted, and that the true key to the na- ture of maHgnant growth Hes in the direction of an explanation of the contiijupus, rather than the rapid, growth of the cancer cell. Thus the power_QfJhis_ceU^tojtnultiply proves but a sorrx_a£com^lishment when compared with the gr owth energy of the bact eria, organisms 1 Arb. a. d. Kaiserl. Gesundheitsamte, 1909, xxx, 441. 2 Centralbl. f. Bakt., etc., erste .\bt., Orig., 1910, Ivi, 325. 3 Proc. Roy. Soc, Series B, 1909-1910, Lxxxii, 297. Lancet, 19 10, i, 789. 84 THE TRAXSPLAXTED TUMOR which Minot ^ has estimated are able to add looo^o to their original weight \Aithin a few hours, while according to the same author the embryos of rabbits and other mammals, no less than those of birds, may be said with safety to grow at least 1000% a dav. A close comparison has been drawn by Bashford,- who weighed mouse embryos at different periods of gestation and estimated the rate at which growth took place after the embryo had reached a LLJl Fig. 3. — Diagrammatic comparison of rate of growth of various tumor strains on the basis of the number of days required to produce i gram of tissue from a meas- ured dose (0.02-0.03 gram) inoculated. The rate of growth of embrj^onic tissue is in- dicated on the basis that the mouse embryo weighs 0.02-0.03 gram at the eleventh day of gestation and at birth 1-1.5 grams. weight (0.02 to 0.03 gram; corresponding to the amount of tumor usually inoculated. When he compared the proliferative acti\dty of the two tissues it became CAddent that, while some tumors might attain the rate of growth possessed by the embryo, most of them fell far below it. ^ The Problem of Age, Growth, and Death, X'ew York and London, igo8, 12 = - Fourth Sci. Report, hnperial Cancer Research Fund, London, 191 1, 199. i: so- 40 30 20- 10 500 DA>f9 eo so- (00 2Q0 300 400 JOOOAYS Rise and fall in success of transplantation in four strains of Jensen's mouse carcinoma. The abscissae represent the number of daj's during which the propagation of the several strains was continued. The ordinates show the percentages of success in consecutive series of sub-inoculations. In each figure the curve bifurcates at the time when two separate tumors of the original strain were used to start new strains. The time when growth was artificially interrupted by transplantation is rep- resented by points and crosses for the two daughter strains of each figure. These curves represent experiments lasting for nearly 500 days, and carried out upon nearlj'' 8000 mice. The results of the first transplantation of Jensen's tumor into English mice are not included. The large fluctuations in the success of transplantation are similar in all four ciu-ves. Success is seen to increase through a series of successive inoculations until a maximum is reached. From this point onward the result diminishes until a minimum is arrived at, from which point the curv^e again rises to a maximiun. The general correspondence in the four series can hardly be pure coin- cidence, and the conclusion appears justified that the power to establish themselves in new hosts varies periodically in the cells of this tumor, from inherent causes. The occurrence of such periodic variations in the growth of tumors must be borne in mind in appraising the value of therapeutic measures. THE TEANSPLANTED TUMOR 85 PLUCTUATIONS IN GROWTH ENERGY When tumor growth was studied over long periods of time in suc- ceeding series of mice by Bashford, Murray, and Cramer/ it was found that : — ''The experimental propagation of mahgnant new growths leads to an apparently continuous proHferation which is merely artificially divided up by the process of transference to successive hosts. The Hmits of growth are not attained in any one animal, and transplanta- tion again becomes necessary after intervals which vary according to the rate of growth of the tumour or to the degree in which the animal suffers from the presence of the tumour or from intercurrent disease. Thus the time of transplantation does not possess the importance of a natural starting-point for the growth of the tumour which follows it, neither does it coincide with a terminal stage of the growth with which transplantation is effected. . . . While for any one sporadic tumour the average percentage of success of a large number of experiments is fairly constant through many generations, considerable variations in success attend the inoculations of one and the same tumour at different times. The variations in success are frequently accompanied by dif- ferences in the rate of growth of the tumour and do not remain con- stant for their descendants in further transplantation. Thus one series of inoculations may give • a small percentage of slow-growing tumours which at a subsequent period may begin to grow rapidly, or on transplantation while still growing slowly give a high percentage of quickly growing tumours. "These variations in the behaviour of different sporadic tumours, therefore, indicate differences in degree rather than in kind, comparable to those which may obtain in the different descendants of the same tumour. . . ." In addition to variations in the histological characters of trans- planted tumours, gradual fluctuations in the rate of growth and per- centage of success were found when the results of long-continued experiments were compared. "These curves represent experiments lasting for nearly 500 days, and give the results obtained on nearly 8000 mice, . . . ^ Sci. Reports, Imperial Cancer Research Fund, London, 1905, No. 2, Part ii, 48 et seq. 86 THE TB_A.XSPLAXTED TUilOR "The large fluctuations in the success of transplantation are similar in all four curves. Success is seen to increase through a series of suc- cessive inoculations till a maximum is reached. From this point on- wards success diminishes till a minimum is arrived at, from which the cur\'e again rises to a maximum. The fall in the curves coincided with the occurrence of numerous negative series of transplantations in daughter strains. . . . "The conclusion appears justified that the power of establisliing themselves in new hosts varies periodically in these tumour cells from inherent causes." In a later paper, Bashford, ^Murray, and Bowen ^ analyzed more minutely the fluctuations in growth energ}'. \Mien all precautions were taken to insure accuracy of dosage, similarity of soil. etc.. fluc- tuations independent of these factors still appeared, and the authors beheved that they were, in all probabiHty. natural features of prolif- eration. They repeatedly saw the inoculation percentage rise to a maximum which could not be maintained, to be followed by a fall, which also was not permanent. They had encountered no exception to this rule in more than six hundred series of inoculations -udth Jensen's tumor, and the rise to a maximum with the subsequent fall had been repeated fifty times in simultaneous series of experiments. Wlien the subsequent beha\'ior of the descendants of several of the daughter tumors from any one batch of inoculations was followed, successive maxima were seen to arise, one after another, at short interA-als. The maximum percentage of success of the experiments as a whole, was maintained continuously at a high level between seventy and ninety. Each strain, after reaching its maximum, fell and made way for another which had pre\-iously presented a lower percentage; and this latter feU in turn after having attained its maximum. The behavior of the component parts of the Jensen tumor, when this growth was propagated in a large number of mice, represented, in the opinion of the authors, what might be regarded as occurring simulta- neously in different parts of a single tumor when it was allowed to pro- liferate for a long period in one mouse. After a time any single tumor 1 Proc. Roy. Soc, Series B, 1906, lxx\'iii, 195. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 284. JOAPft JUN JUL AUG SEP OCT NOV OLC JAIt 07 FLS n>\R APR MAY JUN JUL AUG &EP OCT NOV OLC JMI'08 FES MAB API Percentage curve of a mammary adeno-carcinoma of the mouse. After a short period of depression the curve rises rapidly till the yth generation, and then falls. A second and third rise then follow- in two strains propagated separately. While one ( ) falls again and then rises slowly to a maximum, the other ( ) fluctuates between 35 and 50 per cent for 9 months, after which it also rises. «? JAH07 F(B JUM JUL 7J«rL»4 JUL ' *U0 itP aCT NOV Squamous cell mammary carcinoma of the mouse. Rapid rise in percentage of success at the third transfer (3 A), followed by a fall and a second rise. Hemorrhagic mammary alveolar carcinoma. Slight drop, followed by a gradual rise. Fluctuations of growth energy in three tumor strains. THE TRANSPLANTED TUMOR 87 Fig. 4. — Fluctuations in growth energy. could no longer be regarded as consisting of cells with equal proliferative power, for just as a composite chart of all the strains propagated in- dicated their very different behavior at any one date, so, in any single tumor, growth was proceeding actively in one part while at another it was going on slowly, or had actually ceased. The same heterogeneity was postulated for spontaneous tumors which, in all probability, owed their apparently continuous growth to the simultaneous presence in different areas of numerous growing centers which masked the effects of concomitant immunization, thus accounting for the rarity of spon- taneous absorption among sporadic, as compared with transplanted, tumors. The greater frequency with which growth ceased in propa- jgated tumors, to be followed by spontaneous absorption, seemed to be due to the greater homogeneity resulting from the limited number of growth centers represented in any one implantation. 88 THE TRANSPLANTED TUMOR By choosing a suitable interval for inoculation of tumors selected from series with from 90 to 100 % of success, and especially by in- creasing the dose from o.oi to 0.05 gram, Bashford, Murray, and Cramer^ were able to evade the diminution usually following each maximum, for a considerable number of transferences. Fig. 5. — Graphic record of propagation through ten passages of a single strain of Jensen's carcinoma in which success has not fallen below 85 %. Calkins ^ could not accept the interpretation of these fluctuations advanced by Bashford and his colleagues, and argued that rhythms of growth, to be comparable with those of a free hving, or of a cleavage cell, should be looked for in the individual mouse rather than in succes- sive batches of mice. Again, curves such as had been published by the EngHsh school introduced two factors, one the percentage of "takes" and the other, the time required for the tumor to develop. It was the time factor which really measured growth energy, indicating roughly, as it did, the rate of division of the cells, and what the curves of Bashford had reproduced were merely recurring alternations of infec- 1 Proc. Roy. Soc, Series B, 1907, Lxxix, 174. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 326. 2 Jour. Exp. Med., 1908, x, 283. THE TRANSPLANTED TUMOR 89 tivity (percentage of "takes"). Calkins constructed a curve to repre- sent the variations in infectiveness, and another to indicate the number of days required for the tumors to reach a point where they would kill the mice. His figures showed that the curves were by no means IK/i DOX edJ 40;? ZOX 10% ; F„Cl ■ ,„ d^ irctt L l,U m "'■■■'<• "ii eo i ..-., f.0 1 x> 1 4f) . . .w ?o 10 > I Fig. 6. ninHysrsQffliiiHiiiMayszMZHiMixii - The waves of growth described by Bashford are shown by the regular alternations of infectivity, not in the line showing growth energy. equivalent, and that too much was assumed in choosing the percentage of "takes" as the measure of growth energy. Bashford, Murray, Haaland, and Bowen ^ repHed that, as they had always been aware, the percentage of success could never be regarded as anything but an arbitrary standard by which to measure growth energy. And while, in the comparison of different strains the per- centage of success varied independently of the energy of growth, within the Hmits of a single strain there was a very definite, although not ab- solute, correspondence between them. Either by measuring the area of tumors drawn in silhouette, or by killing large numbers of animals at regular intervals and weighing the amount of tumor produced, it was possible to get a more correct estimate of growth energy than by means of the plan adopted by Calkins ; this was subject to uncontrollable fallacies because the death of the animals was only a secondary result of growth. Nevertheless, the curve constructed by that investigator * Third Sci. Report, It7iperial Cancer Research Fund, London, 1908, 281. QO THE TEANSPLANTED TUMOR showed that the higher was the percentage of takes the lower was the interval within which the mice died, and the authors were, there- fore, at a loss to understand how he had come to a different conclusion. SPONTANEOUS ABSORPTION Transplanted tumors which have become estabHshed may grow pro- giressively until they attain a size equal to or even exceeding that of the animals in which they are growing or, having reached a certain point, may undergo a gradual spontaneous absorption. Spontaneous absorption may be entire or it may be partial, growth setting in again before complete disappearance has been effected. A strain may exhibit among the tumors of one series the phenomena of unin- terrupted growth, retardation and resumption of growth or, finally, spontaneous cure. Such a tumor has been described by Bashford and Russell.-^ Loeb - was the first to observe and to appreciate the significance of the spontaneous disappearance of transplantable tumors. The same phenomenon had been encountered by Jensen,^ but was attributed by him to certain therapeutic experiments then in progress, erroneously, however, as he ^ said in a later article. Michaelis ^ described three tumors of the Jensen strain which had remained stationary or receded after reaching the size of a pea, and Apolant,^ in an article on the effect of radium upon mouse carcinoma, wrote that in two out of eleven controls the tumor had been absorbed spontaneously. Commenting upon the experiment, he said that ces- sation of growth, or even the entire disappearance of nodules, had been observed in every series, although in only a small percentage of the animals. Clowes '' recorded the complete absorption of Jensen's tumor in at least 15 to 20% of the cases and, in a more detailed article with 1 Proc. Roy. Soc, Series B, 1909-1910, Lxxxii, 298. Lancet, 1910, i, 784. ''■Jour. Med. Research, 1901, N.S., i, 34. ^ Centralhl. f. Bakt., etc., erste Abt., Orig., 1903, xxxiv, 30. * Zeitschrift f. Krebsforsch., 1908-1909, vii, 281. 5 Verhandl. d. Komitees f. Krebsforsch., 1903-1904,!^, 38. See Deut. med. Woch., 1904, XXX, 1264. ^ Deut. med. Woch., 1904, xxx, 456. ' Johns Hopkins Hosp. Bull., 1905, xvi, 130. THE TEANSPLANTED TUMOR QI Baeslack,^ in about fifty mice. To exclude early inflammatory swellings as a source of error the authors had adopted for a standard the ab- sorption, not less than three weeks after inoculation, of a tumor which had at some period reached the volume of twenty cubic milhmeters. Gaylord and Clowes ^ later described spontaneous recovery in one hundred and one cases, or about 23% of the inoculated animals, and con eluded that the chances of recovery were inversely proportional to the size of the tumor. Bashford, Murray, and Cramer,^ in their earHer experiments, had seen only one case of spontaneous absorption among three thousand Jensen tumors of over fourteen days' growth, but Bashford^ was later able to confirm the findings of other investigators with the statement that in some series as many as 50% of animals with large tumors were able finally to rid themselves of their growths. Histology of Receding Tumors Bashford, Murray, and Cramer '" examined microscopically a re- ceding tumor of the Jensen strain and found necrosis and a great overgrowth of connective tissue, coupled with the presence of char- acteristic cells containing small nuclei and a coarse protoplasmic re- ticulum, a type which they had ^encountered in growths exposed to radium, and especially about those of animals that had received toxic doses of adrenahn. In some areas a few isolated tumor cells were found lying in a dense connective tissue, and associated with multi- nucleated cell masses like those common in the reaction zone around foreign bodies. As even an extensive necrosis of the parenchyma had never of itself been sufiicient to excite such an active connective tissue proHferation in transplantable tumors, the authors were incHned to connect this fibrosis with the occurrence of hemorrhages in the tumor. Bashford, Murray, and Bowen ^ described a cell with darkly staining ^ Med. News, 1905, Ixxxvii, 968. ^ Surgery, Gynecology, and Obstetrics, 1906, ii, 633. ^ Sci. Reports, Imperial Cancer Research Fund, London, 1905, No. 2, Part ii, 63. '^British Med. Jour., 1906, ii, 209. Lancet, 1906, ii, 315. ^ Sci. Reports, Imperial Cancer Research Futid, London, 1905, No. 2, Part ii, 63. ^ Proc. Roy. Soc, Series B, 1906, Ixxviii, 212. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 302. 92 THE TRANSPLANTED TUMOR nucleus and protoplasm that was often present in spontaneously re- gressing tumors, and Bashford, Murray, and Cramer/ the occurrence of energetic phagocytosis. The histological data recounted and pictured by Gaylord and Clowes - consisted of retrogressive changes in the epithehum accom- panied by the occasional fusion of some of its cells to form pseudo- giant cells, the appearance of hemorrhagic areas, an ingrowth of connective tissue spHtting alveoli into smaller cell groups, and the accumulation of small round cells. CLINICAL COURSE OF THE TRANSPLANTED TUMOR Bashford, Murray, and Cramer,^ after two years' observation of al- most three thousand mice with propagable tumors, concluded that the presence of such a growth, even when it was of greater weight than the mouse itself, did not necessarily involve any disturbance of the normal nutrition which could be regarded as comparable to the cachexia fre- quently associated with mahgnant new growths in the human subject. The animals were able to support large growths for several months, and even massive tumors in the abdominal ca\-ity and extensive metas- tases in the lungs, \vithout visible inconvenience. But when abrasions occurred in the skin overlying a tumor, the hemorrhage and septic infection which followed speedily produced marked constitutional ef- fects, among which emaciation was almost constantly present, and the mice seldom sur\dved for many days. Taken as a whole, the condition of animals mth ulcerated tumors closely reproduced that of cachexia in man, but in the case of subcutaneous tumors which did not involve any important organ, cachexia could be definitely assigned to a position of secondary importance as an occasional accom- paniment of mahgnant growths, to which it had no essential relation whatsoever. Moreschi '* investigated the association between tumor growth and nourishment, using for the purpose a mouse sarcoma of great proHfera- 1 Prcc. Roy. Soc, Series B, 1907, Ixxix, 187. Third Set. Report, Imperial Cancer Research Fund, London, 1908, 340. ^ Surgery, Gynecology, and Obstetrics, 1906, ii, 633. ' Sci. Reports, Imperial Cancer Research Fmid, London, 1905, No. 2, Part ii, 40. * Zeitschrifi f. I mmunitdtsforsch., etc., Orig., 1909, ii, 651. THE TRANSPLANTED TUMOR 93 tive energy. He found that a moderate restriction of the amount of food was followed by a retardation of tumor growth and a lengthen- ing of the life of the animals as compared with controls fed in the usual way, while an extreme condition of malnutrition preceding in- noculation might cause a great decrease in growth, or even prevent it entirely. The death of a tumor-bearing animal Moreschi referred, not to a lack of the ordinary food-stuffs, but to withdrawal by the tumor of a specific, autogenous, nutrient material essential to life. Following up these experiments, Rou s ^ inqui red whether it might not be possible to d elay the cour se of inoperable tumors, or to hinder the development. of metastases after the excision of a primary growth, by restricting the diet. In an attempt to answer these questions a series of experiments was undertaken with the Flexner-Jobling carcinoma of the rat. Contrary to expectation it was found that large tumors continued to grow with the same rapidity in hosts emaci- ating on an insufficient diet as in controls fed on a full one, many of which were gaining weight. With still larger tumors, in hosts already cachectic, the withholdin£_of sufficient food had as little effect upon the tumor as it had upon the frequenQ^_jof metastasis formation. Moreschi's findings were thus corroborated, so far as concerned the in- fluence of limited nourishment upon the development of tumor grafts. When food restriction was begun four days after the introduction of the grafts it was noted that the tumors developed a little more slowly than in the controls. Medigreceanu - estimated the weights of various organs in rats and mice bearing transplantable or spontaneous tumors. The ratio of the ahmentary canal to the body weight (less that of the. tumor) was within the normal limits, while that of the lungs and spleen was vari- able. The relative weights of the kidneys of tumor-bearing (and pregnant) mice showed only such slight variations as were met with in normal animals, but a transplantable rat sarcoma, on the contrary, induced constantly an increase in the size of these organs which was directly proportional to the size of the tumors. However, as mouse sarcomata did not reproduce this phenomenon in mice, the change ^ Proc. Soc. Exp. Biol, and Med., 1910-1911, viii, 128 ^ Proc. Roy. Soc, Series B, 1909-1910, Ixxxii, 286. Berl. klin. Woch., 1910, xlvii, 588. 94 THE TRANSPLANTED TUMOR was not thought to be characteristic of the sarcomata. The heart was enlarged in r ats and mi ce bearing t umors, generally in proportion to the weight_of the^growth, but rapidly proliferating neoplasms did not lead to such marked hypertrophy as did those which were increasing more slowly and were richly suppHed with blood. Hypertrophy of the heart in tumor-bearing animals the author was inclined to refer to the action of mechanical factors. The pregnant normal animals examined did not present a corresponding cardiac enlargement. The important result of the experiment was, however, the demon- stration that the liver was hypertrophied in all animals bearing either transplanted or spontaneous tumors, and a general parallel subsisted between the weight of the tumor and the weight of the liver, whether comparison were made between different growths of the same strain, or tumors of different strains. This disturbance of the normal ratio was due neither to a loss of weight in the other organs nor to the at- tainment by the liver of a weight equivalent to what it would have reached in the same animal during the natural augmentation of the body weight. There were, moreover, differences between the Hxers of normal and tumor animals, such as the increased percentage of water in the latter (as high as 4 to 5% more than in normals), which in- dicated the presence of quaHtative differences as well. Histological studies had not indicated any anatomical changes. In a rat in which a tumor of ten to twelve grams had disappeared by spontaneous ab- sorption, the liver was still enlarged when examined twenty days after complete recession of the growth, a finding which the author contrasted with the fact that the hypertrophy of the Hver found in pregnant mice vanished a short time after the birth of the young. As a corollary to these experiments Medigreceanu ^ undertook to determine whether augmentation in weight were due to an increment in the diet. The simplest explanation of the circumstance would be that the growing tumor attracted nourishment to itself and left the host in a condition of partial starvation, as a logical consequence of which there would follow an increased appetite and the ingestion of a larger amount of food to satisfy the demands of the organism and its growing tumor. This hypothesis, if it were justifiable, would account for the hypertrophy of the Hver, which would have more work to per- 1 Berl. kUn. Woch., 1910, xlvii, 772. THE TRANSPLANTED TUMOR 95 form. Experiments directed toward the solution of this question demonstrated that there was no more food ingested by tumor-bearing rats than by normal ones, that the body weight (less tumor) increased for a time after transplantation, as in normal controls, and then began to diminish, and that during this period of loss in weight the appetite was decreased. The hypertrophy of the liver 'in t umor rats could not, therefore, be explained by an in creased inta ke of food. Amyloid degeneration has been described by several authors in the spleens of mice bearing transplanted tumors. Thus, Albrecht and Hecht,^ in discussing the enlargement of the spleen so commonly found in tumor mice, ascribed it to bacterial infection, and said further that amyloidosis finally made its appearance in these organs. Lubarsch ^ found amyloid degeneration of the spleen, liver, kidney, adrenal, and pancreas, in the order of frequency named, among forty- two mice with transplantable tumors. While the presence of amyloid was not dependent upon the histology of the growths, occurring as it did in animals bearing sarcoma, carcino-sarcoma, and carcinoma, it was of more frequent occurrence and of earlier appearance among mice with sarcoma or carcino-sarcoma than among those with pure car- cinoma. No relation could be demonstrated between amyloid degen- eration and the size of the turr^or. Lubarsch did not beheve that the presence of amyloid could always be referred to an antecedent bacterial infection, and suggested its possible association with the products of the necrosis of tumor tissue. Freytag ^ examined forty mice with non-ulcerated tumors in which bact eria were not microscopically demon strable. In most of the cases the spleeiL.was_erilarged, and in thirty-three it was the seat of amyloid degeneration. The livers in twenty-six showed the same lesion, and the kidneys a trace of it in eleven. Among sixteen mice which had been inoculated unsuccessfully, amyloid degeneration could be demonstrated in the spleen in nine and in the Kver in five, but in only two was it present in the kidneys, and in both of these the amount was very limited. It thus appeared that tne cell destruction resulting from one or two fruitless inoculations might inaugurate an amyloid degenera- 1 Centralbl. f. allg. Path., etc., 1909, xx, 1039. ^ Ceniralbl. f. allg. Path., etc., 1910, xxi, 97. '^ Zeitschrift f. Krebsforsch., 1910-1911, x, 164. 96 THE TRANSPLANTED TUMOR tion, although it was possible that in an animal like the white mouse which, as Davidsohn ^ had already shown, was very prone to amyloid degeneration of its organs, this lesion might occur spontaneously with relative, frequency. Twenty normal mice were, therefore, examined, but without anything of moment being found in either spleen or kidney; although the Hver often showed a sHght degree of fatty degeneration, the amyloid reaction was always negative. Still, these mice were all very young, and experiments were in contemplation which should deal with normal animals that had been kept as long in confinement as the tumor mice, and upon a similar diet. Price- Jones ' studied the blood of sixteen mice with transplantable carcinoma, and found a slight degree of anemia, indicated by a lowered hemoglobin content and a diminution of the number of red blood cells per cubic millimeter. A well-marked leucocytosis was present, due to a great relative and absolute increase in the large mononuclear lymphoid cells and the polymorphonuclear leucocytes. Little change could be found in the marrow of these mice for, owing to the aug- mented myeloid activity of their spleens, any increased vigor on the part of the marrow was either masked or obviated. Thirteen normal mice served as controls. A diniinutiorLin^oxygencapacity with a fall in the p ercentage of hemoglobin parallel to the decrease^ in nuni ber of re d cell s per cubic millimeter, was found by Chisho hn ^ in the blood of rats^Jbearing tranjpJanted_sgxcQmata. The anemia, which was usually accom- panied by a wasting of the body tissues, did not depend on the presence of ulceration nor was it in all cases proportional to the amount of necrosis in the t umo r. Signs of regeneration present in the blood smears indicated that the anemia was due, at least in part, to bloo d destruction, but the mechanism of its production was uncertain. The presence of the tumor was associated with an increase in the total vol- ume of the blood. In certain exceptional cases there was reason to beHeve that the animal increased both its oxygen capacity and blood volume to meet the needs of the tumor, but this occurred only in the case of small growths. ' Verhandl. d. deiitschen path. GcseUsch., 1904, yte Tagung, 41. 2 Arch. Middlesex Hosp., 1911, xxiii, 56. ^ Jour. Path, and Bad., 1911, xvi, 152. THE TRANSPLANTED TUMOR 97 In a review, Cramer ^ said that he had discovered no diminution in the total acidity of the gastric contents in mice that had been suc- cessfully inoculated with the Jensen tumor. Copeman and Hake- found an increase of the physiologically active hydrochloric acid (including both free acid and that combined with proteids and nitrogenous organic bases) in the gastric contents of mice and rats bearing transpLlanted._tumors, and in mice with spon- taneous growths. The stomachs of one hundred and fifty normal mice, where no account was taken of the period of digestion, gave an average of O.I 1 21% hydrochloric acid, while in one hundred and seventy-eight mice with transplanted tumors the average was 0.1752 %. An average of two hundred and forty-five stomachs from normal mice for periods of digestion varying from one to one and one-half hours gave 0.1456% hydrochloric acid, and of two hundred and ninety stomachs from mice with transplanted tumors, 0.1673%. Fifteen stomachs of mice which were the subjects of spontaneous neoplasms showed an average of 0.1929% hydrochloric acid during different uncertain periods of diges- tion. An average of the stomachs of six normal rats after one hour's digestion gave 0.1427% hydrochloric acid, and of seven rats with transplanted growths (varying from 0.3 to 15 grams) 0.1837% after the same period. Regarding experiments on animals bearing transplanted tumors Cramer ^ indicated the importance of recognizing the difference be- tween an animal with a propagable tumor and one the subject of a spontaneous growth, pointing out that the question whether the in- crease in hydrochloric acid in animals bearing transplanted tumors was connected in any way with the decrease of this substance observed clinically in man should still be considered, therefore, an open one. The same author,^ as the result of an inquiry into the gaseous metab- olism of rats bearing the Jensen rat sarcoma, found that the growth of a tumor in a normal organism led at first to physiological changes only, tumor growth causing a distinct increase in the weights of the ' Biochem. Centralbl., 1905, iv, 65. - Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 398. Proc. Roy. Soc, Series B, 1908, Ixxx, 444. ^ Metabolism and Practical Medicine, v. Noorden, London, 1907, vol. iii, 824. ^ Third Sci. Report, Lmperial Cancer Research Fund, London, 1908, 427. gS THE TRANSPLANTED TUMOR animals, and that pathological changes in the general metabolism did not ensue until the physiological resources had become exhausted. He compared the effect exerted on the body by a growing tumor, in so far as it was a problem of nutrition, to that of the growth of a fetus in a pregnant animal. It could not be explained by attributing to the cancer cell the formation of pathogenic substances of a h}^othetical nature, such as a "c ancer f erment" or a "canc er toxi n." Cramer's findings in the case of rats bearing tumors, and his com- parison of such rats to pregnant animals, form an interesting supple- ment to some experiments conducted many years ago by Edelfsen and Hensen.^ These authors demonstrated that female guinea-pigs grew at about the same rate whether they had young or not during their own period of growth. Minot ^ was able fully to confirm these statements, and concluded from his own experiments that gestation did not represent a t ax on the mo ther, but a stimu lus — that it favored, rather than_impeded, growth. Cramer and Pringle ^ determined the nitrogenous metabolism in three rats before and after the implantation of the Jensen rat sarcoma. Their conclusions were appKed only to animals bearing tumors of sufficient size to warrant the assumption that they would reveal any specific property or function which might be possessed by the cells of a neoplasm. The eft'ects which a large tumor must necessarily pro- duce, by \drtue of its mere mass, were not considered. The main out- come of the experiments was stated in the following conclusions : — " I. Less nitrogen is necessary to build up a certain weight of tumour tissue than is necessary to build up an equal weight of the somatic tissues of the host. "2. Animals bearing tumours maintain their positive nitrogen bal- ance, and the nitrogen retention actually increases with the size of the tumour. "3. In our experiments the cells of the new growth derived their nitrogenous material necessary for the building up of new tissue by a sparing action on the protein metaboHsm. The tumour cells do not proliferate at the expense of the tissues of the host, nor is there any 1 Arb. a. d. Kieler Physiol. Inst., 1868. Cited by Minot. ^ Jour. Physiol., 1891, xii, 141. ^ Proc. Roy. Soc, Series B, 1909-1910, Lxxxii, 307. ^? U ~\'- ,^^ G LJ /^ 0, o '^ ^C ^ \ ^jX.r^ cce^ 'o^^ '^<5 J. R. FoTd, del. Mammary chondro-osteo-sarcoma of mouse. Primary tumor. Spindle-cell sarcoma; at left of figure an acinus lined by columnar cells — mouse nearing end of pregnancy. X -j-. WI^^^^S^ J 1 I rd a I Transplanted tumor from first generation of growth pictured above. Illustrates also the structure of the recurrent spontaneous tumor, which contained osteoid nodules, spindle cells, and cartilage. THE TRANSPLANTED TUMOR 99 e\ddence that they have a higher affinity for nutritive material than the growing cells of the host. "4. There is no evidence of the existence of substances secreted by the tumour disturbing the nitrogenous metabolism by means of a toxic action on the tissues of the host." If now, said the authors/ it were true that less nitrogen was needed to build up a given weight of tumor tissue than was necessary to build up an equal weight of the somatic tissues of the host, it would follow that cancer would have a lower nitrogen percentage than the somatic tissues. The estimation of the total amount of nitrogen (pro- tein and non-coagulable nitrogen) in various tissues of the three rats mentioned in the previous article was, therefore, undertaken, and the authors were able to show that, weight for weight, the malignant cells contained only about three-fourths of the protein nitrogen present in the tissues of the host. In other words, with the same amount of pro- tein a larger mass of tumor than of host tissue could be evolved. In a former paper Cramer had indicated the similarity which existed between the growth of cancer and the growth of the fetus, and prelim- inary experiments by Dr. Lochhead had shown that a rapidly growing normal tissue (the fetus) had a lower nitrogen value than the maternal organism. i HISTOLOGICAL VARIATIONS OCCURRING DURING TRANSPLANTATION In the Parenchyma Distinct variations from the tunior_type originally tr ansplan ted may occur under exten ded propagation, and it is po^sible_for^these to be so marke d that, as pointed out by Bashford, Murray, and Cramer,^ if growths with the characters of various transplanted tumors in the same series had occurred sporadically they might well be regarded by pathologists as totaUy_disti nct conditions . Thus an acinous arrange- ment of the parenchyma was a nearly constant feature of tumors ex- amined within the first week after transplantation, while another characteristic was the loose columnar arrangement of tumor cells in large alveoH. At a subsequent stage the central parts of certain al- 1 Proc. Roy. Soc, Series B, 1909-1910, Ixxxii, 315. " Sci. Reports, Imperial Cancer Research Fund, London, 1905, No. 2, Part ii, 52. lOO THE TR.A.XSPLAXTED TUMOR veoli underwent a necrotic change and, indeed, in advanced states of this process the alveoH might consist only of a thin layer of healthy cells and completely necrotic granular central areas. On the other hand, equally large alveoh occurred which were completely filled by healthy cells shoTsdng no traces of necrosis. All the conditions just discussed might be discovered in different areas of the same groAvth. The authors had been able to determine that these different histological appearances were not demonstrable at uniform fixed periods after the date of inoculation, whence they concluded that the time of their pro- duction was not determined by the disturbance of nutrition following transfer from one host to another. ^vlurray ^ encountered a tumor which he at first regarded as a spindle cell sarcoma, recurrences of which contained hyaline cartilage and un- calcified bone. Daughter tumors propagated from recurrences showed the same great variabihty of structure, some consisting of spindle cells, others showing large masses of cartilage either associated with osteoid tissue or alone, while in one instance there had been a deposition of hme salts in the osteoid matrix, so that true bone had actually been formed. The subsequent history of this tumor as given by Bashford^ Avas that, after the fourth generation, cartilage and osteoid tissue disappeared and two strains with divergent characters were obtained, in one of which the growths were soft, consisting of closely packed spindle cells Avith \'ery little interstitial collagenous tissue and large necrotic areas. The second, deriA'ed from a sclerotic tumor of the eighth generation, dift'ered in seA^eral particulars. Its tumors Avere firm, strongly collagenous, Avith httle or no necrosis even AA'hen they had reached a large size, and these dift'erences had been maintained, at the time of AArriting, for over tAvo years. ^Murray ^ reported, further, a mammary tumor of the mouse contain- ing three distinct histological structures — keratinized alveoH, solid ah-eoK, and adenomatous areas. In the first feAv generations of trans- planted tumors the groAvths generally presented a sohd structure Avith only minute areas of keratinization, a condition AA'hich continued Avith a sHght progressive increase in the amount of keratin until ^ Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 78. '^Fourth Sci. Report, Imperial Cancer Research Fund, London, 1911, 178. ^ Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 159. ' i '^« *..,.. \ ^,i n -,\ tM-St^ oj OH ■rf X d rt -d C/3 o 3 fi O H 15 a o d UJ ^ -U '13 c3 »fh o '^^.^ ,*v X 3 r-* ^ bJD rf 'Oh a 1 i ;h 03 O s el .s Ji en 3 ^ bX) a tfl o a, W THE TRANSPLANTED TUMOR lOI the sLxth generation. In the seventh, however, keratinization ap- peared to an even more pronounced degree than in the primary tumor and endured for three or four generations, after which it dimin- ished. At the conclusion of this period the tumors consisted of large alveoH of closely packed, small cells, interrupted only by an occasional capillary. Upon this medullary condition of the paren- chyma, and after it had persisted through several generations, there intruded a well-marked adenomatous structure, which involved some tumors to such an extent that considerable portions of them pre- sented the appearance of an adenoma. Others exhibited the formation of adeno-carcinoma, while still others had preserved their alveolar arrangement. Murray pointed out that separation of the keratinized and adenomatous t}'pes by a long interval of time, during which there had been many successive transferences from animal to animal, proved that the two differentiations, however distinct they might appear, were inherent in cells of one kind. It was of merely academic interest to argue whether the original growth should be considered a squamous cell carcinoma capable of grovvdng as an adeno-carcinoma, or, conversely, an adeno-carcinoma of the mamma in which excessive keratinization had occurred. Bashford ^ has described the subsequent hist ory of this strain, which at the time of writing had been propagated through e ighty- two generati ons. The a cinous struct ure disappeared after the t^/ ^lfth gene ration and since then the tumor had been culti- vate(l as a pure al veolar carci noma without any differentiation whatso- ev-er, except in one very old tumor (eight months) of the fifty-sixth generation, which showed areas of keratinization. A mo ^t com prehensive review of the behavior of the tumor cell during extended^propagation has been written by Bashford,^ based upon observation of spontane ous tumor s in over s ix hun dred and_fifty mice, as well as eighty-five propagable tumors, thirty-five of which had been in cultivation for more than three years. Generally speaking, the results following the continuous propagation of an extremely varied series of tumors showed a remarkable average constancy over long periods among both the carcinomata and sar- comata. The morphological variations which had been observed ^ Fourth Set. Report, Imperial Cancer Research Fund, London, 1911, 155. . "^Fourth Sei. Report, Imperial Cancer Research Fund, London, 1911, 131. I02 THE TR.\NSPLAXTED TUMOR were but slight and, T\dth few exceptions, served only to emphasize this statement. The apparent loss of acinous structure in the later generations of many adeno-carcinomata of the mamma was hardly to be considered as exddence of a real change. In some tumors, examina- tion at a short interval after transplantation revealed a purely alveolar condition, which gradually gave wa}^ to a t>^ical acinous structure as the tumors grew older, and which might even be entirely replaced by perfect acini in very old tumors. This fact showed the necessity for examination of old tumors in all cases where normal histological differentiations present in the primary growth or in the earlier gen- erations of daughter tumors were absent in young tumors of later generations used for transplantation. WTien this precaution was observed, adeno-carcinomata primarily hemorrhagic might be found to reproduce their original hemorrhagic character, even though it had been absent in the periods immediately succeeding transplantation. The same statement could be applied, also, to certain of the squamous cell carcinomata. The differences in histological structure between the spontaneous tumor and its early daughter tumors on the one hand, and the later transplanted growths on the other were, therefore, not necessarily final e^-idence of a real change in the character of the parenchyma in ever}' case in which they were found. As already stated, the general conclusion was drawn that the cells of many tumor parenchymata preserved unaltered and in the most per- sistent manner the majority of the characters with which they were endowed from the earhest period at which they had come imder obser- vation. The various strains did not approximate a common type either in structure, rate of growth, or inoculation percentage ; and while a number of strains might be selected as forming a group homogeneous in naked eye structure and in histology, it could be demonstrated that the similarity was primary and already associated with the sponta- neous tumors, rather than a secondary convergence from an initial heterogeneous condition. The conclusion seemed warranted that the relative permanence of the distinctive characteristics in typical tumor strains had its foundation in the cellular changes by which the non- cancerous tissue cells had passed over into the cancerous state, and this persistence suggested how radical must have been the alterations of which cancerous transformation was the expression. (jiOQci''^ - .•f» -V .a'^ ^cTO C«l%, *^T^ "^tiot <. o\.^\ -SK^^j % • ■CiO Ci *^1 Qj ■tllS ^^ c -a ; o 2 ^'^.tS^s ^o<^ 0& •u> "b O ^J< r£j O be "^ _ 'U T3 ;3 O .2 > ^ 7=; .5 to THE TEANSPLANTED TUMOR lOQ the induction of experimentally increased resistance, structural changes of slight degree, such as a tendency to assume the alveolar type with a loss of the characteristic adenomatous differentiation. This relationship could be explained by the inhibition, in resistant animals, of the stroma reaction induced by the cancer cell in normal ones. It was clear that the construction of a perfect adenomatous con- dition required an intimate relationship between parenchyma and connective tissue, and that if the stimulus of the cancer cells upon the stroma were to be decreased or abohshed, they would remain in masses and the growth would contain a minimum of connective tissue. The reappearance of acini in Apolant's experiment may have been due, according to Bashford,^ to the comparison of an old, slowly growing tumor with more rapidly growing and younger controls." In tlie Stroma Bashford, Murray, and Cramer,- and Apolant^ found that, as a general rule, the character of the stroma in hemorrhagic tumors was preserved during propagation, a fact which Ehrlich ^ attributed to a chemo tactic power resident in the cells of the parenchyma, by virtue of which they were enabled to attract angioblasts. On the other hand, as Bashford '" has observed, the hemorrhagic character of the stroma may be lost or diminished during continued transplantation. An extended consideration of such changes as this may, however, be dismissed as unprofitable while there remains to be reviewed the assumption of sarcomatous properties by stroma cells previously benign. This significant deviation was first observed by Ehrlich and Apolant^ in a transplantable adeno-carcinoma of the mouse, in the tenth generation of which there were found two growths containing both carcinomatous and sarcomatous areas. Exactly at what point the change had occurred the authors were unable to say, but it must have 1 Fourth Sci. Report, Imperial Cancer Research Fund, London, 191 1, 170. 2 Sci. Reports, Imperial Cancer Research Fund, London, 1905, No. 2, Part ii, 30. ' Verhandl. d. deutschen path. Gesellsch., 1908, late Tagung, 7. * Arb. a. d. Kmigl. Inst.f. Exp. Therap., 1906, Heft i, 69. ^ Fourth Sci. Report, Imperial Cancer Research Fund, London, 191 1, 154. ^ Berl. klin. Woch., 1905, xlii, 873. no THE TRANSPLANTED TUMOR been present in the ninth generation, because two of the descendants were of a mixed character. Those of the first, second, and sixth generations, however, were pure adeno-carcinomata. The mixed growths contained carcinoma cell nests separated from one another by strands of large spindle cells in which mitosis was actively progressing, while in some parts of the sections the process was so advanced that the carcinomatous elements had disappeared entirely. In one tumor of the thirteenth, and in all those of the fourteenth and subsequent generations, no trace of carcinoma was discoverable, pure sarcoma having been evolved through the gradual elimination of the epithelial components. The growth had been cultivated as far as the twenty-sixth generation without showing the slightest variation from the characteristic structure of a pure spindle cell sarcoma. Two explanations were advanced to account for the transformation. An alteration in the chemistry of the cancer cell was conceivable, in the course of which there was elaborated a material possessing the power to incite connective tissue cells to malignant growth. This hypothesis was later extended by Ehrlich ^ to include and explain a difference that had been observed between the grade of sarcoma development attained in individual mice of one and the same series. He suggested that this was best elucidated by assuming that the predisposition to the growth of connective tissue varied in different animals as, for instance, certain persons were more prone than others to the development of keloids. To this amendment, however, Russell ^ objected that the tumor with which he himself had been working was able to initiate the sarcomatous change in nearly every mouse and that practically all of them must, therefore, have been "keloid" mice, while Jensen's car- cinoma, which had probably been transplanted into more mice than any other tumor, had never yet happened to be inoculated into a "keloid" mouse. Ehrlich and Apolant's alternative explanation for the sarcomatous change was that repeated transplantation of cancer cells and connective tissue evoked a proliferative energy in the cells of the latter which advanced to the degree of malignancy. This latter hypothesis was excluded by Bashford, Murray, and ^ Zeitschrift f. Krebsforsch., 1907, v, 64. ^ Jour. Path, and Bad., 1910, xiv, 374. THE TRANSPLANTED TUMOR III Cramer ^ on the ground that Jensen, and afterward they themselves, had demonstrated that the stroma of the introduced graft always degenerated. The explanation they therefore considered to be at variance with the facts. To this criticism Ehrlich and Apolant ^ answered that the hypothesis in question did not in the sHghtest degree conflict with the commonly recognized fact that the inoculated stroma usually died out. In the course of many transplantations, and under the influence of the carcinoma, it might occasionally happen that the stroma cells suffered a transformation which allowed them to remain and grow Hke those of a carcinoma. Bashford^ replied that the authors had neglected to eliminate certain sources of error, and until that were accomplished it would be impossible to decide whether the growth had been a mixed tumor from the beginning, whether an infectious granuloma had arisen upon a carcinoma, or whether a true sarcoma had evolved itself from the introduced stroma or the reaction tissue of the host. That a sarcoma might arise during the transplantation of a carcinoma was, a priori, not impossible. EhrKch and Apolant^ retorted that Bashford had never had under observation a tumor of reaUy high virulence, and that in this respect his experience was imperfect. As regarded Bashford's first objection, the authors had already stated that neither in the primary growth nor in the daughter tumors immediately following its transplantation had there been any departure from the usual appearance of carcinoma of the mouse.'^ So far as concerned the dift'erential diagnosis between granuloma and sarcoma, the most eminent pathological judgment in Germany had concurred in favor of the latter ; and, furthermore, such enormous growth energy as was exhibited by the tumor under discus- sion never had been seen in an infectious granuloma. ^ Bed. klin. Woch., 1905, xlii, 1434 (footnote). ^Berl. kliii. Woch., 1906, xliii, 477. ^ Bed. klin. Woch., 1906, xliii, 39 (footnote). ^ Bed. klin. Woch., 1906, xliii, 668. 5 When Apolant demonstrated two of these primary tumors, however (Verhandl. d. deidschen path. Gesellsch., 1905, 9*^ Tagung, 168), v. Hansemann said that he had found spindle cells in the stroma of one of them, and therefore suggested that the second growth be subjected to the closest scrutiny, to ascertain whether they might not be present in other areas. But Apolant thought that the existence of the cells described by v. Hanse- mann would hardly warrant one in calling the growth a mixed tumor. 