COLUMBIA LIBRARIES OFFSITF HEALTH SClCr J*, i -, -•■. ■■■ HX00035874 il !il ;t, ■ I '.I '■ '■ i : ' 'I'll Digitized by the Internet Archive in 2010 with funding from Open Knowledge Commons http://www.archive.org/details/treatiseontumorsOOhert TRI-ATISl- ()\ n MORS BY ARTHUR E. HERTZLER, M.D., Ph.D. ASSOCIATE PROFESSOR OF SURGERY IN THE UNIVERSITY OF KANSAS; FORMERLY SUCCESSIVELY PROFESSOR OF HISTOLOGY, P.'VTHOLOGY, EXPERIMENTAL SURGERY, AND GYNECOLOGY IN THE UNIVERSITY MEDICAL COLLEGE, KANSAS CITY, MISSOURI; ATTENDING SURGEON TO THE HALSTEAD HOSPITAL, HALSTE.^D, KANS.^S, AND CONSULTING SURGEON TO THE SWEDISH HOSPITAL, KANSAS CITY, MISSOURI, AND ST. MARGARET'S HOSPITAL, KANSAS CITY, KANSAS; MEMBER OF THE WESTERN SUR- GICAL ASSOCIATION, THE AMERICAN ACADEMY OF MEDICINE, THE AMERICAN ANATOMI- CAL SOCIETY, ETC. ILLUSTRATED WITH 538 ENGRAVINGS AND 8 PLATES LEA & FEBIGER PHILADELPHIA AND X E W YORK I 9 I 2 ^%s H ^ 'J- Entered according to the Act of Congress, in the year 1912, by LEA & FEBIGER in the Office of the Librarian of Congress. All rights reserved. TO MY WIFE P R K F A C 1{ 1 HIS treatise embodies the experience ot many years jiained in the operating room, the observations of laboratory stud\', and the informa- tion gathered in the library, combined in such proportions as my duties in teaching students and as a surgical consultant have shown to be most advisable. The aim of the book is to give students and practitioners a guide to the proper recognition of tumors. Available books dealing with this subject are either broadly clinical or entirely scientific. A proper comprehension of an\' tumor demands the application of both the scientific viewpoint and clinical observation. 1 he aim has been to suppl\' this need. For this purpose it was thought best to place the chief emphasis upon the practical aspects of the subject and to include as well sufficient of the more abstract considerations to make the prac- tical side broadly intelligible. In view of the essentialh' practical character ot the work the vast field of experimental oncology has been entirely omitted, because, in the first place, the results, promising as they are, have not }et attained a state of practical utility, and would but confuse the student and prac- titioner; and in the second place, those who are engaged in tumor research necessaril\' work in laboratories where experienced personal guidance is available. Citations of literature have been incorporated particularly in those chapters where existent opinion is still divergent. These are presented that the student may follow the subject and as an apolog>- for the unqualified presentation of the author's viewpoints. Papers, with bibliographies, most likely to be accessible to the reader have received preference. Inasmuch as nearly one-fourth ot all surgical literature has to do with tumors, a careful reading of all this vast material is impossible, hence many valuable papers must have been overlooked. Of those read, it is a matter of personal judgment as to which shall be deemed the more valuable. In the preparation of such a work obligations of man\- kinds are incurred by the author. My first debt is to my wife, who through man}- years has kept together the data that have gone to make up the book. Dr. Edward T. Gibson and Dr. William J. Walker revised the manu- vi PREFACE script and corrected the proof. To them I express my appreciation. I feel a keen sense of pleasure at the readiness with which professional friends have placed material at my disposal. To Dr. J. W. Perkins I am under particular obligations for the kind manner in which he offered his vast clinical material and his large collection of photographs. So unstinted was his generosity that his material has become insepar- ably a part of my own. Dr. Frank J. Hall likewise made me a welcome visitor in his laboratory, where I secured much information and many specimens. Many others, the mere mention of whose names would be equivalent to presenting a roster of my professional friends, were equally kind. To all of these I extend my appreciation. Thomas S. Jones and John D. Bigger made nearly all the drawings. The reader will himself be able to estimate my great obligation to them. The excellence of their work is the best expression of the personal interest they exhibited in mine. To my publishers likewise I take pleasure in acknowledging m}^ gratitude for the careful attention bestowed upon every detail of the work. Because of their efforts, seeing the work through the press has been a pleasure rather than a task. A. E. H. Kansas City, Mo., 1912. C () N T U N T S p A R r I THE GENERAL HIOLOG\' OF ILMORS Conception and Definition '7 PART II THE SPECIAL PATHOLOGY OF TUMORS CHAPTER 1 Osteomas 57 CHAPTER II Chondromas and Chordomas 6o CHAPTER III Myxomas 64 CHAPTER IV Lipomas 67 CHAPTER V Myomas 69 CHAPTER VI FiBROMAS 73 CHAPTER VII Gliomas 83 viii CONTENTS CHAPTER VIII Neuromas 88 CHAPTER IX Sarcomas 9^ CHAPTER X Multiple Myelomas . 125 CHAPTER XI Angiomas 127 CHAPTER XII Endotheliomas — Cholesteatomas i3S CHAPTER XIII Adenomas • • • ■ ^45 CHAPTER XIV Papillomas ^5° CHAPTER XV Carcinoma ^59 CHAPTER XVI Cysts 203 CHAPTER XVII Embryoid Tumors ■ 207 CHAPTER XVIIl Granulomas 210 (yj.\n:.\rs ix 1V\ k I I I I ki;c;i()\\i. roNsiDi.k \Ti()\ ()!• iiMok.^ CM \l' I \:\< ,\ IX Biology ok 1 imors and (iimkai. L'i.ink \i. C"t)Nsii)i;KATif)Ns 219 CIIAI'IK K XX TiMORs OK riiK Cranilm 231 CIIAI'TK R XXI I IMORS OF rUE MraIN and SlMNAL CoRD 2^52 CHAPTER XXU Tumors of the Orbit and Lacrymal Glands 267 CHAPTER XXIII Tlmors of the Face 276 CHAPTER XXIV Tlmors of the Nasal Fossa and Pharynx ^09 CHAPTER XXV Tlmors of the Mouth and Tongue -5^5 CHAPTER XXVI Tumors of the Jaws, Gums, Teeth, Periosteum, and Bone . . ^^3 CHAPTER XX\II Tumors of the Parotid and Submaxillary Glands 345 CHAPTER XX\III Tumors of the Neck, Includinc: Tumors of the Thy'roid and Parathyroid Glands, and Carotid Tumors 354 CHAPTER XXIX Tumors of the Chest, Mediastinum, Heart, and Lungs 391 X cox TEXTS CHAPTER XXX Tumors of the Mammary Gland 398 CHAPTER XXXI Tumors of the Abdominal Wall and Umbilicus 439 CHAPTER XXXII Tumors of the Back 446 CHAPTER XXXIII Tumors of the Liver and Gall-bladder 452 CHAPTER XXXIV Tumors of the Pancreas and Retroperitoneum 465 CHAPTER XXXV Tumors of the Digestive Tract 470 CHAPTER XXXVI Tumors of the Urinary Organs 508 CHAPTER XXXVIl Myoma of the Uterus (Fibromas of the Uterus, Fibroids, Fibromyomas of the Uterus), Adenomyoma of the Uterus, Myoma of the Round Ligament, Tumors of the Round Ligament. Pregnancy in Myoma . . 536 CHAPTER XXXVIII Carcinoma of the Uterus 567 CHAPTER XXXIX Syncytioma Malignum (Carcinoma Svncytiale, Deciduoma, Chorio- etithelioma) 591 CHAPTER XL Tumors (Cystic and Solid) of the Ovary, Tumors of the Fallopian Tubes 596 <:(j.\'ij-:.\rs xi CM API Ik \i.i TlMORS (H nil \'l LVA AND XacINA (,26 ClIAl'IKR Xi.ll I'lMDRs 01 nil Mali Cii-MTAL Orcans . . ^^^^ CM \I'TI k XI.I 1 i Tumors of thl L iter KxiRLMrnts (,-(, CHAPTER XLI\' Tl^mors of THt Lower Extremities 6Ji- T U M () R S PART 1 THE GENERAL BIOLOCrV OK TUMORS CONCEPTION AND DEFINITION \'iRCHo\v,' in his great work on tumors, says that if one's life depended upon it one could not define a tumor; but as Ribbert justh' remarks, the difficulty lies not so much in the inabilit)- to define tumors as in the mability to conceive of their nature. W hen the term was first adopted into medical literature it was sufficiently' comprehensive and self- explanator}', in that it expressed the sum total of all knowledge of tumors, viz., the fact of the existence of an enlargement or swelling. As knowledge extended it w^as noted that tumors var}- in life histor\' and in effect upon the host; that some tend to destro\' life while others do not. The real difficult}', however, began with the microscopic stud\' of the structure of tumors. We date all modern conception of the nature of tumors from the enunciation of Virchow's theory, that all tissue comes from preexistent cells. This led to the discovery that not all tumors are due to proliferation of cells. Those which are tumors in form but not in structure were designated false tumors. The true tumors were called neoplasms, to indicate that the}' were the result of tissue prolif- eration. Further stud}' revealed that the neoplasms had to be limited yet further b}' the exclusion of two groups: first, that which comprises the local multiplication of cells otherwise normal, the hypertrophic processes and the like; and second, the granulomas, which resemble the true tumors in structure, but are the result of specific causes. At the present time certain tumors cannot be placed definitel}' among either the true or false tumors. To this group belong notabl}' certain affections of the lymph glands. Since we are unable to limit our conception of neoplasms or true tumors, a comprehensive definition even for this relativel}' narrow group is ' Die Krankhaften Geschwiilste. Hirschwald, Berlin, 1863, vol. i. 18 THE GENERAL BIOLOGY OF TUMORS out of the question. At present any attempt to confine our study to the true tumors will defeat its own purpose, for obviously we can approach a conception of the neoplasms onl}^ by possessing a clear notion of these false tumors, which for one reason or another have been excluded from the dignified group designated as neoplasms or true tumors. Theo- retically, most pathologists now agree that an autonomous growth of ceils characterizes true tumors, and that the determmation of the autono- mous character of any given group of cells usually implies a comparison with cell proliferations not autonomous, such as hyperplastic and reactive processes involving those regions from which the tumors spring. It would seem, therefore, that any attempt to limit the discussion of tumors to the neoplasms is a serious handicap to a clear presentation of the subject, and that Virchow was correct in grouping together all tumors both true and false. From the clinical point of view this is the only possible method of approaching a proper grasp of the subject. The clinician must parallel the development of our knowledge of tumors and in a sense repeat the oncological history, just as a comprehensive under- standing of anatom}' requires a knowledge of embrA^ology. The first question that must be asked in approaching a tumor problem is, "Is it a swelling, or is it an enlargement of an organ, or an inflammation, or a congestion?" Even this question may be difficult to answer. If it is a tumor, is it a true or a false tumor? Only after these questions have been answered is one ready to approach the problem. Frequently, the greatest difficulty is experienced in answering these fundamental questions, for often if one can be certain that the enlargement is a true tumor the problem is simple. As an example ma}' be cited the tumors of the testicle. An enlargement may be due to a hydrocele, which is an exudate; to orchitis, which is an inflammatory process; to syphilis or tuberculosis, which are granulomas; or to a teratoma, which is a devel- opmental anomaly. Finally, the enlargement ma}' be due to a true tumor, a sarcoma, or a carcinoma. The unnatural limitation of the dis- cussion to the true tumors is due in large part to the disposition of pathologists in the past to limit the discussion to dead tissue, disregard- ing in great measure the living manifestations of tumors and their relations to other tissue changes. In the following discussion the broad view of Virchow will be followed and all tumors will be included which might reasonably be brought into question when considering neoplasms, whether it be by clinical or laboratory study. The laboratory student of oncology has as many difficult border-line problems as has the clinician; and, indeed, he must often ask of the clinician the interpretation of the pictures revealed in his slides. No matter how skilled the microscopist, the nature of many tumors can be determined only by their historj^ and subsequent course. cL./ss/j /(:.ir/(j.\ or timors \\\ Not inhi'iiiKiit l\ , intl;mim;if()i\ oi t;i ;iiHil(mi;i i r:ipicll\ t;ir;il uluri such reaction is ahsciu. This is onl)' very hroadi) riiif, toi ruiiiois possissid ot nuicli stroma ma\ metastasize earl\' and recur with pii sisteiice. I he character as well as tht- amount ot con- nective tissue seems to be a tactor. On the other hand, some tumors j^rovv so rapidly that such reaction does not manifest itself, possibly because the tissue is overwhelmed by toxic products. Ibis is similar to what takes place in severe infections where in place of the leulco- c\tosis ordinariI\' expected there is a definite leukopenia. Bloodvessels. I umors have a complete vascular circulation, as was first pointed out by J. Miiller. The arteries and veins with their con- necting capillaries ma\' be demonstrated b\' injection. The sizes of the vessels vary more than in the normal. F^ither set of vessels ma\' predominate throughout the entire tumor or in special parts. T he bloodvessels within a tumor ma\' be either native to the region or newly formed. The first kind are found in the infiltrating tumors only. The expansile tumors have newly formed vessels only, and, therefore, the larger the tumor the farther removed are the vessels from their base of supply. For this reason the centres of expansile tumors are poorly nourished and are frequently' subject to degeneration from lack of nutrition. The vessels are formed by budding, and their number depends upon local conditions, largeh' but not entirely upon the nature of the tissue in which the tumor occurs. The newly formed vessels, like the tumor cells, depart somewhat from the normal. They do not possess the coats of normal vessels, but are formed of an endothelial lining, and at most a fibrous-tissue layer. They vary in size greatly and somewhat abruptly; they may form sinuses which when connected with large vessels ma\' give the sense of pulsation. In some of the more rapidly growing tumors, notabl\- sarcoma, the vessel walls within the tumor, both native and newh' formed, are subject to the various degenerations common to vessels elsewhere, which may lead to rupture or thrombosis. Nerves. • — Nerves have not been satisfactorily demonstrated as grow- ing in expansile tumors, a fact which their painless clinical course, both in growth and in degeneration, confirms. Infiltrating tumors may develop about and include the nerves native to the parts affected, a fact which is attested by the severe pain often attending them. Lymphatics. — Lymph channels have not been demonstrated within expansile tumors. The frequent formation of metastases by these channels make their existence probable in infiltrating tumors, but most pathologists deny the new formation of lymphatics. There is much difference of opinion on this point, and a reinvestigation ot the subject is much needed. ^ ^'oung. Jour. Exper. Med., 1897, ii, i. 24 THE GENERAL BIOLOGY OF TUMORS GROWTH OF TUMORS Benign tumors, because of their slow growth and their resemblance to the normal, are unsuited for the study of the details of growth. There- fore the growth of tumors has been studied chiefly in small epitheliomas or in the advancing border of further developed carcinomas of the same type. When it was first demonstrated by Thiersch^ and Waldeyer^ that all carcinomas develop from epithelium, it was assumed that the neighboring normal cells become carcinomatous by contact with the malignant cells. This was called growth by apposition. Some pathol- ogists believe that new independent foci of development are formed in the neighborhood of the original tumors, and that these foci coalesce (Hansemann, Lubarsch). This has been called multicentric growth. Evidence has accumulated in recent years which has gone far to sub- stantiate these earlier observations. In many carcinomas the cells adjacent to cell nests show alteration in structure. This is seen par- ticularly well in intestinal growths in which the changes in the glands gradually shade off into the normal, as if some stimulus were causing successive glands to undergo abnormal proliferation, the changes being the less marked the further removed the glands are from the source of stimulation. The same is frequently seen in the skin epitheliomas. In transplanted epithelial tumors in animals the connective tissue has been stimulated to sarcoma formation, showing that the power to grow ma}^ be transmitted to other kinds of tissue. Still more con- clusive are the experiments of Borrel and Lewin in which epidermis was stimulated to the formation of squamous epithelioma by the trans- plantation of a glandular tumor beneath it, showing that proliferating stimulus may be conveyed by cells possessed of the power of unlimited growth. In both these instances there could be no confusion