112 THE TRANSPLANTED TUMOR The diagnosis of sarcoma was once again defended by Apolant,^ on the double basis of cHnical behavior and histology. These tumors grew just as rapidly as they had before the change took place, yielding growths the size of a plum within three weeks, and infiltrating the sur- rounding tissues just as seriously. The histology permitted no doubt of their sarcomatous nature, for in addition to the typical arrangement of their elements they presented from their very inception a number of mitoses not exceeded by that found in the most rapidly growing mouse carcinomata. Metastatic deposits, however, had not been discovered. Ehrlich and Apolant ^ reported two further instances, the first of which arose between the twelfth and fourteenth generations of an alveolar carcinoma which had originated from a mixture of four epithelial tumors. Here the growth energy of the new sarcoma was less than in the first case, for the mixed tumor stage persisted through ten generations (six months) without change. The second case occurred during the trans- plantation of one of the four epithehal growths just mentioned. Little alteration had been observed up to the fortieth generation, but at that point an increased cellularity of the stroma made its appearance, although sarcoma development could not be indubitably estabhshed. It was confirmed, however, in the sixty-eighth generation, after two years and a half of propagation. It seemed significant to the authors that the sarcomatous change arose only in the later generations, for although it had occurred relatively early in the first case, after only nine months of transplantation, in the second and third cases it did not supervene until after two and two and a half years respectively. This fact explained why the transformation had not been observed more frequently, for at the time of writing only the Jensen tumor had been cultivated over a period of time at all comparable to these. Its viru- lence, however, was much lower than that of the Frankfort tumors. In later papers the authors returned to this question of the relation between virulence and sarcoma development, and were inchned to attribute great importance to virulence in the initiation of the trans- formation. That the length of time during which a tumor had been under 1 Arb. a. d. Konigl. Inst. f. Exp. Therap., 1906, Heft i, 56. 2 Berl. klin. Woch., 1906, xliii, 38. THE TRANSPLANTED TUMOR II3 cultivation did not deserve the importance at first ascribed to it they were able to record in the following year, after they^ had discovered one instance in which sarcoma development had certainly occurred in the spontaneous tumor, and a second where such a change had in all probability taken place there. A similar example was later recorded by Haaland,^ in which a spontaneous growth, removed by operation, was found to consist of a central carcinomatous part surrounded by a band of spindle cell sarcoma. Useful reviews of the cases of sarcoma development which have been observed at the Frankfort laboratory have been given by both EhrHch^ and Apolant.^ Loeb ^ described a primary non-metastasizing adeno- carcinoma in the submaxillary gland of a female Japanese mouse from six to eight months old. The tumor was of uniformly epithelial structure and belonged to the glandular type. It was possessed of ducts and alveoli ; the lumina of the latter contained colloid material, and the epithelial elements were supported by a diffuse connective tissue. Inoculation was successful in Japanese mice but failed in those of the white variety. In one of two surviving Japanese mice a few slowly growing nodules were found fifty-eight days after inoculation, one of which, removed a few days later by operation, proved to be a spindle-cell sarcoma. When the animal was found dead ninety days after inoculation and twenty- six days after the excision of the nodule, there was a recurrence which had grown rapidly and which consisted almost exclusively of pure sarcoma, and in addition a large glandular tumor in which sarcoma was developing at one point. The tumors of the second mouse, glan- dular growths with some sarcomatous elements in them, were used for transplantation ninety-three days following inoculation. They were ingrafted into six mice, one of which was found later to have developed a pure sarcoma, while the growths in four of the other mice were mixed. ^ Berl. klin. Woch., 1907, xliv, 1399. ^Fourth Sci. Report, Imperial Cancer Research Fund, London, 191 1, 19. ^ Zeitschrift f. Krebsforsch., 1907, v, 62. ^ Handbiich der pathogenen Mikroorganismen, Kolle and Wassermann,erster Erganzungs band, 1906, 452. Zeilschrift f. allg. Physiol., 1909, ix, Sammelreferat, 96. ^ Univ. Pennsylvania Med. Bidl., igo6, xix, 113. Berl. klin. Woch., 1906, xliii, 798. 114 ^^^^ TEAXSPLAMTED TUMOR The sixth had two. one glandular and one a pure sarcoma. In the second generation the sarcoma was in excess of the carcinoma, as it had been in Ehrhch and Apolants hrst case. Because the chances of further transplantation were excluded by a shortage of Japanese mice the experiment was perforce terminated. For the appearance of the sarcoma Loeb offered three explanations — that the tumor was originally a mixed tumor, that the sarcoma was derived from the connective tissue, or, finally, that a transforma- tion of glandular into spindle ceU tissue had taken place. The nxst he thought improbable, and left open the choice between the second and the third. If one accepted the second it would be possible to eliminate the theon.- that sarcoma development was the result of repeated inoculation of stroma, because the change had occurred in the first generation. It seemed most probable that the sarcomatous transformation was the result of a stimulus of u nk nown nature exerted upon the stroma by the carcinomatous moiet}' of the tumor. In a later paper he- said, in discussing the third explanation, that it had been impossible to find any evidence of transition stages between adeno-carcinoma and sarcoma in those areas where sarcomatous tissue was present. The article concluded with a reaffirmation of the h^-pothesis that the probable cause of sarcoma development was a stimulus exercised upon the surrounding connective tissue by the ceUs of the carcinoma. Haaland,^ investigating the occurrence of metastases in the second of EhrKch's sarcoma strains, found that microscopic nodules occurred in the lungs of 60% of his mice with tumors that were from fourteen to forty days old. The im portanc e of this obserx^ation lay in the fact that the method by which a tumor spread through the organism was a sound .criterion for the di stinction between tr ue mahgnant gr owths and the infect ious granulomat a. By examining serial sectio ns Haaland could show that the metastatic embo h reached the lungs through the b lood strea m, determining their relation to the vessels by the aid of Weigert's elastic tissue stain. For the investigation of the stroma either van Gieson's or I^Iallorx^'s method was employed. ^ Zeitsckrift f. Krebsforsch., 1908-1909, \"ii, loi. * Berl. klin. Woch., 1906, idiii, 11 26. Zeitschrijt f. Krebsforsch., 1907, v. 122. THE TEANSPLANTED TUMOR I15 It was probable that a considerable number of emboli were destroyed, or at least did not grow, for he fomid them occasionally, as Sclnmdt^ had described them in human lungs, surrounded by collections of round cells and leucocytes, and the seat of regressive changes. The majority, however, were in active growth, filhng and distending the lumina of the vessels ; in this intravascular stage there was no sign of reaction either in the vessel wall or the surrounding lung tissue. In their further development these emboh assumed an infiltrative growth during which their elements rapidly penetrated the wall and, infiltrating the neighboring tissues, surrounded the vessels with a mantle of tumor cells. Larger nodules were thus evolved which entirely destroyed the lung parenchyma and spared only the elastic tissue. While the stroma of these tumors was sparse, and in most cases hardly recognizable in the early stages, there were a few instances in which MaUory's stain would demonstrate a few very dehcate, sinuous, blue-stained fibrils between the individual cells, even before the penetration of the vessel wall. In the later stages, when the wall had been broken through, the histological picture was compKcated by the fact that portions of the connective tissue of the lung, and of that accompanying the blood vessels, had persisted and were partici- pating in forming the stroma of the new nodule. But even here stroma development remained within certain bounds, being limited to the re- production of a few isolated fibrils between the individual tumor cells. Liepmann- had under observation a mouse carcinoma which, at the eighth transplantation, had assumed the character of a carcinoma sarcomatodes, although as far as the fifth generation the typical carcinomatous structure of the spontaneous tumor had been retained. Bashford, Murray, and Haaland ^ described the appearance of sarcoma during propagation of a mammary adeno-carcinoma ("37 ") of the mouse. The transformation occurred in the seventh and eighth generations in two strains, one of which had always been transplanted with the needle, and the, other, after the fourth generation, with the syringe. Seven tumors out of ten in the seventh generation were transplanted, giving rise to the eighth. The usual systematic examina- 1 Die V erhreitimgswege des Carcinoms, etc., Jena, 1903, 41. ^ Miinch. med. Woch., 1907, liv, 1345. ^ Berl. klin. Woch., 1907, xliv, 1238. Il6 THE TEANSPLANTED TUMOR tion of the growths in each generation had revealed no change in the stroma before the seventh. At this point two again were unchanged, while in two others there was a slight increase in the amount and cellularity of the stroma, and one of them contained a small area of spindle cells. In the remaining three the stroma had increased greatly in amount and now formed broad bands of elongated cells lying among the carcinomatous elements. Transplantation of this gen- eration resulted in a yield of twenty- two tumors, in twenty of which there were found spindle cells arranged in bands and undergoing rela- tively rapid division. The examination of grafts at short intervals after transplantation showed conclusively that the stroma did not die out as it had in the previous generations, but, on the contrary, that it remained alive, and, moreover, possessed the power of independent proHferation. In older tumors of the eighth and in those of the ninth and tenth generations, the carcinomatous portions were sur- rounded by halos of clear polymorphous cells, larger than those of the carcinoma and distinguished from them by their lightly stained protoplasm. That they were not of epithehal origin was certain. In fact, one could find all transitional stages between them and the typical spindle-shaped stroma cells, which appeared to be a more highly differentiated form of the cell now under discussion. Although the zones just mentioned gave the impression of a reaction tissue composed of elements newly arisen from those of the host, the study of young grafts demonstrated that the clear cells did not die out after transplantation, but were able to proliferate and to produce sar- coma. The authors believed that, although endowed with the power of independent growth, these cells had not yet lost entirely their old relation to those of the carcinoma, but had, on the contrary, remained sensitive to influences emanating from these elements and were accordingly arranged about the alveoli in the manner of the old stroma. Only the stroma cells far removed from the carcinoma were exempt from this influence and able to differentiate characteristically into bands of spindle cells. As the stroma of three tumors in the same series showed sarcomatous transformation, it seemed more Hkely that some special stimulus over and above the ordinary stroma-forming stimulus had been exerted upon the fibroblasts by the carcinoma cells, than that the sarcomatous J. R. Ford. del. 3ne of four growths (41 days old) of Series D, 7th generation of tumor 37, which showed first cases of sarcoma development. Abundant cellular interstitial connective tissue, consisting of large spindle cells, runs between the islands of acinous parenchyma throughout the whole tumor. X -^. m>€^ (^ e:§^-> J^. iS. Fortf, del. Iixed tumor of the loth generation, 23 days old. Broad bands of sarcomatous tissue containing mitoses, and separating the acini of the carcinoma. X ^^. THE TRANSPLANTED TUMOR 117 change had been initiated through any particular attribute of the mice in which it had taken place. The stroma, now able to proHferate independently, was no longer subser\'ient to the needs of the epithehal parench}Tna in the matter of pro\dding nutrition for its cells, and, indeed, it even seemed as though the carcinomatous part of the tumor was isolated from the blood vessels by the halos of clear cells. Whether or not this explained the disappearance of the carcinomatous elements, the contrast in size between the succulent cells of the sarcomatous stroma and those of the epithelial parench}Tna was striking. Both the nuclei and proto- plasm of the latter appeared to diminish in size, while mitoses became less and less frequent as the transformation of the stroma proceeded, until finally complete disappearance supervened. An attempt was made, with a fairly successful outcome, to hasten ''purification" of the tumor by introducing it into mice resistant to carcinoma. One tumor showing a pronounced decrease in the amount of carcinoma was obtained from ten inoculations. In the next trans- plantation, which was made into normal mice, only traces of carcinoma could be discovered, and in the next succeeding one a pure sarcoma was found. In none of the controls was there obtained, within the same period, a tumor of purely sarcomatous type. The quantitative relation between carcinoma and sarcoma varied somewhat from tumor to tumor, but as a rule the progress of the sarcoma was uninterrupted, although in isolated cases the carcinoma appeared to obtain the mastery, so that the regaining of a pure car- cinoma seemed possible of achievement. The authors suggested two ways of accounting for the appearance of sarcoma. Either the tumor was from the start a mixed one, or its cells had the power to exert a specific stimulus upon the connective tissue, causing its elements to become capable of continued prohfera- tion and thereby transplantable. Systematic study of early stages led Bashford and his associates to accept the latter explanation, in doing which they were in agreement with the hypothesis advanced by Ehrhch and Apolant. None of the three factors, length of propaga- tion period, virulence, or histological structure of a tumor, seemed to be of moment in determining whether the growth would or would not be able to induce sarcomatous transformation. Il8 THE TRANSPLANTED TUMOR Loeb ^ also was of the opinion that it was not necessary for the carcinoma to be possessed of any particularly high degree of virulence. In the rat, Lewin ^ observed sarcomatous transformation in the fifth generation of a keratinizing adeno-carcinoma, and was inclined to consider that endothelial cells were the ancestors of the sarcoma. Orth,^ who examined Lewin's preparations, was not convinced that the tumors had actually undergone a sarcomatous change, and expressed the behef that the new cells were those of ordinary granu- lation tissue. Sticker ^ advanced the supposition that many cases of so-called sarcoma development were either the issue of the inoculation of a mixed tumor, or else represented the simultaneous occurrence of a trans- plantable carcinoma and a spontaneous sarcoma in the same animal. Haaland^ described the change in other strains of the tumor re- ported by Bashford, Murray, and Haaland, pointing out that sarcoma development had hitherto taken place so unexpectedly that observations had of necessity been made long after the actual occur- rence of the transformation. The material available had, therefore, been incomplete. The present case, however, was encountered under much more favorable auspices, for, as the primary growth had been examined on two occasions and sections from every tumor in the fol- lowing generations were at hand for histological survey, there were no gaps in the continuity of the earlier material. Moreover, it had been possible to supplement the morphological data thus obtained with a comparison of the biological peculiarities of different strains, be- cause the chnical behavior of all the tumors had been recorded on a series of charts. In addition to this investigation of general histologi- cal and biological characters, the processes in play at short intervals after transplantation had been examined, and the beha\dor of the tumor at any period after the moment of inoculation could thus be compared with its behavior at any later stage. Finally, the carci- noma had not been lost during the development of the sarcoma, 1 Deut. med. Woch., 1908, xxxiv, 25. ^ Zeitschrift f. Kre.bsforsch., 1907-1908, vi, 273. ^ Zeitschrift f. Krehsforsch., 1907-1908, vi, 431. ^ Zeitschrift f. Krehsforsch., 1907-1908, vi, 431. ^ Jour. Path, and Bad., 1908, xii, 437. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 175. .? w /T. ^ ^'. ^ . Iff**" V J. R. Ford, del. 24-day-old mixed growth from 9th generation of tumor 37, to show halos aroimd healthy alveoli. X circa ^¥^. THE TEANSPLANTED TUMOR II9 and ten separate strains of purely epithelial tumors had been pre- served. Haaland and his colleagues had been enabled, therefore, to make an exhaustive study of the question from every standpoint — the histology of the primary growth, the progressive advance of the sar- comatous change, the condition of the stroma in old tumors of carcinomatous strains as well as in new grafts, and the biological characteristics of carcinomatous, mixed, and sarcomatous strains. Although the mother material of the four tumors in which sarcoma development had arisen had been examined in the hope that the route taken by the stroma cells could be retraced, the change seemed to have been a sudden one, and no indication could be discovered of the man- ner in which it had been initiated. The progressive advance of the transformation took place by an increase of spindle cell tissue from one generation to the next, and the alveoH became more widely separated from one another in con- sequence, often showing slight necrotic changes due to impairment of nutrition. The next step was the interpolation of an intermediate period where for several generations the spindle elements were replaced by polymorphous cells, and as a constant and characteristic feature of this stage lightly staining and extremely polymorphous elements collected about the carcinomatous acini, forming halos sharply de- marcated from the latter on their inner circumferences, but shading off gradually into the sarcomatous stroma on the outer. The halos varied from a single row of cells to the more usual width of several layers, and the nuclei w ere very large and sometimes mu ltiple. Often the nuclei^ were of the true__gian^t>'pe, v erv^rich irLxhrmnatin . As long as any epithelial components remained in the tumors, so long did polymorphism persist, but with the entire disappearance of the carcinomatous cells those of the sarcoma resumed their primary t>^ical spindle shape. What was the derivation of these Hghtly staining polymor phou s cells ? The idea of an epithehal origin Haaland felt unable to enter- tain and, as they prohferated after transplantation, they could not be the elements of ordinary granulation tissue. It seemed most prob- able that they did not originate from the connective tissue of the new host, but that they were derived from preceding sarcoma cells and 120 THE TRANSPLANTED TUMOR transferred as such with the graft, and this explanation was made even more plausible by the discovery throughout the body of growing metastases made up of similar cells. The alveoh that were surrounded by halos very often showed central necrosis, ordinarily a rather rare occurrence in the alveoH of the tumor under discussion. These degenerative changes Haaland referred to the fact that no trace of capillaries could be found inside the halos, and the cells of the latter thus seemed to be interposed between the alveoh and the capillaries of the stroma. The sarcomata of this strain had a greater initial rapidity of growth than the carcinomata and, on the whole, gave a higher percentage of growing tumors, although spontaneous absorption was of very fre- quent occurrence. Their power of infiltrative growth much exceeded that usually evinced by the carcinomata, and they metastasized readily, producing secondary growths in the lungs, heart, liver, kidney, spleen, and lymph nodes. Mixed growths tended to become pure sarcomata, but the process required a certain length of time and seemed to occur independently of the rapidity of passage from host to host ; and while the disappear- ance of the epithehal part of the tumor was generally accompUshed first in the center, the rate at which the process went on in the different portions was variable. To see whether the transformation were a sudden one Haaland investigated old. and young growths from strains in which the change had occurred, as well as from others which had been consistently free from it. In all the later cases of sarcoma development he found the first alterations at the center of the carci- nomatous tumor, and here sclerotic lesions were habitually present, often accompanied by more or less cellularity of the stroma. The sarcomatous change suggested in some cases a process similar to this, although enhanced in degree. It was not suddenly gained, but ap- peared to be a gradual acquisition, the power of continuous growth and the rate of proliferation becoming more advanced in its later stages. The condition seemed to be, on the whole, the result of a gradual", fluctuating evolution. The dependence of sarcoma development upon virulence, as upheld by EhrUch and Apolant, was not a phenomenon of universal distri- bution, for in the present case the tumor was one of rather slow growth. THE TRANSPLANTED TUMOR 121 OOOQOOOoOOOpQQQO O O 122 THE TR-\XSPLAXTED TUMOR Albrecht and Hecht ^ confessed that they had been unable either to initiate or to influence the transformation of carcinoma into sar- coma. Russell.- during the propagation of a recurrent spontaneous hemor- rhagic mammary adeno-carcinoma of the mouse, observed the assump- tion by its stroma of the biological and morphological characters of sarcoma. The determining factor appeared to be continuous prolifer- ation in one animal for about tifty days, for rapid passage, that is, the transplantation of the tumor at interA'als of about thirty days or less, preserved the purely carcinomatous character of the growth. Once the sarcomatous change had occurred it tended progressively toward the entire ehmination of the carcinomatous component, and the tumors could then be pro_pagated indefinitelv as pure sarcornata. Upon the initiation of the transformation there was no change in the speed of growth, and tumors which had been cultivated through one or more generations as mixed growths retained the rate of the pure carcinomata. But T\-ith the disappearance of the epithehal elements the rate of development increased greatly, whence Russell suggested that they might be capable of exerting a restraining influence upon the proliferation of the sarcomatous areas. Although the sarcomata were of rapid growth the}' were ver\' prone to spontaneous absorption, and mice in which this had taken place were resistant to the inoculation of the purely carcino mato us strains of the tumor. As to the explanation of the transformation the author was in entire agreement with pre\'ious observ^ers, be^e^'ing that it was induced in the connective tissue cells of the host by the parench^mia of the epithehal tumor. The change in the stroma, which was generally unicentric. began toward the middle of the growth in an area containing a relatively enor- mous number of mitoses, and of such small size as to suggest ver}' recent origin. In point of time the transformation usually occurred between the flfty-fifth and sixtieth days, and among one hundred tumors ex- amined only four or five showed no change before the sixtieth day, 1 Coitralbl. f. allg. Path., etc., 1909, xx, 1039. Wien. klin. Woch., 1909, xxii, 1740. ^ Jour. Path, and Bad., 1910, xiv, 344. THE TRANSPLANTED TUMOR 1 23 while one had advanced to pure sarcoma at the end of this period. The length of time required, however, was not uniform, for while in some series the onset of the sarcomatous alteration could not be de- tected until after seventy days, in others the change was present even after thirty-eight. That the variation was not due to individual dif- ferences in the mice was proved by the fact that the initiation of the transformation was uniform for all those of any given series. A stay of longer than sixty days in one animal was difficult to secure — other- wise each tumor would doubtless have progressed to the stage of pure sarcoma. A further case of sarcoma development was recorded by Stahr,^ affecting certain of the tumors in the ninth generation of a fissure- forming carcinoma. Stahr's observations differed somewhat from those of Haaland and Russell as regarded the point at which the change began, in that while these authors had recorded its first appear- ance in the middle of the tumor, Stahr found it earhest at the periphery. He agreed with Haaland, however, that the transformation was in some way connected with a low power of growth on the part of the carcinoma cells, and suggested that possibly it might be set going through the agency of materials elaborated by their death. Clunet ^ discovered still another instance of sarcomatous change in a growth belonging to the third generation of a malignant mam- mary cyst-adenoma of the mouse, where the transition was complete in every tumor of the sixth generation. Since then the growth, which had reached the thirty-eighth generation, had retained its sar- comatous character unaltered. The maHgnancy of the newly pro- duced tissue was manifest not only from its histology, but from its clinical behavior, proliferation continuing indefinitely and leading to death from cachexia. Furthermore, the tumors recurred after opera- tive removal, infiltrated the surrounding tissues, and produced meta- static deposits in the lungs. In discussing the - cause of the mutation, Clunet eHminated any chance that the primary tumor had been a carcinoma sarcomatodes, and accepted a mahgnant change in the stroma as the most probable explanation. He agreed with Haaland's observations on the absence 1 Centralbl. f. allg. Path., etc., 1910, xxi, 108. ^ Reckerches exp. sur les Tumeurs malignes, Paris, 1910, 53. 124 THE TRANSPLANTED TUMOR of any connection between tumor virulence and sarcoma development, for in the first six generations of the tumor, while the sarcomatous change was being initiated, the percentage of takes had been low and negative inoculations or spontaneous regressions frequent. As in Russell's experience, the power of growth increased rapidly after the epithehal components had disappeared. CULTIVATION OF CELLS IN VITRO One of the disadvantages of the ordinary method of investigating cancer was the irregularity inherent in the soil where the tumor cells were grown and in which, therefore, they had to be studied. Follow- ing the lead of Harrison,^ who showed that the nerve tissues of the frog embryo could be grown in lymph outside the body, other ob- servers have applied the method, with certain modifications, to the cultivation of the cells of various normal tissues as well as those of tumors. The use of lymph, however, proved inconvenient for several reasons, and Burrows ^ substituted plasma, which he obtained from blood that had been centrifugalized at a low temperature in paraffined tubes. In the opinion of Carrel and Burrows ^ the pure plasma thus obtained was much superior to oxalated plasma, which could, nevertheless, be used in cases of necessity. The authors described two types of culture — the small hanging drop, and the large plate culture. In making the former, one or two fragments of the tissue which it was desired to cultivate were transferred to a cover-glass and quickly immersed in a drop of plasma, which was spread out in a thin layer be- fore the occurrence of coagulation. The cover-glass was then inverted over a hollow shde to which it was sealed with paraffin to prevent evap- oration, and the preparation immediately transferred to the incubator. Large plate cultures were made by spreading fragments of tissue in a thin layer over the surface of a large, black glass plate, and covering ^ Proc. Soc. Exp. Biol, and Med. 1906-1907, iv, 140. Anat. Record, 1908, ii, 385. Jour. Exp. Zool., 1910, ix, 787. ^ Jour. American Med. Assoc, 1910, Iv, 2057. ^ Jour. Exp. Med., 191 1, xiii, 387. THE TRANSPLANTED TUMOR 125 them quickly with plasma. As soon as coagulation appeared the plates were put into glass boxes containing moist cotton to preserve the proper humidity, and the boxes were sealed with paraffin and kept in the incubator in such a position as to allow the fluid products of the culture to drain to the bottom. The most careful asepsis was, of course, necessary throughout the procedure. For the study of the cultures, a warm stage was employed on the microscope. Growing cells at the edge appeared as fusiform or polyg- onal bodies, the cytoplasm of which was filled with refractile granules. Often the movements of the Hving cells, their modification in shape, and the division of their nuclei could be readily observed. Cultures could be fixed and stained by removing the cover-glass with its ad- herent tissue to appropriate solutions or, in case the plasmatic medium was thick and the cells had grown in many planes, serial sections of the hardened cultures could be made. Carrel ^ reported in a subsequent contribution that cell division had been observed in cultivated tissues washed and placed in fresh media, as long as thirty-one days after their removal from the body, and that cultures had survived even nine transfers to fresh plasma. Lambert and Hanes,^ discussing the growth of rat and mouse tumors in vitro, wrote that the cells of sarcomata possessed the power of wan- dering out separately into the plasma, probably in search of nourish- ment, while those of carcinomata, on the other hand, were denied that function and moved outward in a continuous layer. The nuclei in both types frequently contained division figures, and the cells in the two cases were actively phagocytic. The same authors ^ found that rat sarcoma would grow in the plasma of im mune rats quite as vigor- ously as in that from normal or tumor-bearing animals, an observa- tion affording further proof, they beheved, of the absence of specific cytolytic substances in the body fluids of animals immune toward transplantable tumors. They"* furthermore established the fact that rat and mouse tumors would grow in plasma from ahen species, among which the order of ^ Jour. American Med. Assoc, 1911, Ivii, 1611. "^ Jour. Exp. Med., 1911, xiii, 495. ^ Jour. Exp. Med., 191 1, xiii, 505. * Froc. Sac. Exp. Biol, and Med., 1910-1911, viii, 123. 126 THE TRANSPLANTED TUMOR smtability was : Guinea-pig, rabbit, pigeon, man, dog, goat (no growth) . Those whose interest prompts them to inquire more deeply into the cultivation of cells outside the body should consult the papers of Burrows,^ Carrel,^ Carrel and Burrows,^ and Lambert and Hanes.^ Although all these investigators believed that they had observed cell growth taking place outside of the body, some doubt has been ex- pressed as to the correctness of this view. Jolly ^ did not think that Carrel and Burrows had succeeded in demonstrating the survival of cells, and expressed the view that certain of their descriptions were reminiscent of the phenomena of necrobiosis. And although, as was already known, cellular proliferation appeared to be possible for a time in certain tissues, between such a condition and a culture with continuous and progressive development there was a hiatus which might, perhaps, be conquered at some future day. At the time of writ- ing, however, the author considered it an abuse of language to apply the term "culture" to the results so far obtained. In a later article he ^ said that if the spread of cells in cultures were a true growth and not a mere dissemination, one should be able to hnd evidences of intense cellular multipHcation. These he had not been able to verify in the photographs that he had seen. Ewing '^ was of the opinion that the cells of preparations in vitro were elements which had survived and retained a momentum respon- ^ Compt. rend. Soc. Biol., 1910, Ixix, 291. Jour. Exp. Zoology, 1910, x, 63. 2 Jour. Exp. Med., 1910, xii, 460. Berl. kiln. Woch., 191 1, xlviii, 1364. ^ Compt. rend. Soc. Biol., 1910, Ixix, 293, 298, 299, 328, 332, 365, 367. Compt. rend. Soc. Biol., 191 1, Ixx, 3. Jour. Exp. Med., 191 1, xiii, 416, 562. Jour. American Med. Assoc, 1910, Iv, 1379, iSS4, 1732. Jour. American Med. Assoc, 191 1, Ivi, 32. * Jour. American Med. Assoc, 1911, Ivi, 33, 587, 791. Proc Soc. Exp. Biol, and Med., 1910-1911, viii, 59. Zeitschrift f. Krebsforsch., 1911-1912, xi, 134. Jour. Exp. Med., 191 1, xiv, 129, 453. ^ Compt. rend. Soc. Biol., 1910, Ixix, 470. ^ Compt. rend. Soc. Biol., 1911, Ixx, 4. ''Zeitschrift f. Krebsforsch., 1911-1912, xi, 136. THE TEANSPLANTED TUMOR 12/ sible for certain phenomena, but he was doubtful whether there was any real physiological growth. Dying cells did undergo mitosis and, in fact, this process seemed to be increased in dying tissues. All the phenomena so far recorded might be those of necrobiosis, and while the experiments were of interest, the growth of cells outside the body had not yet been observed. CHAPTER V RESISTANCE Bashford, Murray, and Haaland^ have pointed out that the term immunity, so often used synonymously with resistance, is an un- fortunate one, implying as it does a diminished liability to the de- velopment of spontaneous cancer. It does not necessarily follow that animals resistant to the inoculation of a tumor are protected against the origin of spontaneous cancer, for Bashford, Murray, and Cramer ^ had a mouse which, having been repeatedly inoculated in vain with Jensen's tumor, nevertheless developed a spontaneous growth of differ- ent histological type seven months later. Nor is this an isolated instance. ThoreP later recorded the occurrence of spontaneous tumors in twelve mice that had been inoculated one or more times without success from ten days to nine months previously. The possibility that these growths might have resulted from a postponed development of the grafts was eliminated by their location at a site far removed from that regularly chosen for implantation. Moreover, the tumors all occurred in females. Clunet^ also has reported two spontaneous tumors which arose in mice that had been once unsuccessfully inocu- lated several months previously. Resistance is usually divided into natural and acquired, and the latter, in its turn, into active and passive. NATURAL RESISTANCE Natural resistance is said to be present in consequence of age, race, or ill health and, as it is hardly necessary to state, in species other 1 Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 359. '^ Proc. Roy. Sac, Series B, 1907, Ixxix, 171. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 322, 396. ' Verhandl. d. deidschen path. Gesellsch., 1908, 12*'^ Tagung, 60. * Recherches exp. stir les Tumeurs malignes, Paris, 1910, 24. 128 RESISTANCE I2g than that in which the tumor arose, or, as has been said by Ehrlich^ and others, outside the hmits of bastardy. It is known that sex has no influence in its determination, but the effects of pregnancy have not yet been so clearly defined. Influence of Age Although when experimental transplantations were first under- taken it seemed natural to suppose that as cancer originated most frequently in old animals these would offer the best soil for trans- plantation, it was soon found that the assumption was unwarranted. Thus Loeb - noticed that transfer of a rat sarcoma succeeded in both old and young rats, and Bashford and Murray^ wrote : "Age seems to have no influence on the proportion of successful transplantations, in contrast to its cardinal importance in determining the initiation of the cancer-cycle." The case against the old animal was destined to become even stronger, for in the following year Bashford '^ said : "... transplanted growths flourish as well, or better, in young and vigorous mice/' and this yiew had been consistently expressed since that time in communications from his laboratory.^ In fact, in the paper read before the Royal Society there were described inocula- tions into two series of mice aged four and five days respectively, with an issue of ten tumors in thirteen mice, or 77 % of successful inocula- tions. The conditions for tumor growth being thus more favorable in young animals than in old, Bashford and his colleagues were led to indicate and to emphasize that there must be, in consequence, a differ- ence between the conditions necessary for the origin of a tumor and those required for its continued growth. The experiment was repeated later by Buschke^ with an entirely 1 Arh. a. d. K'onigl. Inst.f. Exp. Therap., Heft i, 1906, 82. ^ Jour. Med. Research, 1901, N.S., i, 36. ' Sci. Reports, Cancer Research Fund, London, 1904, No. i, 14. * Lancet, 1905, ii, 104. ^ British Med. Jour., 1906, ii. 208, 1554. Froc. Roy. Soc, Series B, 1906, Ixxviii, 196. Zeitschrift f. Krebsforsch., 1907, v, 419. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 285. Lancet, 1909, ii, 700. ^ Bcrl. klin. Woch., 1911, xlviii, 215. 130 RESISTANCE confirmatory outcome, rats and mice two, three, and four days old offering a soil favorable for the growth of transplanted cancer. EhrHch and Apolant ^ arrived at similar conclusions and expressed themselves very much as Bashiord had done, namely, that, as trans- planted cancer grew as well in young animals as in old, a difference must exist between the conditions necessary for the inception of a tumor and those requisite for its growth in another organism. Haaland - had found it a matter of daily experience that young mice were, in general, suitable for tumor implantation, and Lewin's find- ings^ in rats coincided with the conditions already defined in mice. Animals from five to eight weeks of age he found susceptible in the highest degree as compared ^^■ith old ones. Gierke ^ fikewise con- sidered that age did not render animals more suitable for transplan- tation but that, on the contrary, young nfice from eight to fourteen weeks old (before puberty had supervened) were as suitable, and in many cases even more suitable, than older animals. The experiments of Albrecht and Hecht^ proved that fully grown mice were fairly resistant to inoculation and that neither these, nor very young animals, were so well adapted as half-grown ones. They finally settled upon healthy animals over four weeks of age as afford- ing the best medium for transplantation. Influence of Race Beside the immunity due to age there has been distinguished a state of resistance determined by racial dift'erences, fully described for the first time by Jensen.*^ In addition to inoculating his turrior into white mice he tried to cultivate it in the common house mouse {Mus muscidiis), meeting, however, with such indift'erent success that in one series only one tumor was obtained in the ten animals that had been inoculated. Once the growth had been transferred from white 1 Berl. klin. Woch., 1905, xlii, 872. ^ Berl. klin. Woch., 1907, xliv, 718. ^ Berl. klin. Woch., 1907, xliv, 1605. Zeitschrift f. Krebsforsch., 1907-1908, vi, 303. ^ Beitr. ziir path. Anat., etc., (Ziegler), 1908, xliii, 343. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 132. ^ Centralbl. f. allg. Path., etc., 1909, xx, 1039. Wien. klin. Woch., 1909, xxii, 1738. ^ Centralbl. f. Bakt., etc., erste Abt., Orig., 1903, xxxiv, 126. RESISTANCE I3I mice to gray, however, its transplantation from gray to gray became easier, and successful implantations could finally be performed in twenty-seven animals out of eighty- four, although growth was fre- quently somewhat slower than in white mice. A transfer from the gray variety back to white was rej^dily _ accom plished. Two attempts to transmit the neoplasm to the long-tailed field mouse {Mus syhati- cus) were un success ful, nor was the tumor transplantable into other varieties of white mice. The same author,^ in a later description of two propagable sarcomata of the tame rat, said that they w ould no t grow at all in wild juts and could be inoculated only occasionally into variegated rats from Hamburg, Berlin, and London. On the contrary in white and variegated rats of the laboratory strain and in Copen- hagen rats (undoubtedly closely related to the laboratory stock) they could be gro wn very succes sfully, the tumor from one rat yielding 87.5 % of daughter tumors and that from the second 57.7 %. The transplantation of both tumors being continued, there occurred during cultivation such an increase in the percentage of successful results that in the fourth, fifth, and sixth generations the yield with the tumor from the first rat had risen to 100 % and with that from the second to 85 %, both calculated in Danish rats. This rise was also evident, however, in foreign rats, for when the first tumor in its fifth genera- tion was inoculated from a Danish rat into five Berlin, three Hamburg, and ten London rats, four, two, and seven, respectively, of these rats, eleven - in all, or 61. i %, developed tumors, while in the first three gen- erations tumors had followed in only 10.5 % of the animals inoculated. The question of the effect of race upon the suitability of the soil occurred also to Loeb,^ but he left its final solution to be determined by future investigations. In a later article he * described the success- ful transfer of a sarcoma of the white rat to a cross between the gray and the white rat, and the failure to transplant a tumor of the Japanese waltzing mouse to ordinary white mice. ^ Zeitschrift. f. Krebsforsch.; 1908-1909, vii, 49. ^ There is a discrepancy in the numbers here : 11 should read 13 and the successful re- sults would, therefore, number 72.2 %. Professor Jensen was kind enough to reply to a letter of inquiry on this point, that 5 Berlin, 5 Hamburg, and 10 London rats were inocu- lated, and that of these 20 animals, 14, or 70 %, developed tumors. Growth had, however, been so slow, that when his paper was published the results had not been final. ^ Jour. Med. Research, 1901, N.S., i, 36. * American Medicine, 1905, x, 265. 132 RESISTANCE Bashford and Murray/ having received Jensen's tumor with the information that in Denmark it had grown best in white mice, tested its growth in Enghsh white mice, which they found yielded less favor- able results than other races of tame Enghsh mice, as yellow, black and white, yellow and white, and black ; in what were known as "blue mice" they had not been able to get the tumor to grow at all. Bash- ford,^ with Murray and Cramer,^ recording later and more elaborate experiments, described attempts to transplant spontaneous or propa- gable tumors of Danish, French, and German mice into Enghsh breeds. The results, which were entirely analogous to those first obtained, demonstrated that the transfer of mahgnant new growths to mice of slightly different race might present considerable difficulties. The apparent unsuitabihty of blue mice for Jensen's tumor, as pubHshed in the earher cornmunication, proved, however, not to be absolute, and in later experiments this breed developed growths which in several cases attained a huge size. The same phenomenon became apparent in other strains of Enghsh mice as concerned the inoculation both of Jensen's tumor and one of German origin, and the percentage of suc- cessful inoculations increased gradually to a point where the results eventually equaled, or promised to equal, those obtained in Danish and German mice. After Jensen's tumor had been growing in Enghsh mice for three years it was inoculated back into Danish mice by Bashford, Murray, Haaland, and Bowen,^ who recorded that its power to grow in the original soil had not been lost. According to Michaelis,^ mice sent him from Denmark provided a better soil for the growth of Jensen's carcinoma than mice obtained in Berhn ; and Clowes ^ found difficulty at first in transplanting Jen- sen's tumor into gray mice of American stock. Hertwig and Poll ^ did not believe that racial resistance actually deserved the importance that had been ascribed to it. The white mouse, they wrote, has been for years an article of commerce, and 1 Sci. Reports, Cancer Research Fund, London, 1904, No. i, 13. ^Lancet, 1905, ii, 1673. ^ Sci. Reports, Imperial Cancer Research Fund, London, 1905, No. 2, Part ii, 22. * Third. Sci. Report, Imperial Cancer Research Fund, London, 1908, 273. ^ Verhandl. d. Komitees f. Krebsforsch., 1903-1904, iii, 38. See Deut. med. Woch., 1904, XXX, 1264. ^ Johns Hopkins Hasp. Bull., 1905, xvi, 130. '' Abhandl. d. Kiinigl. Preuss. Akad. der Wissensch., 1907, 11. RESISTANCE I33 from Berlin, for example, fifty to sixty thousand were exported yearly by dealers who collected them from various small breeding estabhsh- ments. Some years ago, in fact, many of the so-called "Berlin" mice had been imported from Italy. Thus the investigator could never be sure of the origin of his stock, and it might very well be that what had been described by this or that worker as Copenhagen or Frank- fort mice were descended, directly or indirectly, from Berlin mice. On a subsequent page ^ the authors summed up their experience with three tumors that had been transplanted from white to gray mice and conversely. The results were, in the main, so favorable that the authors felt justified in opposing the view that mouse cancer was transferable only between very closely related individuals, and that differences in breed or in diet constituted an obstacle against success- ful inoculation. Haaland ^ found difficulty in getting Jensen's mouse tumor to grow in French and Russian mice, and ^ in transferring Ehrlich's mouse sar- coma to Norwegian mice — so much difficulty, in the latter case, that he was barely able to keep the tumor alive. When, a few months later, he was again in a position to inoculate the latter growth into the strain of mice used in Ehrlicli's. laboratory, he was surprised to see that it flourished with its original vigor. A Berlin stock was sus- ceptible to this tumor, while '■one from Hamburg was much less so, and in the latter case the tumor had not been able to adapt itself even after five passages. Eighteen mice of Danish origin inoculated with this sarcoma did not produce a permanent tumor in one single instance, but a strain of mice that had probably originated in Germany, although it had been several years in Norway, was fairly susceptible. Now it might be imagined, said Haaland, that the susceptibility of certain breeds depended upon general conditions or, in other words, that in such strains cell transplantation would always be more successful. Were this true, one would expect these strains to be susceptible to other tumors as well. But that the case was not so simple Haaland showed by inoculating a mixture of Ehrlich's sarcoma and Jensen's carcinoma. In Berlin mice the sarcomatous element of the mixture 1 Abhandl. d. Konigl. Preuss. Akad. der Wissensck., 1907, 54. ^ Ann. de Vlnst. Past., 1905, xix, 187. ^ Bcii. kliii. Woch., 1907, xliv, 714. 134 RESISTANCE was alone able to proliferate, while in Danish mice, on the contrary, the one tumor that did grow was a pure Jensen carcinoma. Thus it was clear that there were involved, instead of a general condition where certain breeds of mice were susceptible to all growths, special and very specific factors which behaved differently toward different tumors. The Copenhagen mice were at once susceptible to Jensen's carcinoma and resistant to Ehrlich's sarcoma, while conversely, those from BerHn were highly susceptible to the sarcoma and yet resistant toward the carcinoma. When Haaland left Frankfort he took some of the laboratory strain of mice to Norway, and three months later inoculated six of them with ' sarcoma fresh from EhrHch's institute. To this growth, however, they proved themselves highly resistant, although previously mice of the same stock had been susceptible; furthermore, their descendants were also resistant to the tumor. The only possible explanation was that resi dence in Norw ay had made the Frankfort mice an unfavor- able medium, probably because they had received there a d iet differe nt from that to which they were accustomed in Germany. Jensen ^ also could see no other solution, and suggested that the condition was of practical as well as of theoretical interest because of the chance of influ^ncingjnetastasis and recurrence in human beings. One could with perfect justice compare tumors obtained by inocula- tion with those arising by metastasis, and if it should be proved that a radical change of diet w as capable of affecting the susceptibiHty of the mouse to transplantable tumors, the possibility could not be ex- cluded that some influence upon metastasis and recurrence might be achieved in man by the same process. Gierke ^ did not believe that the conclusion of Hertwig and Poll could be extended to include all mouse tumors, for although there might be some which were insensitive to finer differences among mice, the authors had not proved that this was so of all growths. Gierke transplanted three EngHsh tumors into both Enghsh and German mice, with the result that the percentage of success was much higher in the former — 38, 57, and 75%, as com- pared with 9, 14, and 21% in the German animals. The rate of growth in English mice was considerably faster, also, than in the ^ Zeitschrift f. Krebsforsch., 1908-1909, vii, 283. ^ Zeitschrift f . Krebsforsch., 1908-1909, vii, 331. RESISTANCE I35 others. When, however, there was inoculated into English and German mice a strain of the Jensen tumor that had been growing for years in English breeds, no disparity was found either in the percentage of takes or in the rate of growth. These differences in suitability for inoculation Gierke was inclined to ascribe rather to food and surround- ings than to actual racial pecuHarities, for the reason that they had been most marked between the stocks of different countries and usu- ally lacking among those native to the same country. Experiments not controlled by simultaneous inoculation into native animals were always open to the criticism that an unfavorable result might have occurred because the tumor happened at the time to be in the negative phase of growth described by Bashford and Murray. Lewin/ on the contrary, expressed his entire agreement with Hert- wig and Poll, basing his statement upon experiments conducted in both rats and mice. His rat tumor grew in white, variegated, and black rats from many different localities, and he was able to transplant Jensen's rat sarcoma into animals from Berlin and Diisseldorf while, furthermore, a gray mouse tumor had been inoculated with success •into white mice and vice versa. In the experiments of Stahr ^ there were differences in susceptibility between Diisseldorf mice and those from BerKn. The Diisseldorf mice, representing the excess stock of an animal fancier, had been kept for more than half a year in large airy cages and had been fed on hemp seed and milk, but since their arrival at the laboratory they had been given bread soaked in water, food which the laboratory strain of Berhn mice (obtained through an ordinary dealer) had regularly re- ceived. Two tumor strains were employed — one from Berlin and one from Nuremberg. The Berlin tumor grew better at first in the Berhn mice of the labora- tory strain than it did in those from Diisseldorf, which had just entered the laboratory. But after the la tter stra in had been for some months on t he same die t and in the same surroundings as the Berhn stock they proved quite_a^_sensitive to the Berhn tumor, or perhaps even more susceptible. 1 Berl. klin. Woch., 1907, xliv, 1605. Zeitschrift f. Krebsforsch., 1907-1908, vi, 302. 2 Centralbl. f. allg. Path., etc., 1909, xx, 628. 136 RESISTANCE The Nuremberg tumor was inoculated into Berlin mice accustomed to the laboratory food and surroundings, as well as into a fresh supply of Diisseldorf mice from the fancier previously mentioned. In each of the two experiments undertaken, the newly introduced mice were less suitable than those of the laboratory strain. The author concluded, although advising caution on account of the small figures at his com- mand^ that the Diisseldorf mice reacted differently toward tumor in- oculation because they had been kept on a diet different from that of the laboratory strain, and because their health had been impaired by the much less favorable hygienic conditions in the laboratory. Cuenot and Mercier ^ found, between Parisian mice and those of" Nancy, as concerned their susceptibility toward Borrel's tumor "B," a disparity which, exhibiting itself in a smaller yield and a slower growth, rapidly disappeared. The authors, having inoculated many mice of various colors without finding any relation between race and re- sistance, concluded that immunity must rest upon invisible differences. Uhlenhuth and Weidanz,^ in an inquiry regarding the importance of racial differences, determined that Enghsh mice were more suitable for the inoculation of an English tumor than Berlin mice, for not only was there a lower percentage of takes in the latter, but spontaneous cure occurred more frequently. Albrecht and Hecht^ had difficulty in inoculating spontaneous carcinomata from mice of other countries into Vienna mice. Five foreign tumors were ingrafted unsuccessfully or with difficulty, only two such strains being readily transmissible. Tyzzer '^ noted a difference in the susceptibility of certain races of mice to the Jensen tumor and, furthermore,^ that ordinary tame mice were quite insensitive to a growth originating in a Japanese waltzing mouse. Japanese waltzing mice, in their turn, were less suitable than ordinary breeds for the inoculable tumors of the latter. Gay ^ observed throughout his experiments an important variation - Compt. rend, de I'Acad. des Sc, 1908, cxlvii, 1003. ^ Arb. a. d. Kaiserl. Gesundheiisamte, 1909, xxx, 438. ' Centralbl. f. allg. Path., etc., 1909, xx, 1039. Wien. klin. Wock, 1909, xxii, 1738. * Jour. Med. Research, 1907-1908, N.S., xii, 146. ^ Jour. Med. Research, 1909, N.S., xvi, 519. ^ Boston Med. and Surg. Jour., 1909, dxi, 210. Proc. Soc. Exp. Biol, and Med., 1908-1909, vi, 74. RESISTANCE 137 in racial sensitiveness to the Flexner rat tumor, for whereas 100% of the most susceptible strain were inoculable, only 50 % of the rats from a second dealer could be successfully ingrafted, while among those from still another source the yield was even smaller. Influence of Health Mice in poo r condition do not offer so favorable a soil for tumor growth as do healthy ones, according to Bashford^ and Haaland.^ This may serve to explain the results of those who have described the attainment of resistance by treatment with autolyzed tissues, since the possibility of sepsis in the animals of such experiments has not been eliminated. Influence of Sex No observer has yet discovered any difference in susceptibility between males and females — at least this has been the experience of Loeb,^ Ehrlich,* Gierke,^ Lewin,^ and many others. Influence of Pregnancy It has been asserted and denied that the existence of pregnancy ren- dered animals less susceptible to implantation ; and although Morau '' and Herzog ^ had written that gestation accelerated the evolution of tumors, and Bashford and Murray^ that "Pregnancy and full sexual activity in the male (as determined by microscopical examination of the testes) constitute no bar to successful transplantation," Haaland,^" on the contrary, had found that pregnancy often exerted an inhibitory influence upon the prohferation of tumors, the effect of which was to produce a striking retardation of their growth in pregnant animals as compared with animals not bearing young. Uhlenhuth and Weidanz ^^ 1 British Med. Jour., 1907, ii, 28. Lancet, 1907, ii, 32. ^ Berl. klin. Woch., 1907, xliv, 718. ^ Jour. Med. Research, 1901, N.S., i, 36. * Arb. a. d. Konigl. Inst.f. Exp. Therap., 1906, Heft i, 81. ^ Beitr. zur path. 4-nat., etc., (Ziegler), 1908, xliii, 343. Third Set. Report, Imperial Cancer Research Fund, London, 1908, 132. ° Zeitschrift f. Krebsforsch., 1907-1908, vi, 304. ''Arch, de Med. exp. et d'Anat. path., 1894, vi, 693. ^Jour. Med. Research, 1902, N.S., iii, 76. ^ Sci. Reports, Cancer Research Fund, London, 1904, No. i, 14. ^° Berl. klin. Woch., 1907, xliv, 718. ^^ Arb. a. d. Kaiserl. Gesundheitsamte, 1909, xxx, 440. 138 RESISTANCE had also observed this retardation and, furthermore, that spontaneous regression occurred oftener in pregnant mice. Bridre ^ chose males for inoculation whenever it was possible be- cause of the low percentage of positive inoculations occurring in preg- nant females, while Ehrlich ^ had noticed repeatedly that inoculation into animals bearing young was followed with extraordinary frequency by negative results, or was at least attended by the development of tumors in which growth was greatly retarded. Pregnancy, according to Albrecht and Hecht,^ whether already present at the time of inoculation, or commencing afterward, influenced the estabhshment of a tumor or its subsequent growth just as little as the presence of a tumor influenced conception or pregnancy. Cuenot and Mercier^ reported that Borrel's tumor "B" (which rarely underwent spontaneous absorption), if inoculated before fecun- dation, would develop throughout gestation and recede during lacta- tion. But if one mouse only were born and the activity of the mam- mary gland were in consequence at a minimum, the tumor did not regress, nor did absorption take place, even in the presence of several young, if a tumor were so situated that its vascularization was inde- pendent of that of the mammary gland. Fichera ^ explained the inconsistencies that had been observed in the relations between pregnancy and tumor growth by assuming that when many embryos were present the specific food-stuffs were almost wholly demanded by them, while if there were but few the nutrient material was available for the tumor cells as well. ACQUIRED RESISTANCE, ACTIVE AND PASSIVE Active Resistance evolved by Tumor It was reported first by Clowes,^ and later by Gaylord, Clowes, and Baeslack,^ that the Jensen carcinoma sometimes underwent spontane- ^ Ann. de VInst. Past., 1907, xxi, 763. ^ Verhandl. d. deidschen path. Gesellsch., 1908, 12'^ Tagung, 29. ^Wien. klin. Woch., 1909, xxii, 1738. * Compt. re)id. de I'Acad. des Sc, 1909, cxlix, 1012. ^ Cited by Apolant, Jour. Exp. Med., 1911, xiv, 320. ^ Johns Hopkins Hasp. Bull., 1905, xvi, 130. ^ Med. News, 1905, Ixxxvi, 91. RESISTANCE I39 ous regression, and this phenomenon was later described in detail by Gaylord and Clowes.^ Clowes and Baeslack ^ then found that animals in which a cure had taken place were refractory toj^-inoculation, since among thirty mice in which the Jensen carcinoma had been absorbed not one developed a tumor upon re-inoculation with the same growth, although ten of the group were subjected to a third implantation. Flexner and Jobling ^ noted that among seventy rats which had been able to rid themselves of the Flexner- Jobling adeno- carcinoma, only 17 % developed growths upon re-inoculation with that tumor, and Lewin ^ reported that he had see n such resistance freque ntly after the absor^iorijofjiisjy wn rat carcin oma. As a corollary to these observations there arose the question whether a mouse once ingrafted without result could be successfully inoculated at a subsequent trial. Although Jensen '" had observed that in about half his mice the graft failed to grow and that such animals were then resistant, Bashford, Murray, and Cramer ^ were of the opinion that inoculation of mice in which an implantation had been fruitless succeeded in nearly the same percentage as in the first instance. A year later Bashford ''' reiterated this statement, but qualified it with the remark that if the process were repeated, negative mice being discarded each time, animals could be ultimately obtained with a more pronounced power of resistance, affording, for example, 12% of tumors as compared with 68% among controls of the same age. A similar result had been substantiated for mice unsuccessfully injected with two spontaneous sarcomata and various spontaneous carcinomata, when the animals were afterward inoculated with Jensen's tumor. After further investigation he was able to con- firm even more fully the results of others, and in a paper written in conjunction with Murray and Cramer ® said that differences in the size 1 Surgery, Gynecology, and Obstetrics, 1906, ii, 633. ^ Med. News, 1905, Lxxxvii, 969. ^ Proc. Soc. Exp. Biol, a/td Med., 1907-1908, v, 17. * Zeitschrift f. Krehsforsch., 1907-1908, vi, 306. * Centralbl.f. Bakt., etc., erste Abt., Orig., 1903, xxxiv, 126. ^ Sci. Reports, Imperial Cancer Research Fund, London, 1905, No. 2, Part ii, 51. ^ British Med. Jour., 1906, ii, 209. * Proc. Roy. Soc, Series B, 1907, Ixxix, 179. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 331. 140 RESISTANCE of dose would probably explain the apparent contradictions between his own earlier work and that of other observers. In later experi- ments the mass of tumor absorbed in the course of three or four nega- tive inoculations with 0.05 gram had been so great that the animals could not be regarded as comparable with those in earlier experiments, which had received relatively insignificant doses (o.oi to 0.02 gram). In this connection the experiments of Bridre ^ are of importance. Unable at first to discover evidences of resistance in unsuccessfully in- oculated mice, he conceived the idea that his failure might have been due to the employment of an insuf ficient primary dose. Accordingly, he inoculated tumor_ emul sion in amoun ts_fi.ve or six times as jarge as the fragment that had been used in his earher experiments and ob- tained an im munity so perfe ct that no tumors developed in the treated animals, although the controls yielded 55 %. Michaelis, Fleischmann, and Pincussohn ^ had positive results in about 60% of normal mice, while in those which had proved refrac- tory to a first attempt nodules developed after re-implantation in only 12 %. It happened occasionally that a mouse inoculated three times in vain could be successfully ingrafted at a fourth attempt. Borrel ^ noted that while 55 % of mice contracted a tumor after a first implantation, only 12 to 15 % of the negatives developed them after the second, and none after t he thir d. Ehrlich ^ found that in mice unsuccessfully inoculated with a hemorrhagic growth giving but a small percentage of daughter tumors, subsequent inoculation, even with a more vigorous strain, produced but few tumors. Furthermore, the refractory condition was still more distinct in those mice which had been able to resist a neoplasm that would grow in a large percentage of the mice inoculated. Flexner and Jobhng '' reported two hundred and one rats unsuc- cessfully inoculated with their adeno-carcinoma while its virulence was below the maximum, 49 % of which developed tumors upon re-im- plantation with the virulent tumor. ' Ann. de I'lnst. Past., 1907, xxi, 764. ^ Deut. med. Woch., 1907, xxxiii, 827. ^ Bull, de VInst. Past., 1907, v, 603. ^ Arh. a. d. Konigl. Inst.f. Exp. Therap., 1906, Heft i, 90. Zeitschrift f. aerztUche Fortbildimg, 1906, iii, 211. Zeitschrift f. Krebsforsch., 1907, v, 73. ' Proc. Soc. Exp. Biol, and Med., 1907-1908. v, 17. RESISTA^XE 141 Clowes and Baeslack ^ stated that in a series where the mice were intentionally treated with a strain so weak as to afford but about 50 % of abortive tumors, the animals in which the attempt had been fruitless all developed fair-sized growths upon re-inoculation with a more \dgorous strain. And later, Clowes - himself recorded the outcome following re-inoculation of mice refractory to the Jensen tumor. From 30 to 35 % of growths were obtained after the pri- mary inoculation ; the first re-inoculation gave 10 %, and the second nil. The re-inoculation of the sur\-ivors of less vigorous tumor strains with one of more active growth did not indicate, especially when large doses were employed, that there had been conferred as high a de- gree of resistance as Ehrlich had described, although the difference might be due to diversities in the methods of transplantation employed, Hertwig and Poll ^ thought that mice were not made refractory by an unsuccessful inoculation and expressed the behef that in such cases there was involved merely the artificial selection of animals naturally immune, as had already been suggested by Jensen ^ and by MichaeHs.^ Pan-immunity : — Not a Httle attention has been devoted to attempting a definition of the Kmits within which immunity produced by the absorption of tumors is specific. EhrHch ^ believed that, in general, a preliminary unsuccessful inoculation with carcinoma protected not alone against all strains of carcinoma but equally against all sarcomata, and conversely, that preHminary inoculation with sarcoma would protect not only against all sarcomata but against all carcinomata as well. So far as chon- droma was concerned he was of the impression, derived from a series of experiments at that time, however, still unfinished, that in some animals which were resistant to sarcoma or carcinoma the chon- droma grew more slowly than usual or, perhaps, not at all. Still, any considerable res istance to the chondroma was achieved apparently only after the sarcoma-carcinoma resistance had jjeen raised to a 1 Med. News, 1905, IxxxN^ii, 969. ^ British Med. Jour., 1906, ii, 1551. ^ Ahhandl. d. Konigl. Preiiss. Akad. d. Wissensch., 1907, 26. ^Centralbl. f. Bakl., etc., erste Abt., Orig., 1903, xxxiv, 126. ^ Zeitschrift f . Krebsforsch., 1907, v, 191. ^ Zeitschrift f. aerzlliche Fortbildung, 1906, iii, 211. Arh. a. d. Konigl. Inst. f. Exp. Therap., 1906, Heft i, 92. Zeitschrift f. Krebsforsch., 1907, v, 74. 142 RESISTANCE maximum. Hence, as this condition of artificially induced resistance was, to a certain extent at least, not a narrow one, Ehrlich saw no objection to postulating a state of pl uri-immun ity or even pan- immunity. It had been previously recorded by Bashford ^ that mice which had rid themselves of a certain transplantable mammary carcinoma were more resistant to re-inoculation with this tumor than with Jensen's car- cinoma. There was, therefore,^ "... a degree of protection which is common, and a certain degree which is specific." But the common protection was, as Bashford, Murray, and Cramer ^ indicated, con- ferred by the tumor not as tumor, but as mouse tissue, and Bashford, ' Murray, and Haaland ^ could discover no evidence that carcinoma evolved more resistance to sarcoma than did normal tissue, although sarcoma protected to a high degree against carcinoma. They could not, therefore, subscribe to EhrHch's behef that protection between carcinoma and sarcoma was mutual and of equa l degre e. Haaland,^ in earHer experiments, had seen that an unsuccessful inoculation with Jensen's tumor did not protect against the subsequent implantation of sarcoma, but Lewin,^ on the contrary, was convinced that there could exist in the rat c ondition s of r esistan ce co mmon to sarcoma and carcinoma. Rats unsuccessfully inoculated with sarcoma, or in which a carcinoma had undergone spontaneous absorption, had been, in his experience, refractory to both carcinoma and sarcoma. Michaelis, Fleischmann, and Pincussohn ^ found that mice unsuc- cessfully inoculated with Jensen's tumor were not resistant to a BerUn growth, and Gierke,^ that mice negative to the inoculation of a tumor were protected in an extraordinarily high degree against the same, and to a somewhat lower degree against different tumors. 1 British Med. Jour., 1906, ii, 209. Lancet, 1906, ii, 314. 2 Science Progress, 1907, ii, 20. ^ Proc. Roy. Soc, Series B, 1907, Ixxix, 179. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 332. * Jour. Path, and Bact., 1908, xii, 436. ^ Berl. kiln. Woch., 1907, xliv, 716. ^ Berl. klin. Woch., 1907, xliv, 1606. Zeitschrift f. Krebsforsch., 1907-1908, vi, 309. '' Deut. med. Woch., 1907, xxxiii, 827. ^ Beitr. znr path. Anat., etc., (Ziegler), 1908, xliii, 346. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 135. 1-12. CONTROL. NORMAL RATS. Average weight. 53 g-rms. 14 21 28 6 11 18 25 k 1 ' » f f ft 1 2 • t f ^ f • 3 \ 1 f • 4 • f f 5 ' • 6 • \ < f • « 7 i 1 1 # 8 ' • - - - - 9 • ■ - - - - 10 - - - - - - 11 * • 12 • ia-24. 0-1 c.c. MOUSE SARC0JL4. 13.11.07. Average weight. 44 grms. '^ 21 <2 28 ^ fV 18 25 1 13 • • f f t f 14 < • • # • 15 • f • f f 16 • • « t t 17 • • t 18 t f t 19 t 20 ' 21 - - - - - 22 - - - - - 23 24 - - 25-35. 01 c.c. CAT SARCOMA. 18.xi.07 Average weight 42 grmt. 14 12 21 -12 28 12 ^ 11 18 25 I 25 1 • t • f t 26 % • « • # i f 27 • • f • « • • f 28 « « f i f 29 • • t t • • • 30 t • 31 • 32 • • 33 34 35 10 cm. iiiiiiiiiit Comparison of growth of the Flexner-Jobling adeno-carcinoma of the rat, in normal control rats and ii those treated with mouse and cat sarcoma 19 and 14 days respectively before tumor implantation All rats inoculated with 0.02 gram of the Flexner-Jobling tumor 2 — xii — 07. There is no pro^ taction induced by preliminary treatment with tumors of a strange species. I RESISTANCE 143 Albrecht and Hecht ^ noticed that mice in which transplantation had been unsuccessful were, in general, more refractory than normal animals, and in the case of a very virulent carcinoma they had seen this resistance in as many as 50 % of the animals, although the phenom- enon was in no way constant. Such mice were protected against another strain also, but in a lower degree. Active Resistance probably evolved only by Intact Tumor Cells of the Same Species : — ■ Having demonstrated that immunity ensued upon unsuccessful inoculation, those engaged in the study of experimental cancer set themselves the task of producing resistance by treatment with something other than living cancer cells of the same species, for when the mahgnant cell was employed a certain number of animals developed tumors, becoming thereby unavailable for further use. Michaelis ^ tried repeated inoculations of cancer cells killed by chloroform or heat, but reported that his attempts had been fruitless. Searching further, he inoculated white mice with a gray mouse car- cinoma and with a carcinoma from the rat, knowing that by reason of race specificity neither of these tumors would be able to grow in the white mouse ; but not the slightest indication of resistance revealed itself in any of the animals. Heated tumor was injected also by Lewin,^ who could not discpver any evidences of resistance after its inoculation. Although immunity was not evident to Bashfordand his colleagues* in mice inoculated with growths from other species, Lewin^ found that he could produce almost complete resistance in mice by two inocula- tions at short intervals — three to seven days — with rat carcinoma. By preliminary treatment with mouse carcinoma he could protect rats against both carcinoma and sarcoma. Clowes ^ said that more than two years before the date of writing, 1 Centralbl. f. allg. Path., etc., 1909, xx, 1039. Wien. klin. Woch., 1909, xxii, 1740. ^ Med. Klin., 1905, i, 205. Zeitschrift f. Krebsforsch., 1907, v, 192. Bed. klin. Woch., 1907, xliv, 486. Dent. med. Woch., 1906, xxxii, 1728. See also Michaelis, Fleischmann, and Pincus- sohn, Dent. med. Woch., 1907, xxxiii, 827. ^ Zeitschrift. f. Krebsforsch., igoy-igoS, vi, ^oS. ^ British Med. Jour., 1907, ii, 28. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 365, 368. ^Berl. klin. Woch., 1907, xliv, 1606. ^British Med. Jour., 1906, ii, 1550. 144 RESISTANCE experiments had shown him the futiHty of trying to produce immunity with inanimate materials, and that he and his associates had from the first emphasized the role played by the hving cell. On this occasion he repeated his behef that actual ceU growth must take place in order for immunity to be conferred, and placed in CAddence unsuccessful at- tempts to produce resistance \Adth tumor cells that had been de- stroyed by heat or treated with certain chemicals, as mercuric chloride, potassium cyanide, iodine, etc. Furthermore, the refractory state could not be ehcited by inoculating the nucleoproteids or nucleo- histons extracted from tumors. That i ntact ce lls were r equisite for the production of resistance was shown by Bridre,^ who injected the cl ear fl uid obtained by filter - ing and centrifuging a tumor mush. Growths developed in 40 % of the mice thus treated and in 50 % of the controls, while in mice that had been treated -^dth timior fragments heated to points exceeding 50° C. there was likewise a barely appreciable immunity to implan- tation. The general result of his investigations was that the highest re sistan ce followed the inoc ulat ion of uni njured , cells. In perfect agreement "^dth this conclusion was the observation of Haaland,- that the ref ractory st ate did not_supervene in mice inoculated with an e mulsi on of tumor cells de^ dtalig ed by freezing and grinding. Active Resistance Ez'ohed by Xormal Tissue A useful and extremely interesting method of evoking the refractory condition was described from Bashford's laboratory, where it was found that immunity would follow the inoculation of other material than tumor. Thus he ^ and his co-workers discovered that a prelimi- nary injection of from 0.3 to 0.5 cubic centimeter of normal defi- brinated mouse blood induced a definite resistance even in young animals. A repetition of the treatment did not materially increase the refractory condition. Analysis of the phenomenon proved the corpuscles to be the a ctive ag ent, serum alone being impotent, and it was further found that the efl'ect was best brought out when the ^ Ann. de I'lnst. Past., 1907, xsi, 768. ^ Lance*, 1910, i, 7S7. ' British Med. Jour., 1906, ii, 209. Lancet, 1906, ii, 315. Proc. Roy. Soc, Series B, 1907, Ixxix, iSo. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 333, 369. CONTROL 1 19. 20-34 mice treated 35-49 mice treated with 60-67 mice treated with Untreated mice. with 0-] c.c. embryo O'l c.c. mamma emulsion. 005 c.c. ekin emulsioD. Average weight emulsion (skinless). 21 '5 grms. Average weight i2-8 grms. Average weight 23 gnus. Average weight 19 grml 28/ 5/ 12/ 19/ 28/ 2/ 28/ 6/ 12/19/ 26/ 2/ 28/ 6/ 12/ 19/ 26/ 2/ 9/ 16/ 23/ 28/ 5/ 12/ l9/^'/„ V.. V.. 'K. /9/I0/I0/I0/IO/II /9 /lo /lo/io /lo /ii /s /IP /1 0/10 /\o /\\_A>^A \ /I I A At) Ao At /w/\\ /ji/JL I • 9 9 f f 20 , f ^^^ ^ ' » * ^ f^T^ " ' ' ' ' ^ f T T 3r#Cf| 22.»«it»37ff^^O 52 4»t#» :3'**|. -aavtc^t^tvtjs 5 f r 2l-----3a,<#. v___-_^ 6«1tt« 25 ------ 40 •••^^^ 55 ______ 7tt***< 23 ------ 41 ».»|f|^ 5B ------ 8.;i 27 ------ 12 »Q#t»»»" 57 ------ 9,.- 2S ------ 43 ••• 58 ------ IB »..___ 29 ------ 44 59 ______ II r a-. ...-45. ----_-. 60 ------ 12' Sl._----43i.-... 61 ------ 13 ' I - - _ - 32 ------ 47 I* • - 83 ------ 48 IS ------34-- 49 64 11 • _ - 10 cm. I mil J7 - - - - - - II . . - . . * Macroscopic metastases in lungs. Protection against, or hypersensibility toward, inoculation of a transplantable squamous cell carcinoma induced by treatment with normal mouse tissue 15-16 days before tumor implantation. All mice inoculated in right axilla with 0.01-0.02 of tumor. First charting 10 days later. RESISTANCE 145 minimal tumor-forming dose was employed, a dose lying between one and two centigrams. When larger amounts (five centigrams) were introduced, the percentage of growths in the treated animals was greater, although still not so large as in the controls, while in a series where massive doses were inoculated, nodules developed eventually, although their appearance was somewhat delayed. Here, the authors thought, the larger dose of tumor had so exhausted the protection that growth finally became possible. Schone,^ working in EhrHch's laboratory, succeeded in producing resistance by preHminary injections of mouse embryo, liver, or testis. Rather large embryos, emulsified without the addition of fluid, were inoculated in amounts of from 0.3 to 0.7 cubic centimeter. White mice were used in all the experiments, and the testing tumor was a vigorously growing alveolar carcinoma known as "5." In the first experiment, thirteen animals were injected eight times — six times subcutaneously and twice intraperitoneally — during a period of seventy-one days, the last treatment taking place one day before the tumor was inoculated. Among the thirteen treated mice, 46.15 % failed to develop tumors, although inoculation was successful in all but 7.14% of the twenty-eight controls. Not only were there fewer nodules among the treated mice than among the untreated, but those that did occur showed a distinct retardation of growth. In a second experiment eighteen mice received two intraperitoneal injections at an interval of two weeks. Implantation of carcinoma "5," undertaken fourteen days later, was followed by a negative outcome in 66.67 % of the eighteen treated mice but in only 16.67 % among eighteen con- trols, and of the six tumors developing in the refractory animals only two showed no delay in their growth. The presence of a higher degree of resistance after eight injections than after two could not be sub- stantiated. Liver and testis, although producing immunity, were not nearly so efficacious as embryo. Whether the resistance evoked by normal tissues was identical with that following the inoculation of tumors, Schone left an open question. In a later article he ^ said that one single subcutaneous inoculation ^ Miinch. rued. Woch., 1906, liii, 2517. ^ Verhandl. d. GeseUsch. deutscher Naturforsch. u. Aerzte, 1907, 79'"" Versammlung, erste Teil, 304. 146 RESISTANCE of normal tissue would call forth an efficient immunity and that intra- peritoneal inoculations were also active in producing this condition, although here larger doses were preferable (0.5 to i.o cubic centimeter). The immunity consequent upon the inoculation of spontaneous tumors was, in general, more marked than that following prehminary treatment with embryo emulsion, although the two types seemed to have a number of characteristics in common. Investigating the prehminary observations of the authors just cited, MichaeHs, Fleischmann, and Pincussohn^ succeeded in demonstrating a certain amount of resistance after three inoculations of an emulsion of normal mouse Hver. Of twenty animals thus treated, only 30 % developed tumors, while among eighteen controls there were growths in 72 %. As in Schone's experiments, the nodules that did appear in the treated animals were of retarded growth and attained only a mod- erate size, afterward remaining stationary. Proliferation in the control mice, on the other hand, was progressive. Borrel ^ and Bridre ^ inoculated mice with blood and with emulsions of spleen, hver, or testis, in amounts of one cubic centimeter. Two inoculations of blood, of about 0.25 cubic centimeter each, gave but a feeble resistance against the Pasteur Institute tumor "B," 40% of growths developing in treated animals compared with 55 % in the controls. Three inoculations of hver emulsion, at intervals of twelve days, produced a higher degree of immunity, while perfe ct resist ance was obtained by the i noculation of splee n. On the other hand, testis was quite powerless to ehcit the refractory state, 50 % of tumors developing in the treated mice and 55 % in the controls. Lewin ^ extended these studies to rats, and produced resistance against his rat carcinoma and against Jensen's rat sarcoma by a single inoculation with 0.8 to i.o cubic centimeter of normal rat blood. Moreschi,'^ after the inoculation of an emiilsion of actively lactating mouse mamma in amounts of 0.15 to 0.20 cubic centimeter, found conditions ranging from hypersusceptibihty to resistance. The 1 Dent. med. Wock., 1907, xxxiii, 827. 2 Bull, de VInst. Past., 1907, v, 605. ^ Ann. de I'Inst. Past., 1907, xxi, 769. ^ Berl. klin. Woch., 1907, xliv, 1606. ZeUschrift f. Krehsforsch., 1907-1908, vi, 310. ^ Zeitschrifl f. Immiinitdisforsch., etc., Orig., 1909, ii, 675. RESISTANCE 147 immunity varied toward different tumors, a much more refractory state being present for Etulich's carcinoma "ii" than for carcinoma " 5." Toward a sarcoma the resistance was often evidenced merely by a retardation of growth, although at other times 80 to 87.5 % of tumors developed in the resistant animals as compared with 100 % in the controls. The large }'ield was doubtless due to the extraordinary virulence of this growth. The production of resistance by the inoculation of normal, blood and of embryo emulsion was further substantiated by Uhlenhuth and Weidanz/ but they were unable to cause its evolution with mouse lens. Higuchi - recorded the achievement of immunity through the intro- duction into mice of mouse placenta, which was capable of provoking the condition quite independently of its contained blood. It was effectual not only against several carcinomata but against a sarcoma as well. In experiments with tumors other than Jensen's, Bashford and his associates ^ found that injections of blood protected to a lower degree against their tumor "50" and hardly at all against tumors "32'' and "27." In these facts they saw still further eAidence of the truth of their oft-repeated statement that the conditions of growth were spe- cific for different tumors, and th'at factors unfavorable to one particu- lar neoplasm were not necessarily active against aU other growths. Schone.'' indeed, had found the limits of specificity so narrow that the embryos of gray mice were not as suitable as those of white for the production in white mice of immimity toward a white mouse tumor, although -^ith extensive treatment a definite result could be achieved. Rat embryos, as a rule, did not educe resistance, although in isolated instances tumor growth seemed to be held somewhat in abeyance. A lack of specificity in the immune reaction was suggested also by the work of Rous,^ who showed that resistance toward the implantation ^ Arb. a. d. Raised. GesufidheitsanUe, 1909, xxx, 443. ^ Sei-I'Kiuai Medical Journal, Tokio, 191 1, xxx, 91. Fifth Sci. Report, Imperial Cancer Research Fimd, London, 191 2, Sy ' Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 375. ^ Verhandl. d. Gesellsch. deutscher N atiirforsch. u. Aerzte, 1907, 79''''' Versammlung, erste Teil,304. ^ Jour E.xp. Med., 1910, xii, 344. 148 RESISTANCE of fetal cells could be produced by the inoculation of embryonic tissue, and that it was manifest in an absence of the stroma reaction neces- sary for a temporary survival of the ingrafted material. , Active Resistance probably evolved only by Intact Normal Cells of the Same Species : — Just as resistance to subsequent inoculation fails to foUowjDTeKminar}^ treatment with tumor cells robbed of their vitality by heat or other injurious_agents, or -with intact tumor cells of different species, so in the case of normal tissue is the injection of uninjured li\ing cells of the same species an essential for the production of the resistant state. Bashford and his associates ^ were unable to prevent the growth of transplanted cancer in mice by the inoculation of the normal tissues of rats, rabbits, guinea-pigs, or more distantly related species, and Higuchi - discovered little if any inhibitory effect after the introduc- tion into mice of placenta, blood, mammary gland, embryo skin, or spleen of rats and guinea-pigs. In the experiments of Moreschi,'^ although rat mamma produced not a trace of immunity in mice against one mouse carcinoma, it was efficacious against a second one, and he thought that the presence of the refractory condition depended upon the length of time elaps- ing between inoculations, no less than upon certain other factors. Proceeding to test the power of rat sarcoma to effect resistance in mice, he discovered evidence that such a condition had been induced. Lactating guinea-pig mamma was also active, but not in so high a degree as mouse mamma. He concluded, therefore, that Bashford and his associates were not justified in denying the possibihty of producing resistance with the tissues of strange species. Similar results were cited by Apolant,^ who wrote that by treatment with alien blood he had occasionally obtained immunity against mouse carcinoma, and that he was, therefore, unable to agree with Bashford's contention that only tissues of the same species were potent. ^ British Med. Jour., 1907, ii, 28. Proc. Roy. Soc, Series B, 1907, Ixxix, 182. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 334, 365, 376. 2 Sei-I-Kwai Medical Journal, Tokio, 1911, xxx, 91. Fifth Sci. Report, Imperial Cancer Research Fund, London, 191 2, 85. ^ Zeitschrift f. Immunildtsforsch., etc., Orig., 1909, ii, 681. ^ Zeitschrift f. allg. Physiol., 1909, ix, Sammelreferat, 91. . I, o >. SI ^ 3 O « -+^ s> n o US +3 a XJ 1. ^a ^" • 1 •» » a .. 1 — 1 'J a ** ^^ ^c «to * •.• - .« •■ • .. - - -• ■ ?1 2"b ^N." •- •« • %.• •• - • ^ — • • :'T o wm »l CO ■* lO CO l>. 00 ea to (O (C e « o «D •m ^ » - ". .. •• 1 1 1 . 1 1 • • m ^ •■ »• •- - 1 ■ 1 1 f 6 2 o o £ -^ A « ^ A f « • . « - - So -NT CM ^ m m «*. # • « «. •• - CO ■A sX" ^ » «« * • • %. • - • to =-C • « «k - • • t « . • - bo -\" • - •• • • 1 1 <: s< - - - > - - M CM CM CM US C4 CM CM eo CM at ex CO p H O gi cb g f.i •1 1H 8| 12; a < :> ^ »• % •• ••• 4« ^ 1^ • « ^ «^ »# % •l • •• s> «l » % * «. ••• ^» l» •. • • ft.. •!». ^ - • •-•, •« , »• . % 1 « • ■ . •« •• - • - "' 1 • - - 1 - • - ' 1 1 - 1 - .. m ■+ vn ^ s CD O o = - I I I I •_:__ » »• '_^ r I .^ I • - I I 0» O S 04 TT g W lO S IS lO 113 US 9t eg ^ IS CI «0 04 (411 p. O O) 5 !=! O <43 Xi T3 ^ o t3 ^ o SP ^ < 2 XI V c 2 S i-f^^^ ■2 5 rt o c o fL| .M O O rt g a .2 -2 S 3 • s 5 i2 1 1 • 1 1 • 1 1 • 1 1 1 - - X T) .;::; m b -M ji ca is ;-/'".''. S'r^.C'.V-.'-'; •'■''. / J!^00:/ '>! %' 1'/ ^ ^s If: > sr i X > O x a o 15 s a B rs" "G c rt rt U d 'a C/J C! o o I s:^^ C^" - ea 5.#-^^ ^^ ^*n©i3>^2_ ts. — — G^^ ^ cS X 6 >> O .fef* ^~^ i RESISTANCE 175 polyblast cells and polymorphonuclear leucoctyes. There is no evident increase in the vascularity of the tissues. The absorption of the ne- crotic mass takes place very slowly, only after about seven days do the fibroblasts and polyblasts penetrate into it in any number, and about 12 to 14 days is necessary for the whole to be cleared up." The outstanding feature in refractory animals was thus the failure to supply a vascular stroma. The absence of the stroma reaction in resistant animals having been demonstrated, Russell sought to explain the reason for it, and suggested that it might be ascribed to one of two factors. "... either the tissues of the animal have been altered in such a way by the process of immunization that they no longer react to the stimulus of the cancer cell, or else the cancer cell itself becomes robbed of its power of inciting a specific reaction." The author was inclined to believe that the immunity was directed against the chemotactic influences exerted by the cancer cell on the connective tissues of the host. "There must be present in the resistant animals, either in the circulating fluids or in the tissues, something which inhibits this chemo- taxis," although all attempts to demonstrate in vitro the presence of an active cell poison had as yet been inconclusive. "That the induced resistance to inoculation is not due to a very active cell poison may be seen from the power which the cancer cells retain, of continuing their proHferation for 7-10 days in this unfavour- able medium, provided that they can obtain sufficient nourishment. Further, the cells at the periphery of the graft are those which go on growing, and these are the very cells which are most exposed to the influence of any supposed poison, whereas the cells towards the centre of the graft, which are not so exposed to a free supply of this inimical substance, are the ones which die rapidly because of the interference with their food supply." Russell's observations were extended by Woglom^ to include an investigation of the Flexner-Jobling adeno-carcinoma of the rat, and it was shown that here, as in the mouse, the failure of grafts to es- tablish themselves in immune animals could be referred to an absence of the specific stroma reaction. ^ Fifth Sci. Report, Imperial Cancer Research Fund, London, 191 2, 43. 176 RESISTANCE Burgess^ investigated in the same way the phenomenon of natural resistance in mice, working with an adeno-carcinoma discoA'ered by Tyzzer in a Japanese waltzing mouse. This tumor would grow in iiybrid mice of the first generation obtained by cross-breeding Japanese and common mice, and in Japanese waltzing mice, but in all others, including the offspring of hybrids of the first generation, inoculation was invariably -unsuccessful. The study of early stages showed that in susceptible Japanese waltzing mice the tumor receiA'ed a new nbrous and vascular stroma from the tissues of the host. In non-susceptible hybrids of the second generation stroma and blood vessels were supplied in the same way, ' but after about a week the graft became surrounded by an inflamma- tory exudate which impaired its nutrition. Apparently as a part of this reaction there occurred in many of the non-susceptible mice an overproduction of fibrils in the more centrally located portions of the new stroma, but no such inflammatory process was seen in growths of a corresponding age in Japanese mice. In the resistant am'mals, and probably as a result of impaired nutrition, peripheral extension of the nodule ceased and central necrosis advanced, until ultimately the entire tumor had undergone necrosis and absorption. In the case. then, of animals belonging to ver\- closely related varieties capable of interbreeding freely, resistance was only developed after several days of \'igorous growth on the part of the graft. It was not, therefore, a result of the faflure of the specific stroma reaction for, in fact, this eventually surpassed the response in susceptible mice. The active resistance of the non-susceptible animals was apparently due to an inflammaton,' lesion which interfered with the nutrition of the tmnor. AA'hile Russell. Woglom. and Burgess had examined more particularly the phenomena in play in the graft itself, Da Fano - investigated the reactions taking place in the normal tissues of the host. As hmiphocytes appeared in great number about inoculations of immunizing material during the evolution of resistance, their relation to this condition could not be denied. Furthermore, their presence ^ Jour. Med. Research, 1909, X.S., x\"i, 575. "^ Zeitschrift f. Itnmunitatsforsch., etc., Orig., 1910, v, i. Fifth Sci. Report, Imperial Cancer Research Fund, London, 1912, 57. RESISTANCE 177 was independent of the amount of material inoculated. They were absent, or appeared only in small number after the inoculation of dead tumor, or in animals already immune. In growing carcinomata they were to be found only in places where local heaHng was in progress. The carcinoma cells seemed to exert some sort of specific influence on the lymphocytes, and the latter to spread the resistant state through- out the organism. Many of the statements referring to these elements might be applied to the plasma cells, which Da Fano beheved also played an important role in the resistant state. As they were not found normally in the subcutaneous tissue of the mouse, their presence in immune animals might be taken to mean that in the mechanism of resistance there were concerned, beside a local reaction, changes in the entire organism, of which the plasma cells were a morphological expression. CHAPTER VI HYPERSUSCEPTIBILITY Not only has it become possible to affect animals in the direction of making them less susceptible to tumor growth ; the occurrence of a modification in the opposite direction has been suggested by several writers — a transformation of the organism in the course of which it comes to offer a more favorable soil for the proHferation of mahgnant new growths. Flexner and JobHng^ described a series of experiments in which rats were treated intraperitoneally with an emulsion of adeno- carcinoma that had been heated for half an hour at 56° C. Grafts of the same tumor introduced subcutaneously into these rats twenty-four hours afterward grew at the same rate as the grafts implanted in controls. But when tumor was inoculated from ten to thirty days following the injection of heated emulsion, not only did the number of successful inoculations tend to exceed that in the controls, but the individual growths developed with greater_ rapidit y, reached a size more than double that of the control tumors, and showed a far smaller percent- age of retrogressions. The promoting influence was less effective at thirty days than at ten, but indications existed which seemed to show that if the preliminary treatment were repeated once or twice at ten- day intervals, the conditions favoring the growth and persistence of the tumors could be maintained, and, possibly, still further increased. In a second article the authors - recorded an analysis of the phenom- enon previously described. One-half of a group of rats with station- ary or receding growths was inoculated with heated emulsion and ten days later with tumor, as a result of which 60 % developed growths. The other half was inoculated without an antecede nt inject ion of heated emulsion, and in only 36 % was the outcome successful. ^ Proc. Soc. Exp. Biol, and Med., 1906-1907, iv, 156. "^ Proc. Soc. Exp. Biol, and Med., 1907-1908, v, 16. HYPERSUSCEPTIBILITY 1 79 Normal control rats yielded loo %. In a similar experiment carried out on a second group of rats in which the growing tumor was later absorbed, only 9 % of those that had not received heated tumor emulsion were successfully inoculated, while tumors de- veloped in 30 % of those that had been subjected to preliminary treatment. Again, a series of rats injected with heated emulsion and successfully inoculated, the tumors having finally undergone absorption, was divided into two parts. One of these groups received a second injec- tion of heated emulsion, the other was kept for controls. At the expiration of ten days both groups were inoculated with tumor grafts, in consequence of which 30 % of each developed growths, while 100 % occurred in the normal controls. It appeared, therefore, that no promoting effect arose from a second injection of immune rats with heated tumor emulsion. What seemed equally surprising was the high percentage of successful secondary inoculations in this group, as contrasted with the low percentage among the rats in the pre- ceding group which had not received the emulsion — ■ that is, 30 % against 9 %. If, however, this group were compared with the one in which, after spontaneous recovery, heated emulsion was in- jected for the first time and followed by new grafts, the percentage of successful re-inoculations was found to be identical in both, namely, 30 %. The authors thought it premature to attempt a discussion of these results, which seemed to imply that by the injection of heated tumor emulsion a state of susceptibility to implantation could be preserved, while at the same time the tumor originally implanted had undergone absorption. No promotion of growth foUowed previous treatment with heated and unheated emulsions of various organs, or with the sera of normal rats, of naturally resistant rats, of rats that had acquired resistance through the absorption of their tumors, or of rats with growing tumors. Gaylord ^ repeated these experiments with a sarcoma of the rat, but was unable to demonstrate hypersensitiveness, or any definite condition, following the injection of heated tumor material. 1 Jour. American Med. Assoc, 1908, li, 252. l8o ^ HYPERSUSCEPTIBILITY BasMord. ]\Iurray. and Haaland ^ had found that if a dose of spontaneous tumor not exceeding o.i cubic centimeter were used in an attempt to produce resistance the effect was usually not very pronounced, and that when inoculation was performed after a long interval a condition of h}-persu5ceptibiHty might be sho^\'n to exist. They had seen evidence supporting the occurrence of the same condi- tion after the inoculation of mouse mamma- and skinless mouse em- bryo. The treatment of mice ^^dth rat mamma, six weeks before tumor implantation, had also evolved a state of hj'persensibiHty, although where such treatment preceded the inoculation of the tumor by only sixteen days, none was apparent. It seemed to the authors that both the time interval and the dose of normal tissue emplo3"ed were of importance in determining whether protection or h^-persusceptibility would super^'ene. H^-persensibility seemed to be a less specific phenomenon than resistance, and one that could be obtained by prehminarv- treatment with various sub- stances, particularly, however, wdth the tissues of strange species. ]\Ioreschi ■' observed, after two injections of lactating mouse mamma, a distinct stimulation of growth when tumor inoculation was performed from ten to twelve days after the last treatment. PreHminary injection of mice with functionating rat mamma favored proliferation in two different strains of carcinoma, when the interval between treatment and the introduction of the graft was from nine to fifteen days. But if the tumor were inoculated after from thirty to thirty- seven days the growth stimulus was repressed, to be sup- planted by a distinct resistance. Lactating guinea-pig mamma did not produce h\'persusceptibLlity. Gay "^ found that the blood of normally insusceptible rats or of rats unsuccessfully inoculated "^dth the Flexner-Jobhng adeno-carcinoma, injected into normally insusceptible rats before tumor inoculation, or concomitantly wdth it, led to a larger percentage of growths. 1 Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 368, 376, 381. 2 See plate facing page 144. ^ Zeitschrifl f. Immiinitdtsforsch., etc., Orig., 1909, ii, 675. ^ Proc. Soc. Exp. Biol, and Med., 1908-1909, \'i, 75. Jour. Med. Research, 1909, N.S., x\^, 186. Boston Med. atid Surg. Jour., 1909, clxi, 211. HYPERSUSCEPTLBILITY l8l According to Haaland.^ an obvious h}-per5en5ibility followed the inoculation of cancerous or normal cells that had been mechanically disintegrated by freezing and grinding. That this condition was not a phase antecedent to the estabhshment of resistance was demonstrated by var\-ing the time inter^'al between prehminar}' treatment and tumor implantation, from which it became apparent that there was as Httle immunity after thirty days as after ten or twenty. A dose of O.I cubic centimeter of disintegrated tissue appeared to produce the optimum of hypersusceptibihty. Animals could also be h^^ersen- sitized -^^-ith 0.5 cubic centimeter of the expressed fluid of cancer or normal tissue, or through treatment with cells devitahzed by autolysis, heat, or exposure to radium. Leitch- observed h}-persen5ibility in mice that had been treated with the supernatant fluid from an em^ulsion of ground mouse can- cer cells in physiological sahne solution. The fluid was injected into the peritoneum on three occasions at ten-day intervals, one- half a cubic centimeter at the first, and one centimeter at the last two treatments, and tumor inoculation was undertaken twenty-four days after the last injection. The amount introduced represented, in the opinion of the author, an exceechngly minute dose of epithelium. Uhlenhuth, Haendel, and Stefi'enhagen ^ were of the impression that in rats pre^dously treated with the serum of resistant rats, tumor development and growth were favorably influenced. H}'persensi- bihty was noted, also, when tumor emulsion had been mixed before inoculation ^^-ith the serum of resistant rats. Another type of sensibility to tumor cells was described by Yama- nouchi,'* wherein mice bearing tumors were said to react immediately, and with very characteristic s}miptoms, to the intraperitoneal inocula- tion of an emulsion of the same tumor. The pathognomonic S}TQp- toms, a Staring coat and a condition of immobihty followed after twenty-four hours by death, occurred neither in normal mice, nor in ' Proc. Roy. Soc, Series B, 1909-1910, Lxxxii, 293. Lancet, 1910, i, 787. ^Lancel, 1910, i, 991. ^ Centralbl.f. BakL, etc., erste Abt., Ref., 1910, xlvii, Beiheft, 164. Arb. a.d. Kaiserl. Gesiindheitsamte, 191 1, xxxva, 490. * Compt. roid. Soc. Biol., 1909, lx\a, 754. 152 HYPERSUSCEPTIBILITY those whicli had been unsuccessfully inoculated. That mice in the latter category were not affected seemed to demonstrate that the con- dition was not an anaphylaxis (in the usual sense of the word) toward the tissue of the tumor. Apolant ^ was unable to reproduce this phenomenon, and offered the tentative explanation that Yamanouchi's growth had been infected, and that the hypersensibiHty had been directed toward bacteria rather than toward tumor cells. Similarly, the experiments of Rous^ did not support the findings reported by Yamanouchi. ANAPHYLAXIS Pfeiffer and Finsterer ^ found that guinea-pigs which had received forty-eight hours previously an intraperitoneal inoculation of the serum of a carcinomatous individual, showed anaphylactic symptoms (among them a drop of several degrees in the body temperature) upon injection of the expressed juice of the patient's tumor. A similar re- action could be produced, also, in an animal sensitized with the serum of a person other than the one furnishing the tumor, so that a patient with carcinoma seemed to produce anaphylactic antibodies, not only against his own tumor, but against carcinomata in general, Ranzi,^ who said that he had not found the reaction specific, was criticized by Pfeiffer,^ chiefly upon technical grounds. ^ Zeitschrift f. Immu)iitdtsforsch., etc., Orig., 1909, iii, 108. ^ Zeitschrift f . Immunitdtsforsch., etc., Orig., 1910, iv, 238. ' Wien. klin. Woch., 1909, xxii, 989. See also Pfeiffer, Wien. klin. Woch., 1909, xxii, 1227. ^Wien. klin. Woch., 1909, xxii, 1372. ^V/ien. klin. Woch., 1909, xxii, 1375. r^j > *-»- \ \ ''■. ym m ^r^^^9''/^3v^^ '< C?i.^- .' '/ Adeno-carcinoma from small intestine of the mouse. Margin of growth, showing mode of extensior laterally and through muscular wall. X -r • CHAPTER VII THE SPONTANEOUS TUMOR FREQUENCY OF TUMORS AMONG THE LOWER ANIMALS To review fully the occurrence of spontaneous tumors among the lower animals is not within the scope of this volume. It is sufficient for the present purpose merely to state that both benign and malignant tumors are being discovered constantly in mammals, birds, am- FiG. 8. — Grouse : Primary adeno-carcinoma of small intestine. Infiltrative growth in all the coats, great thickening of circular muscular layer, the surface of which at one part forms the base of the ulcer. X Y- 183 184 THE SPONTANEOUS TUMOR phibians, and fish, and that the close degree of relationship with man into which many of the subjects are thrown has not been proved in the slightest degree responsible for the development of the disorder. If malignant disease be more common among the domestic animals it is only because, as Bashford and Murray^ have pointed out, enjoy- ing a longer span of life by reason of man's care more of them reach the cancer-bearing age, and because they are, moreover, under more constant observation than are animals living in a state of nature. The zoological distribution of cancer has been discussed by these two authors together, and by Murray alone ,- and they have recorded, among tumors observed by themselves or others, malignant growths in the cow, dog, horse, sheep, pig, cat, hen, Indian parakeet, macaw, grouse, canary, giant salamander, frog, triton, codfish, gurnard, and trout. According to Murray, no case of malignant tumor has yet been found among the reptiles, although Pettit described a fibroma of the stomach in a python, and Koch a papilloma of the occipital region in a lizard. The occurrence of malignant or benign tumors and the literature germane to this subject have been discussed by Sticker,^ Ehrenreich and Michaelis,^ Plehn,^ Casper,^ Tyzzer and Ordway/ Fiebiger,^ and McCoy.9 Instances of growths apparently malignant have been reported, furthermore, in the vegetable kingdom, none of them, however, so convincing as Jensen's^*' transplantable tumor of the beet. STRUCTURE OE ORIGIN IN THE MOUSE When pathologists began to interest themselves in tumors of the lower animals, preeminently in those of rats and mice, and to attempt transplantation, the first question that had to be answered was whether these growths were actually malignant. The gravest doubts of their ^ Sci. Reports, Cancer Research Fund, London, 1904, No. 1,5. 2 Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 41. ^ Arch. f. klin. Chir., (v. Langenbeck), 1902, Ixv, 616, 1023. ^ Zeitschrift f. Krebsforsch., 1906, iv, 586. ^ Zeitschrift f. Krebsforsch, 1906, iv, 525. ^ Ergebn. d. allg. Path., etc., (Lubarsch and Ostertag), 1907, xi, Abt. 2, 1068. ''Jour. Med. Research, 1909, N.S., xvi, 459. * Zeitschrift f. Krebsforsch., 1908-1909, vii, 165, 371, 382. ' Jour. Med. Research, 1909, N.S., xvi, 285. 