between the normal and the tumor cells. The proliferation excited by Scharlach R and Sudan III likewise show clearly that epithelial cells can be made to proliferate and invade surrounding tissue by the action of certain extrinsic stimuli. It may be regarded as proved, therefore, that nor- mal epithelial cells under certain conditions may be made to proliferate by close contact with malignant cells or by other stimuli. Nevertheless, Ribberf^ is unquestionably right when he maintains that most tumors, particularly epitheliomas, begin at a certain point '^ Der Epithelkrebs, Leipzig, 1865. " Ueber den Krebs, Sammlung klin. Vortr., No. 33, Breitkopf & Haertel, Leipzig, 1872. Die Entstehung des Carcinoms, Cohen, Bonn, 1906. V.ROII III oi 11 MORS 25 (ti^. I) ami j;n)\\ out ti<»m tins point like tin- biaiKlits (W a tree- ( uni- centnc i;ro\\rh). As tlu tuinoi cUxilops these- hiaiiches extend farther and laitlui nuo rlu- surrounding tissue until some of the branches approach the surface epithelium at a distance I'lom the original focus, and may appear to he derived from the epithelium at this |)oint, while Fic. I Schematic drawing of a beginning epithelioma, showing normal epithelium and beginning abnormal proliferation. in fact they are developed from the original focus (Fig. 2). B\- the continued extension of these branches the surface epithelium is destro\ed (Fig. 3). Serial sections and reconstructions have shown this to be true in the great majority of tumors, and it is now generall\- believed to be the usual manner of growth for the more simple t\pe of epithelial Fig. 2 =rr '■^i^^V>-i^ r:C m^ y. (f #- Conrmuation of Fig. i, showing cell columns approaching the normal epithelium. tumors. Cells multiplying from a single centre occupy the line of least resistance, which is usually the connective-tissue spaces. Cells passing into these tissue spaces give rise to new foci which appear at some distance beyond the advancing cell column and which may later coalesce with the mother tumor instead of being a distinct continuous 26 THE GENERAL BIOLOGY OF TUMORS outgrowth from the primary focus. These new foci must be distin- guished from multicentric growths in which new foci arise independently of the original tumor. These foci in the tissue spaces may be seen in many tumors. There is no question, however, but that these cells Fig. 3 T^'^^j-i - By extension of the branches shown in Fig. 2, the surface is destroyed, producing an ulcer. have escaped from the mother tumor and are not normal cells which have become cancerous in that situation from contact with other tumor cells or have arisen independently. The development of other tumors has been studied but little, because the point of origin is difficult to determine. The benign tumors also Fig. 4 Fig. 5 X Schematic drawing of a beginning Continuation of Fig. 4. The develop- tumor. ing cells have pushed aside the connective tissue forming a capsule. probably begin at a single focus, and the development proceeds from this point. Sarcomas probably follow the same course, but in them the problem is more difficult, for very small sarcomas are not presented for examination and the advancing border cannot be certainly identified at MKT.isr./s/s 27 sucli ;iM c;iil\ sr;i^f hccaust- ot tin- rcscnihlanci- to \()unj; coiiiu-ctive- tissLic ami ^laiuilation-tissue cells. In sarcomas developing in henior- rhaj^ic degeneration ot niyonias ot tlu- iittius the malignant process htgins about tin- entire peripluiy of the lieiiiorrhagic mass. In carci- nomas ot tin lactating breast many areas seem simultaneousK' to take on a maliti,nant liiowtii. Hegmmnii at a single point ( Kig. 4), tumors in their mode of develop- nunt follow one of two types that can usualh' he distinguished the expansile and the inHltratmg. Expansile Tumors. In then dexelopnunr the expansile tumors press aside the surrounding tissue. In this wa\' a so-called capsule IS formed hy the condensed displaced tissue (Fig. 5). The tumor is enclosed in this capsule and can be readily shelled out. The tumor cells, in other words, possess the power of autonomous growth, but not of in\ading the surrounding tissue. Infiltrating Tumors. -In contradistinction to the foregoing, while they begin at a single point, as do the expansile type, the cells pass out into the interstices of the connective tissue 'Fig. 6) like the roots of a plant into the soil. Longer or shorter columns of cells may be followed extend- ing from the mother tumor into the surrounding tissue. In this class no capsule is formed. The connective tissue and the contained vessels and nerves of the tissue invaded become a part of the tumor or are destroyed. In certain rapidl}' growing infiltrating tumors, however, the expansile growth is simulated, the surrounding tissue being pushed apart at the circumference of the tumor. This gives rise to a pseudo-encapsulation, but the restraint ordinaril}' imposed b}' the capsule of the surrounding tissue, as observed in the expansile t\'pe, is lacking. This type, in contradistinction to the preceding type, possesses not only the powder of autonomous growth, but that of invading the surrounding tissue as w^ell. Fig. 6 « # Continuation of process in Fig. 4. Instead of pushing tissue aside, as in Fig. 5, the cells have passed into the tissue spaces. METASTASIS W ith the escape of the cells into the interstices of the normal tissue numerous modifications in growth are possible. Cells growing from the peripher\' of malignant tumors invade the 28 THE GENERAL BIOLOGY OF TUMORS surrounding connective-tissue spaces, usually maintaining connection with the tumor. Frequently cells are found lying singly in the tissue clefts. Not infrequently cells escape to greater distances into the connective tissue and begin to proliferate, or they may escape into the lymph vessels and proliferate in neighboring lymph glands, or they may be carried by the bloodvessels to remote parts of the body. When developing in the region of the tumor they are called local metastases, when in the neighboring lymph glands they are called regional meta- stases, and when in distant parts they are called remote metastases. Local Metastasis. — The dividing line between local metastasis and infiltrative growth is not sharp. When cells form perceptible nodules in the region of the mother tumor, local metastasis may be said to occur. Usually these local nodules result from extension into the connective- tissue spaces. Their direction is independent of the lymph current, though development along lymph channels cannot be disproved. The extent to which local metastases take place varies. They are most often noted in the skin, mucous and particularly serous surfaces. They usually indicate an active proliferation of the tumor cells and a corresponding hopeless prognosis. After the removal of the primary tumor multiple nodules often develop in the site of the original tumor. This is seen with particular frequency after breast amputation for carcinoma. Such multiple nodules may develop in the presence of the primary tumor (see carcmoma en cuirasse). Fig. 7 C h Extension of carcinoma cells into a lymph channel: a, nest of carcinoma cells; b, wall of lymph channel; c, connective tissue. Regional Metastases. — Cells gaining access to the Ij/^mph channels (Fig. 7) are transplanted to regional lymph nodes and there develop. Usually the glands affected are proximal to the region of the primary disease. Cells may be stopped in the course of the vessels and form MKT./sr.lSIS 29 iioilulis, (»i (.aiuH I (.tils iii.i\ (l«\clop l)\ coiHiiHiiry alon^ the nirirc- 1\ nipli (.Ikiiii I\ in|ili;inL;it is cai (.iiioiiki tosa). 1 his is partu-iihirl)' likcl\ to ocelli m ilu |iUiiial ami otluT serous cavities. I he chsposi- tion ot the C(.lls to locah/c iii the nei^liliorin^; lymph glands make it possible to piitliet tin- location of such metastases and to anticipate rluiii In rcmo\al of tin- i;laiuls at o|Hration. 1 his tact is made the basis tor modern operations. ( )ccasionally the tirst j^roiip is skij^ped and oiH- ot the next m succession is iinolved. 1 he reason tor this is not deal", hut it has been held that tin- cells jiass through the tirst ^roup of glands and become lodged in a second group. It is nujie probable that channels pass from the site of carcinoma directl\' to the second group of glands. \\ hen the primar\ glands are aftected these in turn may form toci for cells which in\ade more remote groups of glands. Carcinomas usually choose this method of dissemination. Fig. 8 t^-JL'. V* ">'•'. v'"V.*'''"i •.' '•."*"'♦*;■»'» ■'•it', ,':^'.''', Carcinoma growing into a bloodvessel: a, carcinoma cells; b, bloodvessel; c, connective tissue with round-celled infiltration. Remote Metastasis (Systemic Metastasis). — This form is caused by dissemination through the bloodvessels. Tumor cells invade the vessels (Fig. 8) and are then transported to some point impervious to them. The veins, because of their thinner walls, are most frequenth' invaded, though the arteries at the base of the heart may also be attacked. Sarcomas usually metastasize in this manner, but carcinomas sometimes do so. General metastasis may take place by extension of the growth into the thoracic duct. When tumors spread through the blood stream or the thoracic duct an}' region of the body irrespective of the situa- tion of the tumor ma}' be invaded. Retrograde Metastasis. — Besides these three common methods of metastasis, tumors ma}' spread by extension into veins in the direction opposite to the blood current. This is called retrograde metastasis. This ma\' be explained in some instances perhaps b\' a temporar\' reversal of 30 THE GENERAL BIOLOGY OF TUMORS the blood stream, but usually it is a direct extension along the blood channels. This is seen most often in the region of the kidney and about the uterus. Not mfrequently extension takes place upon serous surfaces. It is assumed that cells escapmg from the parent tumor are propelled by the movements of the abdommal or thoracic organs to remote parts of the cavity and there become attached. There is evidence that this takes place because in certain papillomas of the ovary tumors extend directly from the ovarian tumor to the peritoneal surfaces. The more remote situations of such nodules make it appear probable that these have become transplanted m the same way. These lie above and invade the peritoneum. In a large number of cases, nodules generally assumed to have been transported in this manner are seen on section to he beneath the peritoneum and must have reached their location by some channel lying beneath this membrane. Extension by Contact. — Much controversy has arisen about the possibility of so-called contact transplantation. In the ovarian papil- lomas, already mentioned, transmission by direct contact seems cer- tain. In the removal of these tumors fragments broken off and lodged in the abdominal wound ma)^ give rise to similar tumors years after the primary operation. Whether a cancer situated on one surface may excite a like tumor on an opposing surface is more doubtful. It seems plausible that the primary tumor may produce, first, a destruc- tion of the covering epithelium, and then an implantation of the tumor cells or an excitation of the normal cells to multiplication. Instances are on record, particularly about the vagina and perhaps the lip, which compel one to admit the possibility of such transplantation. But when cells from the pelvis of the kidney are said to pass down the ureters and become implanted in the bladder, or from the mouth down the esophagus, and produce tumors in the stomach, we must protest. As Ribbert states, to assume that cells capable of further growth might escape from the ulcerated surface of a carcinoma and withstand the action of the gastric juice to find an eroded spot upon which to alight and propagate surpasses credulity. It is more reasonable to admit the extension by vascular or lymphatic channels or simultaneous multiple origin. Laws Controlling Metastases. — That cells are able to maintain an existence independent of the parent tumor is the strongest argument for the autonomy of the cell growth. That the cells themselves are the origin of the developing metastasis needs only the argument of the similarity or even the identity of cells. When cells in metastases do differ from those of the parent tumor, the difference is only one of degree; usually the change is one of a lessened differentiation. Why cells from certain tumors have the power of developing at a distance, M/rr.tsT.is/s M why soiiu- tlo so early ami orlu-is late-, is clt[Hiiciiiit upou rlu- same factors rhar make- ir possible tor some cells to iii\acle the surroundinji tissue while others do not. it is not a satisfactory explanation to sa\ that some have greater independence of growth than others, hut noth- ing more definite can be stated regarding it. It is of interest to note that cells w hich once gain access to the general circulation do not develop indifferently in an\' region of the body, but do so in particular organs. 1 hose extending by way of the l\ niphatics lodge where the lumen becomes too small to permit their passage, and they develop there or perish. When they gain access to the circulation, other factors must operate. The liver and lung are the favorite seats, though the bone marrow is the site of election for certain tumors. Muscle is peculiarh' immune from all classes of tumors. Clearly these organs of predilec- tion must possess conditions favorable for the developrnent of tumor cells, and those immune must possess some element of protection against them. If we assume with Ribbert that all organs receive a like number of cells, then those must perish which lodge in organs other than those in which they are known to develop. To what extent cells are destro\ed IS but idle speculation. Of greater practical interest but equalh' obscure IS the fact that cells apparently lie dormant for many years after the removal of a tumor and then develop. That these late recurrences are due to the production of a second primary tumor cannot be main- tained, for the new tumor may resemble the original one histologically The latency of these cells may be influenced b\' the removal of the primar\' tumor, the bod\' being able for a long time to hold in check the remaining cells. This must be admitted as a possibility, for in tumors of limited malignancy, as ovarian papillomas and some s\"nc^■ti- omas, nodules not removed at the time of the removal of the primar\' tumor may regress. W hat influence the removal of the primary tumor will have on the existence of inaccessible metastases cannot be predicted. Metastasis of Benign Tumors.' — Occasionall}" tumors believed to be benign, usually of the suprarenal and thyroid glands, form meta- stases, most often by wa\' of the blood channels. The new tumors may take on rapid grow^th. The question of interest is whether or not these conditions should be called benign. If the cells of the organ from which the\' come, whether obviously the seat of tumor forma- tion or not, have the power of invading the vessel walls, they thereb\' show a certain even if limited degree of autonomy and declare their malignancy. In cases in which such cells are believed to have escaped from normal organs, regions of malignancy could undoubtedly be found if the organ were subjected to a complete microscopic examina- ^ Borrmann, Beitr. z. path. Anat. u. z. allg. Path., 1906, xl, 372. 32 THE GENERAL BIOLOGY OF TUMORS tion. It IS just these organs, the thyroid and suprarenal glands, which, even when the site of frank tumor formation, exhibit but slight changes from the normal. RECURRENCE By recurrence is meant the reappearance of a similar growth in the region of the primar}^ tumor after removal. Cells which have travelled to a distance from the mother tumor may easily escape the surgeon's knife and afterward develop at the site of the extirpated tumor. The recurrence usually resembles the tumor extirpated and not the cells of the region where the recurrence takes place. The fact that cells may thus he dormant many years has caused much controversy. The recur- rence may take place in a few months, and usually does so within one to several years, but instances are on record where twelve or more years have elapsed between the removal of the primary tumor and the appearance of a recurrence. In some mstances recurrence does not take place so long as the nutri- tion is maintained at a high degree, but when resistance is lowered the tumor develops. The influence of nutrition on recurrence is exemplified in a case observed b}^ Perkins.