1° Travaux de la deuxieme Conference internat. pour l' Etude du Cancer, Paris, 191 1, 243. ■^% Aa 'J^y Malignant adeno-carcinoma of skin glands from frog. The growth has penetrated beneath the dense lamellar layer of the dermis, a, and infiltrates the subjacent muscles of the thigh, b. X -*r. (5 'J><^ fr ®/ 1^® o '% ao'of '®a« s> «> til Same growth, showing connection with normal skin, the tumor apparently spreading laterally along its deep surface Photograph by Dr. Zarnib. Carcinoma of skin glands from triton. Macroscopic appearance of growth at angle of mouth and scattered nodules on tail. X $. J. R. Ford, del. High power view of above growth. Alveolar part of tumor and commencement of a columnar cell portion (right upper). Note how sharply the alveolar plugs are marked off from the surrounding normal skin. X V. Micropliulugraph by li. Muir. Osteo-sarcoma of operculum from cod. Portions of two osseous nodules are shown, with surrounding vascular spindle cell tissue. X t°- THE SPONTANEOUS TUMOR 185 true neoplastic nature existed on many sides, and the patient collec- tion of data regarding their life history was necessary before the prob- blem could be settled. In addition, one of the most important tasks was to determine the structure giving origin to the tumors, a ques- tion to which the answer at the beginning of the investigation was by no means certain. Eberth and Spude ^ had described tumors in three mice and had at- tempted to transplant these growths, but, puzzled by the lack of apparent connection between the neoplasms and any epithelial structure, con- cluded that they were endotheliomata, in spite of their morphology. As Apolant ^ said later, however, if these authors had only examined a section of normal mouse mamma, they never would have mistaken ducts and alveoli for altered lymph vessels. The ostensible want of relationship between epithelial structures and tumors seemingly of epithelial origin continued, nevertheless, to confuse many of Eberth and Spude's successors. Jensen,^ although not certain of the source of his tumor, was inclined to derive it from the glandular structures of the skin or from the epidermis, while Pick and Poll ^ described an epithelial tumor of the scapular region in a white mouse and suggested that it had originated in the sweat glands. Ac- cording to both Apolant ^ and Murray,^ however, the sweat glands of the mouse are restricted to thte soles of the feet. V. Hansemann "^ expressed the opinion that the tumors found in mice did not proceed from the skin, nor did he consider that they were of glandular origin, and held that the hibernating gland should be con- sidered in any attempt to account for their genesis. This gland, a structure of very extensive distribution extending from between the shoulders forward to the thymus, running thence toward the aorta and widening out on either side to reach finally the region of the kidney, was not found, however, in the neighborhood of the inguinal folds, where ^ Arch. f. path. Anat., etc., (Virchow), 1898, cliii, 60. ^ Arh. a. d. Konigl. Inst.'f. Exp. Therap., 1906, Heft i, 32. ^ Centralbl. f. Bakt., etc., erste Abt., Orig., 1901, xxxiv, i,s- ^ Berl. klin. Woch., 1903, xl, 519. ^ Arb. a. d. Konigl. Inst. f. Exp. Therap., 1906, Heft i, 40. ® Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 85. ' Verhandl. d. Komitees f. Krehsforsch., 1903-1904, iii. 38. See also Deut. med. Woch., 19,04, XXX, 1264. 1 86 THE SPONTANEOUS TUMOR tumors, nevertheless, not infrequently originate. He was inclined, on the whole, to believe that the growths were derived from endothelium, although he confessed his inability to advance proof of this contention. MichaeHs,^ because of the alveolar structure of Jensen's tumor, raised the question whether it might not represent an adenomatous growth springing from a gland without lumina, and suggested the two glands of this type found in the mouse at the sides of the trachea. The studies of Bashford and Murray,^ and of Apolant,^ were the first to fix definitely the tissue from which originated the great majority of the epithelial tumors of the mouse. Apolant, having noticed that nearly all tumors occurred among females on the ventral aspect of the body, was led to inquire what subcutaneous structure was limited to this surface of the animal, extended from the jaw to the region of the genitals, and was best developed in females. The mammary glands met the requirements so adequately that Apolant could not escape accept- ing them as the organs giving origin to the tumors. Of these glands the mouse possessed five on each side, the forward pair of which abutted directly upon the submaxillary glands, while that nearest to the caudal end of the animal was in close proximity to the vulva. Micro- scopic examination, furthermore, demonstrated absolutely a mammary genesis for most of the tumors, and made it extremely probable for the remainder. Similar observations by Bashford and Murray estabhshed, in ad- dition, a mammary origin even for tumors situated upon the back. They found that the five pair of mammary glands occupied the sides and front of the body from the neck to the anus, and, extending at the shoulders around the sides of the thorax both behind and in front of the fore-Kmbs, met dorsally in the middle fine. At the inguinal region there was a similar extension along the crest of the ihum and over the inner aspect of the thigh. Histological proof of the mammary origin of these tumors was not wanting, for it was found that they formed a graduated series, extending from growths built up of small acini indistinguishable from those of the normal resting mamma, 1 Verhandl. d. Kofnitees f. Krehsforsch., 1903-1904, iii, 37. See also Deut. med. Woch., 1904, XXX, 1264. 2 Sci. Reports, Imperial Cancer Research Fund, London, 1905, No. 2, Part ii, 15. Lancet, 1907, i, 798. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 83. ^ Arb. a. d. Konigl. Inst. f. Exp. Therap., 1906, Heft i, 17. THE SPONTANEOUS TUMOR 187 through others in which the acini were dilated into cystic spaces reproducing in exaggerated form the puerperal condition, to alveolar growths built on the plan of the rapidly growing mammary founda- tion of the new-born mouse. Tumors of other types were also en- countered in the mamma — angiomata, sarcomata, and squamous cell Fig. 9. — Mouse : Sites of 142 spontaneous mammary carcinomata, shown by black dots ; nipples by stars. carcinomata, and in a proportion analogous to that met with in mam- mary neoplasms of the human subject. Deton,^ having reconstructed wax models from serial sections of two spontaneous carcinomata from the mammary region in the mouse, concluded that the tumors had not originated in the mamma. Apolant,^ however, could not accept this view, because, in his opinion, the method was not sufficiently accurate to demonstrate the relations existing between tumor and mamma. It has been said frequently that new growths in the mouse oc- curred solely in the mamma, and that their analogy with malignant new growths in the human subject was materially lessened by this 1 Zeitschrijt f. Krebsjorsch., 1909-1910, viii, 459. ^ Arch. f. mikroskop. Anal., 1911, bcxviii, 156. l88 THE SPONTAKEOUS TUMOR observation. That the belief was erroneous, however, will appear from the most cursory review of the Kterature, and it is only necessary to direct attention to a few of the recorded tumors to show that the types of neoplasia found in the mouse have been as divers as in man. Thus Borrel ^ found a squamous cell carcinoma of the floor of the mouth and Haaland ^ four similar cases, while Bashford, Murray and Cramer,^ and Murray ^ alone, have reported a series of growths, among which were an adeno- carcinoma of the small intestine, a squamous cell car- cinoma of the stomach, a spindle cell sarcoma of the kidney region, and an adenoma of the liver, probably malignant. Ehrlich and Apo- lant^ have described a spontaneous carcinoma sarcomatodes, Ehrlich ^ a chondroma, and Haaland "^ a melanoma of the ear, an adeno-carcinoma of the kidney, a hypernephroma ( ?) , an adeno-carcinoma of the ovary (probably primary), a fibro-myoma of the uterus, two adeno-carcino- mata of the perputial gland, several sebaceous adeno-carcinomata, three squamous cell carcinomata of the mouth, a spontaneous car- cinoma sarcomatodes, and several sarcomata. Multiple tumors oc- curred in 17 % of Haaland' s cases, as compared with 15 % in Murray's series, and 12 % in Apolant's. In the rat, sarcomata have been described by Loeb,^ Herzog,^ and Jensen,^° a fibro-epithehoma of the tongue by Stahr,^^ an adeno-carci- noma of the seminal vesicle by Flexner and Jobling,^^ and. a car- cinoma of the mamma by Lewin and Michaelis.-^^ 1 Cited by Haaland, Ann. de VInst. Past., 1905, xix, 188. 2 Ann. de VInst. Past., 1905, xix, 188. ^ Set. Reports, Imperial Cancer Research Fund, London, 1905, No. 2, Part ii, 47. * Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 69. ^ Berl. klin. Woch., 1907, xliv, 1399. ^ Arb. a. d. Konigl. Inst.f. Exp. Therap., 1906, Heft i, 65. '' Fourth Sci. Report, Imperial Cancer Research Fund, London, 1911, i. ^ Jour. Med. Research, 1901, N.S., i, 28. ^ Jour. Med. Research, 1902, N.S., iii, 74. ^° Zeitschrift f. Krebsforsch., 1908-1909, vii, 45. ^^ Centralbl. f. allg. Path., etc., 1903, xiv, i. ^^ Proc. Soc. Exp. Biol, and Med., 1906-1907, iv, 12. Monographs on Medical and Allied Subjects, Rockefeller Institute, New York, 1910, No. I, p. I. Flexner and Jobling at first reported their tumor as a sarcoma. During continued propagation, however, it altered in character; in later articles {Proc. Soc. Exp. Biol, and Med., 1908, V, 52, 91) the authors described the gradual ascendancy of epithelial elements (discovered upon reexamination in the spontaneous tumor) and disappearance of struc- tures previously thought to be sarcomatous. The tumor is now an adeno-carcinoma. 1^ Deut. nied. Woch., 1907, xxxiii, 657. See also Lewin, Berl. klin. Woch., 1907, xHv. 1602. c^ e ^ g •£ ^ .= ■? J — ^ o u ~ Zj ■g) "o 'o O s U3 o a THE SPONTANEOUS TUMOR 189 V. Hansemann,^ while not denying the occurrence of carcinoma in the mouse and, indeed, admitting that Borrel's tumors were true carcinomata, was incHned to place Jensen's among the endothehomata. Certainly, he ^ contended, Bashford was unwarranted in calling it a carcinoma solely because it produced metastases, for secondary deposits were found in connection with many endothehomata. In order, however, to be convinced of the endotheHal nature of these neoplasms it was necessary to examine the spontaneous tumors, for in transplanted growths the evidence was not so striking. Lazarus-Barlow^ concurred in the opinion that the tumors found in mice were more properly classed among the endothehomata. THE QUESTION OF MALIGNANCY Occurrence of Metastases Jensen and all others who believed in the malignancy of his tumor were very severely arraigned by Wilhams,^ who asserted that they had relied too much on histological appearance and not enough on chnical observation, and ascribed to their narrow outlook the "ex- traordinary concatenation of blunders with which the liistory of the experimental study of cancer is cumbered." Such a statement, however, scarcely does justice to the care which the investigators of experimental cancer have bestowed upon the clini- cal course of the disease in the mouse and rat, for one of the most eagerly prosecuted inquiries has been that relating to the presence or absence of metastasis formation and infiltrative growth. As early as 1896, Livingood ° discovered in the shoulder region of a white mouse, an adeno-carcinoma which not only recurred several months after removal, but produced a metastatic deposit in the lung. In another mouse, with a carcinoma on the fore-leg, there were several secondary tumors in the lung and, as in the first case, they were of the same structure as the primary growth. ^ Berl. klin. Woch., 1905, xlii, 315. ^ Zeitschrift f . Krebsforsch., 1905, iii, 570. ^ Proc. Roy. Soc. Med., 1908, i, Path. Section, 171. ■* Trans. Path. Soc. London, 1907, Iviii, 38. The Natural History of Cancer, New York, 1908, 187. ^ Johns Hopkins Hosp. Btdl., 1896, vii, 177. I go THE SPONTANEOUS TUMOR Loeb ^ found metastases in rats with transplanted tumors, and sit- uated in regions which they could have reached only through transpor- tation by the blood or lymph streams. He pointed out that rats bear- ing propagable neoplasms did not live for more than two or three months, and that it was perhaps necessary, as in man, that the growth be present during a longer period in order that secondary deposits might occur. Borrel ^ discovered metastases in the lungs and the lymph nodes of mice bearing tumors, and Haaland ^ concluded, after an examina- tion of serial sections of the organs of mice spontaneously affected with cancer and of those inoculated with the Jensen carcinoma, that metastasis by way of the blood stream was of common occurrence, at least in association with growths which were at all advanced in age. Among six cachectic mice bearing Jensen's tumor he found pulmonary nodules in five, although he was able to confirm Jensen's statement that metastases could not be detected in the Ijnnph nodes, while among spon- taneously affected mice lymphatic deposits were difficult to demon- strate, and might be considered rare. Even after a much more extensive experience with spontaneous tumors, Haaland ^ had not found the lymph nodes involved nearly so often as the lungs, in which metastatic nodules clearly visible to the naked eye were discovered in 38 % among two hundred and seventy-three cases. In other locations, however, secondary deposits were much more rare, occurring but four times in the Hver, once in the kidney, once possibly in the ovary, thrice on the peritoneum, once in the retroperitoneal tissue at the site of the adrenal, once under the diaphragm, once in the posterior medias- tinum, and once in the spleen. In one case the cells of a lung metastasis had grown through the vessels into the heart, where they floated free in the blood stream. Bashford and Murray,^ in the earliest days of their experience with the Jensen tumor, failed to discover secondary nodules, but in the following year, in collaboration with Cramer,^ described them in the lungs of mice inoculated with this growth, 1 Jour. Med. Research, 1902, N.S., iii, 48. ^ Ann. de I'Inst. Past., 1905, xix, 172, 184. ^ Ann. de I'Inst. Past., 1903, xvii, 114. Deut. med. Woch., 1905, xxxi, 1239. * Fourth Sci. Report, Imperial Cancer Research Fund, London, 191 1, 54. ^ Sci. Reports, Cancer Research Fund, London, 1904, No. i, 13 (footnote). ^ Sci. Reports, Imperial Cancer Research Fund, London, 1905, No. 2, Part ii, 38. ;.^'/?' -•'•.-••!••■ J/. ••'•?r /•. • .v. ... . .•■/ •.V-.- .".■ • . '•••'•^jXVi^r'c^r^v;^^ #^&^^^^ ■^;^|^ ■■- •■■■ 1 ■••••.p. X fl. Ford, del. Metastasis in cervical lymph node of the mouse, from a spontaneous mammary carcinoma. X ^. Metastasis in aortic lymph node of the mouse, from another spontaneous mammary carcinoma, i.v.c, inferior vena cava ; u, ureter. The growth has expanded the capsule of the gland and the lymphoid tissue is reduced to a crescent. X ^r. THE SPONTANEOUS TUMOR I9I Murray ^ found metastases in the lung in rather less than 50 % of sixty-eight mice with spontaneous tumors, even though these organs had not been completely investigated in all the cases, and Gierke - dis- covered secondary nodules in the same location in eight out of thirty- five cases. Lymphatic metastases were detected by Murray in three cases of his series, and by Gierke in one. In the rat, Flexner and Jobling^ reported a tumor which metas- tasized extensively throughout the bodies of animals into which it was transplanted and which, in the fifth generation, involved even the regional lymph nodes, although no secondary deposits were found in the animal primarily affected. Michaelis ^ discovered in the same species a carcinoma of the mamma which, although it had not given rise to metastases in the rat bearing the spontaneous growth, pro- duced them after inoculation, nevertheless, in the lungs, Hver, and glands (Drusen) of other rats. Infiltrative Growth The mahgnancy of the tumors of mice has been impugned on the further ground that evidence of infiltrative growth had never been observed. But as early as 1904 Bashford and Murray ^ had investi- gated this type of extension, and had made the statement in a foot- note that the malignancy of the Jensen tumor was demonstrated by its infiltrative growth, while in the following year Bashford, Murray, and Cramer ^ described the phenomenon in more detail. Baeslack,'^ also, had discovered infiltration on the part of some tumors of the Jensen series. Michaelis,^ on the contrary, said that infiltrative growth had never been seen ; and although a year later he^ 1 Third Sci. Report, Imperial Cancer Research Fund, London, igo8, 98. ^ Beitr. zur path. Anat., etc., (Ziegler), 1908, xliii, 331. Third Sci. Report, Imperial Cancer Reicarch Fund, London, 1908, 118. ^ Proc. Soc. Exp. Biol, and Med., 1906-1907, iv, 12. Jour. American Med. Assoc, 1907, xlviii, 420. Joiir. American Med. Assoc. 1908, 1, 66. * Zeitschrift f. Krebsforsch., 1907, v, 190. See also Lewin and Michaelis, Deid. med. Woch., 1907, xxxiii, 657. ^ Sci. Reports, Cancer Research Fund, London, 1904, No. i, 13. ^Sci. Reports, Imperial Cancer Research Fund, 'London, 1905, No. 2, Part ii, 39, 47. ' Dent. med. Woch., 1905, xxxi, 957. ^ Deut. med. Woch., 1904, xxx, 1264. ^ Med. Klin., 1905, i, 203. 192 THE SPONTANEOUS TUMOR still felt that an important characteristic of maHgnancy was lacking while this method of extension remained midemonstrated, a brief note ^ published shortly afterward contained the affirmation that he had observed infiltration in certain stages of his tumors, and that the last distinction betv/een human carcinoma and those found in the mouse had thus disappeared. Metastases he had been able to demon- strate, once in the lungs and once in the lymph nodes in two mice with spontaneous tumors, and in the lungs of a mouse bearing a transplantable growth. The explanation of Apolant^ that infiltrative growth, although it did occur in the mouse, was not so marked as in human beings because the tumors developed in loose connective tissue, where they had an abundance of room in which to grow without involving skin and mus- culature, was entirely in harmony with the observation of Bashford, Murray, and Cramer,^ that the Jensen tumor forsook its expansive growth and assumed the infiltrative type, when dense tissue opposed its progress. Bashford and Murray,^ in a detailed account of sporadic mammary carcinoma in the mouse, described metastases in the lymph nodes and lungs, and invasion of the capsule of the tumor by strands and colunms of cancer cells. The involvement of structures adjacent to the growth made complete surgical ablation possible only when both capsule and surrounding tissue were removed, the mere shelHng out of the tumor from its capsule having been found inadequate. Their observations permitted the conclusion: "... that we are dealing with new growths of the mammary region of the mouse, which grow progressively (recur after incomplete removal) , infiltrate the surrounding normal tissues, and produce metastases of the same histological type in the lungs and lymphatic glands. They lead to the death of the animal." Murray^ found, indeed, that over one-half of the tumors recurred following at- tempts at total excision, after an interval of from two weeks to several months. Furthermore, he had been able to demonstrate by histologi- cal examination that the encapsulation was only apparent. ^ Med. Klin., 1905, i, 496. ^ Arh. a. d. Konigl. Inst. f. Exp. Therap., 1906, Heft i, 60. ^ Sci. Reports, Imperial Cancer Research Fund, London, 1905, No. 2, Part ii, 40. * Lancet, 1907, i, 800. ^ Jour. Path, and Bad., 1908, xii, 437, THE SPONTANEOUS TUMOR I93 The malignancy of mouse tumors was adequately defended by Apolant ^ in a recent reply to v. Hansemann's objections ^ against comparing these growths to those of man. More recently still v. Hansemann^ has receded from his original position, and now admits metastasis to be an occasional event attending spontaneous growths, although still denying that the tumors so widely regarded as mam- mary originate in the mamma. The two main objections urged against the malignancy of mouse and ~at tumors, namely, absence of metastases and of infiltrative growth, have thus been overborne by a mass of evidence gradually collected in laboratories throughout the entire world. Altmann^s Granules as a Criterion of Malignancy The question of the malignancy of these growths was attacked from still another side by Beckton,"* who found that Altmann's granules, although they were demonstrable in most of the varieties of normal cells and in those of inflammatory lesions and benign tumors, were partly or entirely absent from the cells of mahgnant growths in the human subject. In a considerable number of reputedly mahgnant dog tu- mors which this author had the opportunity to examine, Altmann's granules were found in much greater numbers than in mahgnant neoplasms of human origin, while, as concerned normal tissues, the gran- ules were usually missing from the cells of stratified squamous epithe- Hum and unstriped muscle. In the embryos of the mouse, the rat, and the chick, granules of medium or small size were discoverable in moderate numbers in the earher periods of development, but during the later stages these and other granules might be either present, or absent as in some squamous epitheha, unstriped muscle, and the epithehum of the collecting tubules of the kidney. On the basis of these observations the author suggested that an ex- amination for the presence of Altmann's granules might yield a useful diagnostic criterion regarding the mahgnancy of a given tumor, exclud- 1 Berl. klin. Woch., 191 2, xlix, 495. ^ Berl. klin. Woch., 1912, xlix, 224. ^ Berl. klin. Woch., 1913, 1, 81. * Jour. Path, and Bad., 1909, xiii, 185. Arch. Middlesex Hosp., 1909, xv, 182. Lancet, 1909, ii, 391. British Med. Jour., 1909, ii, 859. Jour. Path, and Bad., 1910, xiv, 408. British Med. Jour., 1910, ii, 1422. Arch. Middlesex Hosp., 1910, xix, 103, iii, 115. 194 THE SPONTANEOUS TUMOR ing, of course, such growths as had arisen from tissues that did not normally contain the granules, and having regard to the anomalous position occupied by ovarian and thyroid tumors. Turning his atten- tion to tumors of the mouse, Beckton examined twenty-one unselected examples of propagable sarcomata and carcinomata, two-thirds of which afforded definite e\'idenceof mahgnancy in the absence of granules from their cells, while in the remaining third the granules were present. The latter group, however, Beckton hesitated to stamp as non-malig- nant because of the present uncertain state of available knowledge on the subject. While, then, Altmann's granules were found on the whole in decidedly greater numbers than in mahgnant new growths of human origin, still, in about two-thirds of the cases the mouse tumors were as devoid of granules as were human new growths of un- doubted mahgnancy. Later papers by Colwell and Beckton,^ Beckton and Colwell,^ Beckton and Russ," and Miller/ dealt with various technical matters in relation to the granules under discussion. The experience of Beckton was controverted by Bensley,'^ who beheved that the cells of mahgnant tumors in man did contain Alt- mann's granules, and even in greater number than they could be de- tected in the tissues from which the tumors arose. A disagreement as to the granule content of tumor cells is, however, not a novelty, for while Raum ^ beheved that he had demonstrated the presence of Altmann's granules in the cells of both carcinomata and sarcomata, Burkhardt,^ on the other hand, was convinced that the properties of the cancer cell were altered in such a way as to destroy the granules. In any case, it is doubtful whether these bodies have any sig- nificant relation to the biological quaHties of a tumor. ^ Arch. Middlesex Hasp., 1911, xxiii, 41. 2 Arch. Middlesex Hasp., 191 1, xxiii, 52. ^ Arch. Middlesex Hasp., 191 1, xxiii, 99. * Arch. Middlesex Hosp., 191 1, xxiii, 106. * Trans. Chicago Path. Soc, igio, viii, 78. ^ Arch. f. mikroskop. AnaL, 1892, xxxix, 137. ''Arch.J. kiln. Chlr., (v. Langenbeck), 1902, Ixv, 135. THE SPONTANEOUS TUMOR I95 AETIOLOGY Influence of Infectivity As the investigation of mouse tumors became more widespread, instances of apparent contagion through the medium of *' cancer cages" began to appear in the Hterature. Thus Borrel ^ recorded a breeding estabhshment in which three cases of cancer had occurred in a single month. Upon making a visit to this establishment he was informed that in two years there had been found, in the same cage, more than twenty tumor mice, an enormous percentage compared with the number of animals kept on hand. The dealer had been in the habit of selling the young mice, and the growths had been observed solely in the older ones retained for breeding purposes. Borrel offered as further evidence of infection the instance of a cage in which five or six cancers had arisen during one year, although in other establish- ments, which furnished many hundreds of mice yearly, no case had ever been discovered. All these observations pled for the existence of a cancerous virus, a conception which Borrel ^ has consistently upheld in subsequent monographs. He has always insisted that transplanted tumors represented only the second stage in the development of cancer — that of mahgnant grQwth. Because in spontaneous neo- plasms this secondary period was preceded by a transformation of normal cells into mahgnant elements, it was to the earlier stage that research should be devoted, and especial attention had accordingly been bestowed upon the youngest tumors in the study of his own material. Cysts were not uncommonly found among mice in the groin or axilla, both of which were sites of election for the biting insects. In these cysts, or in the connective and glandular tissue surrounding them, Borrel could frequently discover Cestodes or Nematodes, and these helminths seemed to be without doubt a causative factor in the production of the lesions. In the connective tissue, muscles, or capillaries adjoining young adeno-carcinomata, he had also succeeded in demonstrating nematodes, and in a minute tumor had found the ' Ann. de I'Inst. Past., 1903, xvii, 113. ^ Bull, de rinst. Past., 1907, v, 497. Travaux de la deuxieme Conference internal, pour V Etude du Cancer, Paris, 191 1, 193. 196 THE SPONTANEOUS TUMOR trail of such an organism, with the parasite lying at the end of it. Similar worms had been discovered, moreover, in the Imig and the lymph nodes of its hilum, as well as in the circulation of mice suffer- ing from generaHzed lymphomata, but he had not yet thoroughly ex- amined normal mice for their presence. While the temptation was great to consider the nematodes, introduced by biting insects, as the carriers of a virus, their relation to malignant growth could be no more than suggested. It was possible, however, to express greater certainty in the case of sarcoma of the liver in rats. Four years ago he had reported a rat cancer propagable through three generations, and had attributed the growth to the cysticercus of Taenia crassicola, while in another case there had been discovered a tiny adeno- carcinoma of the kidney developing about a cysticercus. Similar cases have been reported by Regaud,^ Saul," Bridre,^ and by Bridre and Conseil.* But not only were endoparasites found in connection with malig- nant growths. Certain ectoparasites, as the Acaridae, might be at fault, and in the mouse, for example, there was recognized a cutaneous affection which was distinguished by warty excrescences upon various parts of the body. In the early stages of this disease, an adenomatous condition of the sebaceous glands, there could be discovered a still unidentified mite. These organisms, furthermore, played an impor- tant role in the lympho- sarcoma of dogs,^ and one of them, the De- modex, was connected with cancer of the face in man. While an infection with such insects was frequently present in the nipples of cancerous women, Borrel did not wish to draw any conclusions regarding the relation of the parasite to cancer of the breast, and pointed out that the great majority of healthy persons would present the demodex in certain portions of the skin — perhaps hundreds to the square centimeter. Various authors had wrongly ascribed to him ^ Compt. rend. Soc. Biol., 1907, Ixii, 194. ^ Centralbl. f. Bakt., etc., erste Abt., Orig., 1909, xlix, 80. Berl. klin. Woch., 1911, xlviii, 341. Deut. med. Woch., 1911, xxxvii, 233. * Compt. rend. Soc. Biol., 1909, Ixvi, 376. * Bull, de V Assoc, franq. pour V Etude du Cancer, 1909, ii, 171. Bidl de V Assoc, franq. pour VEtude du Cancer, 1910, iii, 318. ^ Compt. rend, de I' Acad, des Sc, 1907, cxliv, 344. THE SPONTANEOUS TUMOR 197 the h}"pothesis that such organisms as these were the parasites of cancer, although he actually believed that they acted only as carriers of a virus. Because a nail occasionally produced tetanus, one would never think of describing it as the parasite of that disease ; it was merely that the bacillus causing tetanus was able to develop after having been carried into the tissues by the foreign body. Attention was directed by Loeb and Jobson ^ to a ranch in Wyoming, upon which there had been discovered one or two cases of carcinoma of the inner canthus of the eye every year for the past ten years, among two thousand head of cattle. This incidence was fifty times greater than the average, and the similarity of the tumors in respect of character and location was remarkable. The animals on the neighboring ranches were free from carcinoma. MichaeHs - mentioned a dealer who had found five tumors arising successively among the mice of a single cage. The last two of these growths were examined, and proved to be maHgnant adenomata. Gaylord and Clowes^ reported that in the spring of 1902 there were brought to the State Cancer Laboratory in Buffalo a number of rats inoculated with a cystic sarcoma of the thyroid, and that for the accommodation of these animals two large cages and a number of smaller ones were constructed. For a certain period of time these cages contained numbers of successfully inoculated rats, but from December, 1902, until the summer of 1903, they were empty. During and after the summer of 1903 the large cages were again put to use, and a year later one of the rats kept in them developed a fibro-sarcoma. Eight others were accordingly placed in this cage, while into the second one there were put six or eight more. Four months later the two sur\dving males in the first cage had developed tumors, one of which was a cystic spindle cell sarcoma of the thyroid, the other a fibro- sarcoma. No tumors appeared either among the inmates of the control cage or of the other cages in the laboratory although there were no less than one hundred rats on hand, nor did any develop among the animals in the smaller cages, which had been sterilized. 1 Jour. Comparative Med. and Yet. Archiv., 1900, xxi, 3SS. See also Loeb, Arch.f. kiln. Chir., (v. Langenbeck), 1903, Ixx, S45,and Centralbl. f.Bakt., etc., erste Abt., Orig., 1904, x.xxvii, 236. ^ Zeitschrijt J. Krcbsforsch., 1906, iv, 2. ^ Jour. American Med. Assoc., 1907, xlviii, 15. 198 THE SPONTANEOUS TUMOR A second observation by the same authors related to the endemic occurrence of cancer among mice. There was discovered in the pos- session of a dealer a cage in which about sixty spontaneous tumors had occurred in the course of three years. The location of this cage had been changed frequently and the stock had been entirely re- moved on at least one occasion. The dealer was of the impression that most of the affected mice had been females, and added that he had never seen a tumor on the back of a mouse. Balancing these, and many similar observations, stands the following quotation from Bashford ^ : "... alleged epidemics have often been re- corded in . . . mice and rats housed together in small cages. Satis-" factory proof that these aggregations of cases were due to infection has not been furnished, and the alternative explanation, that they arise as the result of in-breeding cancerous stock, has naturally sug- gested itself. Our very detailed observations on tens of thousands of mice have not revealed in our laboratory anything which we would call an epidemic. When, however, we take into consideration the manner in which cases of carcinoma mamm« have been sent to us by breeders we find the same kind of e\ddence as that which has led observers in France, America, and Germany to assert that epi- demics of cancer occur in breeding establishments. We may illustrate this kind of evidence by the numbers of tumour-mice sent in by four of the breeders who supply us with mice, under a guarantee that no fresh stock has been introduced. From January i, 1906, to October 31, 1907, Mr. A sent us ten cases, Mr. B, six cases, Mr. C, thirty-five cases, and Mr. D, eighteen cases of carcinoma of the mamma. These figures, which are more remarkable than any others yet pub- Hshed, are no evidence that there was an endemic or epidemic occur- rence of cancer in the breeding-cages of Mr. C or ]\Ir. D. The proportions of mice supphed to us in the same period to cases of cancer were as follows : — Mice with Tumour Total Mice "Mr. A 10 1,302 Mr. B 6' 1,547 Mr. C 35 ... 9,698 Mr. D 18 . 11,842 ^ Proc. Roy. Soc. Med., 1909, ii, General Reports, 72. 'the spontaneous tumor 199 "The numbers of tumours occurring in these stocks of mice have been determined solely by the number of mice of 'cancer age' under obser- vation. This is brought out particularly clearly in the difference between the age constitution of the stock of Mr. C and Mr. D, since the stock of the latter contains constantly a much higher proportion of young animals, and he supplies us with most of our young mice. Further, if we note the dates on which tumours are sent to us and arrange them in columns, we find that the crops of tumours coincide with the ageing of groups of mice. Thus those apparent aggregations of cases, wrongly called epidemics by too enthusiastic advocates of a para- sitic origin for cancer, also give no indication of haphazard in-breed- ing leading to a preponderance of cases of cancer of the mamma. The incidence of the disease for mice continues to obey the laws of age- and sex-distribution, even where in-breeding is proceeding haphazard." Influence of Age and Sex Most investigators are agreed that spontaneous malignant tumors are much more common in female than in male mice, that the most frequent site is the mammary gland, that the tumors are nearly always of epithelial origin, and that, as in man, age plays an important part in preparing the animals for their inception. Malignant growths are found in the mouse with comparative fre- quency if they are sought, and whereas in 1905 Bashford, Murray, and Cramer^ had been able to discover but twelve among nearly thirty thousand tame mice of all ages, Murray,^ six years later, re- corded 8.6 % in two hundred and twenty-three females of non-cancer- ous ancestry, and of all ages from six months upward. Apolant,^ among two hundred and twenty-one tumor mice (two hundred and seven white and fourteen gray), found two hundred and seventy- six growths, thirty-eight of the mice having from two to five each, while in confirmation of observations previously published with Ehrhch "^ not a single case had occurred in a male. Among these two ^ Sci. Reports, Imperial Cancer Research Fund, London, 1905, No. 2, Part ii, 15. & ^ Proc. Roy. Soc, Series B, igii-1912, Ixxxiv, 42. Fourth Set. Report, Imperial Cancer Research Fund, London, 191 1, 129. ' Arb. a. d. Konigl. Inst. f. E.xp. Therap., 1906, Heft i, 15. * Berl. klin. Woch., 1905, xlii, 872. 200 THE SPONTANEOUS TUMOR hundred and seventy-six tumors, two hundred and thirty-three of which were examined and found to be epithehal, two hundred and fifty-nine were on or near the ventral surface of the body. Among two hundred and eighty-eight tumor mice, according to Haaland,^ only six were males, and of the three hundred and fifty- three tumors detected in these mice, three hundred and eleven pre- sented the structure of mammary adeno- carcinoma. Influence of Lactation Apolant ^ was struck by the fact that tumors in the mouse were found most often in an organ of very vigorous functional energy. That, in contrast to man, a single organ should be so uniformly affected, he thought might be partly explained by the injuries inseparable from physiological acti\dty. This interpretation has been discounted con- ,siderably, however, by the later observations of Haaland." "Of 74 mice which had developed mammary tumours, and which were under observation from birth, only 33 are recorded to have Httered previously to the development of the tumour, while no fitter is recorded for 41. . . . Of the 33 ha\dng fittered pre\dously to the development of the tumour : — " I fitter is recorded in 11 cases 2fi tters are '" / 3 ii (C 7 " 4 u cc " 5 5 a a 2 " 7 Cl a I case "The figures show that it is hardly likely that excessive physio- logical demands made upon the mamma play the determining role in the development of tumours of this organ, but that other factors must be looked for." Influence of Heredity One aspect of the investigation of cancer to which the mouse lends itself most readily is that concerning the relation of heredity to the '^Fourth Sci. Report, Imperial Cancer Research Fund, London, 1911, 9. 2 Arh. a. d. Konigl. Inst. f. Exp. Therap., 1906, Heft i, 40. ^Fourth Sci. Report, Imperial Cancer Research Fund, London, 191 1, 43. THE SPONTANEOUS TUMOR 20I occurrence of spontaneous tumors, and this because mice rarely live to be over two years old. The diificulty of settling the question in man is well illustrated by a paragraph from Newsholme : ^ — "The fact that several members in successive generations of a given family have died from cancer is commonly accepted as proof that the disease is hereditary. This is far from being the case. Cancer causes a certain average number of deaths among a given number of persons. . . . Given a certain probabiKty of death from cancer, and knowing the number of a family, it is easy to calculate the probability of one, two, or more of them dying of cancer, quite independently of heredity. Even if heredity were proved to be absolutely inoperative, it is certain that there would be famihes among whom numerous deaths from cancer would occur. It does not prove heredity to show that in one family five deaths, say, occurred from cancer. This might happen from mere chance, and in fact, such cases must occur without heredity at all. De Morgan worked out the probabihty of looo successive heads being thrown in tossing a coin, and he showed that given a sufificient number of people starting to toss coins, it was a certainty that at least one of them would toss looo consecutive heads. So, given a sufficient number of families, it is a certainty, even if there be no such thing as heredity, that of at least one family, say ten members v^^ill die of cancer. The only absolute proof of heredity would be to show that cancer occurred frequently in certain families, and practically nowhere else ; short of this the probability of heredity of cancer would be increased if it could be shown that cancer was much more common in certain famihes than in the average for the whole community, due allowance being made for variations in age and sex-distribution." The suitabihty of the short-Uved domestic animals for the investi- gation of problems in heredity was indicated by Bashford,^ who re- ported later ^ that after three years of failure he had been able to obtain offspring from mice with spontaneous cancer, and expressed the hope that data relating to the existence of hereditary susceptibility might be obtained in time. In the following year '^ breeding experiments were well under way, and by crossing spontaneously affected animals 1 The Elements of Vital Statistics, London, 1899, 248. ^ British Med. Jour., 1903, ii, 128. ^ BritishMed. Jour., 1906, ii, 207. ^British Med. Jour., 1907, ii, 27. 202 THE SPOXTAXEOUS TUMOR with the offspring of cancerous parents, strains of mice were being obtained in which the cancerous heredity was one-hah", three-fourths, fifteen-sixteenths, or even higher. A prehminarj^ account of these experiments was given by Bashford and Murray/ and a completer analysis, founded on more extended ob- servation, by Murray.- The follo-^ing table, abbreAdated from Mur- ray's, shows the ratio of deaths from cancer to deaths from all causes, among five hundred and sixty-three ^ mice of a highly inbred stock with a heredity more or less cancerous, the figures referring to females only. The table "... shows a rapidly increasing proportion of deaths from cancer commencing after sLx months is passed, and attaining a maximum in the three-monthly period ending at i8 months. In the succeeding periods the frequency diminishes, till in mice over 24 months old the frequency is barely t\^dce that found in mice under 9 months old. Similar figures for the human female give a cor- responding curve." TABLE I {after Murray) Age (months) . . ^3 -6 -9 -12 -IS -18 1 -21 -24 Over 24 TOT.\L Total mice — — 100 104 88 94 69 55 53 563 Percentage of mammar}' cancer — — 5-0 10.6 18.2 28.0 14-5 14.0 9.4 14.4 The five hundred and sixty-three mice were then distributed into two groups. The first contained three hundred and forty mice of recent cancerous ancestry, of which the mother, one or both grandmothers, or all three had been cancerous. To the second were assigned two hundred and twenty-three mice of remote cancerous ancestry — mice, that is. where no cancer had occurred in ^ Proc. Roy. Soc, Series B, 1909, Ixsxi, 310. '^ Proc. Roy. Soc, Series B, 1911-1912, Ixxxiv, 42. Fourth Set. Report, Imperial Cancer Research Fund, London, 191 1, 114. ^ Through a clerical error the total was given in Murray's paper as five hundred and sixty-two. THE SPONTANEOUS TUMOR 203 either the mothers or the grandmothers. The majority of the cases of cancer occurred in the members of the first division, as may be seen from tables II and III (abbreviations of Murray's) which show the percentage of cases occurring in the first and second groups respec- tively. TABLE II {after Murray) Age (months) . . 0-3 -6 -9 -12 -15 -18 -21 -24 Over 24 Total Total mice . . . — — 62 63 62 56 40 29 28 340 Percentage of mammary cancer — 6.5 II. I 24.2 32.1 25.0 17.2 10.7 18.2 TABLE III {after Murray) Age (months) . . 0-3 -6 -9 -12 -IS -18 -21 -24 Over 24 Total Total mice . . . — — 38 41 26 38 29 26 25 223 Percentage of mammary cancer — k 2.6 9.8 3.8 21.6 0,0 "•5 8.0 8.6 The figures were submitted, further, to mathematical analysis in- volving a determination of the standard errors of the differences be- tween the cancerous and non-cancerous groups . Taking the crude data, the actual percentages amongst all the offspring were : — Ancestry cancerous 18.2% Ancestry non-cancerous 8.6% When a correction was made for the varying age-distributions of the two groups by calculating corrected percentages based on the age- distribution of all mice as a standard, and reducing the numbers for the five hundred and sixty-three mice to the corresponding proportions per thousand, the difference was increased only from 9.6 to 9.8%. As the standard error of this difference was 2.96, the difference was 3.3 204 THE SPONTANEOUS TUMOR times the standard error, and the chance of its occurring as a mere fluctuation of random sampHng only about one in a thousand. Murray thought, therefore, that the disparity between the two groups was not due to chance but, on the contrary, almost certainly signifi- cant, and that the increased Hability was probably in the nature of a predisposition of one particular tissue or organ system to undergo cancerous transformation under the wear and tear of Hfe. The differ- ence between the two series was apparent at all ages, and the age of maximum incidence did not appear to have been lowered in the predis- posed group. In discussing the application of these results to man, Bashford ^ extended a warning against too pessimistic conclusions, pointing out that the influence of heredity had been demonstrated only for stocks where this factor had been concentrated by careful mating. Such a concentration could occur in man only by hazard as a coincidence of considerable rarity, and it was probable that the influence of heredity in the general population was manifested as an average predisposition of low intensity. Meanwhile the question had been approached by Tyzzer,^ who called attention to the relative frequency of growths in the descen- dants of those mice which had fallen prey to mahgnant tumors. The number which had occurred in three families derived from three female mice with spontaneous growths indicated that tumor development was influenced by inherited quahties. Murray,^ without wishing to minimize the value of Tyzzer's careful work, indicated that the neoplasms recorded by that author belonged mainly to types in which the maUgnancy was not very pronounced. The majority of the tumors were cyst-adenomata of the lung, a vari- ety which it would be hazardous to regard in all cases as mahgnant^ and the multiple lymphomata, which were second in frequency, could not be separated satisfactorily from conditions resembling diffuse hyperplasia, although the type did include new growths of undoubted ^ Fourth Sci. Report, Imperial Cancer Research Fund, London, igii, Introduction, p. xvi. 2 JoiiT. Med. Research, 1907-1908, N.S., xii, 199. Jour. Med. Research, 1909, N.S., xvi, 479. ^ Fourth Sci. Report, Imperial Cancer Research Fund, London, 191 1, 115. J. R. Ford. del. Hypertrophic mairLmar>- nodule. Increased number of acini, on the whole of fairly normal appearance, with increased cellularitj' of the stroma. Concretions in the lumina. At one point the acini have assumed atypical forms, and there are signs of pressure in the surrounding tissues. Xo other sign of active growth. X -i-. J. R. Ford. del. H\-pertrophic mammary nodule from another mouse. The epitheliimi of the acini is distinctly atypi- cal, in that the cells are larger than normal and more crowded, so that several layers are found lining some acini. This appearance is not due to obliquity of the section. A few mitoses are discoverable. X -f^. THE SPONTANEOUS TUMOR 205 malignancy. Furthermore, while cyst-adenomata of the lung oc- curred in males and females with equal frequency, multiple lympho- mata had been found among Tyzzer's material much more often in females, and a serious source of error was introduced when the fre- quency of tumors as a whole was reckoned on males and females together. Jensen ^ was able to bring to maturity among the descendants of a cancerous female about fifty mice in four or five generations, none of which, however, developed a tumor. Another mouse with spontaneous cancer bore four young, one of which developed an enormous intra- abdominal round cell sarcoma. Two males of this litter used for breeding gave rise to four or five hundred descendants in six or seven generations, and even though many of their offspring died at an early age during epidemics, several tumors of the usual type developed among them. Influence of Inflammation Haaland,^ in trying to account for the multicentric origin of mam- mary tumors in mice, examined serial sections of the mammae in both cancerous and non-cancerous animals. The normal condition of this gland after the era of physiological activity had passed appeared to be one of more or less atrophy, acQompanied by sclerotic changes in the connective tissue and arteries. Chronic inflammation of the inter- stitial tissue, either diffuse or locaHzed, was very common, as was dila- tation of the ducts with its consequent cyst formation. The epithe- lium bounding these cavities was usually that of the normal duct or acinus flattened by pressure of the contents, although occasionally the cysts were fined by epithefial cells of the squamous type. In addition to the lesions just mentioned, all stages of nodular hypertrophy of the mammary epithefium were met with, varying from a sfight increase in the number of normal acini in one lobe, to the formation of definite nodules. All gradations might exist between this condition and mafignant growth, and, indeed, there were times when it was impossible to decide between the two. The frequent association in the same gland of all degrees of nodular hypertrophy with veritable ^ Zeitschrift f. Krebsforsch., 1908-1909, vii, 285. 2 Fourth Sci. Report, Imperial Cancer Research Fmid, London, 191 1, 30. 2o6 THE SPONTANEOUS TUMOR new growths, made it probable that the lesion bore some relation to tumor development, and it was intimated that the nodes might be either the base upon which cancerous proliferation started, or true tumors from their very inception. While nodular hypertrophies were Fig. io. — Multiple minute hypertrophic nodules in the mammEe, reflected with the skin. The figure also represents the zone free from mammary gland, chosen for autologous inoculation. found most commonly in mice with cancer in another mamma, Haaland proved that they were not secondary deposits by demonstrating in serial sections their normal connection with the mammary ducts. The inflammatory and sclerotic conditions in the connective tissue were important because many cases of tumor development sug- gested that the process had taken place in an organ diffusely diseased, and, moreover, because the nodular h>TDertrophy just discussed was K ^^1 J. R. Ford. del. Stretched and stained subcutaneous tissue of old normal female mouse. A nematode is seen, con';ainin| large embryos; five extruded embr\^os in the subcutaneous tissue at a considerable distance fromi the mother worm. X ^ . c? r,"^2a. J. R. Ford. del. Oblique section of a nematode embryo free in the interstitial mammar}' tissue of an old normal female mouse. X ^^ THE SPONTANEOUS TUMOR 207 a process too far advanced to be investigated setiologically. It was therefore necessary to examine the mammary glands of old non-can- cerous mice. When this was done, inflammatory changes in the con- nective tissue were discovered to be of common occurrence, and often associated with the presence of nematodes, such as had previously been found by Borrel in young tumors and their surroundings, and in the lungs, mammae, and mediastinal lymph nodes of tumor-bearing mice. In his own cases, Haaland detected the parasites in the subcutaneous tissue of a large proportion of old normal mice from different breeders, in addition to finding them, also, in the pulmonary vessels and once in the pleura. They were usually few in number and discoverable only after long search, although as many as half a dozen or more had been found in one mamma. The worms were all females containing large numbers of eggs or embryos, and the possibility of liberation of the latter in the subcutaneous tissue was demonstrated by their dis- covery in that location. That the nematodes were capable of setting up an intense inflamma- tion was shown, not only by the zone of leucocytes surrounding areas where they had remained, alive, for some time (encapsulated?), but by still other signs marking thei^ path. When they lay dead in the tissues a vigorous leucocytic and lymphocytic reaction supervened, and phagocytosis proceeded until the worm finally became unrecog- nizable. These characteristic lesions accompanying the parasites affected the loose connective tissue alone, and the mammary gland was involved only indirectly and in so far as it was embedded on all sides in this tissue. In several instances the inflammatory condition co-existed with nodular hypertrophy, an association which suggested that the two might be connected and referred to a common cause — the pres- ence of nematodes in the subcutaneous tissue. In half a dozen males examined, the worms were found in the same proportion as in females, while evidence from young or adult mice was still scanty. The observations of Murray ^ on carcinoma of the liver in cows gain an enhanced interest in connection with Haaland's remarks. These growths Murray found very common, and almost invariably asso- 1 Veterinary News, 1910, vii, 563, 2o8 THE SPONTANEOUS TUMOR dated with a severe biliary cirrhosis due to infection with the Distoma hepaticum. The new growths did not arise in the bile ducts, however, but in the parenchyma, and the irritation of the parasites acted only as an indirect cause. The observations of both authors together are of further significance when taken in conjunction with the well-known frequency of epithe- lial growths of the rectum and bladder of man in the presence of in- fection with the Bilharzia haematohia. ^Etiology Theoretically Considered Ehrlich ^ advanced the following explanation of malignant growth, couched in terms of the side-chain theory. The division of nutrient ma- terial throughout the body was regulated by the number of receptors possessed by the various cells, as well as by the avidity of these receptors, so that immoderate growth, like that of a tumor cell, could be achieved only through equipment with receptors having an avid- ity for foodstuffs relatively higher than the normal. When, as often happened, a primary tumor of the mouse was implanted unsuccess- fully into hundreds of other mice, this failure could mean only that the receptors of the transplanted tumor cells possessed an avidity no higher than the average obtaining in the cells of these mice. Hence the origin of a growth must be conditioned, not by an increase in the avidity of its own cells, but by a general lowering of that of the receptor apparatus of the entire body, and the tumor cells possessed, therefore, a relatively but not an actually higher avidity than the normal cells of the organism. The growth of a transplantable tumor in normal mice was, however, in the opinion of Murray,- more a result of the adaptability of its cells to new surroundings. He found that tumors transplanted into sponta- neously affected animals which had been relieved by operation became established in consequence of a great power to accommodate them- selves in new hosts, and that they shared the food supply in common with the normal tissues without starving them. When a spontaneous ^ Zeitschrift f. aerztliche Fortbildung, 1906, iii, 210. Arh. a. d. Konigl. Inst. f. Exp. Therap., 1906, Heft i, 86. ^ Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 108. THE SPONTANEOUS TUMOR 209 growth recurred it increased in size more rapidly, as a rule, than did the transplanted tumor, even though its energy of growth (as tested by inoculation into normal animals) was much inferior to that of the propagable neoplasm. This would be quite inexplicable on the as- sumption that a difference in avidity for food was the factor princi- pally determining their relative rates of proliferation, but became at once intelligible when the importance of adaptability was considered. The recurrent spontaneous tumor was in its native environment and the food materials presented to it were those to which it had always been accustomed, wherefore adaptation was not required, and suc- cessful competition with a transplantable growth was possible. The relation of a spontaneous growth to the animal bearing it could not be described so simply as Ehrlich had supposed, merely by postulat- ing a different avidity for food in accordance with the side-chain theory. The relative affinities for nourishment exhibited by normal tissue and tumor must remain vaguely speculative, as long as the avidity of malignant cells was measured solely by their ability to grow and com- parisons made with normal adult cells no longer engaged in active proliferation. Haaland ^ exchanged grafts between two animals spontaneously affected, each tumor being at the same time inoculated into its bearer and into normal mice. In the course of these experiments two can- croids were transplanted into about a thousand young mice, and although six hundred and eighty- six of them lived more than four weeks, growth did not take place in a single instance. According to EhrKch's hypothe- sis, the body cells of the spontaneously affected mice in which tumors of such low avidity had been allowed to grow should be possessed of a still lower degree of avidity, and such animals should be, in consequence, more susceptible to the transplantation of other tumors. But this was not the case. Grafts were exchanged between the two animals in which these cancroids arose, but no growth was obtained in either ; moreover, the two mice were refractory even to a tumor of higher avidity. These experiments, and others of a similar nature, indicated that the con- ditions for tumor growth were much more complex and specific than could be explained by supposing a difference of avidities. "The condi- tions for which two histologically indistinguishable tumours are adapted 1 Fourth Sci. Report, Imperial Cancer Research Fimd, London, 191 1, 62. 2IO THE SPOXTAXEOUS TUMOR are so different that they can be only very rarely exchanged without the mahgnant mode of growth ceasing altogether."' Bashford ^ represented Ehrhch's hypothesis diagrammatically, and discussed the question as follows : — ''The side-chain theor}^ as appHed to cancer, if stripped of its special technical terminology, amounts simply to assuming that the cells of some malignant new growths take up food-stufls more rapidly and more energetically than do normal tissues. By \-irtue of this property such tumours can be transplanted to fresh hosts ; once estabhshed and grow- ing in these, such tumours can prevent a secondary inoculation or metas- tases from growing by virtue of the rate with which food-stuffs are ■withdrawn. A subsidiary assumption is made that this starvation of a natural or artificial metastasis may be effected by deprivation of special tood-stuft's. Great importance is attached to the assumption that it is possible to increase artificially the a\adity of the tumour-cells for food-stuffs ; in other w^ords, to increase the rapidity of cell-division. " Following the practice observ^ed in the Second Scientific Report, where all the more important theories of cancer were diagram mati- cally depicted, Ehrlich's hApothesis, formulated since then, may also be reduced to a diagram without suggesting that he would approve it as correctly reproducing all the details of his elaborate argument. "A number of lines of equal length may represent, as in the accom- panpng diagram, the normal axddity of the body. Departures from the normal may be depicted by other fines exceeding or faUing short of one passing through the tops of those representing the normal. Thus in the accompamdng diagrams (a) and (a') will represent the normal and (b) will represent the simplest departure from it, namely, an increased aridity ; but this assumption requires no further con- sideration, since, if it were true, then on the basis of Ehrfich's h}'poth- esis every tumour should be easily transplanted into normal animals. This is not the case, and Ehrlich dismisses this assumption, which was put forward by that distinguished pathologist, the late Professor E. Albrecht. Instead of assuming retention of normal a\fidity by the body, and increased a\'idity by the tumour-cells, Ehrfich assumes that as a rule the body-cells lose in avidity with increasing age (b'), (c), (d), while the tumour cefis retain the normal (b') or do not lose it in a degree 1 Fourth Sci. Report, Imperial Cancer Research Fund, London, 1911, 188. THE SPONTANEOUS TUMOR 211 equal to that lost by the body (c). In some cases it is assumed that Albrecht's view (b) may hold, or that while the body-cells lose in avid- ity, the tumour-cells may exceed the normal (d), and it is further as- increased avidity. Normal avidity, — IliiBS iMli (a) (6) Diagrammatic representation of Albrecht's hypothesis. Increased avidity. Normal avidity. • ■■■ I : ::: • • :; • •• '■"■ 'I' I I 1 1 (a') Normal {¥) Transplantable IN Low Percentage (c) Not Trans- plantable (d) Transplantable in High Percentage Fig. II. — Diagrammatic representation of Ehrlich's hypothesis as deduced by him from his transplantation experiments on mice. sumed that this exaggerated avidity, when not present at the first trans- ference to normal animals from the animal in which the growth arose, can be artificially induced by a particular experimental procedure, viz. artificial selection of the rapidly growing tumours and forced passage. 212 THE SPONTANEOUS TUMOR "This hypothesis assigns a very important etiological part to a constitutional change ensuing with advance in years. The ingenuity and the simphcity of the conception are striking. However, sufficient account has not been taken of the differences existing between spon- taneous and propagated cancer, as consistently emphasized from this laboratory, since the hypothesis was based upon the unfavourable results of removing a spontaneous tumour from its native environment and transplanting it into normal animals, without giving due considera- tion to any other possible factor than the assumption of a differential of cell-avidities. At the time when the hypothesis was formulated, the investigations conducted in this laboratory had already fully demon- strated that the hemorrhagic mammary tumours could be transplanted with ease, whereas Ehrlich had concluded that they were practically untransplantable and therefore of low avidity. The transplantation of spontaneous tumours had been shown to give better results in young than in old animals, and, therefore, the peculiar age-incidence of cancer was bound up with the inception and not with the continued growth of cancer. The fact that old animals did not yield a soil uniformly more suitable for growth than did young animals, led as a matter of course to an investigation of the quality of the soil which mice suffer- ing naturally from cancer offered for the growth of their own and other tumours, and to the investigation of the most important question of individuality. ..." A glance at the following table "... will show how it is possible to contrive experimental tests of the hypoth- esis, by ascertaining how tumours behave in animals of normal avidity, and in animals in which departures from it are assumed to have occurred because they have developed cancer. The table, embody- Mice with Spontaneous Cancer Normal Mice Spontaneous Tumour A . . " B . . A B + - + — 0/3 s + 3/35 " C . . " D . . " E . . C D E + - - + - - + o/iooo + 4/26 THE SPONTANEOUS TUMOR 213 ing some of Haaland's experiments with five mice and their spon- taneous tumours, summarizes the kind of results obtained. ... In the first experiment, tumour A, when tested in normal mice, does not grow, therefore the avidity of mouse A is below normal. Tumour B grows in normal mice, therefore it must have a greater avidity than tumour A, and a still greater avidity than the body-cells of mouse A ; but tumour B will not grow in mouse A. It follows that the hypothesis is inadequate to explain the failure of transplantation in this case. In the second experiment, two mice, C and D, have been picked out whose tumours would not grow when tested on a thousand animals of normal avidity ; it follows that tumours C and D were growing in mice C and D in spite of an avidity very much below the normal, and also that mice C and D had an avidity even lower in the scale. Nevertheless, a tumour E, which had so high an avidity that it grew easily in mice of normal avidity, was not able to grow in mice C and D, of which the avidity has been proved to be so low. This is a reductio ad ahsurdum. It follows from these, and from many similar or slightly different experi- ments, that Ehrlich's atreptic hypothesis is inadequate to explain the growth of cancer either when transplanted into a fresh host, or when growing in the animal in which it arose, and it may be inferred is equally inadequate to explain the cause and nature of cancer. Without denying that cancer-cells may' have inter se a different avidity for food-stuffs, the explanation of the development, nature, and preven- tion of cancer is not to be sought along the lines indicated by Ehrlich's atreptic hypothesis." HISTOLOGY During the first few years of the investigation of mouse tumors it was customary to speak of ''Jensen's tumor," or of "Borrel's tumor," an expedient which could from the nature of the case, however, be but temporary. As the number of cases under observation grew larger it became desirable, and even necessary, that some sort of classification be attempted. It was found possible to make one, par- tially satisfactory at least, and, furthermore, that this would follow, although somewhat roughly, those suggested for tumors in man. MichaeKs ^ was, perhaps, the first to attempt to classify the spon- ^ Med. Klin., 1905, i, 204. Zeitschrift f. Krehsforsch. , 1906, iv, 3. 214 THE SPOXTAXEOUS TUMOR taneous tumors, arranging them under three heads. The first two were the simple alveolar carcinoma, which was sometimes tubular, and the adeno-carcinoma, often signalized by the formation of cysts and papills. In those of the third t\pe, the parenchjnna was di\'ided into alveoli, as in the first group, but within each alveolus the cells were wreathed about lumina in single layers, conferring somewhat the appearance of a sweat gland on cross-section. The acini were not in contact, however, but separated by irregularly arranged alveoli. This type, unnamed by ]\Iichaehs, corresponds to the malignant adenomata. Apolant^ suggested the following comprehensive classification for the mammary tumors of the mouse : — I. Adenoma. (a) Adenoma simplex, (h) Cyst-adenoma simplex. (c) Adenoma cysticum cedematosum s. haemorrhagicnm. (d) Cyst-adenoma papilliforme. II. Carcinoma. (a) Carcinoma simplex aheolare, including cysto-carcinoma haemor- rhagicnm. (b) Carcinoma papillare. including Jissure-forming carcinoma. The poverty of the normal mammary gland as concerned a well- developed stroma, made it understandable that connective tissue neoplasms had not so far been discovered in it. The mammary tumors of the mouse were epithehal growths, the ground tA'pe of which was represented by the adenoma simplex. By the term ground type Apolant ^^dshed not only to convey the idea that the adenoma was at once the most simple of all growths and the one most closely allied to the structure of normal mamma, but further, to emphasize its genetic relation to other epithehal tumors, for in spite of a great variety in structure he had seen no growth that could not be traced back to the simple adenoma. The adenoma simplex in its adult condition was more or less lobular in microscopic structure, separated by a capsule from the normal gland and. as a rule, but poorly provided \yiih stroma. The round or oval acini corresponded in general to the normal acini of the mamma, although they lay more closely together and were not ^ Arb. a. d. Konigl. Inst.f. E.xp. Therap., 1906, Heft i, 11. THE SPONTANEOUS TUMOR 215 arranged in groups within the lobules. The cells, arranged in one layer, were usually small, approximately cubical in shape, and were furnished with a relatively large nucleus rich in chromatin. Mitoses were infrequent and never of pathological type. The acini, like those of the normal mamma, usually contained a coUoid-Hke material staining with acid dyes, and either filhng the lumina entirely or irregularly retracted from their walls. Altogether, the histological picture imitated very closely that of a parenchymatous goiter. The blood supply was usually in proportion to the amount of stroma, although some areas might be well provided with vessels while others in the same tumor were supplied but poorly. In very young tumors, below the size of a lentil, a lobular structure was not usually demon- strable, and although these were separated to some extent from the normal gland by strands of connective tissue, at other points there was such a gradual transition that it was impossible to differentiate between the newly formed acini and the mamma itself. The dis- tinction was rendered more difficult since neither the form of the alveoli nor that of its cells departed very far from the normal type. As these immature growths were distinguished from the true adenomata by the possession of ducts which were traceable into those of the nor- mal gland, they might with more justice be described as local hy- pertrophies than as tumors in the strict sense of the word. Never- theless, a sharp distinction between the two was impossible, and not infrequently both forms were present at one and the same time, the hypertrophic moiety still in intimate connection with the mamma, the adenomatous, on the other hand, surrounded by a connective tissue capsule. Occasionally a number of mitoses greater than was ever associated with true adenoma could be discovered in the cells of local hypertrophic areas, and in such cases careful examination would nearly always disclose regions where the glandular epitheHum, having suddenly taken on atypical growth, was progressing directly toward carcinoma without the intervention of an adenomatous stage. Secondary alterations, to which the adenomata always suc- cumbed after any considerable duration, might affect either the parenchyma or the stroma. As an outcome of the former there resulted a cyst-adenoma simplex, and of the latter, an adenoma cysticum cedematosum s. haemorrhagicum. 2l6 THE SPONTANEOUS TUMOR The change in the parenchyma resulting in the evolution of a cyst- adenoma simplex was a dilatation of the acini with the formation of re- tention cysts. The cubical cells lining these cavities were sometimes greatly flattened by pressure and when the process took place in adja- cent alveoli the series of occasionally intercommunicating chambers thus produced recalled vividly_ the structure of lung as seen under the low power. The adenoma cysticum oedematosum s. haemorrhagicum was marked in its early stages through the separation of certain still normal alveoli from their fellows by an edematous stroma, or later, byfinal destruction, of the isolated alveoli from increase of the edema. The edema varied according to whether stasis in the tumor involved more particularly the blood or the lymph vessels. In the former case, dilated capillaries trav- ersed the growth, often conferring the appearance of a cavernoma,, and it was not unusual for the vessel- wall to rupture, giving rise to a more or less extensive hemorrhage. Generally, however, the lymph channels were primarily or concomitantly affected, and an edema was produced which evolved the lesions just described, before stasis in the blood vessels became apparent. The fate of the adenomata was thus a varied one, and depended upon whether secondary degenerations ran their course in the paren- chyma or in the interstitial tissues. Although the lesions just dis- cussed could exist separately to a certain degree, they were sa combined in their later stages that discrimination was impossible. The essential characteristic of the cyst-adenoma papilliferum lay in this, that in contradistinction to the types previously described^ cellular proliferation, keeping pace at least with the dilatation of the cysts, or more usually outstripping it, created a papillary structure. This rather rare tumor owed its peculiar architecture to the energetic stroma development. As the cysts dilated, their form became irregu- lar from introversions of their walls, while at the same time the lining epithelium threw out tiny buds which, aided by the growth of new connective tissue, gradually developed into delicate papillae. In the presence of very vigorous cell growth these papillae could project lateral branches, although Apolant had never seen any very intricate pattern, produced in this way. The carcinoma simplex alveolare originated from adenomata THE SPONTANEOUS TUMOR 217 which had undergone neither cystic nor hemorrhagic transformation, and the mahgnant change was initiated either in circumscribed or more extended locahties. There occurred in the midst of the or- dinary adenomatous tissue one or more sharply defined areas, where the epithehal cells formed a solid nest without any trace of a lumen. Mitoses were seldom lacking even in this earliest stage, and their pres- ence afforded a further distinction between the young carcinoma and the surrounding adenoma. Serial sections demonstrated such entire isolation of areas of carcinomatous transformation from one another, that their origin could be considered multicentric. The mahgnant change seldom developed without the occurrence of distinct morpho- logical alterations, such as a gradual increase in the size of the cells. These elements assumed a characteristic vesicular appearance, while with the increment of protoplasm there was associated an enlarge- ment of the nucleus. Hyperchromatic and other atypical mitoses were abundant in this tumor type as they were in all other rapidly growing mouse carcinomata. As a rule the growth of new blood vessels was unable to keep up with the rapid prohferation of the parenchyma, whence more or less extensive areas of degeneration occurred in the central portions of the tumor even at an early stage of development. Nevertheless, there were alveolar car- cinomata which possessed an adequate blood supply, and these one might be inclined to class among the endotheliomata on account of the seemingly close relation between blood vessels and tumor cells, were it not that an endothelial origin had never yet been indubitably demonstrated for any mouse tumor. Closely connected with the alveolar variety, and a derivative of it rather than a separate type, was the cysto-carcinoma haemorrhagicum. This tumor was, in fact, an alveolar carcinoma that had developed upon a cystic or hemorrhagic adenoma, and a typical example at the height of development had the appearance of a transformed adenoma, with cystic dilatation oi the alveoli and extensive hemorrhages into the interstitial tissue. An adenomatous structure was no longer to be recognized in the outlying parts, and the cystic or hemorrhagic portions were surrounded by cells arranged in the most disorderly manner. In later stages the columns of tumor cells isolated by blood spaces gradually succumbed until finally, the proliferative power of 2l8 THE SPONTANEOUS TUMOR the cells exhausted, the whole process terminated in a cyst filled with thick, brown, bloody debris. The relatively benign nature of this variety was indicated further by a noteworthy lack of mitoses, a condition which was in sharp contrast to the great number seen in true alveolar tumors. Carcinoma papillare was a much less common growth than alveolar carcinoma. While in the latter type an acinous arrangement of the cells with lumen formation occurred but rarely, in that now un- der discussion the presence of lumina was a constant characteristic. If the lumina retained their original size the structure of the tumor corre- sponded very closely to that of malignant adenoma, but cystic enlarge- ment produced the true papillary carcinoma. So long as the growths contained no papillse they might be termed fissure-forming adeno-carci- nomata, for the lumina, instead of being round or oval, were narrow and protracted or, at other times, tortuous or triangular. The epithelial cells at first inclosed these openings in a single layer but evinced their carcinomatous character, notwithstanding this arrangement, by mitos- ing freely and furthermore, by joining with neighboring groups to form an extensive reticulum. Large solid cell nests were not often encoun- tered because of the frank tendency toward lumen formation, and the stroma was developed to a slight extent only. While this fissure- forming variety might persist in pure form, transitions between it and the true papillary type were more usual. The process was quite analogous to that which ran its course in the papillary cyst-adenomata. Coincidently with dilatation of the acini their lining epithelium ex- truded processes into the lumina, and these gradually developed into long branched papillae. The papillary adeno-carcinomata, be- sides being evolved from tumors of the fissure-forming type as just described, occasionally originated in simple adenomata through a sudden widening of the lumina followed by a vigorous ingrowth on the part of the epithelium. Although the several varieties above described were the most charac- teristic forms of mouse carcinoma, Apolant pointed out that the account as given was by no means complete, for malignant tumors of uniform structure were very rare, and the multitude of types was so enriched by manifold combinations that an accurate descrip- tion of them all was hardly possible. THE SPONTANEOUS TUMOR 219 According to Murray/ the histological variations presented by mammary carcinomata could be referred with but few exceptions and by easy gradations to an acinous ground type which, in ks turn, led directly to the structure of the normal mammary gland ; furthermore, the parenchyma of any single tumor might exhibit several of these modilications simultaneously or successively. Although purely acinous growths were extremely rare, spontaneous tumors without any trace of an adenomatous arrangement were seldom met with. Apolant had inferred, from the association of the adenomatous "v\dth other structural t}'pes, that alveolar carcinomata and adeno-carcinomata arose directly from preexisting adenomatous portions, but Murray was of the conviction that the course of transformation was usually in the opposite direction, that is, that an area originally alveolar became spht up into acini by penetrating connective tissue and capillaries. Still, the converse process, by which the acinous was transformed into the alveolar structure, was frequently seen, but its interpretation in the primary tumor was more difficult. Edematous and hemorrhagic changes in the stroma were more apt to occur in those tumors where an extremely delicate connective tissue was associated with thin-walled blood vessels. Fluid exudate accumulated between the walls of the capillaries and the adjacent stroma, so that the vessels appeared to be suspended in wide spaces filled with a light flocculent coagulum. With this condition there was generally associated a dilatation of the capillaries, pointing to a partial stasis of the circulation as the primary change. Hemorrhage into the widened lymph spaces surrounding the vessels frequently occurred, and when the acini of the parenchyma had been dilated into cysts such blood extravasations easily passed through their attenuated walls and distended the ca\'ities. The thin trabeculae of parench}Tna intervening between adjacent cysts were often much compressed when the dilatation was extreme, and the vascular endothehum might be destroyed for long . distances — a condition which, particularly in old growths, produced the appearances that had led v. Hanse- mann and others to place many of the mouse tumors among the endotheliomata. 1 Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 69. 220 THE SPOXTAXEOUS TUMOR Apolant's classification was fully indorsed by Gierke ^ who, however, emphasized the fact that it was purely histological and that it dealt \\ith tumors biologically identical and varying only slightly in their micro- scopic appearance. As regarded the conversion of adenomatous into carcinomatous structure, Gierke was of the opinion that Apolant had been in error, and that exactly the reverse evolution took place, the acinous portions arising from the alveolar. This would correspond to a difl'erentiation (maturation) of the tissue of the tumor, and was com- parable with the formation of Pfliiger's primordial "egg tubes'" in the ovary, or with the development of the embr\"onic or strumous foUicles of the thyroid gland. As Apolant had stated. A'ariations in the rate of growth determined, to a certain degree, variations in architecture, and this observation made it easier to understand the significance of histological structure. The epithehal cells would differentiate faultily when gro^A■ing at a rapid pace and would, in consequence, assume a garb histologically more malignant than that which they possessed when growth was slower. And yet. such relations between structure and growth rate did not always obtain, and there was a whole series of other factors involved, including the behavior of the connective tissue and the general constitution of the animal itseh. CLINICAL COURSE A priman,' growth of the mouse pursues much the same course as does a neoplasm in the human subject. Spontaneous absorption may take place and has, in fact, been recorded by EhrHch^ as an occurrence not particularly rare. ]\Iurray.'^ on the other hand, thought that while absorption was not uncommon as a localized process affecting small areas, the disappearance of an entire tumor was a very unusual event, and although he had frequently observed temporary arrest of growth and in some cases even an actual diminution in size, the usual course was a progressive increase in the dimensions of the nodule. The rate at which growth took place was variable, the most rapid occurring ^ Beitr. zur path. Anat., etc.. I'Ziegler), 1908, xliii, 336. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 124. ^ Arb. a. d. Konigl. Inst.f. Exp. Therap., 1906. Heft i. 82. ^ Third Sci. Report, Imperial Cancer Research Fund, London, 1908, loi. THE SPONTANEOUS TUMOR 221 among the hemorrhagic carcinomata in which sudden increments (although sometimes due to an extravasation of blood) not infrequently super\'ened as the result of an almost explosive proHferation of the parenchyma. Haaland ^ recorded a few mice which had been able to rid themselves of sporadic growths. In one of these animals, although the absorp- tion of a spontaneous tumor was under way, another was growing progressively at the same time and producing large metastases in the lungs and liver. In this case, therefore, the disappearance of a sporadic neoplasm was due to local conditions affecting the tumor cells themselves rather than to the intervention of general constitutional changes. Examined under the microscope, disappearing nodules were found to contain a fair amount of healthy tumor tissue showing no signs of active proKferation and surrounded by a very sclerotic connective tissue. At the periphery, and scattered through the growth, were numbers of l arge phagoc ytes, some of which were filled with brow nish gr anule s while others possessed a va cuolated pr otoplasm Hke that of the ph agocytes seen in spontaneously regressing propagable tumors. These cells appeared not only as a reaction zone outside the alveoli, but also within them, replacing by degrees the parenchymal cells. The resemblance of the picture to that accompanying the absorption of transplantable tumors lay chiefly in the presence of phagocytes and sclerotic connective tissue, whereas the infiltration of small round cells usually encountered in propagable tumors undergoing absorption was less marked. Discussing the effect of a spontaneous growth upon the animal bearing it, Murray ^ said that the weight of the mouse generally in- creased slowly with the growth of the tumor until ulceration or hemorrhage supervened. A diminution in weight then set in, and was always a symptom of the gravest import, for seldom did an animal survive more than two weeks after its inception. The result of operative removal of sporadic growths was discussed by Murray as follows : — ■ "When a large tumour is removed by operation from a mouse, the ' Fourlh Sci. Report, Imperial Cancer Research Fund, London, 191 1, 51. ^ Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 103. 222 THE SPOXTANEOUS TUMOR animal loses weight in excess of that represented by the tumour and blood loss. This is gradually regained in the succeeding week or two weeks, and then the weight remains constant. Minor variations of a half to one gram unless progressive from week to week are of no importance. Inter-current illnesses are always accompanied by loss of weight, sometimes considerable. Should recurrence take place, the animal at first increases in weight, as already noted. Loss in weight generally indicates an approaching lethal issue, either from too great nutritive demands by the recurrent tumour, or from respiratory embarrassment due to extensive pulmonary metastases. The fre- quency with w^hich recurrence supervenes after apparently complete extirpation, is surprising when the encapsulated appearance of these tumours to the naked eye is borne in mind. Early operation, as in man, gives the best promise of lasting freedom from recurrence. Out of 48 animals operated on . . . recurrence took place in 23, two recurring three times, six twice, and the remainder once. The interval of sur\dval between the first operation and death averaged three to six weeks in the later operations, and in five cases was more than 100 days. . . . The number of recurrences and the length of survival are not strictly comparable with similar data from the human subject, since the prolongation of Hfe is of prime importance in man, but in the case of the mouse other considerations than the prolongation of Hfe have such importance that the animal may have to be sacrificed before death would have occurred." Murray's cHnical investigations were continued by Haaland ^ on a material consisting of three hundred and fifty-three primary growths, occurring in two hundred and eighty-eight mice. Roughly speaking, 90% of the tumors were adeno-carcinomata of the mamma corre- sponding very closely in their histology to the descriptions of Apolant and ^Murray. ]\Iost of the ablations were performed as early and as completely as possible, but the difficulty of total removal was increased by the wide distribution of the mammary apparatus and the relatively large size of the tumors. Among one hundred and sevent)^-four operated mice ninety-six exhibited recurrence which, in two-thirds of the cases, took place before the end of the sixth week, and in one-fourth more ^Fourth Sci. Report, Imperial Cancer Research Fund, London, igii, 49. THE SPONTANEOUS TUMOR 223 between the sixth and tenth weeks. Only isolated instances occurred later than this. RELATION BETWEEN TUMOR AND HOST The relation between a spontaneous tumor and the animal bearing it was one of the first problems to receive attention. Loeb ^ found that an adenoma of the mammary gland in a white rat would grow in the rat herself but not in other rats, and Loeb and Leopold- extended this experiment to include the transplan- tation of one of the common mixed mammary tumors of the dog, which they had discovered in a spaniel. Grafts were inoculated into three other dogs (two spaniels and a fox terrier), but did not grow, while all of the pieces implanted into the original animal remained alive, neither increasing nor decreasing in size. Bashford " succeeded in implanting two mice with their own spon- taneous tumors, even though inoculations into normal animals were in vain. In a third experiment the grafting of a growth into the animal herself was fruitless, but fragments grew in two out of one hundred and fifty-six healthy mice as well as in one of the two referred to above as having given a positive result upon inoculation with their own tumors. Bashford, Murray, arid Cramer ^ concluded tentatively in the fol- lowing year, however, that mice in which tumors had originated were not much more susceptible to other grafts than normal animals, and that the subject of a primary tumor could be inoculated only excep- tionally with its own neoplasm. This latter conclusion, as will be seen on a succeeding page, does not agree with the results of more extended investigation by Bashford and his colleagues. Borrel and Petit ^ inoculated a cancer of the horse into the animal in which the growth had arisen, and into a normal horse. Of four auto-inoculated grafts two were successful, although all transplan- tations failed in the normal animal. ^ Jour. Med. Research, 1902, N.S., iii, 46. 2 Jour. Med. Research, 1907-1908, N.S., xii, 299. ^ Briti.'ih Med. Jour., 1906, ii, 208. * Proc. Roy. Soc, Series B, 1907, Ixxix, 170. Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 323. ^ Bull, de rinst. Past., 1907, v, 1033. 2 24 THE SPONTANEOUS TUMOR Apolant ^ ingrafted eight mice each with its own spontaneous tumor and succeeded in obtaining growth in six of them. Eighty normal controls, on the other hand, afforded only three daughter tumors, all of them from the same sporadic growth. Of two inocula- tions into other spontaneously affected mice, one was fruitful and one was not, and the tumor successfully transplanted had been inoculated with a negative outcome into the animal in which it arose. This seemed to Apolant to speak for a considerable predisposition on the part of spontaneously affected animals toward the implantation of other tumors, the more so because such mice were always old, and in old ani- mals the outcome of inoculation was usually much lower than in young. Ribbert ^ inoculated several fragments of a fibroma into a dog from which it had just been removed, and at the end of four months found that each had attained the size of a walnut. Transplantation of grafts from these secondary tumors into the dog itself was again successful, but further attempts carried out in another dog were without result. The question of the relative suitability of an animal for the im- plantation of its own tumor was very fully investigated by Haaland.^ "Out of 59 epithelial tumours of varied histology, 57 have grown on transplantation of the cells into the same spontaneously affected animal. Only two cases have been negative ; of these, one should not be counted, for the mouse was ill practically the whole time it was under observation, and died four weeks after the inoculation. The other negative case occurred in a mouse surviving for 26 weeks after inoculation without exhibiting any growth at the point of inoculation. Of non-epithehal tumours there are 4, three spindle-cell sarcomata positive on re-inoculation, the fourth a melanotic tumour of peculiar structure and of very low power of growth ; no evident increase was noted of the grafts re-inoculated simultaneously in four different places into the mouse in which it had arisen, although the tumour cells remained alive for four months. "In the cases where the grafting has been carried out subcutane- ously, the period elapsing between implantation and the appearance of a palpable nodule showing increase in size is as follows : — ^ Zeitsckrift f. allg. Physiol., 1909, ix, Sammelreferat, 80. ' Centralbl. f. allg. Path., etc., 1910, xxi, 625. 3 Lancet, 1909, ii, 1588. Fourth Sci. Report, Imperial Cancer Research Fund, London, 1911, 57- THE SPONTANEOUS TUMOR 225 End of I St week 6 2d " 23 '' 4th " 8 (in two of these cases the inoculated tumour remained stationary for a long time — in one case for eleven weeks — then grew slowly) End of 5 th week 3 " 6th " I (large dose) " 7th " I ( " "• ) "On the whole the rapidity of growth of the graft corresponds to the rate of growth of the primary tumour, and the same individual characteristics are retained in the histological picture. Two different tumours of the same mouse may exhibit different powers of growth when re-transplanted into the mouse itself. Sometimes, however, it is surprising how rapidly a graft from an apparently stationary haemorrhagic tumour may grow. In the great majority of cases the graft is already developing into a new tumour 2-3 weeks after inocula- tion into the same spontaneously affected mouse. This is on the whole considerably sooner than spontaneous tumours develop when first transplanted into normal mice, and the subsequent rate of growth is much more rapid than that observed in normal animals. This result is most easily accounted for by a large number of cells surviving transplantation, and the circumstance of their being more in concord with their surroundings." The general result of Haaland's experiments was thus that auto- plastic inoculation of a spontaneously attacked mouse was nearly always fruitful, except in the case of tumors of very low growth power. Cells re-introduced into the organism to which they were native found all the necessary conditions for continued existence and growth, but ingrafted into other spontaneously affected mice or into normal ones, were in a less favorable situation, and inoculation was accordingly not so often successful. This difference between the animal in which the tumor originated and any other animal has been well expressed by Bashford, Murray, and Cramer : ^ — ''The influence of the individuahty, i.e. the sum total of changes 1 Proc. Roy. Soc, Series B, 1907, Ixxix, 184. ■ Third Sci. Report, Imperial Cancer Research Fund, London, 1908, 337. 226 THE SPONTANEOUS TUMOE. due to the past life of the organism, will be to make any mouse dif- ferent from all others, and these differences will increase the longer the animal Hves. The difficulty of obtaining success in the primary transplantation of spontaneous tumours would be accounted for by supposing that the new animals provide an environment to the cancer cells so strange, that they cannot survive the interruption to their nutrition. Their failure to grow does not necessarily imply that they would fail to proHferate in their new hosts if the conditions to which they had been accustomed could be immediately suppHed in the experi- ment. Cells which have lived and become accustomed to the body fluids of one mouse for, say, two years, may easily die or fail to adapt themselves when transferred to the bodies of new animals. The frequency, in our experience, of large metastases in animals sponta- neously affected is in marked contrast to the difficulties in obtaining growth in normal animals, and harmonises well with this view." As Haaland did not neglect to indicate, his experiments suggested not only the advantages to be gained by the surgeon in employing tissues from the same individual in all transplantations but, further- more, the peril to that patient in whom cancer cells were disseminated throughout an operation wound. On the other hand, they demon- strated the relatively insignificant danger to be apprehended from cancer cells introduced into individuals other than the one in which they had taken on malignant growth. CHAPTER VIII TUMORS OF A NATURE STILL UNDECIDED A REVIEW of the three following growths has been relegated to a separate chapter because, as the best pathological judgment of the day is still divided over the question of their character, it seemed inexpedient to combine the results following their investigation with those that have been gained through the examination of tumors which have been almost universally accepted as mahgnant. TRANSMISSIBLE LYMPHO-SARCOMA OF THE DOG The transmissible "lympho- sarcoma"' of the dog is, briefly, a tumor which occurs on the genital organs of this animal as the result of infection during coitus. Histologically it is indistinguishable from a small round cell sarcoma, while in its clinical course it reproduces many characteristics of the true tumors ; the different opinions held by various authorities concern chiefly the manner in which the growth develops after having been introduced into a new host. The first investigator to examine the nature of this tumor appears to have been Novinsky,^ who reported its inoculation and the conse- quent production of nodules identical in structure with the primary growth, which was described as a myxo-sarcoma. Metastatic deposits were not found. Wehr ^ considered the growth a carcinoma. He observed the spontaneous absorption of transmitted tumors, and in one instance the presence of metastases in the retroperitoneal and cervical lymph nodes, and the spleen. Duplay and Cazin ^'endeavored for several years to transmit 1 Med. Vestnik, St. Petersburg, 1876, xvi, 289. Cited by Sticker, Arch. f. klin. Chir., (v. Langenbeck), 1906, Ixxviii, 774. 2 Verhandl. d. deiitschen Gesellsch.f. Chir., 1889, xviii, Teil ii, 86. Arch. f. klin. Chir., (v. Langenbeck), 1889, xxxix, 226. ^ Alii dcir xi Congr. med. inlernaz., Rome, 1894, ii, Pat. gen., etc., 103. 227 2 28 TUMORS OF A NATURE STILL UNDECIDED various neoplasms from one dog to another, but sixty animals all proved refractory. Positive results were obtained, however with the growth now under discussion, which the authors described as resembhng very closely the granulomata. A possible metastasis was noted in one case. Geissler ^ recorded the presence of nodules throughout the body in a dog bearing a transmitted tumor, but whether or not they were genuine metastases it was impossible to say, because preservation had miscarried. The tumor was not the typical epithehal carcinoma found in man. In the subsequent discussion of this paper v. Hansemann expressed his con\dction that the tumor was not a carcinoma, although he would not deny either that it was mahgnant or transmissible. Smith and Washbourn - described in detail the cHnical course of the tumor as it occurred sporadically in dogs, and furnished numerous instances of its natural transmission. So far as structure was con- cerned the growths were sarcomata. In two autopsies carried out upon sporadically affected dogs no secondary deposits were found in the viscera, although metastasis had taken place in the inguinal lymph nodes. In no case where the disease was sporadic did regres- sion of the tumor occur. In a later article the authors ^ gave an ac- count of successful inoculation in thirteen out of seventeen dogs ; and while most of these were young, one old animal at least had not proved refractory to the growth. A few dogs appeared to be naturally immune and resisted several successive attempts at transference, while animals which had been able to rid themselves of their growths were invariably refractory to subsequent inoculation. A large part of the investigation of the dog tumor has been prose- cuted by Sticker,^ who has consistently held that the growth was a 1 Vcrhandl. d. dcutschen GcseUsch. f. Chir., 1895, xxiv, Teil i, 87. ^ Trans. Path. Soc. London, 1897, xlviii, 310. Jour. Path, and Bad., 1898, v, 99. ^ British Med. Jour., 1898, ii, 1807. * Zeitschrift f. Krebsforsch., 1903-1904, i, 413. Zeitschrift f. Krebsforsch., 1906, iv, 227. Arch. f. klin. Chir., (v. Langenbeck), 1906, Ixxviii, 773 Berl. klin. Woch., 1907, xliv, 486. Dent. med. Woch., 1907, xxxiii, 867. TUMORS OF A NATURE STILL UNDECIDED 229 round cell sarcoma ; for in addition to its microscopic structure, the recurrence, the infiltrative growth, the formation of metastases in the regional lymph nodes, and the fact that the cells were able to attain the general circulation were all evidences of its sarcomatous nature. The cells differed from the normal lymphocytes of the dog by their larger size, by the possession of a distinct nucleolus, and by their staining reactions. At an early stage of growth the elements of the tumor lay tightly compressed and there was almost no intercellular tissue to be found, although at a later period a delicate reticular connective tissue made its appearance. In older portions of the nodules dilated capillaries were present, and the rupture of their walls produced occasional areas of hemorrhage. Cells exposed to a temperature of — 14° C. for twenty-four hours, or heated at 50° C. for two hours, were still capable of proliferation, but the power to^row was abolished in those which had been kept at — 11° C. for twenty-five_days. The fact that no tumors resulted from the introduction of crushed cells, or of filtered or centrif uged em ulsions, excluded in all probability an extracellular parasite as an etiological factor, and it had been found impossible to demonstrate an organism by means of the usual bacteriological methods. Very few dogs were possessed of a natural immunity and implan- tation could be performed in practically any part of the body, while in two out of three foxes tested, transmission was accomplished. Grafts in dogs produced no inflammatory reaction in their neighbor- hood, and the new tumor was not formed through participation of the surrounding connective tissue. Metastasis took place by way of the blood or lymph streams, but not very frequently in either case. Examination of the blood of tum or-bearing do gs showed an i ncreas e in the number of poly morphonuclear leucocy tes and a decrease in that of the lymphocytes and eosin ophile leucoc ytes, while in the presence of dead, tumor material large mon onuc lear lymphocytes (macrophages) could be demonstrated. Spontaneous healing occurred in about 16 % of all dogs with transmitted tumors, and in some cases the rate of diminution was very striking — - as in one where a tumor one hundred and seven days old, and as large as a hen's egg, van- ished within fourteen days. 230 TUMORS OF A NATUIIE STILL UNDECIDED Dogs in which tumors were regressing or had entirely disappeared were always immune to re-inoculation, and the blood of these resist- ant animals had in two cases brought about a cessation of growth and partial regression in two subcutaneous trnnors, although in common with the serum of normal dogs it had no effect upon tumor cehs in vitro. Animals bearing progressively growing tumors were re-inoculable, but not during the period immediately following the establishment of the first tumor. The tissues surrounding a growth constituted a tumor zone, and the rest of the body an immune zone within which the development of a second graft was inhibited. This condition existed, however, for a certain period of time only, and as the tumor increased in size it extended its sphere of influence at the expense of the immune zone, which grew progressively smaller until finally the organism produced no more immunizing material. A second phase then commenced, during which there supervened the widespread occurrence of metastases. These periods Sticker called the pre- metastatic and the metastatic stages, respectively. White ^ thought that although these growths had been frequently referred to as sarcomata, they differed from that type of neoplasm in the human subject. In the first place, they were highly contagious, while secondly, they never infiltrated the surrounding tissues, gave rise but rarely to metastases, and were very slow in their growth. They had, furthermore, certain analogies with the infectious diseases, such as the high degree of contagiousness and the long incubation period. Still, the author did not beheve that the contagio usnes s of these^rowths could safe ly be use d in support of the argum ent for the paj;asiticjiature of malignantjlisease, since the infective agent niight be the tumor^ell^itself . The fact that the tumors could not be trans- ferred to any animal except the dog pointed to the conclusion that transmission was an example of implantation rather than of infection. White had been informed that the disease was occasionally inherited. Bashford, Murray, and Cramer - believed that in spite of the his- tological similarity between this tumor and the sarcomata there '^British Med. Jour., 1902, ii, 176. 2 Sci. Reports, Imperial Cancer Research Fund, London, 1905, No. 2, Part ii, 2)Z- Berl. klin. Wach., 1905, xlii, 1434. '0h 01 |0 5^ .,.>. ^ rUA* ^S^^; (D -?v^a 4&*i>^ Infective venereal tumor of dog's vagiaa. Pri.nary grovvcli. Transformation of connective tissue corpuscles into tumor cells. X -\-. © 3 ^^ ' Infective venereal tumor of dog. Graft removed 48 hours after transplantation. Degeneration of introduced tissue and proliferation of new capillaries. X ^r^. r ^ '^^M^- a^^i ... (Q ^.'■- : '^. 1.^" ,.^ ^^ - :i i^^ <^'>^ "V,. ' . - *^ y . _ : - - P , :i^^ & W "'■-:'^- - -.■.