^ In a face carcinoma recurrence took place m a plastic flap from the neck, while the skm of the face surrounding the site of the growth remained free from recurrence. Recurrence must be distinguished from multiple origin whether simul- taneous or successive. In epitheliomas of the face (which see) it is not uncommon to see independent tumors spring up m various parts of the face. Immunity. — Closel}^ associated with the question of recurrence is the problem of the natural defence of the body against the aberrant tissue. It has long been noted by clinicians that certain tumors — chiefly secondary nodules on the peritoneum, ovarian cystoma, and metastatic nodules in syncytioma — disappear after the chief mass of the tumor has been removed. The natural assumption is that the organism, being freed from the greater portion of the newgrowth, is able successfully to resist the small remaining portion. The predilec- tion of metastases for certain regions, such as the liver and lungs, and the antipathy to other tissue, notably the striated muscle, indicates a resistance in some tissues not possessed by others. This problem has been the subject of a vast amount of experimental study, with the result that it has been demonstrated that immunity to certain strains of tumors may be produced in animals. The exact nature of this reaction has not been determined. ^ Personal communication. '/■///•; Ei'iEcr ().\ rill, (.em-.k.ii. (:o\srrn tkjx -V-i FORMS OF TUMORS !• 11)111 rlu' foiciioiiiji, It may be tlccliiccil that tumors assume a \aiicty of forms. I lu- expansile tumors naturally are globular when not sub- jected to uneciual pressure. The infiltrating growths, on the other hand, are \ery inejiular in form, sometimes failing in actual tumor formation, producing instead an ulcerating process. If an expansile tumor meets resistance to its growth m any particular direction, it con- forms to the shape of the resisting body; for instance, a lipoma growing in the hilum of the kidney assumes the form of that cavity. On the other hand, an infiltrating tumor or its metastases ma\' present a globular form when growing in tissue offering equal resistance on all sides. The form of the organ is maintained when resistance is offered b\' the capsule of the organ; for instance, tumors of the spleen or kidney often retain the form of these organs. Ordinarih', however, infiltrat- ing tumors form irregular masses, extending along the lines of least resistance. The irregularity in the outline of the tumor ma\" be much increased when local metastases take place, forming nodules about the border of the tumor. It is evident that growing tumors will react on surrounding tissue with varying results. The expansile type exerts a mechanical effect; it displaces the surrounding tissue, and the result depends entirely upon the size of the tumor and the nature of the invaded tissue. An enormous lipoma of the subcutaneous tissue ma\' cause only inconvenience from its weight, while a small intradural fibroma may lead to destruction of the cord hv pressure. The infiltrating growths, on the other hand, lead to local destruction of tissue both from pressure and from some obscure biological property of the malignant cells. THE EFFECT ON THE GENERAL CONSTITUTION' 1 he eftect ma}' be direct from the metabolism of the tumor or second- ary from interference with nutritive processes. The exact cause of this deleterious effect on general nutrition has not been fully deter- mined. It takes place chiefly in carcinomas. The amount of involvement of tissue is not always in proportion to the effect produced. There ma}- be constitutional s3'mptoms when the tumor is small and not the site of an\' secondar}" degenerative processes, and it may be absent in the presence of large tumors. This has given rise to the supposition that some metabolic activity of the cell is the active factor. W hen degenerative changes take place other factors enter. Shutting off ' For literature see Lewin, Die bosartigen Geschwiilste, Kleinkliardr, Leipzig, 1909. 3 34 THE GENERAL BIOLOGY OF TUMORS the nutrition of the tumor may produce a sudden mtoxication from necrosis of portions of the tumor. Secondary bacterial infection may add to the general effect of the intoxication of the necrosed tissue those of a specific infective process. The effect from the metabolic process alone produces a general anemia, loss of weight, and a peculiar bronz- ing of the skin. This symptom complex is known as cachexia. It is not present in all tumors and is variable in its time of onset and its intensity, and may be produced by other diseases. Benign and more expansile tumors are not attended by such constitutional effects un- less there are secondary changes or some interference with nutritive processes. Toxins. — The presence or absence of a specific toxic product asso- ciated with cancer has been much discussed. Some investigators claim to have isolated such substances. A hemolytic extract or enzyme was isolated by Micheli and Donati^ which they believed to be specific. It is generally believed that there is a difference in the biochemical properties of tumor cells, but Lewin energetically denies that these are specific. Carcinoma cells are resistant to pepsin but sensitive to trypsin. Whether these changes in cell chemistry are the cause or the result of the disease has not been demonstrated Blood. — The blood changes in tumors vary greatly. Benign tumors are not accompanied by any change. In the malignant types, partic- ularly in carcinomas, a secondary anemia develops sooner or later, but it possesses no specific characteristics. In general, the blood changes are parallel with the degree of disturbance of nutrition. In some cases, particularly in carcinoma of the stomach, a picture closely resembling that of pernicious anemia is found. The red cells may be reduced to a million or less. These low counts have been noted particularly in carcinoma, but are sometimes observed in sarcoma. Poikilocytes, microcytes, and macrocytes are often present. Small nucleated red cells are often observed, but megaloblasts are rarely present. High hemoglobin percentages are observed when for any reason fluid is abstracted from the blood or fluid cannot be taken ie. g., carcinoma of the esophagus). Leukocytosis may be present in both sarcoma and carcinoma. In carcinoma the increase in the white cells is moderate in degree, rarely exceeding 12,000 unless there is an inflammatory com- plication. Sarcomas, on the other hand, are often attended by leuko- cytic increase, reaching not infrequently 30,000 or 40,000. Virchow regarded the occurrence of increased number of leukocytes in carcinoma as evidence of lymphatic involvement. The increase is usually on the part of the mononuclears,^ particularly in sarcomas. The specific gravit}' ^ Riforma med., 1903, xix, 1037. 2 M^rcotte, These de Paris, 1902. h'/fj/jjc/cn. c/i./k./cr/.R/sr/cs :',:> ot tlu' Mooil IS s;ii(l to 1)1- lisstiud iii (.MirmoiiKi ;iiul tlu ;ilk;ilinit\ ic- diicid. A fcrnunt with jiown to nt;ii(l tin- di^cstioii ot ;ill)imiiii has ht'cn (Itsii ihid, hilt thr s;inu- has hi( n ohsiiAcd in lailuxia Iroiii orhiT causes. BIOLOGICAL CHARACTERISTICS I' loni thi' toiiiiomsi, ir will In- sctn rhar rhcii- is jiicar saiiarioii m rhi- htha\ior ot riiiiiois, souk- tciuhiiL:; to (l(.srio\' rhc siniouiKhiiji tissue, others only to disphice it. In ^ciural, the tormer tend to destro\' the life ot the patient, and are, theretore, called malignant, while the latter do not, and are called benign. These were prima ril\ clinical conci|i- tu)ns with which it was s<)Un l>ui ih.n niiani imnois aic- con- genital, and ni nunuious otiui cases (K\ ciopnunral tictct rs hiinisli tin- stairuiti jionit loi the runiois. I'.nihi \()nias aic trc(|iicnrl\' the site tor rlic (kxtlopnuiit ot saiconia. II \ |)c rn( pluonias, n(\i, and pig- mented moles aic cxampks ot Miipiili'ct dcNelopnuiit ot cells which furnish a ticciucnt jiomt ot tiepartiiic loi tumor tormarion. Hianchio- genic caicmomas arc likewise txamplcs of tumors startmg out tiom developnu nral tests. I he most that can he said tot this Inpothesis is that it accounts tor the origin ot a limited hut detimte gr(;up of tumors. It leaves unaccounted for, however, the stimulus which excites to development the cells thus displaced. The fact that the known c(jn- genital tumors rarely become malignant is evidence against the Cohn- heim theor\'. Many of the sarcomas of young persons cannot be demonstrated as arising from congenital tumors, though such an origin IS otten assumed; but it is particulai"l\' the carcinomas that the theory leaves unexplained. It is only in those rare instances in which malig- nant epithelial proliferation takes place in embryomas that such relationship can be proved. It is impossible to explain by the embryonal theory those carcinomas occurring in the scars of burns or in chronic ulcers. That malignant growths are uncommon in negroes and other inferior races, and in the lower animals in which accidents of develop- ment are as common as in the human subject, is cited b)' Lewin' as an additional argument against the validity of the hj^pothesis. The greatest argument against the h}"pothesis, however, is in the morpho- logical resemblance of the tumor cell to the adult cells. 1 he difference is histologically but slightly apparent, and in man}- instances the tumor cells are able to carr}^ on the function of the normal organ. On the contrar}^, they bear no resemblance to embryonal cells, which would be the case if they sprang from them. This fact alone is sutficient to disprove the general applicability of the hA'pothesis. While limited in its applicability, the value of Cohnheim's theor\' has been great, in that It was the first plausible explanation of the development of tumors which took into account cellular pathology. It pointed the direction in which subsequent studies should be pursued and gave a stimulus to that histological study of tumors which has done so much to discredit it. It paved the wa}* and still serves as the basis for the ever broadening theories of cell autonomw It is for these reasons worth\" of a place in classic tumor literature. Many attempts have been made to remedy the defects of Cohn- heim's theory and to extend its application. Borrmann- assumed that ^ Die bosartigen GeschwLilsten, Kleinkhurdt. Leipzig, 1909. ^ Ergeb. d. allg. Path. u. path. Anat., 1900-1901, vii, 833. 40 THE GENERAL BIOLOGY OF TUMORS carcinoma developing without connection with epitheHum must have begun in isolated nests of epithelium. This addition to Cohnheim's theory aims to make it applicable to all classes of tumors at all periods of life. Wilms, ^ in the light of his researches on mixed tumors, attempted to explain the late spontaneous manifestation of activity of mixed tumors upon the ground that being independent of the surrounding tissue they are not constrained by it. He attempted to substantiate this hypothesis by experimental evidence, but without definite results. Rather more noteworthy is the attempt of Ribbert^ to extend the validity of Cohnheim's theory. He assumed that the dislocation of tissue was operative not only during embryonal, but also adult life. By the proliferation of connective tissue about epithelial cell columns the cells are separated from the remaining cells. When thus separated they are freed from the restraint such union imposes, and are then capable of further proliferation. The primary process, therefore, was in the connective tissue which brought about a liberation of epithelium. He later amended this by assuming that an actual isolation of epithelium was not necessaiy, but that some change in the connective tissue pre- ceded the epithelial proliferation. In this assumption he approaches more nearly the theory of Thiersch than that of Cohnheim. Ribbert's theory takes into account many conditions under which cancers develop, but it leaves without explanation the actual cause of the beginning of the process. Epithelium experimentally isolated does not go on to proliferation. Other factors must, therefore, be at work. The sum total of the efforts at the establishment of the theory of congenital origin of tumors consists in showing that certain tumors arise as the result of error in development. It leaves unexplained why these displaced tissues proliferate under certain conditions. The material only is identified, the causative factors are not indicated. For the vast majority of tumors the source of the malignant cells is not explained in a satisfactory manner. The attempts at the extension of Cohnheim's theory have served to emphasize rather than to remove its limitations. The Acquired Tumors. — The congenital theories offer no satisfac- tory explanation for the vast majority of tumors, and especially for those of epithelial origin. Since the establishment of the cellular pathol- ogy, which culminated in Virchow's^ work, it has been recognized that all cells come from preexistent cells. If the cells of the tumors do not arise from congenital cells they must come from adult cells. The cytol- ogy has been worked out carefully. The character of the changes is 1 Die Mischgeschwiilste, Georgi, Leipzig, 1900. 2 Beitr. z. Entsteh. d. Geschwiilste, Cohen, Bonn, 1906-1907. 3 Die Cellularpathologie, Hirschwald, Berlin, 1871, 4th ed. KTlOUniY or r I MORS 41 imi'tortant ami tlu sinculation as lo tin- caust- is iiitirisnn^. I he changes which taki- plact iii ci lis \shcn uiulcrgoinji the transition tioni normal to mahi^nant ha\i- hciii chisseil iiiulcr several heads. Metaplasia. \\\ metaplasia is meant a change of the cell t\ pe. it has luin luietotoie Ih1ii\ iil that such changes could take place ireel\ . Rihheit has insisted that this supposition does violence to biological laws, lie helicNes that \anation in cell t\pe is held in as close bounds as is \aiiation in tlu animal species. Most pathologists believe that changes withm nanow limits are possible, and a tew men of prominence are disposed to accept very broad limits. Changes have been most certainly demonstrated in epithelial cells. 1 he deviation is from a higher to a lower order. Columnar cells may become cuboidal or sc|ua- mous and take on the structure and function of epidermoid cells but squamous epidermoid cells never become columnar and acquire a secre- tory function. When squamous cells become cuboidal it is evidence of reversion of developed squamous cells to an earlier less differentiated state and not of an approach to the columnar. What has been often regarded as evidence of change in cell type has been an e.xample of the replacement b\' a suitable type of cells which, being unsuited for the changed conditions, have disappeared. This is well shown in the cervix of the uterus. The columnar cells frequently extend out much farther than the usual line of demarcation. This comes about not by a metaplasia of the squamous cells, but by a destruc- tion of the squamous epithelium and a replacement at its expense by the columnar cells. In other regions, as in the urinar;- bladder, areas are frequenth' covered with squamous epithelium. This mzy be due to a congenital dislocation. In the trachea islands of squamous cells are normally found, and it is possible that such islands may be dis- covered in other regions where cells have heretofore been considered to undergo metaplasia. Experimental metamorphosis has been pro- duced by ligation of the outlet ducts of one of the sahvary glands. Here the columnar cells of the duct after a few months give place to squa- mous epithelium. The change in form here is, more properly speaking, due to pressure rather than to a change in structure. On the other hand, the occurrence of squamous-celled cancer in the gall-bladder can be explained only on the ground of metaplasia, as even Ribbert admits. The same is unquestionabl\' true for squamous-celled metas- tasis in the lung from carcinoma of the stomach. In hemorrhoids long prolapsed the columnar epithelium of the rectum may be replaced by squamous epithelium, and it is eas}' to conceive the development of carcinoma from it. While it has been definitely proved that epithelial cells of one t}"pe may change into epithelial cells of another type, it is difficult to decide 42 THE GENERAL BIOLOGY OF TUMORS if one kind of cell may be changed into a cell of another kind. For instance, the question which presents itself most often is whether bone may be developed at some distance from any known anlage of bone. In certain muscles true bone formation occurs. It may be assumed either that the connective tissue of the region has the power to undergo such change, or that embryonal bone cells have been misplaced to this region during development. Both lack final proof. Many changes observed in a tumor ma}^ be ascribed to degenerative processes, e. g., the formation of myxoid tissue in libromas and the development of fat in the same tumor, or to the simultaneous proliferation of