*■ ■^■■: ' ;^' Infective venereal tumor of dog. Graft removed 4 days after transplantation. Transformation of connectiv^e tissue corpuscles into tumor cells. Mitotic figures in many of the latter. TUMORS OF A NATURE STILL UNDECIDED 23 1 were several facts which mihtated against its inclusion under that class of growth. In the examination of a case of natural infection it was found that where the tumor adjoined the areolar tissue of the host there was apparent, unless growth had been so rapid as to pro- duce marked pressure, a gradual alteration of the surrounding connec- tive tissue corpuscles. While these cells were normally elongated and flattened structures with densely staining nuclei, a gradual transition occurred between this type and that characteristic for the parenchyma of the fully developed tumor. In the course of this transformation the cytoplasm of the connective tissue cells became more voluminous and the nuclei of these elements increased in size, but stained more faintly. Although additions to the tumor mass oc- curred by this conversion, the increase in size of the growth was mainly due to the active division of cells which already possessed the distinc- tive characteristics of tumor elements; hence in its later stages the tumor grew almost entirely from its own resources, closely imitating in this respect a true malignant new growth. When grafts were examined after removal at varying intervals following implantation, it could be seen that the establishment of a new tumor was accomplished by a transformation of the connective tissue cells at the inoculation site into elements indistinguishable from those of the primary growth. The introduced tumor cells, however, degenerated very rapidly, only a few retaining their vitality for three or four days. Another reason against including this tumor among the sarcomata was the observation that it never appeared naturally in animals before sexual congress, and was uncommon in old age. It showed, therefore, an age incidence different from that of sarcoma, the curve for which was similar to that of carcinoma. Bashford and his collaborators were inclined to describe the tumor as a connective tissue reaction educed by a Hving virus still undis- covered, notwithstanding the fact that the growth resembled the neo- plasms in its histological features, local mode of origin, partial in- crease from its own resources, and a Kmit of transmission within one species. The transfer of this tumor to dogs the authors had found nearly always successful, no matter what breed was inoculated. A small 232 TUMORS OF A NATURE STILL UNDECIDED proportion of the transmitted growths regressed and ultimately dis- appeared, and in cases where this had occurred, as well as in dogs bearing large tumors, re-inoculation had succeeded. Beebe and Ewing ^ discussed several diseases of the dog which might readily be confounded with lympho-sarcoma, and suggested that more than one malady had been included in the investigations of those who had worked with dog tumors, and that certain conflicting obser- vations might find herein an explanation. They had been unable exactly to classify the growth While it most nearly resembled a large cell lympho-sarcoma, the elements were polygonal more often than round ; moreover, the fine reticulum of a lympho-sarcoma was missing, the growth was much less malignant than large cell lympho- sarcoma in man, the arrangement of the cells in alveoh without lumina was sometimes distinct, the protoplasm was nearly homogeneous, and a mucinous discharge from the cut surface of the tumor was char- acteristic. The general histological features suggested the diagnosis of alveolar sarcoma or endothelioma. The crucial point which these authors attempted to settle was the mode_oLorigin of the transmitted tumor — whether the nodules re- sulted from the proliferation j3fjnx)c^^ cells or were produced by sjtimulatim of^^ tissue elements. For the purpose of deciding this question, grafts were removed at intervals of from one to twenty-one days. After two days peripheral islands of tumor cells, chiefly normal, were found in contact with small blood vessels which seemed to belong to the original tumor, since they con- tained healthy red blood cells, as though circulation had been restored in them. Most of these islands were sharply separated from the host's tissues by fibrin or leucocytes. After three days the graft had become well fused with the surrounding subcutaneous tissue, and the peripheral islands of intact cells were usually incorporated into a definite layer surrounding the necrotic center. In this zone many blood vessels were to be found, some of which must have belonged to the graft, as they were too large to have developed in three days. The outer margin of the layer was sharply separated by strands of fibrin from the neighbor- ing tissues, and no trace could be found of the transformation of fibro- blasts, or any other elements of the host, into tumor cells. Many of ^ Jour. Med. Research, 1906, N.S., x, 209. TUMORS OF A NATURE STILL UNDECIDED 233 the parenchymal cells of the graft were undergoing mitosis at this stage. In six-day specimens areas were occasionally encountered in which a minghng of fibroblasts and tumor cells had taken place, but no evidence could be discovered of the conversion of fibroblasts into malig- nant cells. Such areas were, however, rare, and all but small portions of the circumference of the graft Was from the very first separated from the host's tissues by fibrin, fat, or connective tissue. In by far the greater part of the growing mass the authors found it impossible to conceive how the malignant cells could have been derived from the elements of the host. They were forced to the conclusion, therefore, that the infectious lympho- sarcoma of dogs was a true malignant neoplasm, the cells of which showed, perhaps, a greater capacity for independent existence and infectivity than was seen in any other known tumor process. The same authors,^ in a discussion of the biology of the tumor cell, gave a preliminary account of a series of experiments designed to test the growth power of the cells of this tumor in various media. In physiological saline solution death supervened after forty-eight hours at room temperature. The serum of a rabbit immunized against pure nucleo-proteid prepared from the growth caused pronounced agglutina- tion, but preserved the cells considerably longer than did salt solution. In sterile defibrinated blood frorh a dog with growing tumors the cells remained aHve for as long as seventy-two hours, and at the end of this period they were apparently still capable of growth. At the end of forty-eight hours they presented figures highly suggestive of karyo- kinesis. When the medium was changed every twelve hours, viability was preserved for as long as ninety-six hours, although no evidence of cell division was discernible after forty-eight. In all the experiments, the blood from dogs with growing tumors seemed to be a more favorable medium than that from animals in which the tumor had regressed. Because of the location of the primary growths, and the trans- missibiUty of the dog tumor, Apolant ^ considered it a granuloma rather than a true malignant growth, and suggested as a possible etiological factor a virus analogous to that of lues. 1 British Med. Jour., 1906, ii, 1559. - Handbuch d. path. Mikroorganismen, Kolle u. Wassermann, Jena, 1907, erste Erganz- ungsbaud, Heft 2, 446. 234 TUMORS OF A NATURE STILL UNDECLDED Wade ^ recorded the transference of the growth to two foxes and a number of dogs. In normal animals of the latter species transmission was invariably successful, but those which had been able to rid them- selves of a tumor were no longer susceptible. Nodules excised at various intervals after introduction showed that the tumor was formed both from the cells inserted and the ele- ments of the surrounding connective tissue, the latter, under the influence of the tumor cells, remaining immature and acquiring the characteristics of tumor cells instead of developing into finished fibroblasts. Examination of a receding tumor demonstrated a smaller number- of mitoses than that generally seen in this growth, a peripheral lami- nated border of fibrous tissue, and several hemorrhages. A nodule removed from a dog in which a number of tumors had just disap- peared, was surrounded by a thick, laminated border of dense, fibrous tissue containing many polyblasts and damaged tumor elements, while in the central part there was found an enormous number of polyblasts, and an occasional degenerating parenchymal cell. The lymphocyte and its derivatives having been found to play such an important part in the hfe cycle of the tumor, examination of the blood was undertaken, and daily estimations made in eight animals during a period of two months showed that establishment and growth were accompanied by a steady increase in the percentage of lympho- cytes in the circulating blood. The kidneys of twenty- three animals were examined. Sixteen were dogs which had been inoculated with portions of the tumor, five were animals into which a filtrate had been injected, one was a dog that had recovered from a sporadic growth, and one a fox in which a tumor had developed as the result of implantation. In all these animals, with but three exceptions, evidence of inflammatory change was present, the apparent age of the lesion corresponding closely with the length of the period during which the virus had been exerting its action. The kidneys of six apparently healthy dogs showed, on the contrary, no sign of inflammation. Discussing the classification of the dog tumor, Wade expressed a preference for the name infective sarcoma, in the belief that the 1 J OUT. Path, and Bad., 1908, xii, 384. TUMORS OF A NATURE STILL UNDECIDED 235 growth lay in the borderland between the infective granulomata and the true neoplasms. V. Dungern ^ attempted to elucidate the nature of the tumor by a method employed in conjunction with Coca^ in earlier investigations of a hare sarcoma. This tumor was able to proHferate in rabbits also, where it produced antibodies against hare blood, a proof that the tissue growing in the rabbit had been derived from hare cells. A similar procedure appHed to the dog tumor growing in the fox showed that in all probability the nodule was composed of elements produced by the latter host, rather than of cells from the dog. The growth was accordingly not a true blastoma, but a granuloma, and presumably the product of a micro-organism. However, this conception could not be used to support directly a theory of the parasitic etiology of neoplasms, for in spite of a similarity in histological structure the type of growth exhibited by the dog tumor was enough to distinguish it from the true blastomata. Still, it was evident that there could exist an infective agent able specifically to excite the growth of certain cells and, therefore, that the distinction between the genuine blastomata and the granulomata need not be necessarily absolute. As for therapeutic investigations upon the dog tumor, Sticker ^ brought about partial regression through the inoculation of immune serum, and cure by the repeated intravenous inoculation of hve tumor cells. Salvarsan also exerted a definite curative action, but the effect of atoxyl was only temporary. Bergell and Sticker '^ announced the disappearance of a tumor after treatment with liver ferments. Beebe •' and Crile and Beebe ^ were able to cure tumor-bearing dogs by bleeding them as completely as possible and transfusing them with the blood of immune dogs, while Beebe and Tracy "^ found that certain bacterial toxins exerted a destructive action upon the tumors. ^ Munch, med. Woch., 191 2, lix, 238. ^ Zeitschrift f. Immiinitdtsforsch., etc., Orig., 1909, ii, 395. ^ Zeitschriftf. Krebsforsch., 1906, iv, 269, 310. Berl. klin. Woch., 1908, xlv, 1391. Centralbl. f. dig. Bakt., etc., Erste Abt., Orig., 1911, lix, 464. * Deut. med. Woch., 1907, xxxiii, 1521. ^ Jour. American Med. Assoc, 1907, xlix, 1492. Jour. Med. Research, 1910, N.S., xvii, 389. ^ Proc. Soc. Exp. Biol, and Med., 1906-1907, iv, 118. Jour. Med. Research, 1908, N.S., xiii, 385. '^ Jour. American Med. Assoc, 1907, xlix, 1493. 236 TUMORS OF A NATURE STILL UNDECIDED Gaylord ^ reported the successful treatment of the growths by the in- jection into them of uncoagulated blood from immune dogs, suggesting that the apparent importance of preventing the blood from clotting indicated that leucocytes might play an important role in preser\'ing the immune characteristics of the injected serum. THYROID ADENO-CARCINOMA OF THE TROUT Bonnet^ was among the first to describe this disease, not, however, as a mahgnant growth. Among the trout in a hatchery at Torbole on the Gardasee he observed an epidemic which in four months and a half • destroyed not less than three thousand fish. In the trout affected, tumors appeared on the floor of the mouth and the gills, grew with great rapidity, and determined the death of the fish. In two cases Bonnet found soft, bluish red, smooth nodules on both sides of the tongue at the first and second gill-arches, and similar, although smaller, confluent tumors ventrally situated on the gill-arches, while in three others the lesions were hmited to the gills, and in one case were unilateral. Micro- scopic examination, which oft"ered "nothing characteristic," showed an epithehal, often tubular structure, and demonstrated the presence of bodies resembhng lymphoid cells, which Bonnet took for gregarines. Purvis,^ in 1888-1889, discovered columnar cell carcinomata affect- ing the pharynx in two trout from the Solway fisheries in Scotland. According to Murray "* the last epidemic of the disease in the British Isles occurred in 1888. The disorder was next reported from New Zealand by Scott, Ayson, Gilruth, and Wilkie.'^ Scott discussed a form of tumor prevalent among the American brook trout in the New Zealand hatchery, which, originating in the middle line of the ventral wall of the pharynx, ex- tensively involved the gill-arches. Upon microscopic examination Scott found that there took place at first a prohferation of the cells in 1 Jour. American Med. Assoc., 1909, lii, 411. 2 Bayerische Fischereizcitung, 1883, Xo. 6, 79. (As this publication was inaccessible in this country, Bonnet's account has been taken from Pick's excellent monograph.) ^ Cited by Bashford and ^lurray, Sci. Reports, Cancer Research Fund, London, 1904, No. I, 9. * Third Sci. Report, Imperial Cancer Research Ftind, London, 1908, 52. 5 Reports of the New Zealand Department of Agriculture, Division of Veterinary Science, 1891 and 1901-1902 ; cited by Pick. TUMORS OF A NATURE STILL UNDECIDED 237 the acini (the organ was not mentioned), and secondarily a destruction of boundaries which permitted an outgrowth of cells, and that these elements, infiltrating the stroma, finally produced a true carcinoma. Ayson encountered three cases of this "gill disease" in 1890 among American brook trout, and later among rainbow and other trout. The disorder occurred invariably in certain hatcheries and always among certain kinds of salmonoid fish. Gilruth received from Ayson three rainbow trout, each with a reddish soft tumor the size of a large walnut and involving the first and second gill-arches. The smaller alveoli were clothed with cylindrical or cubical epithelium, while the larger were lined with a single layer of columnar epithelium and con- tained cubical, more or less degenerated elements in the center. The capsule and the stroma carried wide, thin-walled blood vessels, and hemorrhages were frequently encountered. Marianne Plehn ^ described the disease in the trout and the char, several affected specimens of which had been discovered in hatcheries and sent into the laboratory during the preceding few years. The tumors were in the neighborhood of the lower jaw, involving both mouth and throat, and were firm while still of small size, but became softer as the process advanced. In fish bearing tumors of consider- able size the jaws were not infrequently forced apart, a condition which led to grave interference with both respiration and the intake of nourishment. These growths, which microscopic examination showed to be adeno-carcinomata, arose in the thyroid and presented at first a structure simulating the normal gland, the cancerous nature of the affection remaining masked until a later period when it displayed itself by invasion and destruction of bone and muscle. Hofer ^ described the disorder under the name of malignant goiter, or adeno-carcinoma of the thyroid, and reported instances among lake trout living in a state of nature in the Mondsee. Of the causation of this rare disease nothing was known, and no parasite had been dis- covered which could be brought into etiological relationship with it. Jaboulay ^ studied six trout with thyroid adeno-carcinomata invading ^ Allg. Fischerei-Zeitung, 1902, xxvii, 117. ^ Handbuch der Fischkrankheiten, Munich, 1904, 191. ^ Lyon med., 1908, ex, 335. Province med., 1908, xix, 186. 238 TUMORS OF A XATOIE STILL L"XDECLDED all the tissues in the neighborhood of the gland and. in an advanced stage, invohdng even distant organs. In the opinion of this author the disease, which was both hereditary and contagious, was the out- come of infection ^ith mA-xosporidia. Pick and Poll ^ described a thyroid tumor of the trout, glandular in structure throughout its early stages, but possessing the character- istics of a medullary carcinoma when more advanced. Pick himself ^ published a much fuller description of the disease about two years later. That it was not of equal distribution was shown by the fact that, in spite of reported epidemics in other parts of the world, the disorder was totally unknown in Xorth and ^Middle Germany, while in the southern part of the country only sporadic cases had been en- countered. The tumors were indubitably carcinoma ta with the power of infiltrating and de5tro}-ing neighboring tissues, although they varied in structure from parench}'matous goiter to medullary or scirrhous carcinoma. Histological appearance was. however, no guide to chnical behaAior. and those growths which looked most benign were not infrequently found to infiltrate most seriousl}'. Pick had never seen any metastatic deposits, if one doubtful instance were excepted, although he confessed that the search had not been by any means so careful as that which had been prosecuted upon neo- plasms of the mouse. The tumors originated in the thyroid gland, as was clea'.iy shown by their location and. above all, by their architec- ture, and no variety of salmonoid fish hA-ing in hatcheries was exempt, although the growths were often curiously limited to one species in a hatchery, as when, for example, where Salmo iridens and S. fontinalis were both kept, only the former was affected. On the other hand, in a neighboring hatchery neither might be spared. Again, certain ponds would be found to contain affected fish while in others the trout remained free from the disorder. The disease attacked chiefly fish over two years old. and from 2% to 7% of those exposed fell vic- tims to it.^ Pick did not see any necessity for implicating parasites in the etiology of the disease or for assuming that the carcinoma, as ^ Berl.klin. Woch., 1903, xl, 547. ^ Berl. Mill. Woch., 1905, xlii, 1435, 1477, 149S, 1532. ' According toPlehn {Travaux dela deuxieme Conference internal, pour F Etude du Cancer, Paris, 1911, 227) 70% of the fish in a pond ma} be affected. TUMORS OF A XATUEE STILL UNDECIDED 239 such, was endemic. Reasoning from what was known of human cases it was very probable that in the trout the primary lesion was a simple endemic goiter evolved, perhaps, by some condition of the water, and that the mahgnant prohferation was secondary to this lesion. V. Hansemann ^ expressed the opinion that, in spite of the absence of metastases, there could be but Httle doubt of the carcinomatous nature of the tumor which Pick had described. Plehn - was unable to accept the water of the hatchery as an etio- logical factor, for only one of two kinds of trout in the same pond might be affected, while, furthermore, the disease might disappear after ha\Tng involved a hatchery for seA-eral years, and this ^^•ithout any change ha\"ing been made in the water. To explain the causation of the disease there remained, in addition to a parasitic h\pothesis, the assump- tion that all affected fish were the descendants of a few groups of parents, and that hereditary influence was a powerful predisposing factor. Gaylord ^ pubhshed a prehminary description of an epidemic among two-year-old brook trout and bro\\m trout. At the hatchery in ques- tion, water was supphed from a spring issuing from a hillside and emptying into a pond, to be piped from there to a small reservoir and finally through a series of tanks. Carcinoma of the thyroid had been discovered among the fish in this pond two years pre\'iously. and a year later the pond was emptied and re-stocked \sath young fish. One of the tanks fed from the water passing through this pond, and hold- ing nearly four thousand two-year-old brook trout raised from eggs procured at a hatchery where the chsease was not known to exist, con- tained seven hundred fish in various stages of the disorder. In an adjoining tank, which had no connection whatsoever \\dth the one in question, there were two hundred bro^^^l trout reared from eggs hatched on the premises ; of these, from 3% to 4% were affected. The infected fish had at no time come into direct contact \\-ith those in the upper pond, where it was knowm that the disease had existed, neither had the, brook trout and the bro\\Ti trout been at any time in contact \\dth each other. Gaylord believed that the condi- tions discovered in the hatchery under discussion pointed very strongly to the infectious nature of this form of cancer, and to the suggestion ^ Berl. Mill. Woch., 1905, xlii, 1542. - Zeitschrift f. Krebsforsch., 1906, iv, 560. ^ Jour. American Med. Assoc, 1909, Hi, 411. 240 TUMORS OF A NATURE STILL UNDECIDED that the contagion was water borne. It was possible that in this in- stance the feeding of liver was in some way connected with the out- break, for in another hatchery, where this material had been replaced by chopped sea fish, the disease had disappeared entirely, although formerly it had been endemic. Continuing his report, Gaylord ^ described a fish with carcinoma of the th}Toid and a similar growth on the lower jaw, a distribution which made it plain either that the tumor could metastasize or that it was transplantable. Analysis showed that trout occupying ponds which received water from those containing cases of the disease might be- come affected, although a great many, and more particularly hybrids, were immune. The disease attacked fish both large and small and, especially in the former, spread rapidly, eroding the bone, destroying the cartilage, and infiltrating the muscle. The tumors showed varying characteristics, frequently retaining the alveolar type with colloid or, again, shomng a structure strictly adenomatous; but in all cases there were areas which presented the features of soHd carcinoma. In the discussion of this paper Stockard ^ pointed out that many of the infiltrative phenomena in the aft'ected thyroid might be due to the fact that the gland was not encapsulated, and that smaU foUicles often appeared among the muscle fibres and loose tissues of the branchial region. On the same occasion Gudernatsch offered an account of the normal thyroid gland of the Teleostei. This was not a compact, uniform organ as it was in mammals, but was broken up into numerous single folhcles, the distribution varying not only with the species but with the individual. The folhcles were generally most densely packed around the ventral aorta and its branches to the gills, while toward the periphery the arrangement became less close until the foUicles lay completely separated. Their distribution extended as far as the neigh- boring structures would allow and they even invaded other tissues, reaching out laterally along the gill arteries and sometimes even pene- trating the gills. A more detailed description of this gland in the fish may be found in two papers by Gudernatsch ^ and in the article by 1 Jour. American Med. Assoc, 1910, liv, 227. '^ Jour. American Med. Assoc, 1910, liv, 227. ^ Jour. Morphol., 1910, xxi, 709. Johns Hopkins Hasp. Bull., 1911, xxii, 152. TUMORS OF A NATURE STILL UNDECIDED 24 1 Maurer/ who was the first fully to describe the diffuse arrangement of the thyroid gland in the fish, although Baber ^ had previously men- tioned a few detached vesicles in the thyroid of the skate. According to Ewing,^ while the diffuse condition of the gland did not affect the conclusion that genuine thyroid tumors did arise in some fish, it indicated that the local extensions of these growths were not always to be regarded as signs of metastasis and malignancy. He ^ considered it highly improbable that the disease was parasitic in its origin. Gaylord ^ thought that the non-encapsulation of the gland would not account for the infiltration of vessel walls, bone, and cartilage seen when the growth had reached the status of carcinoma. He had found the disease endemic in probably not less than 75% of the hatch- eries containing salmonoids throughout the United States, and epi- demic from time to time, and had observed a certain variation in the severity of the disease and in the characteristics of the lesions in differ- ent epidemics. Thus, in one hatchery, more than 50% of the tumors presented the structure of carcinoma, while in an epidemic encountered the following year in another locaHty the growths were more like simple goiters. If the disease started in the uppermost tanks, it progressed with the course of the water through the lower ones, and the percentage of infected fish increased from above downward ; but where it occurred low down it had never been found to proceed upward against the stream. About 50% of fish taken from infected pools and placed under better hygienic conditions would recover, although spontaneous cure occurred also in tanks where the disease was in progress. Fish in which tumors had disappeared did not acquire the disease a second time, and entire groups had been observed to resist attack, but whether through possession of natural immunity or through a condition of resistance conferred by previous recovery, had not yet been deter- mined. Muddy water from the most infected ponds produced goiter in dogs after five weeks and caused marked enlargement of the thy- roid in rats. ^ Mdrphol. Jahrb., 1886, xi, 129. ^ Philosoph. Trans. Roy. Soc. London, 1881, clxxii, 580. ' Jour. American Med. Assoc, 1910, liv, 228. "• Arch. Internal Med., 1908, i, 176. ^ TravaiLv de la deuxieme Conference internal, pour V Elude du Cancer, Paris, 1911, 787. 242 ' TUMORS OF A NATURE STILL UNDECIDED Marine and Lenhart ^ undertook experiments designed to discover any possible connection between the ordinary goiter (active thyroid hyperplasia) of fish and animals, and the so-called cancerous affection of the thyroid. For the reasons that young fish were more affected than old, that the hyperplasia was checked by the addition of Lugol's solution to the water in the tank, that the removal of fish to an open brook effected a cure, that the chnical incidence of tumors was in close relation with the supply of water, and finally, because the severity of the disease as determined by histological examination was also directly related to the water supply, these authors could not accept the pre- vailing opinion that the disease was true cancer. On the contrary, they believed it to be an extreme illustration of endemic goiter, the end stage of which was cretinism. All phases of the lesion reacted to iodine, the mild degrees more rapidly than the severer, and the dis- order in its earlier stages underwent involution after from two to three weeks of treatment, although at later periods one or two months might be necessary. The iodine reaction was a specific test for functional hyperplasia, since it did not occur in the case of true tumors. As the authors had been unable to find any point in the course of the disease under consideration at which it did not yield to iodine, they concluded that there was no stage which could be looked on biologically as cancer. Although they had been unable to transmit the disease by grafts from one fish to another, they did not consider that their technic had reached a degree of perfection sufficient to permit the conclusion that the affected tissue was not transplantable. They had never observed secondary growths, with the exception of two doubtful tumors in the gills and in the lower lip, but it was questionable whether nodules in these locations might not be due to hyperplasia of existing anlagen. There was no evidence that the disorder was either infectious or con- tagious, but much in favor of the view that it was the symptomatic manifestation of a metabolic and nutritional disturbance. There were three major conditions which, in some way still obscure^ ^ Jour. Exp. Med., 1910, xii, 311. Johns Hopkins Hasp. Bull., 1910, xxi, 95. Bull. No. 7, Dept. Fisheries, Pennsylvania, Harrisburg, 1910. Jour. Exp. Med., 191 1, xxii, 455. Bull. No. 8, Dept. Fisheries, Pennsylvania, Harrisburg, 1911. TUMORS OF A NATURE STILL UNDECIDED 243 influenced the thyroid growth : Limited water supply, overcrowding, and overfeeding with a highly artificial and incomplete food. The water of the hatchery was not intrinsically goiter-producing, because fish would not develop the disease unless at least the factor of over- feeding with an incomplete food were in operation, and because they recovered if the overfeeding and overcrowding were corrected, even though remaining in the same pond. Therefore it seemed probable that food was the major factor in bringing about some fault of nutrition favorable to goiter development, although it was impossible to suggest what elements in the diet were impHcated. That the thyroid was an extremely labile organ, reacting quickly to relatively sHght variations in the body metaboHsm and even showing a slight daily histological change, has been frequently emphasized by these authors ; ^ and while they found that thyroid changes did not take place in fish so rapidly as in mammals, the range of histological variation was nevertheless equally wide. Gaylord - was able to substantiate the action of iodine upon the thyroid as described by Marine and Lenhart, and found that affected fish kept in running water to which potassium iodide was continu- ously added in a concentration of one to five milHon, showed distinct histological changes after a few days. They were similar to those described by the authors just cited, and consisted in e\^dences of ces- sation of active proHferation on the part of the glandular disease, flat- tening of the epithelium, and the presence of colloid within the alveoli. But Gaylord discovered that bichloride of mercury in the same con- centration would produce all these alterations more effectively and in less time, and described a tumor of seven millimeters diameter which disappeared completely in the course of forty days' exposure to mer- cury. Hence he concluded that the action of iodine upon the thyroid was not specific. In the observations that the disease was of epidemic occurrence, that its incidence increased with the flowing of water from pond to pond, that it was ''created" in certain ponds and that it was not a result of the original constituents of the water, that fish recovered spontaneously and were then immune, that the disorder could be markedly influenced and perhaps cured by iodine and mercury, both ^ Johns Hopkins Hasp. Bull., 1911, xxii, 217. ^ Travanx de la deuxieme Conference internal, pour V Etude du Cancer, Paris, 191 1, 789. 244 TUMORS OF A NATURE STILL UNDECIDED well-known antiseptics, Gaylord saw evidence to support the conten- tion that the disease was infectious. TRANSMISSIBLE SARCOMA OF THE FOWL Fujinami and Inamoto ^ described in 1 910 a spontaneous connective tissue neoplasm of the hen, probably a myxoma or a myxo-sarcoma, which was so readily transplantable that inoculation had been almost invariably successful. Spontaneous absorption was relatively com- mon and seemed to depend upon some special condition of the host._ In the case of transplanted tumors metastases were discovered in the skin, eyeHd, lung, heart, Hver, kidney, and intestinal wall. At the time of writing, the neoplasm had reached its twentieth generation without any essential change in its histology having occurred, and there was no evidence to support the hypothesis of parasitic etiology. Growth power was abolished by boiling, but not by a two hours' ex- posure to 50° C. or -15° to -20° C, or by grinding the cells in a mortar with sand. The action of antiseptics for from one to eighty minutes was fatal. Another transmissible tumor has been described since then by Rous - in a young barred Plymouth Rock hen of Hght color and pure blood. First noticed when the bird was about fifteen months old and approximately two months before she was brought into the laboratory, the neoplasm had developed slowly until, when first seen by Rous, it appeared as a large, irregularly lobular mass projecting sharply from the right breast. This growth was removed by operation and frag- 1 Verhandl. dcr japanischen path. Gesellsch., 1911, erste Tagung, 114. "^ Jour. Exp. Med., 1910, xii, 696. Jour. American Med. Assoc, 1910, Iv, 1805. Jour. American Med. Assoc, 1911, Ivi, 198. Proc. American Philosophical Soc, 1912, li, 201. See also Rous and Murphy. Jour. American Med. Assoc, 191 1, Ivi, 741. Jour. Exp. Med., 191 2, xv, 119. Jour. Exp. Med., 1912, xv, 270. Jour. American Med. Assoc, 191 2, Iviii, 1938; and Rous, Murphy, and Tytler. Jour American Med. Assoc, 1912, Iviii, 1751. Jour. American Med. Assoc, 1912, Iviii, 1840. TUMORS OF A NATURE STILL UNDECIDED 245 ments of it were at once inoculated with a trocar into the opposite breast and the peritoneal cavity of the host, and into two other hens from the same setting of eggs. Upon section the tumor was found to have undergone widespread coagulation necrosis at the center, although it had a margin of translucent, rather friable, yellowish pink tissue with a glistening, finely striated surface. Microscopically the tumor sug- gested sarcoma. Thirty-five days after implantation, the host was dead of intraperitoneal growths, and in the breast of one of the in- oculated fowls a large nodule had developed. Microscopic examination of the original growth and of the nodules occurring elsewhere in the host following implantation, showed them to be sarcomata composed of loose bundles of spindle cells, crossing in every direction and separated from the smaller blood vessels only by endothelium, a structure which was reproduced with great fidelity in all the growths of subsequent generations. Intercellular fibrils were demonstrable with Mallory's phosphotungstic acid stain, but they were rare in the more cellular portions of the growth. The areas of necrosis seemed to be dependent in general upon insufficient vas- cularization, although hemorrhages from thin-walled blood vessels were occasionally responsible for their production. In spite of a fairly well marked capsule the growth showed a certain tendency to infiltrate, not so actively, however, as did its daughter tumors. Re- peated bacteriological examination yielded negative results, and the growth was not transmissible to pigeons, ducks, rats, mice, guinea- pigs, or rabbits. The daughter tumors throve at first only in intimately related fowls of the pure stock in which the spontaneous neoplasm had been dis- covered, but later transplantations into similar fowls of pure blood from another source were successful, and the tumor finally came to grow in about 80 to 100% of pure bred, hght-barred Plymouth Rocks, and in an occasional individual that showed by its plumage the shght admixture of some darker strain. Contrasted with this outcome was that obtained in the first four generations, where the number of suc- cessful transplantations was but three in twelve. It was never pos- sible to propagate the tumor in any breed except Plymouth Rocks, and proliferation was especially active in young birds. The re-inoculation of fowls that had proved refractory to implantation was uniformly 246 TUMORS OF A NATURE STILL UNDECIDED unsuccessful. Growth became so vigorous in the later generations that a fragment of from one to two millimeters in diameter might give rise within three weeks to a nodule having an average diameter of about eight millimeters, and retrogression of a developed tumor, fairly frequent at first, became a rare event save in cases where the host was ill. While the growths obtained from the first inoculation required seventy-one days to reach an average diameter of about four centimeters, and to affect seriously the health of the host, tumors be- longing to later generations, and the product of a similar mode of inocu- lation, attained an average diameter of about nine centimeters three weeks after the implantation of a fragment two millimeters in diameter^ and frequently determined a fatal issue in from twenty-six to thirty days. In the seventh generation, the average time required after inoculation for the development and dissemination of secondary nodules was approximately one-half of that demanded in the third. Metastasis, which occurred by way of the blood stream and but rarely through the lymphatics, was more frequently observed in the later than in earUer generations, and secondary nodules were encoun- tered less often in the heart than previously, but more commonly in the hver, while the spleen, exempt until the seventh generation, con- tained secondary deposits at that period in two cases out of twenty- five, and was thereafter frequently involved. In a study of early stages it was found that during the first forty- eight hours the tumor fragment, unchanged in appearance, occupied a rent in the host's tissues surrounded by exudation and a moderate accumulation of polymorphonuclear leucocytes. The connective tissue of the host was edematous at the torn margin, its blood vessels dilated, and toward the end of "the period under description fibroblasts and macrophages appeared. During the third day the graft often united at one or more points with the host's tissues, and although it still remained un vascularized, prohferation was in active progress, cell division taking place chiefly by amitosis, and the fragment was not infrequently much increased in size. Its loose structure facihtated nourishment by a direct interchange of fluids, but unless the graft were broken or of very small size necrosis took place at its center. The polymorphonuclear leucocytes commenced to disappear at this time and fibroblasts, macrophages, and an abundant supply of newly TUMORS OF A NATURE STILL UNDECIDED 247 formed capillaries took their place. In some fowls small round cells, histologically identical with the lymphocyte, collected in Hmited num- ber, localizing at points where the implantation had become attached to the surrounding connective tissue, but in very susceptible hosts this, like all cellular reaction, might be absent. By the fourth day or later the fragment had increased greatly in size, vascularization was accomplished, and establishment of the nodule might be considered perfect. Even before this period, however, tumor cells were often discovered invading the tissues of the host at points where these were in contact with the graft, proliferating strands of spindle cells extending into the normal structures and utilizing for their support the blood vessels and connective tissue already present. The activity of the neoplastic cells, once the tumor had become established, was startling, and as many as 46% had been found in process of segmentation at one time, of which 42% were in amitotic and 4% in mitotic division. It was not necessary for the development of tumor fragments that they be brought directly into contact with the connective tissue of the host, for grafts placed in the peritoneal cavity might establish themselves in spite of its endothelial lining, proliferating sometimes in a way to suggest cell cultures in a fluid medium and at length obtaining a blood supply from the underlying tissues. In naturally immune fowls, or in those with acquired resistance, the phenomena in action about the tumor fragment did not differ during the first forty-eight hours from those in susceptible hosts. Following this period death of the graft took place in one of two ways. (a) The fragment often remained for many days unvascularized and almost unattached to the surrounding tissues, and underwent a gradual necrosis in which the marginal cells were the last to succumb. These elements might live for a week or ten days and might even multiply at first to some extent, but they did not invade the connective tissues at those points where they were in contact with them. Meanwhile leucocytes disappeared ^ fibroblasts and macrophages were substituted, lymphocytes in small numbers infiltrated the zone of living tumor, and by the end of two weeks, as a rule, the graft was wholly necrotic and reorganization under way. (b) More frequently the implantation increased in size, invaded the tissues of the host, and became vascu- larized with the same rapidity as in susceptible fowls, but imme- 248 TUMORS OF A NATURE STILL UNDECIDED diately upon its obtaining a foothold a lymphocytic reaction supervened in the surrounding tissues, so intense in character that even by the fifth day the graft might be embedded in a mass of small round cells six to eight times its own diameter. A certain degree of infiltration of the fragment occurred during this period, and despite its blood supply the tumor tissue was already degenerated. These two sets of phenomena leading to the death of the implanted fragment might both be discovered in one host, an observation sug- gesting that their incidence was largely determined by local conditions. The death of the implantations could not be ascribed to their fail- ure to acquire a supporting and vascular stroma, as Bashford and Rus- sell had assumed in explanation of the death of grafts in hosts with acquired resistance, for in by far the greater number of unsuccessful transplantations vascularization actually occurred, although it ex- isted for a brief period only. The authors agreed with DaFano re- garding the significance of the lymphocyte for immunity. In the case of the fowl, however, the lymphocytic response was more vigorous than in mammals, possibly for the reason that the lymphocyte not infrequently constituted over 40% of the white cells in normal indi- viduals and, furthermore, because this type of cell responded to cer- tain processes more actively in the fowl than in mammals. No gen- eral lymphocytosis accompanied the local reaction surrounding the tumor. It v/as found that the sarcoma could be transmitted after an emul- sion had been passed through a Berkefeld filter (No. 5 medium) im- permeable to B. prodigiosus, or by means of the dried and powdered neoplastic tissue. Growths resulting from the inoculation of a fil- trate were much longer in making their appearance than where grafts had been deposited in the ordinary way, not a trace of tumor being palpable for from ten days to three weeks, while subsequent prolifera- tion went on more slowly than in control tumors. The growths following upon injection of a filtrate, which were first noticeable as minute nodules at some point in the need^a track, de- veloped in only a small number of inoculated fowls ; but when the causative agent was introduced in the form of dried powder suspended in Ringer's solution the sarcomata appeared as more or less diffuse masses at the site of injection, and developed in a much larger pro- TUMORS OF A NATURE STILL UNDECIDED 249 portion of the inoculated fowls. These observations led Rous and his collaborators to assume that the causative agent required for its ac- tion a cell proHferation or derangement such as that induced by the needle prick or the presence of dried tissue, and experiment proved this hypothesis tenable. A number of susceptible fowls having been injected in one breast with a large quantity of active filtrate and in the other with an equal amount of filtrate to which had been added a httle sterile washed diatomaceous earth, it was found that in the latter case a larger number of growths developed than where the hltrate had been injected alone, and also, that in the one instance the tumor regu- larly arose as a diffuse mass owing to simultaneous proHferation from many foci, whereas in the other it slowly appeared as a small discrete nodule in the needle track. The importance of injury was demonstrated again in the observa- tion that when large quantities of an active Berkefeld filtrate free of foreign particles were injected into the circulation, only four tumors resulted among seventeen chickens, while in contrast to this propor- tion, seven out of twenty developed growths after a Httle diatomaceous earth had been added to the filtrate. Apart from these figures, how- ever, even the site of those tumors arising after the injection of filtrate free of foreign particles demonstrated the importance of ceU derange- ment, for in three cases the growth had its primary seat in the func- tioning ovary, an organ where injury and proHferation are of daily occurrence, while in the fourth the growth was in a Hver where the authors felt that some focal derangement must have been present. Significant though trauma might be, its absence would not of itself suffice to explain the striking lack of infectivity displayed by the sar- coma under ordinary circumstances. Thus, for three years over twelve hundred chickens, many of them bearing the sarcoma, had been confined at one time or another in relatively close quarters. To some of them fresh sarcomatous tissue had been fed, and many others must have been contaminated with the dried tumor, in which it was known that the causative agent would remain alive for over seven months, and yet, although trauma and other types of injury had been frequent among the chickens so exposed, not one had developed the sarcoma except after direct inoculation. Furthermore, the growth was not naturally endemic among fowls. 250 TUMORS OF A NATURE STILL UNDECIDED The nature of the other factors conditioning tumor development had not yet been determined, but it was e\'ident that they were both local and general. It was still inexpHcable why, even with the agent in its active form, and with the factor of injury suppHed, many of the fowls injected with a large amount of filtrate should fail to develop a growth, and why the tumors following intravenous injection were seldom primarily multiple, despite the numerous injuries ever^^vhere caused by the infusorial earth. The conditions governing the appearance or non-appearance of growths possessed considerable interest in \'iew of the close resemblance of the tumor to some malignant mammalian neoplasms, not only in regard to growth and general behavior, but also in the obvious lack of inf ectivity under ordinary circumstances ; and that injury should have a share in determining tumor development was noteworthy in view of the importance of this factor as a contributory cause of mammalian sarcomata, including those of man. To determine whether pulmonary metastases were invariably due to the proliferation of transported cells, or whether they might not sometimes follow localization of the causative agent as such, a sus- pension of fresh tumor in Ringer's solution was injected intravenously. It was found, however, that secondary deposits arose uniformly by a sur\dval and growth of the inoculated tumor cells, and this observa- tion, no less than the one that visceral metastases occurred by far most frequently in the -ungs, led to the assumption that secondary nodules were a product of the proliferation of cells sieved out of the circulation by the pulmonary capillaries in the usual way. But to explain the frequent involvement of the lungs there might be enter- tained the alternate h}q3othesis that the extrinsic agent, as such, engendered tumors more easily in these than in other organs. That this was not true, however, was demonstrated by the outcome of in- travenous injections of a Berkefeld filtrate containing the causative agent, for in only two out of eleven fowls developing the growth after such treatment was the sarcoma primary in the lungs, despite the fact that the agent had been directly carried there. In seven of the cases diatomaceous earth had been added to the filtrate and had lodged for the most part in the pulmonary capillaries, producing injuries such as had been found to favor the action of the agent. Although it was thus TUMORS OF A NATURE STILL UNDECIDED 25 1 certain that metastases were referable in general to a development of cell emboli, there still remained the question whether secondary tumors might not occasionally be caused by the filterable agent as such, for that this was able to enter the circulation was shown in some instances by the appearance of a sarcoma after the injection into susceptible hosts of large quantities of centrifugalized plasma from fowls moribund with metastasizing growths. The general importance of injury in determining the action of the agent led the authors to at- tempt the achievement of secondary tumors through trauma, and under conditions which would show them to be independent of transported cells in their origin. Tissue derangements in tumor-bearing fowls were produced by the injection of Scharlach R or diatomaceous earth, or by means of incisions allowed to heal by secondary intention, but numerous experiments gave results almost entirely negative, the sar- coma failing to localize at the seat of injury. In only a single case was there suggested the direct action of the extrinsic agent, where, in a chicken bearing a very large sarcoma, but with no discoverable metastases in the viscera, a small secondary growth was found post- mortem at the site of injection of diatomaceous earth. The heart was not examined for a patent foramen ovale. Two other instances of localization at the point of injury were less significant, since there was demonstrated the presence of a large number of metastases in the lungs and elsewhere. A fourth case came to attention during autopsy on a fowl with a large inoculated tumor and many metastases in the lungs, the remainder of the body having been spared with the excep- tion of the oviduct, which contained an inspissated egg and exhibited a reactive thickening and increased vascularization of its walls. Whether in any or all of these tumor-bearing fowls the extrinsic agent, as such, had been active in evolving a growth at the site of an injury, the authors were unwilling to declare. The conditions governing the curious relationship between the tumor and its cause were of great interest. How did it happen that the sarcoma, although ultimately dependent upon an extrinsic agent, was dominated in its behavior by the cells composing it ? In answer to this question the author suggested several replies. The production of a neoplastic change by the causative factor might take place with such extreme slowness, as compared with the proliferation of cells 252 TUMORS OF A NATURE STILL UNDECIDED which had once become neoplastic, that growth might seem to occur entirely through the multiphcation of the cells in an initial focus. Tumors did doubtless result wholly in this latter way in many cases, because of a second pecuHarity of the agent — its dependence on a special set of conditions in order that it might initiate neoplastic change. The necessity for these conditions would go far to explain the failure of the agent to take an active part in the dissemination of the tumor throughout the body, granting that the agent was present in the cir- culation before the fowl was moribund. The possibility of the exist- ence of immune processes effective against the agent when it was sep- arated from the cells should also be kept in mind, although the ' evidences of resistance thus far recognized had been directed against the cells themselves, as in the case of mammalian tumors. Regarding the nature of the filterable agent causing the sarcoma, the most important question for decision was whether it was Uving, to which the most direct reply would be to grow and transfer it in culture ; but investigation in this direction had not been particularly successful. Differential filtration showed that in a dilute tumor ex- tract the agent would pass through Berkefeld filters which would hold back at the same filtration B. fluorescens liquefaciens, an organ- ism measuring 0.5 micron by i.o to 1.5 micra, while on the other hand, Chamberland bougies (F) were impermeable. Were it not known that filters of fine texture would hold back complex proteins these findings might indicate that the agent was organized, and per- haps even visible, although repeated attempts to demonstrate it with special stains or with the dark field had been in vain. The agent would retain its activity in dried tissue for seven months, and for at least one month in tissues placed in 50% glycerine, under- going gradual attenuation, however, in both instances. Repeated rapid freezing and thawing, which reduced the tumor to a pulp, did not greatly lessen the activity of the associated agent, although the resistance of the latter to heat was but little greater than than of the neoplasm itself. The sarcoma, when submitted to a heat of 50° C. for fifteen minutes, failed absolutely to grow in vitro where its proKf- eration was ordinarily very active, in spite of the fact that it often gave rise to tumors when inoculated into susceptible fowls, even after exposure to 53° C. Whether growth in vitro was a real index of viabil- TUMORS OF A NATURE STILL UNDECIDED 253 ity was uncertain, but the ability of the filterable agent to withstand a temperature of 50° C. for fifteen minutes was demonstrated a second time by the production of tumors with tissue which had been dried, ground, and heated after suspension in Ringer's solution. Material that had been raised to a temperature of 55° C. for fifteen minutes, however, never developed into a tumor. In sarcomatous tissue autolyzing at the temperature of the chicken's body (41° C), the agent remained active for less than forty-eight hours, while toluol and chloroform in the proportions employed to prevent bacterial growth during autolysis destroyed it in less than two hours, as did 50% alcohol or 2% phenol. Unlike the virus of poKomyeHtis the agent could not resist exposure to 0.5% phenol, and like the animal organisms, in distinction to most of the vegetable (v. Prowacek), it was rapidly destroyed by bile and by saponin in high dilutions. Although no single attribute among those determined would suffice to show the nature of the agent, its char- acters were in general those ordinarily associated with micro-organisms. It was found that the sarcoma could be transplanted into the de- veloping embryos of barred Plymouth Rock eggs or into the mem- branes, either by the deposition of tumor fragments or by injection of a cell-free filtrate. That growth under these conditions was ex- cellent was shown by the obseryation that, in tumors of the embryo, more cells on the average were in process of division at one time than in those of the adult. To regulate the amount actually implanted in any tissue was, how- ever, difficult, for that which it was intended to place in an extra- embryonic membrane largely escaped at either side, while that im- planted in the embryo itself not infrequently followed the needle to the surface. The majority of the growths occurred, in the outer membrane (fused chorion and allantois) for the obvious reason that this was necessarily pierced in reaching other structures, so that its inoculation could hardly be prevented during withdrawal of the needle. Tumors were discovered also in the fused allantoic and amniotic membranes and the extra-embryonic body cavity, according to the point at which puncture had been undertaken. Although no attempt was made to plant the growth in special regions of the embryo, since the instability of the latter under pressure of the needle rendered the procedure difficult, by indiscriminate implantations growths had 254 TUMORS OF A XATimE STILL tJNDECLDED been obtained in the chest wah, heart, h^^er, peritoneal surface of the gizzard, and the soft tissue of the thigh and of various other parts, all appearing in the track of the needle. In the adult fowl it was not always necessary for the development of tumor fragments that they be brought directly into contact with connective tissue, since pieces intro- duced into the peritoneal cavity might estabhsh themselves despite the endothehal Hning, but it was not so in the extra-embryonic membranes of the chick. Here a layer of ectoderm or endoderm but one cell thick constituted an absolute protection against surface implantation, and only where a mesodermal layer was exposed, as, for example, on the. outer side of the allantoic membrane, did such implantations occur. When incubation was far advanced at the time of inoculation, and the chick was allowed to hatch, tumors that had developed in the outer membranes were cast off \\dth them. Growths in the yolk-sac were drawn with it into the body Avhere they continued to develop. ■ Among one hundred and forty-seven inoculated eggs, one or more tumors resulted in one hundred and eight ; but of the thirty-nine nega- tive attempts, inoculation had been done at a very early period in seventeen, in nine others a Berkefeld filtrate alone had been intro- duced, and in five, dried material, leaxdng but eight that gave negative results after the most eft'ective method of transplantation, as com- pared with one hundred and eight positive results from inoculations of all sorts. Histologically the tumors difi'ered but httle from those in adult fowls ; although the structure of the sarcoma was much looser and the spindles might be so attenuated as to suggest myoma, necrosis was a rare occurrence, and infiltration seldom well marked except where the growths lay in structures which, hke the heart or liver, opposed a certain amount of resistance. Regional metastases, pre- sumably distributed by the lymphatics, were fairly common in the outer allantoid membrane, assuming the form of small nodules near the main growth and along the larger vessels supplying it. Remote metastases, however, had never been encountered, probably because of the structure of the membrane. By using the outer allantoid mem- brane as the injection site, and emplopng the growths found there after seven days, the authors had succeeded in transplanting the sar- coma to three successive sets of eggs, only to lose it on the fourth transfer by the death of all the embryos. The same event terminated TUMORS OF A NATURE STILL UNDECIDED 255 more quickly two other series. During continued propagation in the egg no change was noted in the behavior of the tumor, and inoculation back into the adult yielded the characteristic sarcoma. While the growth did not proliferate at all in pigeons or ducks, and grew but slowly in chickens of a variety other than that in which it originally occurred, with the embryos of these alien fowls the results were quite different ; thus, in four of nine pigeon eggs inoculated and developing, and in six of sixteen duck eggs, tumors were obtained. Inoculations were made into the membranes alone, and the resulting growths re- sembled those arising in similar situations in Plymouth Rock eggs, which was true also of the tumors developing in the eggs of an alien variety of chicken (Brahma). When portions were transplanted to adult pigeons or ducks, implantation was always unsuccessful. The growth occurring in the eggs of pigeons or ducks was not analogous to the brisk temporary proliferation often seen in alien adults, be- cause there, even in so short a time, an immense accumulation of lymphocytes took place about the graft, which itself showed well marked degenerative changes. In the embryo, however, neither phenomenon was present. Certain embryos failed to develop tumors, a circumstance probably due entirely to chance causes rather than to the possession of natural immunity, for extensive search failed to disclose in any instance those histological signs about and within the graft which the authors were accustomed to associate with the display of resistance in the adult. On the whole, young embryos (tenth day of incubation) seemed to offer a more favorable soil for tumor growth than adult fowls, and Rous and his collaborators could see no reason against extending this con- clusion to embryos at even earlier periods of developn.ent, since the less striking results obtained before the tenth day could be ascribed in large part to the method of inoculation and to the relative unfitness of the host to nourish the implanted tissue, because of small size and simple structure. CHAPTER IX THERAPEUTICS The discovery that cancer was not at all of infrequent occurrence among animals and that it could be transplanted from one to the other, placed within the hands of the biologist a means of investigating this disease at once so easily available and so inexhaustible as to en- courage the hope that the scourge could not continue much longer to defy those who were attempting its subjugation. Indeed, the con- quest of transplantable cancer has frequently seemed an accomplished fact, but only because it was not fully appreciated that such tumors often undergo spontaneous regression. Hope has run the higher, too, since a sharp distinction has not always been drawn between an autochthonous new growth on the one hand, and, on the other, a prop- agable neoplasm situated in alien soil. Thus many observers have failed to reahze that, even though it should prove possible to influence detrimentally the growth of a transplanted neoplasm, there would be no reason to assume that the curative agent would be equally effica- cious against the proHferation of cancer cells in the organism to which they are native. However, investigators have begun to perceive the distinction, and to appreciate that the cure of spontaneous cancer offers this therapeutic puzzle : To find a method of damaging a certain area of native cells while at the same time their neighbors within the organism are left undisturbed. Although Jensen ^ at first beheved that the serum of rabbits treated with cancer could bring about the disappearance of transplanted tumors in mice, he - wrote later that when he had credited this serum with curative power it had not been known that propagable tumors were prone to retrogress independently of any treatment whatsoever. While he now reahzed that spontaneous heahng would explain most of these 1 Centralbl. f. Bakt., etc., erste Abt., Orig., 1903, xxxiv, 30. ^ Zeitschrift f. Krehsforsch., 1908-1909, vii, 281. 256 THERAPEUTICS 257 earlier cures, it would not account for the disappearance of several very large tumors in his series, and he did not despair, therefore, of the final discovery of a specific cytotoxic serum against mouse cancer. V. Leyden and Blumenthal ^ practised an analogous method in dogs, successfully treating a cancerous animal with the serum of a rabbit that had been injected with dog cancer. Another was almost entirely cured of carcinoma by the inoculation of juice expressed from a dog carcinoma. Furthermore, alleviation of the disease in man was said to have been brought about in several instances after the injection of fluids expressed from human cancer, a result which, according to V. Leyden,^ was confirmed by later investigations. Clowes ^ treated mice bearing propagable tumors with the serum of mice in which Jensen's carcinoma had been absorbed, and found that while small growths were restrained in development or even cured, tumors of more than three or four grams in weight were not influenced in any way. Commenting upon these results, Clowes * said that the inhibitory effect could scarcely be interpreted to mean anything but that the serum contained an antibody directly antag- onistic to the development of neoplastic cells ; and since it had been shown that the serum in question possessed neither hemolytic nor cytolytic characteristics, it appeared probable that the effect was exerted directly upon some intracellular virus rather than upon the cell itself. Bashford,^ on the contrary, found that although the action of immune serum outside the organism had in some cases deprived tumor cells of their power to grow, no definite evidence of such action had been obtained in the Uving body, while neither Borrel ^ nor Bridre ^ was able to affect tumors in mice by injecting the animals with the serum of sheep ' Deut. med. Woch., 1902, xxviii, 637. 2 Zeitschrift f. Krebsforsch., 1907, v, 164. ^ Johns Hopkins Hosp. Bull., 1905, xvi, 130. Med. News, 1905, Ixxxvi, 477. British Med. Jour., 1906, ii, 1550; see also Gaylord, Clowes, and Baeslack, Med. News, igos, Ixxxvi, 91. '' British Med. Jour., 1906, ii, 1552. ^ British Med. Jour., 1905, ii, 96. Lancet, 1905, ii, 104. ^ Bull, de Vlnst. Past., 1907, v, 607. '' 4nn. de Vlnst. Past., 1907, xxi, 774. 258 THER-\PEUTICS or fowls previously treated •u'ith mouse cancer. Furthermore. Bridre could see no conclusive outcome from inoculating tumor-bearing mice with fresh macerated cancer, dried or heated tumor being even less successful. Le\\dn ^ recorded the failure of all attempts to obtain a curative serum by the injection of mouse cancer into animals of other species. On the other hand, the serum of rats in which tumors had regressed spontaneously produced a distinct retardation of the growth of tumors in other rats, although the cases were too few to be convincing. Walker.- arguing that certain phenomena proper to reproductive cells took place also in those of mahgnant growths, and that tumor cells seemed to be. hke reproductive cells, out of coordination -uith the rest of the body, attempted to cure transplanted carcinomata in mice by means of the serum of rats inoculated with mouse testis. It was found that tumors in animals so treated grew with only one- third the rapidit}- evinced by those in control mice injected vnth normal rat serum, and. furthermore, that favorable results could be obtained also when tumor-bearing mice were treated v^ith the serum of rats inoculated vvdth the same growth. The results in all cases seemed to be much less satisfactory,' where large tumors were treated than where the growths were small. Gay '^ found that when rats bearing the Flexner-Jobhng adeno- carcinoma were re-inoculated with the same growth not less than three or four weeks after the primary implantation, the second graft failed to grow, and in about 50*^"^ of the animals the original tumor was cured and metastases were prevented. Freytag "* inoculated tumor mice vdih the serum of immunized rabbits, with mouse blood, or with the blood or serum of alien species. Although regressive changes were evident within the tumors, complete disappearance had never been observed. The condition of parabiosis.'" introduced by Sauerbruch and ^ Zeiischrift f. Krebsforsch., 1907-1908, ^-i, 308. ^ Laiicet, 1908, ii, 797, 1299, 1474. Lancet, 1910, i, 990. Lancet, 1911, i, 1275. ^ Boston Med. and Surg. Jour., 1909, clxi, 211. ^ Zeitschrift f. Krebsforsch., 1910-1911, x, 157. ' Albrecht and Hecht (Ceniralbl. f. allg. Path., etc., 1909, xx, 1039) were led to the pre- THEEAPEUTICS 259 Heyde/ was utilized by Rous - for the demonstration of the presence or absence of antibodies in the circulation of rats that had been three times unsuccessfully inoculated with the Flexner-Jobhng adeno- carcinoma. Such resistant rats were united in parabiosis to rats bear- ing the same tumor, but the union had not the shghtest effect upon the course of the growths, neither retarded development nor retrogression having been observed. Uhlenhuth, Haendel, and Steffenhagen ^ prepared a rat antiserum by injecting rabbits with rat serum. Half a cubic centimeter was in- troduced into rats three times at six day intervals, beginning just after tumor inoculation, but in spite of these attempts to confer passive resistance, vigorous tumor development took place. The serum^ of a rabbit that had been injected intraperitoneally with tumor emulsion was equally ineffective, as were organ extracts and the serum of cattle. Blumenthal ^ treated rats with autolyzed transplantable sarcoma. Tumors the size of a hen's egg were made to disappear after one treat- ment, in spite of the fact that nodules larger than a pigeon's egg had never been known to recede in untreated rats. Lane-Claypon and Starhng ^ having postulated the existence of hormones,^ or substances able to stimulate normal growth, A. and H. Griinbaum '^ suggested that,^ given an excess of a hormone together Vv^ith a lesion or irritation of the tissue complementary to the hormone, liminary conclusion that parabiosis in animals otherwise normal exerted a striking inhibi- tory effect upon tiunor growth as well in unilateral as in bilateral inoculation of the para- biotic pair. Such tumors as grew in parabiotic animals were retarded — a circumstance which the authors considered analogous to immunization with blood, embr>'onal tissue, and skin. A short time after the separation of parabiotic animals with young tiunors, a distinctly rapid and vigorous growth of the tiunors set in, which might be explained by the removal of inhibitory factors. Goldmann {Beitr. zur klin. Chirurgie, [v. Bruns], 1911, Ixxii, 82), however, was imable to convince himself that parabiosis, as such, had any inhibitory effect upon tumor growth. 1 Milnch. med. Woch., 1908, Iv, 153. Zeitschrift f. exp. Path. u. Therap., 1909, vi, 33. - Proc. Soc. Exp. Biol, and Med., 1909-1910, vii, 12. Jour. Exp. Med., 1909, xi, 810. ' Arb. a. d. Kaiserl. Gesundheitsamte, 1911, xxxvi, 491. ^ Med. Klin., 1910, vi, 1982. ^ Proc. Roy. Soc., Series B., 1906, Ixxvii, 505 ; see also Starling, Lancet, 1905, ii, 579. ^ For a refutation of Starling and Lane-Claypon's results, however, see Frank and Unger, Arch. Internal Med., 1911, vii, 812. ''Jour. Path, and Bad., 1911, xv, 289. 26o THERAPEUTICS unlimited growth might take place. When immune rats were inocu- lated with rat sarcoma and parotid gland together, the result indicated that the gland was able to assist the growth of sarcoma in animals otherwise insusceptible, while, on the other hand, removal of the parotid gland from a tumor-bearing rat was followed by the occurrence within the nodule of fatty and fibrotic changes, although growth continued progressively. Other experiments with the parotid gland, or the ovary, were indecisive. The same authors ^ found it possible to effect the regression of small transplanted rat tumors by injecting antivenom horse serum, and found further ^ that by the simultaneous administration of adrenalin they could bring about the disappearance of still larger growths in about 70% of the animals inoculated. Beard ^ recorded the successful treatment of Jensen's mouse tumor with trypsin, and ascribed it to destruction of the less powerful fer- ment of the cancer, malignin, by the more active trypsin. Vidal ^ had observed the arrested evolution of tumors in four pa- tients with a temperature above 40° C. In order to be sure that this was not a quadruple coincidence Vidal daily exposed tumor-bearing mice to a temperature above that of the normal organism, and found that they Hved longer than their controls and that their tumors were the seat of degenerative changes. In a bitch the subject of spontan- eous lympho-sarcoma the "zone of Richet" was punctured, in con- sequence of which the temperature rose to 40.8° C, while the tumor diminished rapidly in size, finally to disappear. The action of X-rays upon the new growths of animals has been in- vestigated by several observers. Clunet ^ found sHght hyperplasia and accentuation of the cystic structure in transplantable and spontaneous carcinomata of mice exposed to these rays, and discovered at a later period still more serious lesions which were coincident with macro- scopic alterations. Marie and Clunet,^ in summing up the changes which they had observed both in human carcinomata and in those of 1 Lancet, 191 1, i, 879. 2 Lancet, 1912, i, 644. 3 British Med. Jour., 1906, i, 140. '^ Travaux de la deuxieme Conference internal, pour l' Etude du Cancer, Paris, 191 1, 302. 5 Recherches exp. sur les Tumeurs malignes, Paris, 1910, 250. 6 Travaux de la deuxieme Conference internal, pour V Etude du Cancer, Paris, 191 1, 160. THER.\PEUTICS 26 1 the mouse, described a latent stage signalized by the occurrence of a large number of abnormal cells and followed by one of ceU death. In the third stage groups of necrotic cells were replaced by poly- morphonuclear leucocytes and young connective tissue, while in the fourth there occurred the formation of a scar, which for a long time inclosed in its meshes cells of ''diminished \dtality." Contamin ^ found that the younger the growth the more readily could it be brought to absorption "^nth X-rays, and that irradiation before inoculation exerted a direct action upon maHgnant cells, affect- ing more their growth energy than their ability to A^ithstand trans- plantation. As for the effect of radium. Apolant - reported that manv mice bearing transplantable carcinomata had been entirely cured, while in the remainder the tumors had been very materially reduced in size. During the earher periods of healing there were to be found in the con- nective tissue a series of changes presenting all the signs of inflamma- tion, among which the presence of large iibroblasts was especially noticeable. These ceUs penetrated the tissues in all directions, and w^andering cells continued to the interior of the alveoli instead of halt- ing at the edges ; the alveoH themselves were continually under sub- division by advancing fibroblasts, and the epithehal cells within them lost their sharp contours in the process and came to resemble a syn- cytium. The changes were essentially similar to those described in man, in that there was a disappearance of epithehal elements and an overgrowth of connective tissue. In Apolant's opinion it was hardly to be doubted that radium possessed a specific action upon the malig- nant cells because these elements were entirely absorbed under its influence, while those in a tumor undergoing spontaneous regression were, on the contrary, transformed into necrotic masses which never underwent absorption. Bashiord, ^Murray, and Cramer ^ also described the spHtting up of alveoH into smaller cell groups by proHferating connective tissue, but denied that radium had any selective action upon tumor cells. On the contrary, when the connective tissue proHferation was at its ^ Bull, de r Assoc, franq. pour VEtude du Cancer, 1910, iii, 160. ^ Deut. med. Woch., 1904, .x.xx, 454, 11 26. ^.Sci. Reports, Imperial Cancer Research Fund, London, 1905, Xo. 2, Part ii, 56. 262 THERAPEUTICS height tumor cells could be found in active division and, furthermore, tumors could be successfully transplanted even after a prolonged exposure to radium. As great engorgement of the blood vessels, and even hemorrhage, were prominent features in growths that had been treated, the attempt was made by these authors to produce bleeding by other agencies in order to determine its significance in the earlier stages of the radium reaction. For this purpose mice with gromng tumors were injected with one cubic centimeter of a one to ten thousand solution of adrenaHn in the hope that the great rise in blood pressure would cause an extravasation of blood from the capillaries of the tumor, since they possessed neither vasomotor nerves nor the muscular mechanism by which these act. Mice dying within the first twelve hours after injection showed widespread hemorrhages into the lung and into the substance of the tumor. Three animals sur\dved these large doses of adrenahn, and in one of them the growth, which by the fifth day had diminished in size, presented the same histological ap- pearance as resulted from the action of radium. Reicher ^ reported that he had succeeded in curing transplantable sarcomata and carcinomata in rats and mice through injection of adrenahn into the tissues surrounding the growths. Microscopic ex- amination showed a central necrosis surrounded by a secondary in- flammatory reaction. Spiess,- in continuation of his previous investigations upon the effect of anesthesia on inflammation, injected anesthetics into the sarcomata or carcinomata of mice, employing for the purpose nirvanin, novo- caine, and a third substance ("337 ") not upon the market. In a pre- liminary experiment the last-named proved to be the most active, and the other two were accordingly discarded. Seventy-eight mice in all were treated ^^ith "337," their tumors varpng from the size of a pea to that of a hazel-nut, and of the seventy-four appearing in the final reckoning, fifty-two were favorably influenced ; of these, twenty-two were entirely cured, sixteen showed great, and fourteen less, though still distinct, improvement. The more mahgnant were the tumors, the less favorable was the outcome, and only tumors of slowly growing strains could be cured. That spontaneous healing need not be taken 1 Deut. med. Woch., 1910, xxxvi, 1356. ^ Zeitschrift f. Krehsforsch., 1907, v, 81. THERAPEUTICS 263 into account was shown by the different course of events in treated mice, by the observation that control tumors grew rapidly and led to the death of the animals bearing them, and finally, by the fact that animals which had been cured were not refractory to re-inoculation. Uhlenhuth and Weidanz/ investigating the action of atoxyl on mouse tumors, found that growth was so much more rapid in treated animals than in their controls that after a month's medication the tumors were nearly twice as large. Pyocyanase, on the contrary, when injected directly into rat sarcomata, brought about the disappearance of tumors even after they had attained the size of a walnut. Uhlenhuth, Haendel, and Steff enhagen - employed quinine, sodium taurocholate, ox-gall, antiformin, arsenophenylglycin, ferments, ad- renalin, alcohol, and the appHcation of cold. None of these was so effective as pyocyanase which, however, was found to be poorly borne by rats. Beck^ investigated the action of arsenic, atoxyl or some of its prep- arations, and quinine, finding them all without effect. More favorable results followed the use of certain substances prepared from bacteria or their metabohc products (if injections were made directly into the tumor), and filtrates and extracts from B. prodigiosus and 5. pyogenes aureus were employed w4th success although other bacterial prepara- tions, as tuberculin, were without curative action. V. Wassermann, Keysser, and Wassermann ^ outhned the difficulty of attacking the mahgnant cell somewhat as follows, employing the terminology of Ehrlich. While in treating the protozoal diseases the aim of the investigator was to prepare a medicament which would be parasitotropic but not organotropic, the problem in cancer was to obtain a material which should be organotropic, but whose action should be Hmited to a certain area in the organ — the tumor cells. To affect the very mahgnant neoplasms of the mouse it would be necessary either to employ an unusually powerful agent or to find a means of attacking the nucleus, the most vital point of the tumor cell ; it would be necessary, in other words, to discover a nucleotropic 1 Arb. a. d. Kaiserl. Gesundheitsamte, 1909, xxx, 444. 2 Centralbl. f. Bakt., etc., erste Abt., Ref., 1910, xlvii, Beiheft, 159. Arb. a. d. Kaiserl. Gesimdheitsamte, 1911, xxxvi, 492. ^ Zeitschrift f. Krebsforsch., 1910-1911, x, 153. * Deut. med. Woch., 1911, xxxvii, 2389. 264 THERAPEUTICS substance. This would have to be at the same time a material capable of exerting its curative effect when introduced by way of the blood stream, for all agents, when locally employed, lacked the neces- sary elective power to make their own way to the neoplastic cell. Gosio ^ had discovered that tellurium and selenium were reduced by living bacteria as black and red precipitates respectively, and the pres- ent authors found that in the case of the cancer cell this reduction, in vitro at least, was most intense about the nuclei. It was then es- sential to know whether the specific reaction would occur in the living organism. While preparations of selenium or tellurium, injected into a tumor, were found to have a destructive effect upon its cells, this was especially true of tellurium. But as these compounds re- mained entirely inert after introduction into the body through the blood stream because they had not succeeded in reaching the malig- nant growth, it was necessary, in the words of the authors, to con- struct a railway by means of which the selenium and tellurium could be carried to their destination. Because in mouse tumors the blood supply was Hmited almost entirely to the periphery, it was necessary to use for this purpose a rapidly diffusible substance, and after many trials it was found that eosin fulfilled the requirements. This dye, combined with selenium and injected into the circulation of mice bearing transplantable growths, produced an immediate reddening of the entire animal, although no effect upon the tumor was noticeable until three daily inoculations had been given. After the third treat- ment a distinct softening was demonstrable, while after the fourth the tumor felt more like a fluctuating cyst than a sohd nodule ; and after this condition had been obtained absorption set in until, within ten days, the cure was complete. Often, however, recovery was not so uneventful and, as was especially the case with large tumors, the mice were killed by a. toxemia resulting from the absorption of their own growths. Two mice with spontaneous tumors had also been cured, and four strains of transplantable carcinoma and one propagable sarcoma had all proved equally susceptible to treatment. Tumors in control mice, on the contrary, continued to grow progressively. Re- covered animals had been kept under observation for months, but no suggestion of recurrence had been observed. In mice autopsied dur- 1 Zeitschrift f. Hyg., etc., 1905, li, 65. THERAPEUTICS 265 ing the first few days of treatment the growths were stained an intense red as contrasted with the surrounding tissues, showing that the neo- plastic cells had exerted a specific attraction upon the injected material. As the eosin-selenium compound was extremely toxic, and as it was un- fortunately necessary to give almost the lethal dose in order to achieve a cure, many mice perished merely as the result of the injection. V. Hansemann,^ in conjunction with the three authors previously mentioned, discussed the histological appearance of treated tumors, and found evidence to support the selective action of selenium upon the nuclei. Examinations of the organs of injected mice showed no serious lesion attributable to the treatment. All four investigators united in declaring that the treatment was still in the experimental stage, and that it marked rather the intro- duction of a new era of chemotherapy than the discovery of a cure for cancer. In discussing the work of v. Wassermann and his collaborators Klemperer ^ said that he also, in conjunction with Emil Fischer, had been able to cause the regression of mouse tumors with selenium. Walker ^ began a series of experiments with selenium immediately after the publication of v. Wassermann's article. The first preparation to be investigated was a combination of colloidal selenium and eosin, but this proved to be entirely Vithout effect upon the transplantable* tumors of rats and mice. It was, however, non-toxic, in contrast to the highly poisonous preparation of the German investigator. Combinations of selenium and eosin of doubtful composition were also tried, but although they were highly poisonous none of them showed the slightest selective action upon tumor cells. Neuberg and Caspari,^ starting with the observations that tumors were characterized by increased fermentive activity, particularly autolysis (in the widest sense of the word), and that certain substances like radium and compounds of the heavy metals in colloid form were capable of increasing autolysis in vitro, tried to influence tumor growth by intravenous injections of compounds of the heavy metals. They were able to find several materials possessed of an affinity for malignant cells — compounds of gold, platinum, silver, rhodium, ruthenium, ^ Berl. klin. Woch., 191 2, xlix, 4. ^ Deiot. med. Woch., 191 2, xxxviii, 89. ^Lancet, 1912, i, 1337. * Deut. med. Woch., 1912, xxxviii, 375. 266 THERAPEUTICS iridium, lead, and especially copper and tin. Other substances, arsenic and vanadium, also exerted an action upon autolysis, although a less certain one. Sulphur, and iodine and its compounds were, on the contrary, entirely inert. In a later article Neuberg, Caspari, and Lohe ^ added compounds of antimony, mercury, cobalt, osmium, and palladium to the list of active materials. Microscopic deposits of the metals employed were to be found in the tumors cf animals that had been subjected to treatment. The attack of these elements was extraordinarily sudden, beginning, sometimes after as short an interval as one minute, with extreme con- gestion, which was shortly followed by the hemorrhage initiating the ■ cure. Within the first twenty-four hours a tumor of considerable size might become almost entirely softened, while repeated injections resulted in an entire liquefaction and absorption of the nodule. Several strains of transplanted tumors, as well as sporadic growths, proved equally amenable to treatment, and as spontaneous recoveries from the transmissible growths employed were very rarely encoun- tered, the authors rejected the possibility that healing might not have been the result of their treatment. The materials used were all toxic and, as in v. Wassermann's experiments, the effective dose lay very near the lethal amount, wherefore these observers were unwilling to forecast the relation of their investigations to the treatment of cancer in the human subject. The question of chemotropism in neoplastic tissue has been re- viewed by von den Velden.^ Simple though the conception might be that one material could act as a pilot or a railway to direct or trans- fer another substance to its destination, the discovery of such a com- bination was an entirely different matter. The union must not be so unstable as to be broken down in the stomach, the tissues, or the blood, nor yet so firm that the active principle could be deterred from exert- ing its effect through remaining coupled after arrival at its goal. The clearest instance of specific affinity was offered by iodine, and was best expressed negatively by saying that fat, as well as the central nervous system with its large lipoid content, was practically barren of iodine. Oswald Loeb had shown, nevertheless, that iodine could be made to penetrate both fat and the central nervous system if it were first com- 1 Berl. Mill. Woch., 1912, xlix, 1405. 2 ^^j-i j^n^^^ Woch., 1912, xlix, 825. THERAPEUTICS 267 bined with a lipoid. Thus was afforded an example of an exogenous factor with power to alter the distribution of a chemical substance within the body. Contrasted with this instance there were certain endogeneous factors. A change in the relative bulk of the various tissues might act in this way, as was demonstrated by the difference in what occurred when a material soluble in lipoids was administered to a very thin or a very fat person. The endogenous factors were most distinct, however, under pathological conditions, as Jacoby had shown by his work on saHcylic acid. Thus, in a rabbit with experimentally produced general sepsis, sahcyHc acid would be taken up in large amounts by the joints although previously they held none of this drug ; furthermore, Oswald Loeb and Michaud had shown that in turpentine abscesses and in the tuberculous eyes, lymph nodes, and lungs of guinea-pigs and rabbits, there occurred a very appreciable storing of iodine when potassium iodide had been introduced into the body, von den Velden himself had been able clearly to demonstrate iodine in metastatic carcinomata of the liver and pancreas after intro- ducing into a moribund cancer patient one gram of sodium iodide for every twenty kilograms of body weight, although the tissues them- selves in which the secondary nodules were embedded contained no iodine. This finding had, furthermore, been confirmed by Takemura, who found that mouse carcinobiata and rat sarcomata were capable of storing up iodine, while Blumenthal had discovered a tropism for neoplastic tissues in the case of arsenic. The affinity of various chemicals for certain organs had been re- ferred by Oswald to changes in the diffusion processes caused by altera- tions in the colloid structure of the membrane and peripheral portions of the cells ; and as, in the opinion of that author, similar alterations were to be found not only in inflammatory processes but in new growths as well, the "membrane problem" thus suggested became of such wide significance that it might transfer the entire field of in- vestigation to the realm of physical chemistry. Therapeutic experiments have been undertaken not infrequently in man also, certain of them inspired without doubt by the early endeavors of Hericourt and Richet ^ to obtain an immune serum from 1 Compt. rend, de I' Acad, des Sc, 1895, cxx, 948. Compt. rend, de VAcad. des Sc, 1895, cxxi, 567. 268 THERAPEUTICS animals injected with human cancer, although others, falling within the domain of immunity, have been attempts to transfer to man such knowledge as has been gathered by observation upon the lower ani- mals. The investigations which have been prosecuted in the labora- tory within the last decennium have, however, yielded no result ger- mane to therapeutics other than to show that an animal can be made resistant to the transplantation of malignant cells from another or- ganism, and the fallacy of appl^dng these findings to the treatment of tumors autochthonous in origin and firmly estabHshed, has been in- dicated by Bashford ^ as follows, in a discussion of the refractory con- dition in animals: "It is necessary to point out emphatically, that- nothing but harm can result from ignoring the fact that the above experiments only indicate the possibility of rendering normal mice unsuitable for the growth of experimental cancer. They have not enabled us to arrest the progress of experimental tumours with cer- tainty, far less to affect the cure of the disease occurring naturally." Fichera ^ attempted to cure cancer in rats and in the human subject by injecting autolyzed rat embryo in the one case, and autolyzed hu- man embryo in the other. Twenty-five patients with tumors of various kinds continued treatment long enough to allow the method a fair trial. While nine of them showed no improvement whatever, in the remainder there occurred arrest of development, involution, or some- times temporary disappearance of the growth. Coca and Oilman ^ injected a series of patients, each with his own carcinoma after it had been thoroughly ground and emulsified, and found the results encouraging. A footnote stated that a carcinoma of the same kind taken from another individual might be substituted. In a later article, however. Coca, Dorrance, and Lebredo ^ concluded from a study of seventy-nine cases that, except for the rapid relief of cachectic symptoms, they had seen no evidence that the injection of tumor tissue exerted any specific influence over malignant growths. Rovsing,^ who introduced heated tumor emulsions, gained the im- ^ British Med. Jour., 1906, ii, 209. ^ Tumori, Turin, 1911 ; abstr. in Bull, de Vlnst. Past., 1911, ix, 272 ; see also Lancet, 1911, ii, 1194. ^Philippine Jour. Sci., 1909, iv, B, 391. * Zeitschrift f. Imniunitdtsforsch., etc., Orig., 1912, xiii, 543. * Travaux de la deuxieme Conference internal, pour I'Etude du Cancer, Paris, 1911, 562. THERAPEUTICS 269 pression that although this treatment was without effect upon car- cinomata it might possibly prove valuable against sarcomata. Hodenpyl ^ offered a preliminary report regarding the treatment of cancer by means of ascitic fluid withdrawn from a patient in whom both primary and recurrent carcinomata had disappeared. The fluid was tested first upon mice bearing transplantable carcinomata, and when it had been found that these growths diminished in size or even disappeared, the experiment was extended to include forty-seven in- operable cases in man. In general the effect of the injection was to cause a temporary redness, tenderness, and swelling in the neighbor- hood of the growth, this reaction being followed by necrosis of the tumor and a decrease in size or, in some instances, by complete disappearance. Ill and Miningham ^ obtained ascitic fluid from a somewhat similar case and introduced it into more than thirty patients with malignant disease. Although not a single case was cured, pain was mitigated, hemorrhage was decreased or entirely stopped, and there occurred a transient gain in weight. The psychic element was eliminated from this striking improvement in the general condition by the fact that only one patient knew the nature of her disorder. A useful resume of the serum therapy of cancer, which contains nearly one hundred and fifty references, has been compiled by Vidal,^ while Coca ^ has but recently prepared a critical review of the ''vaccination" treatment for human cancer in its relation to the ex- perimental data bearing upon normal tissue and tumor immunity. ^ Med. Record, 1910, Ixxvii, 359. ^ Jour. American Med. Assoc, 191 2, lix, 497. * Travaux de la deuxieme Conference internal, pour V Etude du Cancer, Paris, 191 1, 293. ^ Zeitschriftf. Immunitdtsforsck., etc., Orig., 1912, xiii, 524. CHAPTER X GENERAL SUMMARY Although the literature of experimental cancer may seem to be a maze of conflicting statements, there are nevertheless a few threads which may be fixed upon and followed through. These are repre- sented by a number of findings which, ha^'ing received the support of a majority of those engaged in cancer research, may be accepted as probably rehable. Nevertheless, it has happened occasionally that even the most competent observ^ers have come to diametrically opposed conclusions in spite of the most careful work, a situation suggesting that differences still unrecognized may have existed between the animals used in the several experiments. The mouse is not a test tube where so much of one substance and so much of another will alv/ays produce so much of a third, but is, on the contrary, a highly complex organism exhibiting protean reactions. \\Tiat takes place, therefore, in one mouse may not happen in the next — nay, what happens in two groups of ten mice each may not, and often does not, occur in a third series. Malignant growths have been found throughout the animal kingdom down to, and including, the fish — except that no certain instance has yet been reported among the reptiles. By far the greater amount of attention has, however, been bestowed upon the disease in the mouse, not only on account of the ease with which this animal can be procured and the cheapness and facihty with which it can be kept under ob- servation, but also because the subcutaneous situation of its tumors makes for early and ready recognition. The fact that information has been accumulating chiefly in regard to mouse cancer must not, therefore, be understood to imply that the findings might not hold good for malignant new growths of other species, were it convenient or possible to investigate them in the same way. There can be scarcely a doubt that the neoplasms of the mouse are entirely comparable to 270 GENERAL SUMMARY 271 those of man, since they exhibit continuous, infiltrative, growth, no less than the power to metastasize and to recur after incomplete re- moval. Evidence has been adduced that there is some hereditary influence at work, making mice of recently cancerous ancestry more prone to develop cancer than are those in which the taint is more remote. But whether or not this observation may be appHed to man it is at present impossible to determine, nor would it be safe to venture an opinion until data as accurate as those relating to the mouse become available. It has been ascertained that almost any spontaneous new growth of the mouse can be transplanted into other mice, and that although the first inoculation may not be attended by any great degree of success, propagation generally becomes easier in succeeding generations. Why this should be so is not known ; some observers ascribe it to an in- crease of virulence on the part of the tumor cell, while others see in it merely an increasing power of this element to adapt itself to altered con- ditions. It has been the experience of the majority of investigators that for transmission to be eft'ective, living and intact cells must be introduced into an animal of the same species as that in which the primary tumor took its origin, or at least into a species very closely related. In this connection there has been discovered a phenomenon of the highest importance. It is that cancer is more readily transplantable into the animal in which it arose spontaneously than into any other^ This, and the complementary observation^ that an animal cannot be immunized in any manner yet at command against the inoculation of its own primary tumor, should never be lost sight of in evaluating descriptions of therapeutic experiments, for it demonstrates beyond reasonable doubt that cancer cells proliferating in an animal to which they are not native are doing so under a certain disadvantage. A pro- cedure, therefore, able to turn the tide against them might still be ineffectual when appHed against malignant cells more favorably situated in an animal to which they are autochthonous. Furthermore, the almost uniform success with which the cancer cell can be ingrafted in the organism within which it first acquired malignant properties, as contrasted with any other organism, mihtates very strongly against any parasitic theory of tumor genesis. At the same time, the observa- 272 GENERAL SUMMARY tion that the bearer of a tumor offers a particularly favorable soil for the cells of that neoplasm will intimate to the surgeon the necessity for a continuance of the most extreme care against the distribution of cancer cells throughout the operation wound, a precaution already suggested in the past by the frequency of local recurrence following the excision of mahgnant tumors. Secondary growth along the needle track after exploratory puncture of a tumor furnishes an even more striking example from human pathology of this extreme susceptibility of the individual to inoculation with his own neoplasm. Finally, these considerations demonstrate how minimal is the chance of the transfer of cancer from one individual to another in the ordinary course of events. The proliferation of a tumor after its introduction into another host is an example, not of infection, but of transplantation; the paren- chymal cells of the resulting growth are derived, not from the elements of the new host, but from those of the transferred graft. These cells are supported and nourished by a stroma which, built up by the con- nective tissues and blood vessels of the new organism, is well on its way toward completion by the third or fourth day following trans- plantation. Propagable tumors are subject to fluctuations in growth energy. They may grow progressively, or disappear spontaneously. Even under optimum conditions, however, the proUferative power of the cancer cell falls far short of that possessed by the bacteria, and per- haps even of that displayed by the cells of the developing embryo. The key to the nature of the malignant cell lies, therefore, not in the explanation of augmented growth power, but rather in an elucidation of the capacity of this cell for continuous growth. A tumor under cultivation usually retains unaltered the characteris- tics of the primary growth. While most transplantable neoplasms show a marked deviation from the normal structure of the organ which gave them origin, this is by no means an invariable rule, since com- plete loss, or even latency, of differentiation is not necessary for con- tinuous growth, and tumors which, under the microscope, are dis- tinguishable from their mother tissue only with difficulty, have been proved capable of unlimited propagation. So far, then, as the re- tention of normal structure is a criterion of innocence its significance GENERAL SUMMARY 273 has been lost, a situation parallel to that obtaining among certain tumors of the thyroid gland and the ovary in man. The cells of a few transplantable carcinomata have been found capable of causing a transmutation of the stroma in the course of which, acquiring malignant properities, it becomes indistinguishable from other sarcomata of the mouse. Mice in which tumors have undergone partial or complete absorption are refactory to subsequent transplantation, either of the same or of another growth. ' This condition of resistance is brought about also by the absorption of hving and intact cells of the same species, but not those of the animal itself. Immunity, once induced, cannot be transferred passively to other animals either by the cells or the body fluids, although it is probably these agents which are responsible for the spread of the condition throughout the organism within which it has been engendered. And while there is a certain amount of evidence in favor of the fact that natural resistance may be the outcome of hereditary transmission, there is none to show that acquired immunity can be so transferred. Immunity has been attributed to a lack of the specific nutritive materials necessary for tumor growth. On the other hand, it has been described as a suppression of the stroma reaction which, under more favorable conditions, would have undertaken the support of the tumor in its new host. The latter supposition has more evidence in its favor. It must be frankly confessed that so far, at least, the study of cancer, instead of affording an understanding of the nature of the disease, has but opened up new problems which were formerly not even conceived. Hence earnest students have had to content themselves with slowly and laboriously collecting data, in the hope that at some future time these may be combined into a coherent whole. Slow at the best, this is nevertheless the only way in which the goal can be surely attained, for the empirical method, in spite of its centuries of effort, has pro- duced nothing nearer to a cure than surgical removal. 'The hope of substituting some plan of specific treatment for this crude and only too often inefficient means, rests on the fact that in many mice trans- planted tumors, and in a few, spontaneous tumors, have vanished completely, although it must be admitted that we have not yet the slightest insight into the mechanism by which this occurs. Still, the 274 GENERAL SUMMARY balance must be very, delicate, j the margin between continuous proKf- eration and cessation 'very narrow, and the newer conceptions of malignant growth encourage the hope that at some time, far in the future though it may be, the nature of this difference will be unriddled, and the dream of effectual interference with incessantly proliferating cells will become an ac tuaU ty. (,;^^^ ,^ m . jj, ,y^^^^^ ^ ^ ^^ j INDEX OF AUTHORS Adam, on homoplastic transplantation, 42 Adami, on cancer houses, 43 ; on cancerous de- generation, 12 Alberts, on heteroplastic transplantation, 51 Albrecht and Hecht, on age and susceptibiHty, 130 ; on amyloid degeneration, 95 ; on para- biosis, 25S (footnote) ; on pregnancy and susceptibility, 138 ; on race and susceptibiHty, 136; on sarcoma development, 122; on Scharlach R, 37; on the specificity of im- munity, 143. Alibert, on homoplastic transplantation, 43 Apolant, on the action of radium, 261; on athrepsia, 163, 166; on autoplastic trans- plantation, 224; on classification, 214; on the frequency of cancer in mice, 199; on the genesis of mouse tumors, 185, 186, 187; on the hare tumor of v. Dungern, 156 ; on hyper- sensibiUty, 182; on immunity with alien tissues, 148; on immunity and histology, 107 ; on infiltrative growth, 192; on lactation and aetiology, 200; on transmissible lympho- sarcoma of the dog, 233 ; on the malignancy of mouse tumope, 193; on sarcoma develop- ment, 112, 113; on spindle-shaped epithehal cells, 106; on spontaneous absorption, 90; on the stroma of hemorrhagic tumors, 109; on sweat glands in the mouse, 185 ; on the transplantation of timior mixtures, 79, 171 ; on variations during transplantation, 105 ; on virulence, 62 ; see also Ehrlich and Apolant Apolant and Ehrlich, on the transplantation of tumor mixtures, 79 Apolant and Marks, on immunity with autol- ogous tissues, 150 Arndt, on Egyptian medicine, i Askanazy, on pregnancy and growth, 28; on the production of tumors, 27 AuBERTiN, see Marie and Aubertin Ayson, on thyroid adeno-carcinoma in the trout, 237 Baber, on non-encapsulation of the thyroid, 241 Baeslack, on infiltrative growth, 191 ; see also Gaylord, Clowes, and Baeslack, and Clowes and Baeslack Basheord, on acquired immunity, 139, 144; on aetiology, 6, 9, 210; on age and susceptibility, 129; on autoplastic transplantation, 223; on the distribution of immunity, 153; on dosage, 69, 73 ; on the duration of immunity, 154; on growth rate, 84; on health and susceptibility, 137; on heredity and aetiology, 201, 204; on histological structure, 109; on immunity with ahen tissues, 143, 148 ; on im- munity with devitalized tissues, 149 (foot- note) ; on the infectivity of cancer, 198; on inoculation site, 76 ; on race and susceptibility, 132; on sarcoma development, in; on the specificity of immunity, 142, 147 ; on spon- taneous absorption, 91 ; on the stroma of hemorrhagic tumors, 109; on the technic of inoculation, 70; on treatment, 257, 268; on variations during transplantation, 100, loi ; on virulence, 65 Bashford and Mi-rray, on age and suscepti- biUty, 129 ; on early stages, 59 ; on the genesis of mouse tumors, 186; on heredity and aetiol- ogy, 202; on heterotypical mitosis, 18; on infiltrative growth, 191; on the malignancy of mouse tumors, 192; on metastasis, 190; on pregnancy and susceptibiHty, 137; on race and susceptibility, 132 ; on the re-inocu- lation of tumor-bearing animals, 167; on the zoological distribution of cancer, 184 Basheord, Ml-rray, and Bowen, on age, race, and susceptibiHty, 76; on diminished trans- plantability, 78 ; on fluctuations in growth energy, 86; on spontaneous absorption, 91 Basheord, Mltrr.