different tissues. Some observers are disposed to accept much greater possi- bihties than the opinion above indicated, believing that it is possible not only for tissue types to change, but also for one group to change into another. Lubarsch remarks that he would not shrmk from admit- ting the possibility of epithelium changing mto connective tissue. The evidence he cites is the appearance of epithelial cells in some carcinomas so closely resembling connective tissue that a differentiation may be impossible. Changes of epithelial tumors into sarcoma in experi- mental tumor transplantation have been regarded as additional evidence. We are ignorant of the influence operative in determining the formation of any kind of tissue, and we can best ascribe it to hereditary influences. Those changes which do take place Ribbert believes happen because the cells first assume a lower order and then re-differentiate into a new higher type. It is probable that metaplasia in tumors is confined within rather narrow limits. The problem finds its importance not onl}-^ in the etiology of tumors primarily, but also in the metamiorphosis of tumors of changing type. Some pathologists contend that instead of a change in the type of tumor, a so-called malignant degeneration, the tumor was malignant from the beginning, and that when a benign tumor takes on malignant growth it is a newgrowth superimposed upon the primary benign growth. It would seem that this assumption ignores much definite observation for the purpose of maintaining the integrity of a theorv. In myomas of the uterus the transition from the myoma cells to malignant cells often may be traced definitely. The transi- tion appears to be direct without the intermediate reversion to a lower type. A local agent of unknown character seems to be able to stimu- late an adult resting cell to a development that is atypical and un- limited. Regression (Riickschlag). — By regression, Ribbert means a process of cell change whereby cells relapse from a more differentiated plane to a lower, less differentiated plane. It differs from metaplasia, with which it is frequently confused, in that it assumes a change from a /■rr/o/.ocr oi timors 43 hi^lur to a l()\Mr onlci ot chrtt-rc-iuiarioii, wink- nR-ra|')lasia assunic-s a cliaii^c from oiu- t\ jic of cell into ancjther t}pc of an ecjiial decree of tlirtcrcntiation. Regression finds its best example in the changes which connective-tissue cells undergo in inHammaror\ and healing processes. I he ceils react h\ heconung larger and rouiuli r, ajipai- ently resuming the partially de\elo|ied stage. In like manner, eiuhjthe- lial cells may become cuboidal and resemble in a general way epithelial cells, without, however, taking on an\- ot the biological characters of epithelium. Epithelium likewise undergoes similar changes. Epidermal epithelium may lose its prickles and the c\toplasm may alter its struc- ture. In the more highly developed cells changes are less obvious if the\ occur at all. Muscle cells undergo certain changes in form and struc- ture, but in specialized cells generally regression is confined to the cytoplasm. This slight disposition to change may account for the rarity of primar}' tumors in these tissues. The changes in cells are never abrupt, and it is necessar\- in man\- cases to trace them through the various stages to ascertain their origin. Eor instance, in epitheliomas the cells first lose their prickles and become rounder, keratohyalin is no longer formed, and as the deviation be- comes greater the cells become smaller, due to a lessened amount of protoplasm. Anaplasia. — By anaplasia is meant the process, both morphological and physiological, whereb}' cells change their character and become converted into a type not normal to any period of development, form- ing a cell with characters entireh' distinct from normal cells. Morpho- logically the change consists in a deviation from the normal cell division. Hypochromatic and hyperchromatic cells are observed. The chief difference lies in the fact that the number of chromosomes in the poles varies so that daughter cells of varying size are produced and the resemblance to the mother cell is thereby lost. Physiologicallv this deviation finds expression in the increased capacity of proliferation of the cells of a lower order which the atypical mitosis has produced. The greater the morphological deviation the correspondingly greater is the physiological. Hansemann,' who originated the term "anaplasia," regarded these observations of etiological importance. Such changes are often observed, but they are not confined to malignant processes, being observed in regenerative processes and often absent in beginning carcinomas. Most pathologists regard atypical mitosis as an expres- sion of bewildered growth and not a cause of it. Changes within the cells above noted are structural, but in tumors ' Die niikroskopische Diagnose der bosartigen Geschwiilste, Hirschwald, Berlin, 1897. 44 THE GENERAL BIOLOGY OF TUMORS the chief factors are physiological. The power of proliferation beyond normal limits is the essential, and this does not run parallel with the morphological changes. Both are alwaj^s present, but in varying degrees. It is important to remember that structure alone is not sufficient to determine the character and potentiality of the tumor. Changes in the Tissue Balance. — Attempts at explaining the actual causation of the cell proliferation have been made. The most note- worthy IS the theory of Thiersch,^ that the primar\' change is in the connective tissue, whereby the lessened resistance of the senile connec- tive tissue permits the epithelium to proliferate. This is based upon the assumption that there is an antagonism between the epithelium and connective tissue. The epithelium, which predominates in the embryo, is held m check in the adult by the connective tissue. When the connective tissue, notably the basement membrane, undergoes degeneration, the epithelium, released from this restraint, resumes the proliferative capacity of the embryonal cells. This explains the fre- quent occurrence of carcinoma in advanced life, but it does not take account of the local occurrence of carcinoma nor its occurrence in earl}^ life. Acidobasic Tissue Reaction Theory.- — I have attempted to develop the theory of Thiersch by a tinctorial study of the tissues in order to deter- mine the relation of connective-tissue changes as revealed b}^ tinctorial chemistry. In the senile subepithelial connective-tissue changes of slight and inconstant degree may be demonstrated, but these changes are not local nor are they present at the places and time of maximum frequency of tumor formation. There is much of interest, however, that seems to explain the conditions under which tumors develop. The most common accompaniment of carcinoma, chronic irritation, is attended by changes in the connective tissue. Round-celled infiltra- tion, when existent for any length of time, is attended by a decrease in the acidophilic reaction of the connective tissue. In these regions it is common to observe a degree of epithelial proliferation. These changes are seen about all ulcers. Parallel with the changes produced by irritation are those produced by the injection of certain d3^es (Scharlach R, Sudan III). An active epithelial proliferation takes place which Fischer^ ascribed to an "attrac- tion" for the epithelial cells. He called the dye, therefore, an "attraxin." The changes are primarily in the connective tissue and resemble the ^ Der Epithelkrebs, Leipzig, 1865. ■^ For a more complete presentation see Trans. Western Surg. Assoc, 1907; Jour. Amer. Med. Assoc, 1908, 1, 425, and 1910, Iv, 2290. ^ Miinch. med. Wochenschr., 1906, liii, 2041. ETiOLoc.y or riwfORS 45 ch;in<;fs piodiucil 1>\ iliioim- iiit];imm;i t ion. \\ lun Sudan III or Scharlach R is injccrctl iiiiiiudiarcl) lulow tin- cpit luliuiii, the coniu-c- tivf rissiu- Hist losi-s its reaction to acid dyes, lollowiii^ this, tin- cpi- tlulnini luiiins to pioliferate. The epitht-lial prolittratioii is coiihnt-d to the- :\WA in which thr comucti\c tissue has undergone its modified reaction to acid d\es. The ejiithehiini retains its normal characters, hut the mitotic huures may he ohserved and pearl formatmn is common. Thi- (.pitht, hum ch)es not invade regions where prehmmary changes in the connective tissue have not occurred. In .v-ra\- carcinomas all writers' who have studied the lesion are agreed that the piimar\ changes are in the connective tissue and that epi- thelium hegins to proliferate only after such changes have taken place. These changes in the connective tissue seem to be identical with those produced by Sudan III. Unlike the proliferations in the Sudan experi- ments, however, the epithelium after a time invades regions where the tissue is still normal and metastases may be formed. 1 he influence of the .v-ra\s upon the tissue is much more widely diffused than in the Sudan injection, and much longer periods are passed before the growth of the epithelium in the v-ray tumor begins. In beginning epithelioma of the same class of tissue there is a change in connective tissue closely resembling the changes in both the Sudan experiment and in the x-r2Ly carcinoma. The connective tissue is less cellular and it loses in part its affinity for acid dyes. Like the latter and unlike the former, the cells invade regions beyond which the con- nective-tissue changes occur. In both the spontaneous and .v-ra}- proliferations there is a change in cell t^pe. The fact that the changes in the Sudan experiments and in the .v-ra}- cancer are primarily in the connective tissue makes it seem probable that the changes in the sponta- neous proliferation are likewise in the connective tissue. Chronic irrita- tion is attended b}- round-cell infiltration. In an\- acute inflammatory process the connective tissue loses temporaril}- a part ot its acidophilic property, reverting to a state approaching tinctoriall}- that of fibrin. When the inflammation subsides the normal affinity for acid d\es re- turns. In chronic irritation in some instances the reaction to certain d\'es is lessened and is permanent. In ascribing the pnmary change to connective tissue the theor\' of Ribbert is suggested. In none of these instances, however, is epithelium "liberated" from the adjacent epithelium. On the contrary, the continuous growth ma}- easily be demonstrated. The onl}- tangible explanation is that the lessened acidophilic propert\- of the connective tissue permits the invasion of the basophilic epithelium. In certain regions the changes 1 S. B. Wolbach, Jour. Med. Research, 1909, xxi, 415. 46 THE GENERAL BIOLOGY OF TUMORS in the epithelium seems to be primary inasmuch as changes in cell type and active proliferation may occur and yet the surrounding tissue be not invaded. This is shown in polypoid proliferation of the uterine mucosa. The cells change in type, become strongly chromophilic, but the surrounding uterine muscle is not invaded. The balance of the tissue is maintained so long as there is not a preponderance of basic over the acidophilic power, and we may say, therefore, that invasion occurs whenever the basic power is excessive or the acidophihc power is reduced. The very strongly acidophilic basement membrane may be placed about epithelial (and endothelial) cells as a barrier to inva- sion. For the formation of metastases it may be assumed that it is necessary that the basic power of the tumor cells is excessive or that they reach a region where the acidophilic power is normally less than in the basement membrane of the region from which the epithelium was derived. Carcinoma metastasis has a predilection for regions where the acidophilic power is normally less than that of ordinary connective tissue. It is interesting in this connection to note that the basement membrane is more strongly acidophihc than ordinary con- nective tissue. Parasitic Theories. — During the period when the blasteme theory was accepted generally as accounting for the prime etiological factor in carcinoma the contagiousness of the disease was maintained. When the ulcerative stage was reached, gases were supposed to escape which carried the infective elements. Cases were reported to substantiate this view. When the cellular theory was established, cases were reported in which infection by direct contact was believed to have taken place. Cases in which carcinomas developed on opposing surfaces, such as the lips and vulva, were reported and regarded as evidence of infectious- ness. In the reported cases in which the diagnosis was certain the possibility of multiple primary origin cannot be excluded. Instances of epitheliomas on opposing surfaces of the vulva are open to the same objection. I doubt if there is an authentic case on record. In most of those recorded the diagnosis was clinical. A case of contact infec- tion reported by Hartman and Lecene met the requirement that the inoculated tumors shall be of the same structure as the primary one, but the location was such that other doubts enter. The case was one in which a glandular carcinoma of the cervix produced a glandular ulcer of the vagina. In this situation extension by retrograde metastasis through the vaginal veins seems more probable, smce the new tumor was distinctly stated to be an ulcerous process. I saw recently a case in which the entire vaginal plexus of veins was infiltrated from a gland- ular carcinoma of the cervix. The perivascular tissue in many places ETioiJH.Y or rrMoRs I7 \v;is iiihlti ;i ticl, aiul iiisr ihskIc tin \iil\;i one ot tlusi- IkkI iilrci ;ir<(l. Such extt'iisioii is nor iiiuisual in s\ ik\ tioiiias. rhosi- cases calltd iniicfr a dfiix hy the French, in which hiishand aiul wife htconu- sinuiltaneoiisly or successivtl\ attiictcd, \scre citccl as evidence. Ir is {luesrionalde whither siiniiliaiuous occurrence ot carcinoma ot the peins and cervix is more than a coincidence. I hus, I)emaiiiua\ reported 134 cases of carcinoma of the penis, and in only one was tliere a possibility- of contact infection, which seems rather remarkable, considering the frequency of carcinoma of the cervix. Nearly all the cases cited in substantiation of the possibility of contact infection rest upon insufficient evidence. 1 he nature of the disease in man\' of the cases reported is not clear, and m tliose in which the diagnosis may be accepted as correct the instances are so rare, as com- pared with the enormous frequency of carcinoma, that they hardly impress one as being more than interesting coincidences. 1 he endemic occurrence of carcinoma has been another factor in the argument in favor of its infectiousness. It has been noted that in certain communities or even houses carcinoma occurred with unusual frequenc}^. Some writers have sought to connect this with the nature of the soil or drainage. Thus, Behia' collected many statistics tending to show that in some instances carcinoma occurs with greater frequency in regions where the soil is badi)' drained. The evidence in the cases reported is insufficient to permit of judgment. The number of persons alone is mentioned without reference to age and the presence of other factors with which carcinoma is recognized as being more or less asso- ciated. As Williams well remarks, these epidemics have occurred in remote rural communities, and not in crowded unsanitary tenements which ordinarih' invite the transmission of communicable diseases. More recently the occurrence of epidemics of tumors in animals in certain cages have been reported. Gay lord' reports an instance in which from a single cage 60 tumor mice were removed within three years. The location of the cage was repeatedly changed and the stock of mice once. It would seem, therefore, that the cage itself was the factor at fault. This places these observations in the class of epidemic occurrence in moist communities. More suggestive still is the occur- rence of peculiar tumors in fishes in certain hatcheries. Thus, Pick reports an instance in which 7 per cent, in a certain hatchery were affected. Foreign fish procured to replenish the stock w'ere likewise affected. Pick,'^ therefore, ascribes the effect to the water. The thyroid gland alone was affected. He concludes that first a simple goitre w^as ' Deutscli. nied. \\ ochenschr., 1901, .x.wii, 427. -Jour. Amer. Med. Assoc, 1907, xlviii, 15. ^ Berl. klin. Wochenschr., 1905, xlii, 1435. 48 THE GENERAL BIOLOGY OF TUMORS produced which later become carcinomatous. The instances above quoted are sufficient to show that these observations are more than coincidences and that some truth ma)' he in the mass of the reported instances. When the bacterial origin of manj^ diseases became definitely estab- lished it was no wonder that investigators turned to bacteriology for the solution of the carcinoma riddle. It has been the absolute con- fidence in the correctness of the assumption which explains the laxity with which experiments were carried out and the disregard of the ordi- nary rules of logic with which conclusions were formed. Much of the evidence adduced was the result not of impartial scientific examina- tion into the possible or even probable parasitic nature of cancer, but of violent efforts to prove a prejudged conclusion. Much of the argu- ment against the parasitic origin of cancer has been made, a priori, an equally unsound procedure. Many of our proved scientific facts have previously been proved impossible upon just such carpet philos- ophy. While I readily admit that the bulk of the literature on the subject should be relegated to the museum of scientific folly, ytt there is much work by careful men that should be known to workers in this field, and should be kept in mind now that the pendulum is swinging violently in another direction. It is worth while, therefore, to present in brief outline the ground covered by these investigators. Three classes of organisms have been advanced as the cause of tumor formation; the bacteria, the protozoa, and the blastomycetes. They may be considered in this order, for it represents in a rough way the historical sequence. The Bacteria. — Even before Koch's epoch-making discoveries bacteria were described in tumors, but it was not until later that they were con- fidently announced as the cause of the disease. None of them could meet the requirements of even one of Koch's laws. It was Scheurlen's^ paper that received the most attention. He produced a culture which when injected into the breast of bitches produced a tumor in a month, which he declared to be of an epithelial nature. It was proved by Baumgarten- and others that the organism was a potato bacillus. Moreover, Shattock and Balance^ showed that non-ulcerated tumors were sterile and could be kept so for a long period. Doyen's^ announce- ment of Micrococcus neoformans is still fresh in our minds. He claimed that his microbe, a diplococcus, was always present in malignant tumors, could be demonstrated in sections, and cultivated in special media; ^ Deutsch. med. Wochenschr., 1887, xiu, 1033. 2 Centralbl. f. BakterioL, 1888, iii, 397. ^ Trans. Path. Soc, London, 1887, p. 412, and 1888, p. 419. ^ Etiologie et Traitement du Cancer, Maloine, Paris, 1904. ETIOI.OC.Y or TIMORS 49 ;iiul that tumois coiiKl l>r pi otliu id iii animals l)\ inoiiilat ion uirh piiii- culriiiis. Monoxn, In rlu- pii|iarari()ii ot a sixiial roxin he ciainu-cl to haw iiirtii 42 of 242 paticius attictccl with carcinonia. His icsiarclus win- consuKiiil ot siirticuiit mi|)ortaiUH- to wairaiu rlu- appoiiuiiKiu ot a spctial coniniittc-e in tiie Scjcicte de Chirurf^ie. This conmiirrft'' cxaniiiuil the- patients treated by Doyen over a period ot ti\e months, ami tailed to tind a single- mstaiue ot cure or e\en a single case ot amehoiation ot the- s\ miHoms. I hese may serve as e'xamples of the exideiice so far produced in the attempt to establish the direct relation of bacteria to the etiology of tumors. I p to the present time it ma\' safely be said that bacteria tound in tumors are accidental infections, and while they are present in all ulcerative tumors, no proof has been adduced to show any direct etiological relationship. The Protozoa. — While some contentions for the protozoic origin of carcinoma antedate the microbic theories, they were held in check by the statement of Virchow- that tumors were endogenous cell forma- tions, and the real prominence of this theory began onl}' after bacterial theories were proved untenable. Each investigator has advanced his own organism, which he has proved to his own satisfaction as a specific for carcinoma. Man\' of these have been dignified with special sonorous cognomens, which may, however, be left to he in the special literature. It is pureh' a matter of present preference which of the investigations one shall select for presentation as example. Those which have gained particular prominence have been due to the persistence or prominence of the authors rather than the character of the organisms present or any great variation from those presented by other investigators. The works of Schaudin,^ Sjobring,^ Schiiller,'^ GaAlord," and Feinberg," together with the elaborate work of Pianese,' will give the student the best notion of the field covered by these investigations. Nearly all of these works can be obtained in monograph form. That of Feinberg,^ ma\' be regarded as the final effort to establish an etiological relation- ship between this type of parasite and carcinoma. He describes the histor}' of sporozoa, and compares them with normal cells. The inclu- sions found in certain tumors are then compared with sporozoa and .are declared identical. The findings are familiar, and the argument ' Compt. rend. Soc. de cliir., July 15, 1905. - Virchow's Arch. f. path. Anat., 1892, cxxvii, 188. ' Sitz d. Berl. Acad. d. \\ issensch., 1896, xxxix, 951. •• Centralbi. Bakteriol., 1900, xxvii, 129. •''Die Parasiten in Krebs und Sarcom., Fischer, Jena. 1901. " Amer. Jour. Med. Sci., 1901, cxxi, 503. ^ Die Erreger und der Bau der Geschwiilste, Friedlandcr. licrlin. 1907. ^ Beitr. z. path. Anat. u. z. allg. Path., Fischer, Jena, Suppl., 1906. ^ Loc. cit. 4 50 THE GENERAL BIOLOGY OF TUMORS in favor of their parasitic nature is purely academic. The relation of the "parasites" and the tumor, as a whole, is not discussed. The facts that have made a parasitic origin of tumors doubtful are not considered. It is extremely difficult to express in brief space a comprehensive resume of the researches presented m this voluminous literature. The evidence is largel}^ histological, and is based mainly on tinctorial chem- istry. Some of the investigations were made, it is true, on fresh un- stained material, where motile organisms resembling ameba have been described, but it is impossible to gain from the descriptions a clear notion of the author's findings. Generally stained specimens were studied which showed curious protozooid structures within or between the cells. These have been for the most, part highly refractile, strongly acidophilic bodies occurring within the cells of carcinomas. At the present time these bodies are not proved to be parasites; even among their advocates no harmony exists as to their classification or source. It is particularly significant that protozoologists not inter- ested particularly in carcinoma theories do not admit the protozoon nature of the bodies in question.^ The majority of pathologists regard them as cell products or evidence of cell degeneration. The latter assumption has much to favor it, since they occur, at least chiefly, in the older degenerating portion of the tumors. The lack of unanimity has already been noted, and the occurrence of the same bodies in con- ditions other than carcinoma are strong arguments against the etiological relationship. The fact that no experimental tumor has been produced, and that diseases demonstrably of protozoon origin do not resemble carcinoma either clinically or pathologically, have caused the question to lose the interest of clinicians. The conclusion of Williams, that "the more carefully the subject is considered the clearer it appears that the 'protozoon of cancer' has failed to make good its identity," expresses the view of most of those who have studied the problem. The Blastomycetes. — As a direct corollar}^ of the foregoing are the so-called Russel bodies first described by Russel.' It was studies of Sanfehce'* however, that brought the structures into prominence. The organisms this investigator employed were obtained from extraneous sources, though similar organisms have been obtained from tumors of man. He claims to have produced carcinoma by them. The mass of evidence indicates that the tumors produced by these organisms are granulomatous in character.'* The problem is somewhat confused by 1 Doflein, Die Protozoen als Parasiten und Krankheitserreger, Fischer, Jena, 1909. 2 Brit. Med. Jour., 1892, ii, 1356. ^ Centralbl. f. BakterioL, xvii, 113, 625; xviii, 521; xxi, 254; xxvii, 155, 156. * Peterson and Exner, Beitr. z. klin. Chir., 1899, xxv, 769. KTioijH.y oi n MORS :.] the fact rhar ^t-mmu- Mastoniycotic disease- does occur in the luiinaii subjecr. I In usriiihlaiue to carcinoiiia, however, is sh^ht and explains the contusion \shuh has arisin in inocuhition experiments, and serves rather to coiniiui thi- ohseivii ol the non-spteiHcit\ ot the or^anisni. Spiit)chetes were cliseiducl in tumors by CJaylord' and Calkins.- I lust- were umtormh lound m Imast tumors of a series of mice examined and from tiansplaiHid tuniois. I he oi<;amsms wtie ohser\ed in the borders ot j^rowing tumors, thus tultillmg some ot the retjuirements lacking in other organisms advanced as the causative factors in carcinoma. I yz/er,^ however, described tumors of mice free from them, and. what is more conclusixf, he h)inul spirochetes in the nuchastinuni in mice- m which inoculations were unsuccessful. 1 he spirochete theory, theietore, seems to share the same fate as other parasitic explanations. Such a cursory review of some of the more important theories of the causation of tumors and of the conditions under w Inch the\' develop leaves the problem in an unsatisfactory state. The most striking fact noted in the literature is the collapse of the parasitic theories which were once regarded as so full of promise. Pathologists have almost with- out notable exception discarded all extraneous factors as the cause of tumors. Microorganisms have been rejected with decision, often with impatience. The cause of this is that most investigators who have gone over the whole ground ha\e come to the conclusion that the carcinoma parasite is the carcinoma cell itself. The conviction that the cell itself contains the secret of the endless development w hich characterizes these tumors has become nearl\- universal. I his power of the cell to limitless growth has been repeatedly referred to as autonomy'. Ewing^ has summarized the more important problems in a masterl\' manner. He correctly remarks that an appreciation of the full meaning of the theory of cell autonomy is acquired only after prolonged stud\' of the factors which go to make it up. It is the com- posite of many theories advanced by many of the foremost pathologists. The tumor cell is one which is able to proliferate in an environment in which a normal cell does not multipl}'. What is known about the conditions which bring about this increased power of proliferation has already been discussed. Cohnheim's theor\- furnishes a limited but a positive basis tor our belief. Certain tumors unquestionably arise in misplaced embryonal cells. No satisfactory reasons can be given to account for the develop- ment of the cell in later life, but it is safe to say that parasitic theories failed utterh' to offer an explanation why a parasite should select an ^ Jour. Infect. Dis., 1907, iv, 155. -Ibid., 171. ^ Soc. of Exper. Biol., New "^'ork, 1907, iv, 85. ^ Cancer Problems, Arch. Int. Med., 1908, i, 175. 52 THE GENERAL BIOLOGY OF TUMORS isolated group of cells for the exercise of its activity; but what is more convincing, the new tumor cells are observed to be identical with the embryonal cells. In other situations the facts are equally striking. The tumor cells may be but little altered from the mother cell. This is true not only of the cells which lie adjoining to those from which they develop, but also the cells of metastatic nodules may imitate the mother cells faithfully. Adenomas of the digestive tract may form metastases within the liver which retain a distinct gland structure, and lymph gland metastases from glandular breast tumors may show a like struc- ture. Epithelial tumors often retain their characteristics in metas- tases sufficiently to permit the recognition of their source with cer- tainty. These observations are so incompatible with any known action of microorganisms that it is impossible to regard the argument of the adherents of the parasitic theories seriously. Satisfactory as is the basis of our belief that it is the cells of the body, embryonal or adult, which, having acquired the power of autonomous growth, continue to develop, it is much more difficult to explain the methods of their operation. Embryonal cells which lie dormant, it may be assumed, retain a greater power of development than cells that have become fully differentiated, though this is mere assumption. Various explanations have been advanced why at some period these cells begin to assert their power of limitless development. Injury and increased nutrition have been the chief causes assigned. It would be nearer the truth to assume that it is the increased nutrition brought about by injury, together with probably the absence of an acidophilic basement membrane. It is to explain malignancy in fully developed cells that our chief argument must be directed. We have acquired some knowledge as to why cells develop, thanks to the labors of Loeb,^ Hertwig,2 and others; but we are ignorant as to why they develop in certain directions in the embryo and why their development ceases at a certain point. Tissue tension is a term which is employed to express the balance of the tissue in what we may call the resting state, which controls the multipHcation of cells and maintains their physiological balance. Direct mechanical control of cell growth in some instances seems probable, since Loeb^ found that in animals tumors which had ceased to grow began again to grow when a portion was transplanted to another region. The validity of this argument is somewhat lessened by the fact that not only was the tumor released from the mechanical 1 Resume and lit. Die cliemische Entwicklungserregung des tierischen Eies, Springer, Berlin, 1909. ^ Die Zelle und die Gewebe, Fischer, Jena, 1893. ^ Jour. Med. Research, 1901, vi, 28. ETIOI.OC.Y or TIMfJkS 58 i'i'sti;iiiil lull .ilsd lioin the iiillii(in( (jI Iiithn estimates thai about oiu-tbird give a carcinomatous family histoi\ . Recent fij^ures l)\ the Middlesex Hospital' fi;ive only S per cent. This ti<;uii' represents the percentage m which a direct transmission from either parent might be assumed. It IS worth noting that in the histories of patients where tumors are not m (|uesti()n a considerable proportion will be found to possess tumorous ancestrw I lowever, there are families in which carcinomas occur with striking frequency. If the probabilit\- of error in diagnosis be taken into account, the percentage of the positive existence of carci- noma in ancestry probably would be reduced. The occurrence of non- malignant tumors, especialh' moles, Hbromas, and neuromas, in certain families has been noted. It ma\- be stated tentatively that heredit\ exerts a slight influence in the production of tumors. It might be due to direct displacement of embryonal tissue (embryomas) or to trans- mission of the exciting factors (Wolf), or to heightened \ita!it\ of epithelial cells. - Influence on Nutrition of Tumor Formation. It is interesting to note that while approximately 2 per cent, of malignant tumors are found on the extremities, their occurrence upon paralyzed limbs is prac- tically unknown. Statistics relative to general nutrition are conflicting, but in general it ma}^ be stated that it is the physicall)' vigorous that are particular!}' predisposed. The imbecile and insane are but little subject to malignant disease. THE RELATION OF SPECIFIC INFECTION TO TUMORS Tuberculosis. — Carcinomas have been reported as occurring in lupous ulcers and in tuberculous scars. Rokitansk}' regarded the two processes as mutuall}' exclusive, while Ribbert believes that tuberculous processes act directly as an excitant of epithelial proliferation by isolating epi- thelial islands. Not infrequentl\', giant cells are seen in or about epitheliomas, but these are more satisfactorily explained as foreign body giant cells. They resemble foreign bod}' giant cells histologicall}', and tubercle bacilli have not been demonstrated in them. Ihat two ' Arch. Middlesex Hospital, 1904, ii, 104; 190^, \', 103. 'Alberts, Das Carcinom in Histologischer und experinieiitell pathoiotiisclier Bezie- luing, Fischer, Jena, 1887, p. 53. 