\y, and Cr.4mer, on acquired immunity, 139; on adaptation, 225 ; on auto- plastic transplantation, 223; on the cHnical course of transplanted tumors, 92 ; on the distinction between tumors and inflammatory swelUngs, 78; on the distribution of im- munity, 153; on the duration of immunity, 154; on early stages, 59; on fluctuations in growth energy, 85 ; on the frequency of cancer in mice, 199; on variations during trans- plantation, 99; on the histology- of receding tumors, 91; on infiltrative growth, 191, 192; on the transmissible lympho-sarcoma of the dog, 230; on metastasis, 190; on the optimum conditions of transplantation, 88 ; on phagocytosis in spontaneous absorption, 92 ; on the propagation of hemorrhagic tumors, 72; on race and susceptibiHty, 132; on the relative importance of soil and graft, 72; on sarcoma development, no; on the specificity of immunity, 142; on spontaneous absorption, 91 ; on the development of spon- taneous tumors during immunity, 128; on the stroma of hemorrhagic tumors, 109; 275 -^ 276 INDEX OF AUTHORS on the action of radium, 261 ; on tumors of the mouse, 188; on virulence, 66 Bashford, Murray, and Haaland, on aher- nations in biological quahties of the tumor cell, 75 ; on athrepsia, 167 ; on the hereditary transmission of immunity, 158; on hyper- susceptibility, 180; on immunity in tumor- bearing animals, 151; on passive resistance, 156; on sarcoma development, 115; on the specificity of immunity, 142 ; on the develop- ment of spontaneous tumors during immunity, 128 Bashford, Murray, Haaland, and Bo\ven, on adaptation, 65 ; on growth energy, 8g ; on race and susceptibiHty, 132; on the technic of in- oculation, 70 Bashford and Russell, on athrepsia, 168; on spontaneous absorption, go Beard, on aetiology, 18; on treatment, 260 Beck, on treatment, 263 Beckton, on Altmann's granules, 193; see also CoLWELL and Beckton Beckton and Colwell, on Altmann's granules, 194 Beckton and Russ, on Altmann's granules, 194 Beebe, on treatment, 235 ; see also Crile and Beebe Beebe, and Ewing, on the transmissible lympho-sarcoma of the dog, 232, 233 Beebe and Tracy, on treatment, 235 Behla, on homoplastic transplantation, 42 Beneke, on kataplasia, 16 Bensley, on Altmann's granules, 194 Benthin, on atypical epithelial growth, 36 Bergel, on a fat-sphtting ferment in lym- phocytes, 37 Bergell and Sticker, on treatment, 235 V. Bergmann, on autoplastic transplantation, 45 Billroth, on heteroplastic transplantation, 51 BiRCH-HiRSCHFELD and Garten, on the pro- duction of tumors, 22 Blumenthal, on treatment, 259; on the depo- sition of arsem'c, 267 ; see also v. Leyden and Blumenthal Bonnet, on thyroid adeno-carcinoma of the trout, 236 BoRREL, on athrepsia, 165; on immunity, 140, 146; on the infective nature of tumors, 195; on the interval after which growth becomes apparent, 77; on metastasis, 190; on para- sites and ffitiology, 195, 196 ; on the per- centage of success, 71 ; on squamous cell carcinoma, 188; on treatment, 257 ; on trans- plantation, 57 BoRREL and Petit, on autoplastic transplanta- tion, 223 BoRRMANN, on setiology, 13 Borst, on setiology, 13 BowEN, see Bashford, Murray, and Bowen, and Bashford, Murray, Haaland, and Bowen Boycott, on uterine tumors in the rabbit, 22 Bridre, on athrepsia, 166; on immunity, 140, 144, 146, 154; on the interval after which growth becomes apparent, 77 ; on parasites and Eetiology, 196; on passive resistance 156; on pregnancy and susceptibility, 138; on the premetastatic stage, 152; on the re- sistance of the cancer cell, 82 ; on treatment, 257 ; on the technic of inoculation, 68 Bridre and Conseil, on parasites and aetiology, 196 Brodie-Mills, on horn core, 6 Brosch, on atypical epithelial growth, 22 Brown, on the nucleus, 3 BuDD, on homoplastic transplantation, 41 Burgess, on the stroma reaction in immunity, 176 BuRKHAEDT, on Altmann's granules, 194 Burrows, on tissue cultivation in mtro, 124, 126 ; see also Carrel and Burrows Buschke, on age and susceptibihty, 129 BuTLiN, on autoplastic transplantation, 44, 46 ; on homoplastic transplantation, 39 Calkins, on rhythms of growth, 88 Carmalt, on the ameboid motion of cancer cells, S Carrel, on tissue cultivation in vitro, 125, 126 Carrel and Burrows, on tissue cultivation in vitro, 124, 126 Caspari, see Neuberg and Caspari, and Neu- BERG, Caspari, and Lohe Casper, on the zoological distribution of cancer, 184 Cazin, see Duplay and Cazin Celsus, on classification, i Chalmers and Perry, on aetiology, 6 Chisholm, on the blood of rats bearing trans- plantable sarcoma, 96 Clowes, on the hereditary transmission of immunity, 157 (footnote) ; on immunity, 141, 143 ; on race and susceptibihty, 132 ; on re- inoculation, 164; on the resistance of the cancer cell, 81 ; on spontaneous absorption, 90, 138; on the technic of inoculation, 71 ; on treatment, 257; see also Gaylord and Clowes, and Gaylord, Clowes, and Baes- lack Clowt.s and Baeslack, on age, race, and sus- ceptibility, 76 ; on the distinction between tumors and inflammatory swelUngs, 78; on dosage, 73; on immunity, 139, 141, 155; on the interval after which growth becomes apparent, 77 ; on spontaneous absorption, 91; on the stimulation of growth energy, 68 Clunet, on the histology of tumors after ex- posure to X-rays, 260; on sarcoma develop- ment, 123; on sarcoma developing after ex- INDEX OF AUTHORS 277 posure to X-rays, 30 ; on the development of spontaneous tumors during immunity, 128; see also !Marie and Clunet Coca, on treatment, 26g ; see also v. Dungern and Coca Coca, Dorrance, and Lebredo, on treatment, 268 Coca and Oilman, on treatment, 268 CoH>'HEiii, on aetiolog>', 8 CoLWELL, see Beckton and Col-well CoLWELL and Beckton", on Altmann's granules, 194 CoxsEiL, see Brjdre and Conseil CoNT.AJinN, on the action of X-rays, 261 CoPEiLA.x and Hake, on the acidity of the gastric contents in tumor-bearing animals, 97 CoRNiL, on autoplastic transplantation, 49 Cramer, on metaboHsm in tumor-bearing rats, 97 ; see also Basheord, Mltibay, and CR.AilER Cramer and Prixgle, on metabohsm in tumor- bearing rats, 98 Crile and Beebe, on treatment, 235 CRITZM-A.XX, on the relation between twin births and cancer, 18 CuEN'OT and Mercler, on the hereditary trans- mission of immunity, 159; on the hereditary transmission of susceptibiUty, 158; on preg- nancj' and susceptibility, 138; on race and susceptibilitj% 136 Da\idsohn, on amyloid degeneration in the white mouse, 96 Deton, on the genesis of mouse tumors, 187 DoRR-ANCE, see Coca, Dorr-\nce, and Lebredo DtDGEON, see Shattock, Seligm.a.nx, and DlT)GE0N V. Dl"xgerx, on immunity, 155; on the trans- missible lympho-sarcoma of the dog, 235 ; on passive resistance, 155 V. Dl^stgerx and Coca, on immunity, 172; on a transmissible sarcoma of the hare, 235 Duplay and Cazln, on heteroplastic transplan- tation, 52 ; on the transmissible lympho- sarcoma of the dog, 227 Dltuytrex, on heteroplastic transplantation, 50 Eberth and Splde, on the genesis of mouse tumors, 185 Edelfsen and Hensen, oh growth and preg- nancy, 98 Ehrexreich and jMich-AELis, on the zoological distribution of cancer, 184 Ehrlich, on the resistance of the cancer cell, 82 ; on inoculation site, 77 ; on the percentage of success, 71; on the technic of inoculation, 69 ; on the stimulation of growth energy, 68 ; on ■ the stroma reaction, 61 ; on increase of \-irulence, 62 ; on the stroma in hemorrhagic tumors, 109; on sarcoma development, no, 113; on passive resistance, 156; on immunitj% 140, 141, 154, 159; on the transplantation of tumor mixtures, 79 ; on pregnancy and sus- ceptibility, 138; on sex and susceptibiUty, 137; on the limits of transplantation, 129; on athrepsia, 159, 169; on spontaneous absorp- tion, 220; on aetiology, 208; on chondroma in the mouse, 188; see also Apolant and Ehrlich Ehrlich and Apolant, on age and susceptibility, 130 ; on the frequency of cancer in mice, 199 ; on sarcoma development, 109, in, 112, 113; on spontaneous carcinoma sarcomatodes, 188 Ehrlich, H., see Kr.\us, R.anzi, and H. Ehrlich V. Eiselsberg, on a transplantable sarcoma of the rat, 54 EwTNG, on autoplastic transplantation, 44, 48 ; . on thjToid adeno-carcinoma of the trout, 241 ; on tissue cultivation in vitro, 126; see also Beebe and Ewing Fallopius, on diagnosis and treatment, 3 Da Fano, on the relation of plasma cells and Ij-mphocytes to immunity, 176; on the dis- tribution of immunity, 153 Farmer, Moore, and Walker, on heterotj-pi- cal mitosis, 17 Fere, on the growth of transplanted tissues in brooding hens, 28 FiCHERA, on pregnancy and susceptibility, 28, 138 ; on treatment, 268 ; on the production of tumors, 28 Fiebiger, on the zoological distribution of cancer, 184 FixsTERER, see Pfeiffer and Finsterer FiRKET, on a transplantable sarcoma of the rat, 55 Fischer, on atypical epithelial growth, 30 Fischer, E., on treatment, 265 Fleischmaxx, see Michaelis, Fleischmann, and PiNCUSSOHN Flexner, on heteroplastic transplantation, 52 (footnote) Flexner and Joblixg, on adeno-carcinoma in the rat, 188; on the duration of immunity, 155; on hypersusceptibihty, 178; on immu- nitj', 139, 140; on inoculation site, 77; on metastasis, 191 ; on the premetastatic stage, 153 FoLLiN, on heteroplastic transplantation, 51 Fraenkel, on the production of tumors. 22 Frank and Unger, on hormones, 259 (footnote) Frel'N'd, on pregnancy and growth, 29; on the production of tumors, 27 Freytag, on am3doid degeneration, 95 ; on treatment, 258 FujiNAin and Inamoto, on a transplantable tumor of the fowl, 244 278 INDEX OF AUTHORS Galen, on aetiology and treatment, 2 Garten, see Birch-Hirschfeld and Garten Gay, on hypersusceptibility, 180; on passive re- sistance, 156 ; on the premetastatic stage, 153 ; on race and susceptibility, 136 ; on the trans- plantation of metastases, yg ; on treatment, 258 Gaylord, on hypersusceptibility, 179; on passive resistance, 157 ; on the resistance of the cancer cell, 81 ; on thyroid adeno-carcinoma of the trout, 239, 240, 241, 243 ; on treatment, 236 Gaylord and Clowes, on cancer cages, 197 ; on dosage, 73 ; on spontaneous absorption, 91, 92, 139 Gaylord, Clowes, and Baeslack, on sponta- neous absorption, 138 Geissler, on the transmissible lympho-sarcoma of the dog, 228 Gierke, on age and susceptibility, 130; on athrepsia, 162, 168; on classification, 220; on metastasis, 191 ; on race and susceptibility, 134; on histology and malignancy, 107; on sex and susceptibility, 137; on the specificity of immunity, 142 ; on the stroma reaction, 61 ; on the technic of inoculation, 70 GiLMAN, see Coca and Gilman Gilruth, on thyroid adeno-carcinoma of the trout, 237 Goldmann, on parabiosis, 259 (footnote) ; on inoculation site, 77 Gosio, on the reduction of selenium and tel- lurium by bacteria, 264 Gougerot and Laroche, on the production of tumors, 29 Grawitz, on slumber cells, 18 Greischer, on atypical epithelial growth, 35 GrtJnbaum, a. and H., on treatment, 259, 260 Gudernatsch, on non-encapsulation of the trout thyroid, 240 Guelliot, on homoplastic transplantation, 40 Haaland, on age and susceptibility, 130; on autoplastic transplantation, 209, 224; on the resistance of the cancer cell, 83 ; on the frequency of cancer in mice, 200; on health and susceptibility, 137 ; on hypersusceptibility, 181 ; on immunity with devitalized tissues, 144, 149 ; on immunity in tumor-bearing mice, 151, 152; on inflammation and aetiology, 205; on lactation and aetiology, 200 ; on metastasis, 1 14, 190 ; on operative removal and recurrence, 222 ; on passive resistance, 156 ; on pregnancy and susceptibility, 137 ; on race and suscep- tibility, 133 ; on the relative resistance of tumors to heat, 82 ; on sarcoma development, 113, 118; on the specificity of immunity, 142; on spontaneous absorption, 221 ; on squam- ous cell carcinoma, 188; on the technic of inoculation, 68, 69 ; on the transplantation of tumor mixtures, 79 ; on tumors of the mouse, 188; see also Bashford, Mxjrray, and Haaland, and Bashford, Murray, Haaland, and Bowen Haendel, see Uhlenhuth, Haendel, and Steffenhagen Hahn, on autoplastic transplantation, 48 Hake, see Copeman and Hake Hamburger, on autoplastic transplantation, 46 Hanau, on atypical epithelial growth, 21 ; on a transplantable carcinoma of the rat, 53 Hanes, see Lambert and Hanes V. Hansemann, on aetiology, 7 ; on the analogy between human and mouse cancer, 193 ; on anaplasia, 14, 16; on atypical mitosis, 14; on the genesis of mouse tumors, 185, 189; on heterotypical mitosis, 17; on the trans- missible lympho-sarcoma of the dog, 228- on metastasis, 103, 193 ; on sarcoma develop- ment, III (footnote); on the selective action of selenium, 265 ; on thyroid adeno-carcinoma of the trout, 239 ; on heteroplastic trans- plantation, 52 Harrison, on tissue cultivation in vitro, 124 Hauser, on aetiology, 7, 12, 13 Hecht, see Albrecht and Hecht Helmholz, on atypical epithelial growth, 32 Hensen, see Edelfsen and Hensen Hericourt and Richet, on treatment, 267 Hertwig, on aetiology, 9 Hertwig and Poll, on athrepsia, 165; on im- munity, 141; on race and susceptibility, 132, 133 ; on the technic of inoculation, 71 Herxheimer and Reinke, on the relation of lipoids to cell division, 37 Herzog, on a transplantable sarcoma of the rat, 188; on pregnancy and susceptibility, 137 Heyde, see Sauerbruch and Heyde HiGUCHi, on immunity, 147, 148 v. HippEL, on the production of tumors, 27 Hippocrates, on cancer of the breast, i Hodenpyl, on treatment, 269 Hofer, on thyroid adeno-carcinoma of the trout, 237 Hooke, on the cellular structure of cork, 3 III and Miningham, on treatment, 269 Inamoto, see Fujinami and Inamoto Jaboulay, on thyroid adeno-carcinoma of the trout, 237 Jacoby, on the deposition of salicylic acid, 267' Janeway, on aetiology, 14 Jenny, on the histology of Hanau's rat car- cinoma, 53 Jensen, on acquired immunity, 139; on diet and susceptibiHty, 134; on early stages, 58;: on the genesis of mouse tumors, 185; oa heredity and eetiology, 205 ; on the interval after which growth becomes apparent, 77; on INDEX OF AUTHORS 279 race and susceptibility, 130, 131 ; on the resist- ance of the cancer cell, 80 ; on selection, 141 ; on spontaneous absorption, 90 ; on the technic of inoculation, 68; on treatment, 256; on a transplantable carcinoma of the mouse, 56 ; on a transplantable sarcoma of the rat, 131, 188; on a transplantable tumor of the beet, 184; on transplantation, 71 Jentzer, on pregnancy and growth, 28 Joachim, on Egyptian medicine, i JOBUNG, on the premetastatic stage, iS3 ; on the re-inoculation of tumor-bearing animals, 168; on heteroplastic transplantation, 52 ; see also Flexner and Jobling JOBSON, see Loeb and Jobson Jolly, on tissue cultivation in vitro, 126 JOREs', on atypical epitheHal growth, 31 JUNCKER, on homoplastic transplantation, 40 Kautmann, on autoplastic transplantation, 45 ; on the production of tumors, 21 Keeling, on the production of tumors, 23 Keysser, see v. Wassermann, Keysser, and Wassermann . Klebs, on the conjugation of leucocytes with epithelial cells, 17 ; on autoplastic transplanta- tion, 45 ; on mitosis, 14; on heteroplastic transplantation, 52 Klemperer, on treatment, 265 KoKUBO, on aetiology, 13 E.RASKE, on autoplastic transplantation, 45 _ Kraus, Ranzi, and H. Ehrlich, on the distri- bution of immunity, 154 Lack, on autoplastic transplantation, 47 ; on th{ production of tumors, 21 Lambert, on immunity with autologous tissues, 150 Lambert and Hanes, on the ameboid motion of cancer cells, 5, 125; on tissue cultivation in vitro, 125, 126 Lane-Claypon and Starling, on hormones, 259 Langenbeck, on heteroplastic transplantation, 50 Lanz, on homoplastic transplantation, 44 Laro'che, see Gougerot and Laroche Lazarus-Barlow, on the genesis of mouse tumors, 189 Lebert, on heteroplastic transplantation, 51 Lebredo, see Coca, Dorrance, and Lebredo Leitch, on hypersusceptibility, 181 ; on uterine carcinoma in the rabbit, 22 Lenhart, see Marine and Lenhart Leonides, on treatment, 2 Leopold, on the production of tumors, 21 Leopold, see Loeb and Leopold Levin, on immunity, 149, i5o, iS4; on the in- tense connective tissue reaction of rats, 33 ; • on the production of tumors, 26 Lewin, on age and susceptibihty, 130; on athrepsia, 165; on immunity, 139, 142, i43, 146; on race and susceptibility, 135; on the resistance of the cancer ceU, 82 ; on sarcoma development, 118; on sex and susceptibility, 137 ; on treatment, 258 ; on variations during transplantation, 107; see also Michaelis and Lewin Lewin and Michaelis, on carcinoma mammae in the rat, 188 ; on metastasis, 191 (footnote) v. Leyden, on parasites and etiology, 59; on treatment, 257 V. Leyden and Blumenthal, on treatment, 257 Liepmann, on sarcoma development, 115 LwiNGOOD, on spontaneous tumors of the mouse, 189 , Loeb, L., on etiology, 25; on age and susceptibihty, 129; on autoplastic trans- plantation, 223; on dosage, 73; on early stages, 59 ; on the importance of soil, 72 ; on metastasis, 190; on race and susceptibihty, 131 ; on the resistance of the cancer cell, 81 ; on sarcoma development, 113, ii4. "8; on sex and susceptibility, 137; on spontaneous absorption, 90; on a transplantable sarcoma of the rat, S7, 188; on the production of tumors, 24; on the transplantation of stationary or receding tumors, 78 ; see also White and Loeb Loeb and Jobson, on carcinoma in cattle, 197 Loeb and Leopold, on autoplastic transplanta- tion, 223 , , , Loeb and White, on the effect of_ heat upon growth, latent period, and absorption, 83 Loeb O., on the deposition of iodine, 267 Lohe', see Neuberg, Caspari, and Lohe LUBARSCH, on amyloid degeneration, 95 Marie and Aubertin, on uterine carcinoma in the rabbit, 22 , i ^ Marie and Clunet, on the histology of tumors after exposure to X-rays, 260 Marine and Lenhart, on thyroid adeno-carci- noma in the trout, 242 ; on thyroid histology, 243 , - , ' Marks, see Apolant and Marks Martin, on the production of tumors 21 Maurer, on non-encapsulation of the thyroid in fish, 241 . , ,. , *u ,T • McConnell, on atypical epithehal growth, 31, on heteroplastic transplantation, 37 McCoy, on the zoological distribution of cancer, MsmGRECEANU, on the weights of organs in tumor-bearing animals, 93 ; on the_ amount of food ingested by tumor-bearing animals, 94 Mercier, see Cltenot and Mercier_ Metchnikofe, on parasites and stiologj', 52, see also Roux and Metchnikofe Meyer, on atypical epithelial growth, 35 28o INDEX OF AUTHORS MiCHAELis, on athrepsia, i66; on cancer cages, 197; on classification, 213 ; on the distinction between tumors and inflammatory' swellings, 78; on the genesis of mouse tumors, 186; on immunity, 143 ; on the stimulation of growth energy, 68 ; on infiltrative growth, igi, 192 ; on the inter^-al after which growth becomes apparent, 77 ; on metastasis, 191 ; on passive resistance, 156; on race and susceptibility, 132 ; on the resistance of the cancer cell, 81 ; on selection, 141 ; on spon- taneous absorption, 90 ; see also Ehrenreich and MiCHAELis, and Lewin and Michaelis MiCHAELis Fleischmann, and Pincussohn, on immunity, 140, 142, 146 Michaelis and Lewin, on variations during transplantation, 107 MiCHAtT), on the deposition of iodine, 267 Miller, on Altmann's granules, 194 MiLXER, on implantation cancer, 49 MixiNGHAM, see III and Miningham MiNOT, on growth rate, 84 ; on pregnancy and growth, 98 Moore, see Farmer, IMoore, and Walker Moore and Walker, on the resistance of the cancer cell, 81 MoRAD, on the hereditary transmission of sus- ceptibihty, 157; on pregnancy and suscepti- bUity, 137 ; on a transplantable carcinoma of the mouse, 54 MoRESCHi, on hypersusceptibUity, 180; on unmunity ■n'ith alien tissues, 148 ; on immunity ■nith normal tissues, 146; on tumor growth and nourishment, 92 De Morgan, on jetiology, 18 MuLLER, on the cellular structure of tumors, 3 Mlt^jhy, see Rous and Mltlphy Murray, on adaptation, 65, 208; on the chnical course of spontaneous tumors, 221; on the frequency of cancer in mice, 199; on hepatic carcinoma in the cow, 207 ; on heredity and ffitiologj', 202, 204; on the histology of car- cinoma mammje, 219; on immunity and histologj% 108 ; on the inter\-al after which growth becomes apparent, 77 (footnote) ; on the malignancy of mouse tumors, 192 ; on metastasis, 191 ; on spontaneous absorption, 220; on sweat glands in the mouse, 185; on the technic of inoculation, 70; on thinroid adeno-carcinoma of the trout, 236; on a transplantable keratinizing adeno-carcinoma of the mouse, 100 ; on a transplantable osteo- chondro-sarcoma, 100 ; on the transplantation of metastases, 79 ; on tumors of the mouse, 188; on the zoological distribution of cancer, 184; see also Bashford and Mi^rray, Basheord, Murray, and Bowen, Bashford, Murray, and Cr.auer, Basheord, IMurray, and H.^\LAXD, and Bashford, Mutrray, Haaland, and Bowen Neuberg and Caspari, on treatment, 265 Neuberg, Caspari, andLoHE, on treatment, 266 Neve, on the Kangri basket, 6 Newsholme, on cancer houses, 43 ; on heredity and aetiologj^ 201 Nichols, on the production of tumors, 25 NoviNSKY, on the transmissible Ijrmpho-sar- coma of the dog, 227 Oefele, on Egj'ptian medicine, i Oertel, on cancerous degeneration, 12 Ordway, see Tyzzer and Ordway Orth, on sarcoma development, 118 Oswald, on diffusion, 267 Paracelsus, on Galen's hypothesis, 3 Pare, on treatment, 3 Perry, see Chalmers and Perry Petersen, on aetiologj^, 13 ; on retrograde metastasis, 44 Petit, see Borrel and Petit Peytulhe, on heteroplastic transplantation, SO Pfelffer, on a transplantable melano-carci- noma of the mouse, 53; on anaphylaxis, 182 Pfelffer and Finsterer, on anaphylaxis, 182 Pforringer, on aetiologj', 13 Pick, on thyroid a,deno-carcinoma of the trout, 238 Pick and Poll, on the genesis of mouse tumors, 18s ; on thyroid adeno-carcinoma of the trout, 238 Pincussohn, see Michaelis, Fleischm.a.nn, and Pincussohn Plehn, on thyroid adeno-carcinoma of the trput, 237, 238 (footnote), 239; on the zoological distribution of cancer, 184 Plicque, on irritation and aetiologj^ 6 PoDWYSSOZKi, on asymmetrical mitosis, 15 Poll, see Hert'wig and Poll, and Pick and Poll Price-Jones, on the blood of tumor-bearing mice, 96 Pringle, see Cr.amer and Pringle Purvis, on thj^roid adeno-carcinoma of the trout, 236 Quincke, on autoplastic transplantation, 47 Ranzi, on anaphylaxis, 182; see also K1r.\us, R.ANZI, and H. Ehrlich Rautm, on Altmann's granules, 194 Regaud, on parasites and aetiology, 196 Reicher, on treatment, 262 Reinke, on atypical epithelial growth, 26; see also HERXHEiiiER and Reinke Remak, on cell division, 4 Ribbert, on aetiology, 10; on anaplasia, 16; on asymmetrical rnitosis, 15 ; on autoplastic transplantation, 224; on heteroplastic trans- INDEX OF AUTHORS 281 plantation, 52, 224 ; on the production of tumors, 23 Richardson, on autoplastic transplantation, 47 RrcHET, see Heeicourt and Richet Rous, on hypersensibility, 182; on parabiosis, 259; on pregnancy and transplantation, 28; on the specificity of immunity, 147 ; on a trans- missible sarcoma of the fowl, 244 ; on the production of tumors, 29; on tumor growth and nourishment, 93 Rous and Murphy, on a transmissible sarcoma of the fowl, 244 Rous, Murphy, and Tytler, on a transmissible sarcoma of the fowl, 244 Roux and Metchnikoff, on heteroplastic transplantation, 52 RovsiNG, on treatment, 268 Russ, see Beckton and Russ Russell, on extirpation and immunity, 163; on the stroma reaction in immunity, 174; on the immunization of tumor-bearing animals, 151; on the duration of immunity, 155; on antibodies in immunity, 157 ; on sarcoma de- velopment, no, 122; see also Bashford and Russell Sailer, on the inoculability of carcinoma, 57 Sauerbruch and Heyde, on parabiosis, 258 Saul, on parasites and aetiology, 196 Schleiden, on the cellular structure of plant tissues, 3 Schmidt, on pulmonary tumor emboli in man, IIS Schmincke, on atypical epithelial growth, 35 ; see also Wacker and Schmincke ' ScHONE, on athrepsia, 161 ; on immunity with embryo, 14s ; on the specificity of immunity, 147 ; on immunity in tumor-bearing animals, ISO Schwann, on the cellular structure of animal tissues, 3 Scott, on thyroid adeno-carcinoma of the trout, 236 Seligmann, see Shattock, Seligmann, and Dudgeon Senn, on homoplastic transplantation, 43 ; on autoplastic transplantation, 49 Seyberth, on bladder tumors in aniline dye workers, 33 Shattock, on uterine carcinoma in the rabbit, 22 Shattock, Seligmann, and Dudgeon, on preg- nancy and growth, 28 , Smith, on homoplastic transplantation, 42 Smith and Washbourn, on the transmissible lympho-sarcoma of the dog, 228 Snow, on atypical epithelial growth, 32 Spiess, on treatment, 262 Spude, see Eberth and Spltde Stahr, on atypical epithelial growth, 33 ; on the interval after which growth becomes apparent, 77; on the production of tumors, 23, 188; on race and susceptibility, 135; on sarcoma development, 123; on the technic of inoculation, 71 Starling, see Lane-Claypon and Starling Steffenhagen, see Uhlenhuth, Haendel, and Steffenhagen Sticker, on the transmissible lympho-sarcoma of the dog, 228 ; on the premetastatic stage, 152 ; on sarcoma development, 118; on treatment, 23s ; on the zoological distribution of cancer, 184; see also Bergell and Sticker Stockard, on non-encapsulation of the trout thyroid, 240 Stoeber, on atypical epithelial growth, 34 Stoeber and Wacker, on atypical epithelial growth, 34 Stroebe, on asymmetrical mitosis, 16 Takemura, on the deposition of iodine, 267 Thiersch, on aetiology, s, 6; on metastasis, 5 Thorel, on the development of spontaneous tumors during immunity, 128 Thorn, on autoplastic transplantation, 46, 47, 49 V. Tiesenhausen, on the production of tumors, 27 TR.ACY, see Beebe and Tracy TuLPius, on homoplastic transplantation, 40 Tytler, see Rous, Murphy, and Tytler Tyzzer, on the hereditary transmission of im- munity and susceptibility, 157; on heredity and aetiology, 204 ; on race and susceptibility, 136 Tyzzer and Ordway, on the zoological dis- tribution of cancer, 184 Uhlenhuth, Haendel, and Steffenhagen, on athrepsia, 162; on the distribution of im- munity, 154; on dosage, 75 ; on the duration of immunity, 155 ; on hypersusceptibiHty, 181 ; on passive resistance, 157 ; on treatment, 259, 263 Uhlenhuth and Weidanz, on immunity, 147, 149; on pregnancy, and susceptibiHty, 137; on race and susceptibility, 136; on the re- sistance of the cancer cell, 83 ; on the trans- plantation of metastases, 79 ; on treatment, 263 Unger, see Frank and Unger von den Velden, on treatment, 266 Velich, on a transplantable sarcoma of the rat, 55 _ Verse, on aetiology, 13 Vesalius, on Galen's hypothesis, 3 ViDAL, on treatment, 260, 269 Viel-Hautmesnil, on homoplastic transplanta- tion, 41, 50 Virchow, on aetiology, 4; on the analogy be- tween tumor cell and ovum, 1 7 ; on autoplastic 282 INDEX OF AUTHORS transplantation, 48 ; on heteroplastic trans- plantation, SI ; on metastasis, 4 Wacker, see Stoeber and Wacker Wacker and Schmincke, on atypical epithelial growth, 36 Wade, on the transmissible lympho-sarcoma of the dog, 234 Wakasugi, on autoplastic transplantation, 47 Waldeyer, on aetiology, 5, 7 Walker, on treatment, 258, 265 ; see also Farmer, Moore, and Walker, and Moore and Walker Walker and Whittingham, on heterotypical mitosis, 18 Washbourn, see Smith and Washbourn Wassermann, see v. Wassermann, Keysser, and Wassermann V. Wassermann, Keysser, and Wassermann, on treatment, 263 Wehr, on the transmissible lympho-sarcoma of the dog, 227 Weidanz, see Uhlenhuth and Weidanz Weigert, on aetiology, 11 Werner, on Scharlach R, 37 White, on atypical epithelial growth, 35 ; on the transmissible lympho-sarcoma of the dog, 230 White and Loeb, on the transplantation of stationary or receding tumors, 78 Whitehead, on homoplastic transplantation, 41 Whittingham, see Walker and Whittingham WiLKiE, on thyroid adeno-carcinoma of the trout, 236 Williams, on the malignancy of mouse tumors, i8q Wilms, on the production of tumors, 25 WoLFLER, on medicine in India, i WoGLOM, on immunity, 150, 154, 155, 175 WOLPF, on Celsus, i Wyss, on X-ray carcinoma, 33 Yamanouchi, on hypersensibility, 181 Zacutus, on homoplastic transplantation, 39 Zahn, on the production of tumors, 21 INDEX OF SUBJECTS Absorption, spontaneous, 90; connective tissue in, gi, 92, 221, 234; in foreign race, 136; giant cells in, gi, g2 ; heated tumor, 83; hemorrhage in, gi, g2 ; histology, 91, 221, 234; during hypersusceptibility, 178; during lacta- tion, 138; lymphocytes in, 92, 177, 221; mitosis, 234; phagocytes, 92, 221; during pregnancy, 138; re-inoculation after, i3g; of spontaneous tumor, 87, 220, 228, 241, 243; of transplanted tumor, 83, 87, 90, 120, 122, 136, 138, 177, 178, 221, 228, 22g, 232, 244, 246 Absorption of graft in immunity, 175 Acquired resistance, 138, 228, 230, 234, 241, 243, 24s Active resistance, 138 Adaptation, 61, 208, 226 Adrenalin, gi, 260, 262, 263 /Etiology, 195; age, 6, 7, g, 11, i2g, 130, 184, 199, 202, 231, 238; alien cells, 23; anaplasia, 14, 17; asexual phase of development, ig; avidity for food-stuffs, 7, ss, 64, 208; bile, I, 2; cancer cages, 195; cancerous de- generation, 12, 14; conjugation of cells, 17; embryonal rests, 8, 9, 10, 11, 13 ; function, 200 ; heredity, 200, 230, 238, 239 ; heterotypical mitosis, 17, 18; in-breeding, 198; infection', 195, 22g, 230, 23s, 237, 238, 239, 241, 242, 244, 252; insects, 76, 195, ig6; irritation, in- flammation, and trauma, 4, 5, 6, 7, 8, 10, 11, 13, 14, 18, 21, 23, 24, 34, 200, 205, 250; lactation, 200; lipoids, 38; parasites, igs, 207, 238; sen- sitization and stimulation, 25; sex, igg; side chain theory, 208; slumber cells, 18; soil, 11; stimulus, s, II, 13, 17, 18, 25, 34; trauma, 6, 8, 250; twin pregnancy, 18; X-rays, 30, 33 Age, in aetiology, 6, 7, 9, 11, 129, 130, 184, 199, 202, 231, 238; limit of, in mouse, 201; as affecting transplantation, 76, 129, 212, 228, 24s, 255 ; of tumor and differentiation, gg, 102 ; of tumor and metastasis, igo AUergie, 172 Altmann's granules, ig3 Alveolar carcinoma, 105 Ameboid motion of cancer cell, s, 125 Amitosis, sg, 60, 246, 247 Amyloid degeneration, 95 Anaphylaxis, 182 Anaplasia, 14, 16, 103 Angioblasts, chemotaxis of cancer cell for, 109 Angioplastic stroma reaction, 61 Aniline dye workers, tumors of the bladder in, 33 Animals, frequency of tumors among, 183 Antibodies, 63, 125, 155, 159, 163, 165, 166, 171, 182, 257, 259, 267 Appetite in tumor-bearing animals, 94 Arsenic, i, 3, 263 Athrepsia, 159, 208 Atoxyl, 23s, 263 Attraxin, 31 Autoplastic transplantation, 27, 44, 151, 209, 212, 223, 24s Avidity for food-stuffs, 7, g, 2g, 64, gg, 161, 208 Bed bugs, attempted transmission by, 76 Beet, transplantable tumor of, 184 Betel-nut, 6, 10 Bile as an aetiological factor, i, 2 Biological qualities of tumor cells, al ;ernations in, 75 Bladder, tumors of, in aniline dye workers, 33 Blastosis, 38 Blood in tumor-bearing animals, 96, 2 2g, 234 Blood stream, metastasis by, igo, 229, 246 Body, growth during pregnancy, g8 Cachexia, g2 Cancer a deux and cancer a trois, 42 Cancer cages, 195 Cancer cell, alternations in biological qualities, 75; ameboid motion, s, 125; chemotaxis, 61, log, 173, 17s; growth energy, 83; resistance of, 56, 80, 163, 171, 229, 248, 252, 261 ; serum proof, 163, 170 Cancer houses, 42 ; see also "Cancer cages" Cancer, ovarian, production of, 21; and twin . births, 18; zoological distribution of, 184 Carcinoma, alveolar, 105 ; sarcomatodes, 80 ; X-ray, s:^ ; in the young, 7 Cattle, carcinoma inner canthus, ig7 Cell as a tissue unit, 3 Cell genesis and intercellular substance, 4 Chemotaxis, by cancer cell, 61, log, 173, 175 ; by fat stains, 31, 32, 33, 35 Chemotherapy, 263 Chromosomes, in asymmetrical mitosis, 15 Classification of mouse tumors, 213 Clinical course of transplanted tumors, g2 ; of spontaneous tumors, 220 Collagen, 60 283 284 INDEX OF SUBJECTS Comparative growth rate of the malignant cell, 83 _ _ Conditions, of growth, 129, 130, 147 ; of origin, 129, 130 Conjugation of cells, 17 Connective tissue, growth energy of, 6, 7, 10, II, 13; intense reaction in rats, 33; in irradiated tumors, 261; mitosis in, 60; re- moval of from tumor emulsions, 71; in spontaneous absorption, 91, 92, 221, 234 Contact transplantation, 44 Contagion, 39, 195, 230, 242 Cow, hepatic carcinoma in, 207 Cultivation of cells in vitro, 124, 233 Cure, see "Absorption, spontaneous," and "Treatment" Cysts, in mamma, 205 ; production of, 21, 26 Deciduoma, production of, 24 Degeneration, amyloid, 95 ; cancerous, 12, 14 Dermoid cyst, production of, 26 Diet, amount in tumor-bearing animals, 94 ; and metastasis, 93, 134; and recurrence, 134; and susceptibility, 133, 135 ; and tumor growth, 92 Differentiation, age of tumor, 99, s.02 ; cyclical variability of, 108; dual, loi, 104; in embryo and tumor, 9 ; and growth energy, 14, 105, 220; latent, 104; lost, 11, 104; and meta- plasia, 103 ; in metastases, 103 ; in thyroid tumors, 108; in transplanted tissues, 22, 24, 25 _ Distribution, of active resistance, 153 ; of mamma in the mouse, 186 Dog, transmissible lympho-sarcoma, 227 Dosage, and athrepsia, 165, 166, 167, 168, 170; and growth energy, 65 ; and hypersuscepti- bility, 180, 181; and immunity, 75, 140, 141, 14s, 146, 149, 150, 151 ; importance of ac- curate, 65, 69, 73, 75, 86 ; small and large, 67, 72, 88 Dose, minimal tumor-forming, 75, 76, 145 Duration of active resistance, 154 Dye workers, tumors of bladder in, 33 Early stages, see "Stroma reaction" Edematous changes in the stroma, 219 Emaciation as a terminal event, 92 Embryo, growth energy of, 84 ; transplantation into, 253 Embryonal rests, 8, 9, 10, 11, 13 Emulsion, transplantation by, 68, 71, 165 Encapsulation, of trout thyroid, 240; of tumors, 192 Endothelioma, 185, 186, 189, 217, 219 Epithelioma, production of papillary, 23 Epithelium, conjugation of leucocytes with, 17; growth energy of, 6, 7, 10, 11, 13; spindle shaped, 106 Equilibrium between tissues, 6, 7, 10, 11, 12, 13 Ereption, 170 Ferments, in metabolism, 98; treatment with, 23s, 263 Fetal cells, immunity against, 148 ; immunity with, 14s Fetus, nitrogen value of, 99; and tumor com- pared, 98 Fibroblasts, penetration of graft by, 58, 60, 174, 246 Fibroplastic stroma reaction, 61 First appearance of active resistance, 154 Flies, attempted transmission by, 76 Fluctuations in growth energy, 67, 85, 220 Food-stuffs, avidity for, 7, 9, 29, 64, 99, 161, 208; for mouse tumors in the rat, 160; spe- cific, 138, 160, 164; specific, in pregnancy, 138 Fowl, transmissible sarcoma of, 244 Fox, transmission of dog tumor to, 229, 234, 235 Frequency of tumors among animals, 183 Function, and growth energy, 16 ; and aetiology, 200 Gametoid neoplasms, 17 Gastric contents, 97 Generation stages, 15 Gestation, see "Pregnancy" Giant cells in spontaneous absorption, 91, 92 Gland, hibernating, 185 Graft, absorption during immunity, 175; and soil, relative importance of, 71; transplanta- tion by, 68, 71 ; vascularization of, 58, 60, 168, 174, 246, 247 ; see also "Stroma reaction " Growth, of body during pregnancy, 98; cyto- typic and organotypic, 9 ; expansive, 192 ; in- filtrative, 120, 123, 191, 229, 230, 237, 238, 240, 241, 247; interval after which it becomes apparent, 77, 83, 225 ; necessary for immunity, 144, 149; negative phase, 67, 135; of normal tissues during pregnancy, 28; and origin, con- ditions of, 129, 130; positive phase, 67; pro- duction of atypical epithehal, 21, 22, 26, 30; its rapidity in hypersusceptibihty, 178; re- tarded in immunity, 145, 146, 147, 172 ; rhythms of, 88 ; specific conditions for, 147 ; substance, 64; of transplanted tissues, 20; of tumors during pregnancy, 55, 137 ; of tumor, relation of nutrition to, 92 Growth energy, in athrepsia, 161, 164, 169, 171 ; in autoplastic transplantation, 225 ; and avid- ity, 7, 208; of the cancer cell, 83; compara- tive in same mouse, 225; decreased, SS', and differentiation, 14, 105, 220; and dosage, 65; effect of cold upon, 80; effect of heat upon, 80; effect of X-rays upon, 261; of embryo, 84; of embryonic rests, 8, 9, 10, 11; of epithelium and connective tissue, 6, 7, 10, II, 13; fluctuations in, 67, 85, 220; and function, 16; in grafts from stationary or INDEX OF SUBJECTS 28s receding tumors, 78; and histology, 105, 220; inhibition of, 68 ; in lympho-sarcoma of the dog, 230; of ovum, 17, 84; in recurrences, 162, 164, 209; in sarcoma development, 120, 122, 123, 124; of spontaneous tumor, 220; stimu- lation of, 68 ; ill transmissible sarcoma of the fowl, 246; transplantability and virulence, 169 Hare, tumors of, 155, 172 Health as affecting transplantation, 137, 167 Hemolysis, 38 Hemorrhage, in irradiated tumors, 262 ; in spontaneous absorption, gi, 92 ; as a terminal event, 92 Hemorrhagic, changes in the stroma, 219; tu- mours, stroma of, 109 ; tumors, transplantation of, 72, 212 Heredity, in aetiology, 200, 230, 238, 239 ; trans- mission of immunity and susceptibility by, 55, 157 Heteroplastic transplantation, 37, 50, 229, 234, 235. 25s Hibernating gland, 185 Histological, expression of growth energj', 105, 220; variations during transplantation, 99 Histology, and immunity, 107 ; of irradiated tumors, 261; and malignancy, 104, 107, 217, 238; of mammary tumors, 186, 213; of receding tumors, 91, 221, 234; of spontaneous tumors, 184, 213 ; and stroma reaction, 109 Historical review, i Homoplastic transplantation, 27, 39, 53 Hormones, 259 Horn core, 6 . Host and tumor, relation between, 223 Hybrids, susceptibility and immunity, 158, 240 .Hyperemia, 8, 10 Hypersensibility, 181 Hypersusceptibility, 178; and dosage, 180, 181 ; and immunity, 179, 181; local, 173; optimum of, 181 ; produced by normal tissues, 146, 151, 156, 157, 179, 180, 181; produced by tumor, 151, 178, 179, 180, 181 ; rapidity of growth in, 178; specificity of, 180; spon- taneous absorption in, 178 ; and time interval, 178, 180, 181 Hypertrophy, nodular, 205 ; of organs in tumor- bearing animals, 93 Immune zone, 230 Immunity, 128; abrogation by operation, 162; absorption of graft in, 175 ; acquired, 138, 228, 230, 234, 241, 243, 24s; active, 138; and age, 76, 129, 228, 245, 255; and allergic, 172; athreptic, 159, 208; with autologous tissue, 150; with autolyzed tissue, 137, 149, 150; to autoplastic transplantation, 151 ; and chemo- taxis, 173, 175; common and specific, 142; concomitant, 67, 70, 74, 87, 162, 165, 167; de- velopment of spontaneous tumors during, 128 ; distribution of, 153; and dosage, 75, 140, 141, 14s, 146, 149, 150, 151; duration of, 154; against fetal cells, 148; with fetal cells, 145; first appearance of, 154; growth necessary for, 144, 149; growth retarded m, 145, 146, 147, 172; as affected by health, 137, 167; hereditary transmission of, 55, 157; and his- tology, 107; in hybrids, 158, 240; and hypersusceptibility, 179, 181 ; with intact homologous normal cells, 148; with intact homologous tumor cells, 143 ; leucocyte in, 236; lymphocyte in, 176, 234, 247, 248: lymphoid cells in, 173; macrophage reaction in, 173; mitosis in, 174; natural, 128, 228, 229, 241, 247; nature of, 159; with normal tissue, 144; and operation, 162; and para- biosis, 258; survival of parenchyma in, 174, 17s; partial, 107, 172; passive, 155; plasma cell in, 173, 177; and pregnancy, 137; pre- metastatic stage of, 152, 230; and race, 55, 76, 79, 130, 231, 245, 255 ; resistance of timior cell to, 163 ; and selection, 141 ; and sex, 137 ; specificity of, 122, 141, 147; stroma reaction in, 168, 174, 247; with tumor, 122, 138; in tumor-bearing animals, 150, 152, 161; tumor mixtures in, 133; X-stuff in, 160 In-breeding, 198 Incitant, specific, 160 IndividuaHty, 212, 225 Infection, in astiology, 195, 229, 230, 235, 237, 238, 239, 241, 242, 244, 252; as a terminal event, 92; or transplantation? 58, 229, 230, 231, 232, 234 Infectiveness, variations in, 89 Infectivity, 195 Infiltrative growth, 120, 123, 191, 229, 230, 237, 238, 240, 241, 247 Inflammation, in etiology, 7, 8, 10, 13, 24, 205; chronic, of mamma, 205 Inheritance, Mendelian, and susceptibility, 158 Inoculation, site, 76 ; of stationary or receding tumors, 78 ; of tumor mixtures, 79 ; see also " Transplantation " Insects in aetiology, 76, 195, 196 Intercellular substance and cell genesis, 4 Interval after which growth becomes apparent, 77, 83, 225 Intervals between successive inoculations, 71 Inunction, transplantation by, 75 Iodine, 242, 243 Irritation in aetiology, 4, 5, 6, 7, 10, 18, 21, 23 Kangri basket, 6, 10 Karyokinesis, see " Mitosis " Kataplasia, 16 Keloid, production of, 29 Keratin, 83, 100, 107 Lactation, as an aetiological factor, 200; re- gression of tumors during, 138 286 INDEX OF SUBJECTS Latent period, effect of heat upon, 83 Leeches, attempted transmission by, 76 Leucocyte, conjugation with epithelium, 17; in immunity, 236 Lipoids, significance of, 38, 266 Lymph nodes, enlargement recognized by Celsus, 2 Lymphocyte, in immunity, 176, 234, 247, 248; in spontaneous absorption, 92, 177, 221 Lymphoid cells in immunity, 173 Lympho-sarcoma of the dog, 227 Lymph stream, metastasis by, 53, 57, 190, 192, 227, 229, 246 Macrophage reaction in immunity, 173 Malignancy, 189; and histology, 104, 107, 217, 238; and metastasis, 161; and mitosis, 215, 218; of mouse tumors, 189 Malignin, 260 Mamma, chronic inflammation of, 205 ; cyst formation in, 205 ; distribution in the mouse, 186; histology of tumors, 186, 213; nodular hypertrophy of, 205 ; normal, 205 ; in old mice, 207 ; origin of tumors in, 184, 193, 200; sclerosis of, 205 Marrow in tumor-bearing animals, 96 Membrane in relation to therapeutics, 267 Mendelian inheritance and susceptibility, 158 Mercury, 243 Metabolism in tumor-bearing animals, 97 Metaplasia and differentiation, 103 Metastases, transplantation of, 79 Metastasis, 189; age of tumor, 190; and athrep- sia, 161; by blood stream, 190, 229, 246; and diet, 93, 134 ; differentiation in, 103 ; in the dog, 227, 228, 229, 230; early stages in, 114, 250; in the eyelid, 244 ; in the fowl, 244, 246, 250 ; in the heart, 244, 246 ; immunity against, 134, 151 ; in the intestinal wall, 244 ; in the kidney, 190, 244; in the liver, igo, 191, 244, 246; in the lungs, 92, 114, i8g, 192, 222, 244; by lymph stream, 53, 57, 190, 192, 227, 229, 246; and malignancy, 161 ; in the mediastinum, 190; metastatic stage, 152, 230; mode of origin, 4, 5, 114, 250; in the ovary, 190; on the peritoneimi, 190; premetastatic stage, 152, 230 ; prevention of, 93, 134, 151, 258 ; ret- rograde, 44; in the retroperitoneal tissue, 190 ; in sarcoma development, 112, 114, 120, 123; in the skin, 244; in the spleen, 190, 227, 246; in spontaneous tumors, 53, 57, 189, 192, 193, 226; stroma reaction in, 114, 250; in trans- plantable tumors, 190, 192 ; in the trout, 238, 239, 240, 241, 242 Metastatic stage, 152, 230 Mincing machine, 69 Mitosis, asymmetrical, 15, 68; in connective tis- sue surrounding graft, 60; in young grafts, 60, 174, 233, 247; in the fowl, 247; hetero- typical, 17, 18; hyperchromatic, 14, 217; hy- pochromatic, 14; in immunity, 174; and ma- lignancy, 215, 2i8; in spontaneous absorption, 234; stimulation of, 26; in vitro, 125, 126, 233 Mixed tumors, 109; purification of, 82, 117, 120; see also "Tumor mixtures" and "Sarcoma development" Mixtures, see "Tumor mixtures " Morphology, see "Histology " Mouse, age limit in, 201 ; distribution of mamma in, 186; tumors of, 184; carcinoma, growth of, in rat, 159 Multiple, spontaneous tumors, 188, 199, 200; transplantation, 72 ; transplanted tumors, 72, 150, 152, 161, 208 Natural resistance, 128, 228, 229, 241, 247 Nature of the resistant state, 159 Needle for inoculation, 70 (footnote) Nipple, retraction of, 2 Nitrogen balance, 98 Nitrogen value of fetus, 99 Normal tissues, growth of, 21 ; growth during pregnancy, 28 ; hypersusceptibility produced by, 146, 151, 156, 157, 179, 180, 181; resist- ance produced by, 144 Nucleus, discovery of, 3 Nutriceptors, 64, 208 Nutrition and tumor growth, 92 Nutritive capacity, limit of, 167 Operation, and immunity. 162; recurrence after, 189, 192, 222, 229; re-inoculation after, 161; results of, 221 ; transplantation during, 47, 226 Organs, transplantation into, 77, 154; hyper- trophy of, in tumor-bearing animals, 93 . Origin, conditions of, 1 29, 130 ; of metastases, 4, s, 114, 250; multicentric, 205, 217; site of, 184, 193, 200 Osteo-chondro-sarcoma, transplantable, 100 Ovum, growth energy of, 17, 84 Pan-immunity, 141 Papilloma, production of, 23 Parabiosis, 258 Parasites in setiology, 19s, 207 Parenchyma, histological variations during trans- plantation, 99 ; secondary changes in, 215; survival in immune animals, 174, 17s; sur- vival after transplantation, 58, 59, 168, 174; variations during transplantation, 99 Partial immunity, 107, 172 Passage, rapid, 71 Passive immunity, 155 Phagocytes, in spontaneous absorption, 92, 221 ; in vitro, 125 Pin prick, transplantation by, 76 Plasma cell in immunity, 173, 177 Pluri-immunity, 142 Pregnancy, and athrepsia, 172; growth of body during, 98; growth of normal tissues during. INDEX or SUBJECTS 287 28; growth of tumors during, 55, 137; as affecting transplantation, 137 ; specific food- stuffs in, 1 38; spontaneous absorption during, 138; weight of organs during, 93, 94 Premetastatic stage, 152, 230 Prevention, of metastasis, 93, 134, 151, 258; of recurrence, i, 134, 151, 164, 192 Proliferative energy, 169 ; see also " Growth energy" Purification of mixed tumors, 82, 117, 120 Pabbit, uterine carcinoma in, 22 Kace as affecting transplantation, 55, 76, 79, 130, 231, 24s, 25s Radium, 83, 91, 149, 181, 261 Rat, food-stuffs for mouse tumors in, 160 ; growth of mouse carcinoma in, 159 ; intense connective tissue reaction in, :is ; tumors of, 1S8 Receptors, 64, 171, 208 Recession, see "Absorption, spontaneous," and "Treatment" Recovery, see "Absorption, spontaneous" and "Treatment" Recurrence, and diet, 134 ; growth energ>' in, 162, 164, 209; prevention of, i, 134, 151, 164, 192; after operation, 189, 192, 222, 229; in relation to re-inoculation, 164; after sarcoma develop- ment, 113, 123 ; increase in weight during, 222 Refractory condition, see "Immunity" Regression, see "Absorption, spontaneous" and "Treatment" Reinoculation, after operation, 161 ; in relation to recurrence, 164; after spontaneous ab- sorption, 139; of tumor-bearing animads, 150, 152, 161, 164, 166, 167, 168, 212, 223, 230, 232; and vdrulence, 140, 143, 171 Relation between tumor and host, 223 Resistance, 128 ; see also "Immunity" Resistance of cancer cell, 56, 80, 163, 171, 229, 248, 252, 261 Retraction of nipple, 2 Retrograde metastasis, 44 Rhythms of growth, 88 Salvarsan, 235 Sarcoma, development, 79, 106, 109; production of, 23, 30; stroma reaction in, 59, 116 ; X-ray, 30 Scharlach R, 27, 30 Sclerosis of mamma, 205 Selection and immunity, 141 Selenium, 263 Senility, see "Age" Sensitization, 25 Serum-proof cancer cells, 163, 170 Sex, in etiology, 199; as affecting transplanta- tion, 137 Side chain theory in jetiology, 208 Size attained by tumors, 57, 62, 192 Slumber cell hypothesis, 18 Soil, in setiologA', 11; and graft, relative im- portance of, 71; and sarcoma development, no, 116 Spontaneous, absorption, 90; tumors, 183 Stimulation, of growth power, 68; of mitosis, 26 Stimulus in aetiologj', 5, 11, 13, 17, 18, 25, 34 Stroma, death following transplantation, 58, 59, 168, 174; edematous changes in, 219; hem- orrhagic changes in, 219; in hemorrhagic tumors, 109 ; histological changes during transplantation, log ; relation to sarcoma development, in; sarcoma development in, 109 ; secondary changes in, 215 Stroma reaction, 58 ; amitosis in, 60, 246, 247 ; angioplastic and fibroplastic, 61 ; in athrepsia, 168; in the dog, 229, 231, 232, 234; in the fowl, 246 ; and histological structure, 109 ; in immunity, 16S, 174, 247; in metastases, 114, 250; mitosis in, 60, 174, 233, 247; in the mouse, 58, 168, 174, 176; in the rat, 59, 175; in sarcoma, 59, 116; specificity of, 60 Structure of origin in the mouse, 184 193, 200 Structure, see "Histology" Sudan III, 30 Summarj', 270 Susceptibility, and age, 76, 129, 212, 228, 245, 255 ; and diet, 133, 135 ; and health, 137, 167 ; hereditary' transmission of, 55, 157; of hybrids, 158, 240; increased, 165, 166, 168, 178, 223; and pregnancy, 137 ; and race, 55, 76, 79, 130, 231, 24s, 255; and sex, 137; in tumor-bear- ing animals, 165, 166, 168, 223; for txmior mixtures, 79, 133 S^'ringe for inoculation, 69 Technic of inoculation, 68 Tellurium, 263 Teratoma, production of, 25, 26 Therapeutics, see "Absorption, spontaneous," and "Treatment" Thyroid, adeno-carcinoma of the trout, 236; gland, non-encapsulation in the trout, 240 Toxins, in treatment, 235 ; in metaboUsm, 98 Transfer, of tumors from one person to another, 39 ; of human tumors to animals, 50 Transmissibility, earher observations on, 39 Transmissible sarcoma of the fowl, 244 Transplantabihty, 62 ; diminished, 78 (see also " Resistance of cancer cell ") ; growth energy, and \arulence, 169; of thyroid adeno-carcinoma of the trout, 240, 242 Transplantation, as affected by age, 76, 129, 212, 228, 24s, 255; autoplastic, 27, 44, 151, 209, 212, 223, 245; and avidity, 208; into bearer, 44 ; behaNdor of tumor cell during, loi ; by contact, 44; by contagion, 39, 195, 230, 242 ; into embryo, 253 ; emulsion and graft methods, 68, 71 ; first appearance of tumor after, 77, S3, 225 ; as affected by health, I37> 167; of hemorrhagic tumors, 72, 212; heteroplastic, 37, 50, 229, 234, 235, 255; his- INDEX OF SUBJECTS tological variations during, 99; homoplastic, 27, 39, 53; or infection? 58, 229, 230, 231, 232, 234; of injured cells, 80 (see also "Resistance of cancer cell") ; through insects, etc., 76, 195, 196; by inunction, 75; of me- tastases, 79; multiple, 72; needle for, 70 (footnote); during operation, 47, 226; optimum conditions for, 70, 88; into organs, 77, 154; sur\'ival of parenchyma after, 58, 59, 168, 174; by pin prick, 76; as affected by pregnancy, 137 ; as affected by race, 55, 76, 79, 130, 231, 245, 255 ; as affected by sex, 137 ; site of election for, 76 ; within the same species, 27, 39, S3 ; of spontaneous tumors, 65, 71, 72, 75, 212, 223, 226; of stationary or receding tu- mors, 78, 225 ; death of stroma after, 58, 59, 168, 174; increased susceptibihty to, 165, 166, 168, 178, 223; syringe for, 69; technic of , 68; as affected by trauma, 249, 251 ; into tumor- bearing animals, 150, 152, 161, 164, 166, 167, 168, 212, 223, 230, 232; of tmnor mixtures, 79; of warts, 44; zigzag, 159 Transplantations, inter\^als between successive, 71 Transplanted tissues, differentiation in, 22, 24, 25 growth of, 20; growth diudng pregnancj', 28 Transplanted tumors, 58 Trauma, in relation to aetiology, 6, 8, 250; in relation to transplantation, 249, 251 Treatment, i, 2, 3, 19, 230, 235, 236, 256 Trout, th>Toid adeno-carcinoma in, 236; non- encapsulation of thjToid gland, 240 Tumor, hrst appearance after transplantation, 77, 83, 225; cHnical course of, 92, 220; and fetus compared, 98 ; and host, relation be- tween, 223 ; hypersusceptibihty produced by, 151, 178, 179, 180, 181; immunity produced by, 122, 138; mixtures and athrepsia, 171; mixtures, susceptibility and immunity for, 79, 133; mixtures, transplantation of, 79; reced- ing, transplantation of, 78; size attained by, 57, 62, 192 ; the spontaneous, 183 ; sponta- neous, development during immunity, 128; spontaneous, distinguished from transplant- able, 97, 212; spontaneous, fluctuations in growth energ>', 87 ; spontaneous, frequency of occurrence, 183, 199; spontaneous, growth energy, 220; spontaneous, histology, 184, 213 ; spontaneous, metastasis, 53, 57, 189, 192, 193, 226; spontaneous, multiple, 188, 199, 200; spwntaneous, recurrence, 189, 222 ; spontane- ous, sarcoma development in, 113; sponta- neous, spontaneous absorption of, 87, 220, 228, 241, 243; spontaneous, temporary arrest, 220; spontaneous, transplantation of, 65, 71, 72, 75, 212, 223,226; stationarj^ transplantation of, 78, 225 ; transplantable, distinguished from spontaneous, 97, 212 ; transplantable, metas- tasis, 190, 192 ; transplantable, spontaneous absorption of, 83, 87, 90, 120, 122, 136, 138, 177, 178, 221, 228, 229, 232, 244, 246; the transplanted, 58 ; zone, 230 Tumors, attempts to produce, 20; of the dog, 227; of the fowl, 244; of a nature still un- decided, 227; of bladder in aniline dye workers, 33 ; growth energy in stationarj' or re- ceding, 78; of the hare, 155, 172; mixed, 109; of the mouse, 184 ; origin of mammary, 184, 193, 200; of the rabbit, 22; of the rat, 188; transplantable, multiple, 72, 150, 152, 161, 208 ; of the trout, 236 Twin births in relation to cancer, 18 Variations, in infectiveness, 89 ; in parenchyma during transplantation, 99 Vascularization of graft, 58, 60, 168, 174, 246, 247 Virulence, and adaptation, 61 ; and athrepsia, 161, 164, 165, 166, 167, 169, 170, 171; in- creased, 79, 210; maximal, 170; and re- inoculation, 140, 141, 143, 171 ; and sarcoma development, 112, 117, 118, 120, 124; trans- plantability, and growth energy, 169 Warts, transplantation of, 44 Weight, increase during recurrence, 222; of organs in tumor-bearing animals, 93 ; of timior- bearing animals, 94, 97, 221, 222 X-ray, carcinoma, 33 ; sarcoma, 30 X-rays, 30, 35, 260, 261 X-stuff, 160 Yoimg, carcinoma in the, 7 Zigzag inoculation, 159 Zoological distribution of cancer, 184 RC26i Woglom 1913 • •• The study of experinien- c.l tal cancer.