56 THE GENERAL BIOLOGY OF TUMORS such common maladies as carcmoma and tuberculosis should occur simultaneously cannot create surprise. The actual occurrence of the two diseases together is, however, surprisingly rare, probably because the maxima of frequency of the two diseases fall at different ages. Frequently at autopsies of carcinoma patients, evidence of healed tuberculosis is found. That carcinoma is less frequent in tuberculous subjects than in non-tuberculous, as is shown by statistics of Lubarsch,' is not fully explained by their different age incidence, and it seems probable that there is some antagonism between the two diseases. It might be assumed that the tuberculous process excites an active resistance on the part of the connective tissues which tends to restrict rather than to invite the proliferation of epithelium. The relation of sarcoma to tuberculosis is more difficult to define because of the close resemblance of granulation tissue to sarcoma and because of the clinical similarity of the two diseases. This latter feature makes it doubly important that the scientific study of this problem should be undertaken with caution. Confusion between the two has been particularly great in certain lymph gland diseases. To prove any relationship there must be demanded, on the one hand, the demon- stration of the tubercle bacillus, and on the other hand, evidence of the sarcomatous nature as manifested in the clinical course. Animal inoculation of questionable material should demonstrate the bacilli if present. In the face of negative results a positive opinion should be held in reserve irrespective of the microscopic appearance of the lesion. If this extreme caution be observed, the causal relationship between tuberculosis and sarcoma cannot be admitted. Syphilis. — The possible causal relation between carcinoma and syphihs is limited to the occurrence of epitheliomas in syphilitic scars. In these cases there is no evidence that the specific nature of the scar played any part, but rather that the scar tissue acted merely as such tissue occasionally does, irrespective of the cause of the primary wound. Extensive epithelial proliferation is sometimes seen about syphilitic processes, but it differs in no way from similar processes about any ulcer. The histological resemblance of syphilis to sarcoma is m certain rare instances very perplexing; in fact, the microscopic evidence in such cases cannot be regarded as decisive. The therapeutic test for syphilis is distinctive in most cases and the recent specific reactions for this disease may serve to clear up some doubtful cases. At present no relationship between the two conditions can be accepted. Other disturbances, such as bilharzia, are sometimes followed by carcinoma. The action of the parasite is probably merelj^ that of a chronic irritant. ^ Virchow's Arch. f. path. Anat., 1888, cxi, 305. P A R r II THE SPECIAL PATHOLOGY OF TUMORS CHAPTER I OSTEOMAS General Conception. 1 umors which develop from bone are, like those from cartilage, usually reactive in nature, the result of some directl}' traceable irritation. The border-line between these and true tumors is often difficult to draw. In addition, bone formation is some- times found in true tumors which are for the most part of a different character, notably sarcomatous. When all these are excluded there remains a group of rare tumors which may rightl}' be regarded as true osteomas. True Osteomas. — These may be derived from the periosteum or from the medulla. In structure the}' may be dense or cancellated. When periosteal they form nodular or globular projections from the surface of the bone and are covered by a vascular periosteum. Those arising in the medulla expand the bone or cause it to disappear by pressure atrophy. Usuall}' in response to the expanding bone the periosteum proliferates actively so that new bone is formed over the expanding tumor. This may disappear and the tumor ma}' project into the soft tissues. This type occurs most frequentl}" in the long bones and in the bones of the face. Macroscopic Appearance, — The surface when denuded of periosteum does not differ essentiall}' from normal bone. The dense varieties are vet}" white in color, while the more spong}' varieties are reddish brown. In the former there are no openings for the passage of vessels, while in the latter, numerous and often large foramina ma}' be seen. The t\pe of tumor ma}' thus be distinguished b}' inspecting the unbroken surface. On section, the dense variet}' closel}' resembles in appearance sawed ivor}', while in the softer t}'pes the cortex often presents the appearance of cancellated bone, while the centre ma}' be formed b\' a pulp}' mass resembling medullar}' tissue, with here and there bone spicules. 58 OSTEOMAS Microscopic Appearance. — Sections of decalcified osteomas closely re- semble normal bone (Fig. 9). Haversian canals are present, but they are not arranged in systems (Fig. 10). The cells vary more than in normal bone; they are often larger, and many are multinuclear. In Fig. 9 V ,V.)- .f ■> :>■" Osteoma. the cancellated variety, cells resembling those of normal medulla may be seen in great numbers, and the intertrabecular spaces are larger than in the normal. Growth. — Osteomas are expansile and benign. When they are asso- ciated with mahgnant conditions, as is often the case, microscopic Fig. 10 Osteoma of the humerus, showing Haversian canals. Study of their character may be inconclusive; but the rate and char- acter of growth often makes an accurate judgment possible. Varieties. — Osteomas are often associated with cartilage, and in many situations are usually covered by it; but here the presence of the cartilage is as much a part of the developmental process as is the I'lLSE osriioM.is 50 fiu ilopiiiL!, |)ii lositiim m oilui Ml ii;i I lolls. hi nilni cases the cliai- acttM" ot tin- i;,i()\\ili is that dI romhincd rlioiulioma and osteoma ami rile teiin chomlKx'isrioma shoiiM \n- lesei \e(l lot this eomhrion. I his illsrimtion is of impoiraiuc-, siiuc thi- eomhim-d rumor is of (httereiu im|>ort from those m wimh eaitihii^e a|)peais meiil\ as a sta^eol descl- opmeiu. The eoml^ined cartilage and bony tumors ai( olreii associared wirh activeh pi oHferarinj; connecrive rissue, and the chondro(")sreo- sarcoma iisulrs. The demonstrarion of rhe Hrsr rwo elements m a lajiidh uiowini:, tumor ofrtn wananrs the assumprion ot rhe jiresence of the third element. When Hhious tissue is present m an osteoma, it ma\- he regarded as an accidental admixture, while myxoid tissue is the resulr of degeneration. False Osteomas. — These are bony outgrowths which are not the result of true neoplastic bone formation. They ma\- result from trauma or from chronic irritation, either mechanical or inflammator\-. Dis- turbances in normal development may result in the formation ot bon\ excrescences about the epiphyseal lines. These are often multiple, and when symmetrical are described under the term exostosis cartila- genia. This condition is brought about by the direction of growth at the epiph^'seal line being transverse instead of parallel to the long axis of the bone, with the result that the bone fails to develop in length. These growths ma^' appear less typically in other situations. 1 hex- have been observed at birth, and frequently the\- appear or undergo more rapid development at pubert}'. Dwarf development or mal- development of a part is often found in connection with them. In other situations, bony tumors are obviously the result of metaplasia of displaced cartilage, which is most often found about the trachea and its branches and in the deeper parts of the lungs, about the ears, in the tonsils, and about the site of the gill clefts. Exostoses are observed about joints in a great variety of joint lesions. They may occur about the vertebral disks. A special name has been applied to exostosis \n the sphenooccipital articulation (the clivus Blumenbachii), namely, chordomas. Bone occasionally develops as a result of inflammatory process in tissues normall\- free from bone, as in muscle in myositis ossificans. Diagnosis. — The character of the tissue is easily recognized. '1 he differentiation between true and false osteomas is often difficult and subject to different interpretations. The character and circumstances of the growth may permit a distinction. The microscopic differentia- tion between these and osteosarcomas is often ditlicult or impossible. Treatment. — Excision is indicated when their situation or size causes annoyance. Osteomas of the medullary t\pe ma)- require the removal of some normal bone before access can be had to them. CH APTE R I I CHONDROMAS AND CHORDOMAS CHONDROMAS General Conception. — Tumors composed wholly or in part of cartilage are not uncommon. Cartilage occurs most frequently as a component of certain mixed tumors which are best considered as a separate group. Trauma sometimes results in cartilage proliferation of limited growth. This IS, strictly speaking, an hj^pertrophy and is usually grouped sepa- rately under the term ecchondrosis. When these cartilaginous tumors, because of their origin or nature, are excluded, there still remains a group which is classed under the head of true chondromas or enchon- droses. False Chondromas. — The ecchondroses or ecchondromas, as they are more properly called when they are large enough to form tumor masses, are, strictly speaking, local hypertrophies. They seldom attain a size large enough to make them of practical significance. They occur at the site of the normal cartilage and are caused by some disturbance in normal development, or perhaps by injury which produces a local overgrowth analogous to hypertrophied scars in wounds of the con- nective tissue. They are composed. of tissue approaching that of normal cartilage and may contain bone at their base. True Chondromas (Enchondromas). — Tumors formed entirely of cartilage have been described at a distance from bone. Those which occur in organs where mixed tumors are known to exist, as in the salivary glands, testicles, and breasts, are most satisfactorily considered as mixed tumors in which the other elements have been either compressed by the growing cartilage, or overlooked in the examination. Many of those described in the older literature were not subjected to micro- scopic examination at all. Other purely cartilaginous tumors have been described which were connected neither with bone nor with organs in which mixed tumors arise. In contradistinction to these are the tumors of cartilage which spring from bone or cartilage or from their immediate vicinity. When not directly continuous with cartilage or bone they are derived from meta- morphosed connective tissue or from misplaced embryonal tissue. Their origin from the latter source seems more probable, both on account CllOSDROM.IH CI of tlu- {ifciiKiu association ot tartihi^.f witli rumors known ro l)c con- genital, and also on account of tluir frc-qiKnt aj^iHarance after injur) . These tumors are found most frequently on the diaph\ses of the ion^ bones near the epiphyseal lines, especially of the hands, feet, clavicle, pelvic hones, and rihs, though none of the bones are invariablv spared. Macroscopic Appearance. Chondromas are dense tumors, usualh' with smooth or nodulated surfaces, and are generally encapsulated. Their cut surface is smooth, glistening, and hluish in color. Frequently, fibrous bands extend for some distance into the tumor and convex* bloodvessels. Not infrequently ossihed areas and various degenera- tions are encountered. Fig. II e \ k 1 " c • V *s ■■•J Chondroma of finger. Microscopic Appearance. — A section presents the appearance usually quite t3'pical of cartilage i Fig. ii). The interstitial tissue stains blue with hematow'lin and reddish \'ellow with safranin. The cells may occur singh' or in groups of two or more; frequently in large groups. They are often of various sizes and may lack a capsule. The variation in cells is the most important deviation, and forms the chief index to the biological character of the tumor (Fig. 12). Large areas ma}' be seen devoid of cells (Fig. 12). Fibrils are often apparent in the inter- cellular substance, and are often continuous with fibrils about the cells. The microscopic appearance of chondroma is often little indicative of the character of the tumor. Tumors containing few cells may grow- rapidly and may invade the surrounding tissue. Growth. — The chondromas frequently become stationary-, but may attain a considerable size, so as to become a serious menace to the activit}- of the individual. The growth takes place centralh' and the 62 CHONDROMAS AND CHORDOMAS surrounding tissue is displaced. Occasionally outgrowths take place at the periphery, which may grow into a vein and metastasis may result. These metastases are like the mother tumor, composed of cartilage, and are especially likely to occur in the lung. The growths may be so extensive that the skin covering it becomes necrosed, thus exposing the tumor to infection. The entire bone about which they grow may be displaced. Fig. 12 V ' ■; \. Metastatic chondroma of the lung, with large cells and very small nuclei. Secondary Changes. — The intercellular substance often calcifies, indi- cating the existence of a previous degeneration. Often there is softening of the intercellular substance, which may give place to liquefaction and thus produce a cyst. Varieties. — Combinations with other tissue are not infrequenth' seen. Myxochondromas are not rare, and osteochondromas are often seen. The natural disposition for cartilage to form bone appears hardly to warrant the classification of these as compound tumors, but often they tend more than simple chondromas to malignant transformation. The most important combination of chondromas is with sarcoma in the chondrosarcoma. The disposition to sarcomatous proliferation is CHORDOM.IS 03 b\' no iiuaiis rare-; it is iliaiactti i/cd h\ i;i|)icl jiiowrh with :i rtndenc)' ro unasioii i set.- Saiconia ). Developmental Malformations. Nor inhcciucnrlv , as a rtsiilr ot some disriirhaiKt in (growth, cartilaginous excrescences are formed about joints, limiting their movements. The most marked of these is the exostosis cartilagenia, which was mentioned also under the discussion of osteomas. Rickets and s\phihs may give rise to similar conditions. Congenital Displacements. In many of the mixed tumors cartilage is an important element. Ihis is true particularly of those occurring in the salivarx' glands and in the testicle, as already mentioned. Isolated islands of cartilage, without the usual accompaniments characterizing mixed tumors, have been found and the question has arisen as to their nature. Some pathologists classify them w^ith congenital displace- ments, while others regard them as examples of metaplasia. 1 he fact that the\" occur in regions in which cartilage normally occurs (trachea) or where mixed tumors are common (neck, ear) makes it seem probable that thev are congenital displacements. It is rarely that the\' undergo further development. Prognosis. — Chondromas are benign tumors, but their frequent asso- ciation with mixed tumors, which are liable to malignant change, and with sarcomas makes the determination of the exact nature of a cartilage tumor a matter of some concern. Those which occur on the fingers and toes are usual!)' permanenth' innocent, but those which arise about the epiph^'ses, and particularly- those which appear following trauma, often necessitate a guarded prognosis. Those which develop rapidlw even if careful examination fails to show an\- evidence of malignance", must be regarded with suspicion. Treatment. — The rapidly growing tumors require removal because of the uncertainty of their character and because of their size. Those of the hands and feet require removal because of the interference with the function of these parts. CHORDOMAS According to Ribbert,' tumors springing from the remains of the noto- chord are to be noted in about 2 per cent, of all autopsies at the spheno- occipital synchondrosis (the clivus Blumenbachii). The dura is some- times perforated and the tumor lies upon or near the basilar artery. The cells are spheroidal with deeply staining nuclei and with clear protoplasm. Minot regards them as epiblastic in origin, while Adami- regards them as hypoblastic. 1 Centralbl. f. Path., 1894, v, 457. ^The Principles of Pathology, Lea & Febiger, Philadelphia, vol. i, p. 761. CHAPTER III MYXOMAS Myxomas are the most embryonal of the benign connective-tissue tumors just as myxoid tissue is the preliminary stage of connective tissue. Topography. — These tumors have been observed wherever there is connective tissue, particularly in the subcutaneous tissue of the back and thigh, less often in the face, genitals, nerves, and medullary cavities of bones. Usually they occur in conjunction with other tissues, particularly in various tumors where the nutrition is insufficient, such as myomas and fibromas; or in tumors growing from mucous surfaces, uterus, nose, or pharynx; or develop in the presence of moisture, as in the spinal cord; or are the results of developmental anomalies, as in mixed tumors of the breast and parotid. Macroscopic Appearance. — Myxomas form spheroidal tumors when small, but when large are often lobulated. They are soft, semifluctuating, and on section are pinkish white and moist or succulent, often with vessels apparent. Microscopic Structure. — The matrix is clear and translucent, and yields a granular or fibrillar precipitate upon the addition of acetic acid. The typical cells are stellate with anastomosing branches, but some may be round or spindle-shaped (Fig. 13). The vessels are large and numerous and are often surrounded by groups of round cells (Fig. 14) believed by some to be the mj^xoid cells in the course of development. Mixed Forms. — Pure myxomas are very rare. They are usually associated with fibrous tissue and less often with fatty tissue. With these tissues they form combinations in varying proportions, but usually the myxoid tissue predominates. When the reverse is true, the myxoid tissue is probably a degeneration product. Embryonal fibrous tissue particularly is Ukely to be associated with myxoid tissue. Such tumors are especially likely to develop malignant tendencies marked by per- sistent local recurrence and finally by metastatic formations. It is only when the myxoid tissue is an active factor that a compound term is applicable, as myxofibroma, myxolipoma. The presence of stellate cells, particularly if the branches anastomose, has been said to be evi- dence that the myxoid tissue is taking an active part, but this is by no means an infallible sign. MIXIJ) KjRMS 05 Myxoid tissue, as most frecjiit-ntly observed in tumors, is the result of def^eneration, and is ^enerall\' observed in tumors where nutrition IS detective and often is but one step in the regressive prcjcess, beinji a 4 Fig. 13 9 q> si ^ t «r ^' • ■^' # ^T-',y-: \>^0^'~ :m^^ Myxoma of the elbow. Fig. 14 4->®l? ;.y%v .rS.^J^T^~' A / i'- - . . . .-j< - * * ' Myxoma showing cells about bloodvessels. followed in turn by hemorrhage and cyst formation. In other cases the occurrence of myxoid tissue is but the result of the conditions of growth, as in nasal polyps, and it is often difficult to determine if the 5 66 MYXOMAS changes in the tumor are really myxoid or merely the result of edema. Often the clinical character of the tumor, and the gross appearance, may aid in arriving at a conclusion; but in some instances, as in the uterus, when edema and myxomatous degeneration are often present, these considerations give little aid; indeed, it seems that a prolonged edema may institute myxoid changes within the connective tissue. The determination of these points is often more than an academic problem; for instance, the myxoid tissue may be but the preliminary step in a series of changes resulting ultimately in the development of sarcoma. Degeneration. — Hemorrhage is said to be a frequent complication. This is true when myxomatous degeneration takes place in tumors of other composition, but m tumors in which the myxoid tissue is an active factor, hemorrhage does not take place with readiness because the vessels, though large, have well-defined walls. Diagnosis. — The soft circum.scribed encapsulated tumors are often distinguished from fibromas and lipomas by the aid of palpation, but usually a gross section of the tumor is required before a dijfferentiation is positive. The succulent jelly-like tissue is then easily identified. If the myxoid tissue is an active factor the entire tumor is of like com- position, while if it is but a degeneration product, only part of the cut surface presents that appearance. Whether a pure myxoma or a compound tumor is in hand the microscope must often be relied upon to decide. Treatment. — The small myxomas are innocent and a simple enu- cleation is sufficient. When large or rapidly growing, the}^ must be regarded with suspicion; their capsule should be removed with the tumor, and every care exercised that the small lobulations and pro- longations be thoroughly removed. This is particularly imperative in tumors of the face, back, and thigh. r II A VJ V. R I V LIPOMAS Tumors composed of fatty tissue are common objects. iliey are found wherever fatty tissue occurs normal!}', particularl\ in the sub- cutaneous tissue, but also in unusual refi;ions, as in the marrow ot bones, beneath rhi periosteum ;in(l in the meninges. In these exceptional situations the\- are merely pathological curiosities. Cjenerall\' increased fatt\- deposits are sometimes noted, the most familiar ot these being lipomatosis dolorosa, a condition first described b\- Dercum in which the general increase of fat is attended by pain. Multiple fatty tumors in the course of certain nerves are sometimes noted. Nervous mflu- ences, therefore, as in the case of those hbromas which are smidarly situated, seem to be associated with their genesis. Macroscopic Appearance. — Lipomas are spheroidal or lobulated tumors varying from microscopic size to enormous masses w'hich are exceeded by no other solid tumors. As ordinarily encountered the\- var}- from the size of a hazelnut to that of an adult head. They are usually composed of a number of lobules each more or less perfectly encapsulated, while the entire tumor is similarly encapsulated agamst the surrounding tissue. In some instances lipomas are more infiltra- tive in character, and fine tongues of fattA' tissue are sent m between muscle bundles or between bloodvessels. On section, lipomas are moist and \ellow, and globules of fat stand out prominently over the cut surface. The tumors are divided off into separate lobules by fibrous septa in which the bloodvessels run. In some instances, fibrous tissue maA' be abundant. Microscopic Appearance.— The fat globules in lipomas are char- acterized, as opposed to normal fat, b\ a variation in size of the mdi- vidual globules. Otherwise the relations of cell, nucleus, and fat globule conform to the normal; as also do their relations to the fibrous tissue and bloodvessels. Course. — Lipomas tend to increase in size. The growth is usually slow but may be quite rapid. They usually are of no clinical impor- tance except from their size. In most cases even large ones may be well borne, since these usually occur about the shoulders or back. But even if they do occur in the neighborhood of important structures their softness prevents compression. 68 LIPOMAS Diagnosis. — Lipomas are the easiest tumors to recognize, whether by the naked eye or with the microscope. The only question is to determine if the fatty tissue is primary or the result of degeneration of other tissue. Usually if other tissue is present the latter is true; for fatty tissue rarely develops along with other tissue. Treatment. — Local enucleation is best performed as soon as the tumors are recognized, since they tend to increase in size. C II A PT F. R \' MYOMAS Each variet\' of muscle Hhers is siihitcr to tumor formation. I he vast majority of myomas are the product of tlu- non-stnated muscle and are called leiomyomas. Rhabdomyomas, tumors which arise from striated muscle, are very rare. Leiomyomas. In these tumors the normal non-striated muscle cell is frequentl)- accurately reproduced. They develop in an\ region where non-striated muscle normally occurs, hut show an esprcial pii-dikction for the uterus. The gastro-intestmal tract, the prostate ^land, and the skin are much more rarely affected. Macroscopic Appearance. -The tumors are spherical oi noduhir m out- line, sharpl\ circumscribed, and usually very dense. On section the knife creaks in the denser varieties, and in even the softer varieties meets more resistance than in normal non-striated muscle. The reason for this is that pure niA'omas are very rare, there being usually an accom- panying increase in the fibrous tissue, and it ma^- even predominate; whence the term fibromyomas or fibroids often applied to such tumors in the uterus. The cut section is whitish or pinkish, dependent upon the proportion of fibrous and muscular tissue respectively. The peripherA' is usually formed largel\' or entirel\- of fibrous tissue in w'horls or bundles which run in various directions. For this reason some bundles are cut longitudinally, others transversely, and others obliquely. Occasionally the lumina of bloodvessels can be made out. Microscopic Appearance. — The cells of these tumors resemble normal non-striated muscle cells very closel)'. The nuclei are long and spindle- shaped, and are surrounded by the protoplasm of the elongated muscle fibers. The outline of the fibers is best appreciated in the teased speci- mens. When a muscle bundle is cut obliquel}- the cells appear as short spindles, and when cut transversely the}- simulate round cells (Fig. 15). In the more rapidh' growing tumors the nuclei are broader and larger and stain less deeply. The cells are arranged parallel to the smaller vessels, but indifferentl}- in respect to the larger vessels. 1 he proportion of fibrous tissue to muscle tissue varies considerably, but is usually greater than in normal muscle. The differentiation may not be plain unless special stains are empkned. With the van Gieson stain the connective tissue appears as a brilliant red, while the muscle 70 MYOMAS stains yellow. The difference is even more strikingly presented with Mallory's stain, which colors the connective tissue a deep blue, while the muscle is stained red. The border of the tumor is always sharply circumscribed, even in the very small ones. This has led some patholo- gists to believe that they originate in a congenital displacement which never has been in organic union with the normal tissue about. The frequent presence of a small vessel in the centre of these small tumors has been thought by some to indicate that the vessel walls are the source of the tumor formation. Fig. is \'N^'= ':SX^X X &\-''- \-'^-^ ■■\ ': \^ -v- ■ \ ^- -V"'; ■, ' ■ ^1'"^ ■' "^> '-^^ ■' ■ ''."' ' • '~7 ' -' '. !■ ' ~~ i.-\ '" 'X\ Myoma of the uterus, showing muscle bundles cut longitudinally and transversely. Growth. — The disposition of myomas is to grow slowly but persist- ently, displacing the neighboring tissue but not invading it. The rate of growth, however, varies considerably. The more fibrous variety may become stationary, while others may grow with considerable rapidity. Very rapidly growing types are sometimes seen and metas- tases in distant organs have been observed. The border-line between these and the much more frequent myosarcoma is difficult to place. Spontaneous limitation of growth may take place by secondary degen- eration. Frequently growth ceases when the organ in which they lie undergoes regressive changes (uterus), but even after such changes renewed growth is possible. When their original source of blood supply is jeopardized they may form attachments to convenient organs and thus receive a new supply of nutriment. Degeneration. — Because of their precarious blood supply they are subject to the various degenerations resulting from a limitation of nutriment. The most frequent is a myxoid degeneration (Fig. i6). Many tumors show this in some region. Hyaline degeneration may MIM.l) I IKII/niiS 71 result. Fatty defeneration is rarcl\ ohsirvid. An\ ot these t\pes mav liquefy and j^ivc rise to c\st toiinarions or ni:i\ form cavities info wliiili lunioriha .■ . ' f^. Edematous fibroid of the uterus. Secondary Chaiiges. — -Myxoid degeneration is the most frequent retrograde change in these fibromas. It occurs in the older parts ot the tumor where the blood supply is least perfect. Fatt\' degenera- tion is sometimes seen, often associated with the myxoid degeneration. Liquefaction in such degenerated areas ma\' occur, giving rise to cysts, though this is rare. Hemorrhage into the tumor and the formation ot cholesterin crystals have been noted. Calcareous infiltration is often associated with other degenerations as an end result. The lime salts are deposited among the fiber bundles so as to form spicules which otten resemble bone. True bone formation has been noted, but evidently many of the cases reported as such have really been calcareous infiltra- tion. ^ For the finer structure of fibromas, see Mallory, A Contribution to the Classifica- tion of Tumors, Jour. Med. Research, 1905, xiii, 113. 76 FIBROMAS Soft Fibromas (Fig. 21). — The soft fibromas resemble closest the submucous reticular tissue and are usually found just beneath the mucous or cutaneous surfaces. They may be either circumscribed or diffuse. Those which arise from the deeper tissues are usually cir- cumscribed and form irregular masses. The subcutaneous variety is inclined to be diffuse, so much so that in many cases it is difficult to distinguish them from inflammatory thickenings, especially since the conditions often co-exist. Both types are frequently edematous because of the disturbance of circulation. They are soft and semifluctuating on palpation. A cut section is white or pink in color, and usually glistening. A network of fibers can usually be made out, and in some areas the fibers may be so abundant as to cause resemblance to the hard variety. Fig. 21 i^/ '-e: #'-' Soft fibroma of the subcutaneous tissue, showing loose reticular tissue, round connective-tissue cells, and bloodvessels. Microscopically the fibers form a network containing connective- tissue cells within its meshes. These cells are round, with darkly stain- ing nuclei and a considerable amount of protoplasm (Fig. 5). They are most abundant about bloodvessels. Stellate cells are frequently observed as well as large cells, with opaque, faintly staining nuclei, with one or more deeply staining nucleoli. In regions of inflammatory reaction polynuclear leukocytes may be seen. Bloodvessels may be present in great abundance, and thin-walled capillaries are seen every- where coursing through the loose meshwork of fibers. The presence of these vessels and the stellate cells frequently lead to a diagnosis of myxoma. A chemical test will obviate this error. In the fibromas involving the skin the elastic fibers are quite uniformly increased.^ ^ Baldwin, Vaughn Anniversary Contributions, Wahr, Ann Arbor, 1903, p. 495. spKc/./L I jki/rriKs or i ihrumis 77 Secondary Chaugt's.-Vh*:^^: an limited to in\ xoniatous and fatty changes and cdcnia and an- most h((|iuiit I\ the result of inflaniniatory processes. So otiiii ;iic intlaininatoi\ piocesses assf)ciared with the orhei ihmiui at loiis that Hu\ woiihl stiiii to he in inaii\' msfanees the pi imar\ (.'oniiihca t ion. Mixed Varieties. .\ tumorous mt lease ot hhi(;us tissue is iieciuently associated with othtr r\ pes of tissue. These tissues mav he of lower, hke, or hit^hei" order. I nless the hhrous tissue takes an active part in the tumor forma- tion it should not he represented in the name. Fre(juentl\' a tumor IS made up laii;el\ of fihrous tissue without conn ihurin<; an\ thm^ to the essential nature of the tumor, as m so-called Hhrocarcinoma, in which the connective tissue is merely the supporting framework; nor should a compt)und term be employed when the associated tissue is the product of a degeneration, as in my.xoid degeneration of a fibroma. There is in general a disposition to employ the prefi.x "fibro" when- ever there is an easil\' perceptible amount of fibrous tissue, irrespective of its histogenic importance. The essential nature of the tumor is often obscured because of such fault}^ nomenclature. Some of the more common mixed fibrous tissue tumors are: Myxofibrom.a.s. — Myxoid tissue is usually found with the soft type of fibroma in the deeper tissues, especially of the nervous system. LiPOFiBROMAS. — The fibrous element of these tumors is generally of the hard variet}'. The\' are oftenest situated in the subcutaneous tissue and in parenchymatous organs. Fibromas growing about and enclosing lobules of preexisting fat and lipomas containing thick fibrous septa are not properl\' included in this variet\'. True fibro- lipomas are rare and of little importance. FiBROMYOMAS. — Non-striated muscle fibers form the most common combination with fibrous tissue. All degrees of admixture are found. This t\'pe occurs most frequently in the uterus, and is discussed under m^'omas. Adenofibromas. — These tumors are usuall\' found in the mam- mar}' gland. The^' are hard fibromas which have included in their growth relativeh' normal gland acini. The tumors in which the glands are active participants will be discussed under adenomas. Fibrosarcomas. — This term is properl\' reserved tor sarcomas of low degree of malignanc}- in which the fibers nearl}- reach full develop- ment. Often some regions of the tumor present completeh' developed fibrous tissue. Special Varieties. — There are a number of new formations made up of fibrous tissue which form distinct types, and are best described as varieties of fibromas. Manv of these are not true tumors, but since they appear 78 FIBROMAS clinically as such they are included in this discussion. This is partic- ularly necessary, since some of them may not appear as tumors under some conditions, but do appear as such under other conditions, e. g., elephantiasis. Keloid (Fig. 22). — By keloid is meant a dense fibrous growth which occurs in the subcutaneous tissue usually at the site of a scar. Two types were formerly recognized — the true, in which no previously exist- ing scar was present, and the false, which developed from a scar. It is now assumed that a very small scar may be the point of origin of the true variety as well, and therefore both types are genetically the same. Fig. 22 Keloid, showing large connective-tissue fibers and nuclei of fibrous-tissue cells. These varieties cannot be differentiated pathologicall}' or clinically, whatever may be the etiological factors. They ma}^ be round, lobu- lated, or irregular in outline, projecting a centimeter or more above the surface. When they grow from a scar the}' usually take the form of the scar itself, but they ma^' extend irregularly into the surrounding tissue. They are dense to the touch and may be more or less fixed to the underlying structure, particularly if they go out from scars involv- ing the deeper tissues. When arising spontaneously or from slight injuries, they are usually movable over the underlying tissue. They are found most frequently on the chest, back, face, and lobes of the ear. The negro race is especially predisposed. Histologicalh" they are characterized by very large homogeneous E I. Ill' 1 1. 1 \ ri.isis 79 fibers iiUtispt Tstcl with tew coniuttiNc-tissue cells li;i\ inu mikiII inttiiseU staining luick-i i I" ij;. ii). Iluii liisrolojiN' ^l\•l■s no cliu- lo rlic cause of tluii' (Jisi^osition to recur. 1' iiiiuciulv ui tin- luahnti, of wouiujs excessi\c- scar tissue is forimd and masses reseiiihlin^ keloids may he produced. liie)' are distin- jiuished from riie lattir hy the ^:reater numher of cells and by the smaller fibers ( I'i^. 23), ami h\ rheir association with recently healed wounds ot somr ma<;mriidc, |iai riciilaily burns. I hey tend to iirroji;ressive chanj^es. Sonuwhar anal()n:5 ^% -_. ^ - "4 ■~1—- '^-^ Hypertrophied scar, showing large connective-tissue ceils and fibrillae. These are made up of more or less cellular tissue with large fiber bundles. This term is properly' applied to tumors of this character occurring in the abdominal wall, which will be considered in the discussion of the tumors of that region. Elephantiasis. — Elephantiasis is a diffuse fibrosis of the subcuta- neous connective tissue, and to a less degree of the fibrous tissue of the deeper lying structures, the intermuscular septa, perivascular sheaths, etc. Two types have generally been distinguished, the con- genital and the acc]uired. Congenital. — Usualh' onh' the anlage of the growth is present at birth, but soon afterward it begins to develop rapidly. In acephalic monsters it is sometimes well developed at birth, and it may e.xtend over the entire body. This variety has been proved in recent }-ears to be formed from the nerve sheaths (elephantiasis neuromatosa, v. 80 FIBROMAS Recklinghausen's disease^. It is, in other words, a diffuse fibromatosis of the nerve sheaths. Later the process is beheved to extend to other fibrous tissue, especially that of the blood and lymph vessels. A further indication of intimate connection with the nerves is the fact that elephantiasis is frequently combined with true tumors in other parts of the nerves. Pachydermatocele, a similar process, has likewise been proved by Herczel^ to go out from nerve sheaths. Acquired. — This type occurs endemically in the tropics and spor- adically in other climates. In the tropical form the whole limb is affected, owing to occlusion of the main lymph trunk by the Filaria sanguinis hominis. It is very rare in temperate regions, and interest in it is scientific rather than clinical. In this country a localized form in the genital region not uncommonly follows the operative removal of the inguinal glands or their destruction by inflammation. It may also be produced hy diffuse inflammation of the cellular tissue. Often Fig. 24 Elephantiasis, showing fiber bundles and cells. the localized form in other parts, such as the scalp and hips, has no obvious cause. It is characterized by soft flabby masses which pro- ject or hang from the point of origin; often the epidermis is irritated and a rough encrusted or weeping surface results. Pathologically the earliest change is in the skin, the corium first becoming thickened, and then the papillae becoming elongated, hypertrophied, and branched. The epidermis is thickened and the interpapillary processes proliferate. The principal change is in the subcutaneous connective tissue, which ^ Von Recklinghausen, Ueber die multiplen Fibrome der Haut und ihre Beziehung zu den multiplen Neuromen, Berlin, 1882. ^ Ueber Fibrome und Sarkome der peripheren Nerven, Beitr. z. path. Anat. u. z. allg. Path., 1890, viii, 38. /)i.i(..\osjs or iiHko.M.is si becomes nuieli thitkeiutl, the filnis Inconiin^ laijie aiul, unless etleina- tous, dense. Round cells max In- diHused over lar^e areas or collecred into groups (Fig. 24). I lie connecriN e-tissue spaces contain a in\ xoid tissue which renders tin mass firm ami elastic to the t(juch. 1 he \eins are often the seat of noilular hhrous thickening or saccular dilatation. The arteries show marked thickenmji of the walls ami, in spite ot this, an iiu iiase m calilni. I lu l\ nipli \essels and spaces are maikiilU thiated, and their endothelial lining cells often proliferate, prohahh on account ot a chemical change in the composition of the l\mph. Xanthoma. — Ihese are peculiar yellow or brown tumors which occur in the skill. .More than three-fourths of them are on the e\ elids. rhe\' are treiiuenrly multiple, and are then arranged along the C(jurse of a cutaneous nerve. I hey form flat nodules or patches, but sometimes may be large enough to become pol\'poid. While the\' seem alwa\'s to begin in the cutis, they extend deeper into the connective tissue and even into the muscle and periosteum. They appear frequenth- where the skin is subjected to pressure or over bony prominences. In some congenital cases the}' are located more deepl\' under the skin. They are composed of fibrocellular nests which extend into the Kmph spaces. There is always a la^er of cutis between the tumor and the epi- dermis. The color of the tumor is due to a \ellow pigment in the cells. The chief characteristic of the tumor is the occurrence of large cells which look like fat cells, and which led \ irchow to call these tumors fibrolipomas. Some of them have giant cells resembling those of sarcoma, and are frequenth' associated with icterus and other hepatic disturb- ances. Diagnosis. — Macroscopic. — The hard tibromas are well-defined, en- capsulated tumors, dense to the touch and on section. The}' are pearl\' white in color and show wavy bands and whorls. In the ver\' vascular t\pes the lumen of the vessels may sometimes be detected b\' the naked eye. The softer varieties are less sharpl\' circumscribed except in cases where the\' are pedunculated. Lsuall}' on section these show bands of fibers in a glistening homogeneous matrix. Their slow growth and the features detailed above make the diagnosis eas}'. Microscopic. — In man\' instances the microscopic diagnosis is more difficult than the clinical. The sparseness ot the nuclei, their uniform size, oblong shape, and arrangement parallel to the fibers ma\' be regarded as the most important signs of fibroma so far as the nucleus is concerned. The parallel arrangement of the fiber bundles and the bloodvessels is evidence of a benign tumor. However, in some fibromas the nuclei are less regular in form, particularl}' if there has been some secondar\' change, and the fiber bundles ma\ var\' from the normal arrangement. In such cases the diagnosis depends upon the histor\ ot 6 82 FIBROMAS the growth and upon special considerations which will be taken up in connection with particular regions. Prognosis. — The prognosis in all varieties of fibromas as regards Hfe is uniformly good. There is no disposition to form metastases, and when simple fibroids are once removed they do not tend to recur. In some of the soft types which extend diffusely into the tissue new tumors may recur in the site of operation. With keloids local recur- rence is the rule. Treatment. — Fibroids when the}^ cause annoyance by their size should be removed. The hard type may be readily shelled out by splitting the skin. In the soft type, particularly when pedunculated, an elliptical incision about the base may be made in order to remove the redundant skin. The keloids tend to return when removed, and had better be left unmolested unless located on an exposed surface, when excision may be practised under a guarded prognosis. Elephan- tiasis when localized may be excised. When an entire limb is involved, ligation of the chief arterial supply may be undertaken. Recently, reports have been made of keloids treated by the A;-rays by Pusey,^ Morton,- and others. ^ Jour. Amer. Med. Assoc, 1902, xxxviii, 916. 2 Med. Record, New York, 1903, Ixiv, 124. C II A 1' r K k \ I I GLIOMAS Conception and Definition, (iliomws aw tuiiiois toinu-d tioin tlic supportiNf tranuwoik. of the central lurxous system, and are found not onh' in the brain and spuial cord, hut also in the e\e. An exact pathological conception of glioma is difficult; the structure of the glia cells is not sufficiently- understood. Ihere is still a question as to the relation of the fibrillar network to the cells. W'eigert believed that b\' the use of his stain he had demonstrated their independence of the cell. Others' regard the fibrilhe as prolongations of the cell proto- plasm, and therefore as a part of it. Unlike other sustentacular tissue, glia cells are derived from the epiblast. Gliomas are distinctive enough in their clinical course to warrant a separate consideration. They are the most frequent true tumor involving the central nervous system. The\' occur most frequentl)' in earl\- life, usually in children. Macroscopic Appearance. — Gliomas present the physical characters of the sustentacular tissue of the central nervous system from which they spring. They are little different in color from the gray substance and are gra^'er than the white substance. Most cases of so-called brain hypertrophy are due to such tumors (Oppenheim). Frequently, they are differentiated from the surrounding tissue by their greater vas- cularity, the individual vessels being often distinguishable. The}' are usually infiltrative in growth and have no sharply defined border. They can usually be distinguished from the surrounding brain substance b\' their greater softness, which is sometimes increased by secondarx' changes. In some cases, however, and this is particularly true of the smaller circumscribed tumors near the corte.x and in the ventricles, the\- are denser than the surrounding tissue. Microscopic Appearance. — Normal glia cells have round, deep-stain- ing nuclei of uniform size and very little protoplasm. In gliomas the cells vary in structure; they approach the normal at the peripher\- of slowly growing tumors, and deviate most toward the centre of the rapidly growing tumors. According to the degree of variation the cytoplasm is increased, and its nuclei become larger and ovoid; in 1 Taylor, Jour. Exper. Med., 1897, ii, 611; Golgi, Glioma des Cehirns, Fischer, Jena, 1884; V. Lenhossek, Der feinere Bau des Nervens^stems, Fischer, Berlin, 1895. 84 GLIOMAS Fig. 25 c> M"^ i^\ & /y' Glioma of cord, showing glia cells with processes and connective tissue about a bloodvessel. Fig. 26 ^M'H W-^i% . "^ fcf' ?\.' : . <%. f d ^-f. ^'^,^ f ;, :# © ^ \(^ (B Glioma of brain, showing glia cells without processes, and connective tissue. modi: (>i (.Ron rii or (./.io.m.is s5 rlu- i;i|Mill\ ^louinii i:,ln»iu;is wiili ;i hii;li di-^iii- ol \ ;ii i;i t loii tlu- fll)rils becoiiu' it-l:iti\ cl\ lew. In sp(rinuns wliuli ;ii( teased or staiiucl hy special nurlKuls, tin- s|Mclii-likt loiin ol tlu <;li;t nils l^ccoincs apparent (Fiji;. 25). Ill tlu- tastii i:i()\\m<; t\ jus tlu- cills aic less perfectl\' formed ami tlu sullati- aiul spiiullc-shapt-d \aiic-tus pii-doiiiinare (Hfi;. 26). 1 lu nils iiKi\ In- so hum- as to ii-scnd^K- ^aii^hon tx-ils. Differentiation from Sarcoma. ilu- naiiu- fiiiosarcoiiia has Ixcn apjdu-d to lapulh ^ row 1111; tiiinois ot tlu-st- lai^i-i fells. Ilu- tiansi- tion ot a ulioiiia iiUo a sarcoma is unlikely, for ^lia cells have been .shown to be of cpiblastic orij^in. Ilu- chitf itscniblaiui to sarcoma is in rhe relation of the cells to the vessel \salls; rlu- iniitation is \ei\ close-, aiiel h\ histological nuans often inipossihk- to eletect. Mallor\' is contidenr that differentiation is possible b\' means of his technujue, but untortu- nateh' material Hxed in Zenker's Huiel only permits ot such stuel\ , a fixative not often used in the material presented tor examination. Borst has sought to escape this difficulty by dividing such tumors into two classes: (i) The verv cellular gliomas he calls glioma sarcoma- toides, indicating thereby only a rapidly growing glioma; (2) the glio- sarcoma, in which both kinds of tumor are growing side b\- side. Against this classification one may say that the name of the first type is mis- leading and that the second type has not been observed. The resem- blance between glioma and sarcoma is increased by the presence of intercellular fibrils in both;' and to establish the existence of glia cells the fibrils must be shown to be connected with the c^-toplasm. But in rapidl\- growing tumors man}' glia cells are free from processes; the presence of small, round or oval, deeply staining nuclei with a small amount of protoplasm, so suggestive of sarcoma, renders the problem ver)' difficult. In obscure cases other points ma\- aid in diagnosis: The presence of giant cells, degenerated nerve fibers, and corpora am\lacea (Stroebe) speaks for glioma; in glioma, cells are not arranged so definitely' about the vessels as in sarcoma; sarcomas devele)p from the meninges and the vessels of the brain only. Mode of Growth. — Gliomas are essentially inhlrrative in growth, though expansile growth may occur rarely in the smaller tumors, which are then quite sharply defined from the surrounding tissue. It is par- ticularly noteworthy that these tumors do not grow beyond the border of the neural tissue in which they develop. Ihe spaces may be mhl- trated, but the surrounding membranes are never penetrated and destroyed. Metastasis does not occur in primary glioma of the brain, but it has been observed in gliomas arising from developmental di.s- placements (Mallory) and from the retina. Secondar\- gnnvths from ' Stroebe, Centralbl. f. path. .-Xnar.. 181)4, v, 855; 1806, vii, 643. 86 GLIOMAS the retina have been found in the brain, the meninges, and the spinal cord (Arnold, v. Recklinghausen). Although gliomas lack many of the biological characters which ordinarily belong to malignant tumors, they are of grave danger to the life of the patient because of their posi- tion. Some of the very small ones, particularly those growing in the lateral ventricles, are innocuous; and a number in the cortex have grown so slowl}^ as to appear only at postmortem exammation. Ordinarily, however, from their size and location they produce compression and destroy the function of the organ in which the}^ lie. Secondary Changes. — Edema and myxoid changes sometimes occur. The former is more often seen at postmortem than at operation, and, indeed, it seems possible that edema is a terminal manifestation and not a factor of importance in the life history of the tumor. Myxomatous changes are most likely to occur in those types in which the fibers predominate. The term gliomyxoma has been applied to these tumors, but the myxoid element is probably always due to secondary changes. Hemorrhage is the most important accident to which gliomas are liable; this is easily understood when the number and histology of the vessels are kept in mind. Hemorrhages which cause death may be so extensive as entirely to obscure the tumor, and it may remain un- detected, unless a microscopic examination of the region of the hemor- rhage is made. The occurrence of hemorrhage into the brain substance remote from parts ordinarily subject to apoplectic accidents should alw^ays suggest the possibility of ghoma and lead to a careful micro- scopic investigation. After hemorrhage the blood may be absorbed and leave cysts filled with a straw-colored fluid, the origin of which is indicated by the presence of cholesterin crystals. Necrosis and softening of the tumor area sometimes takes place, with subsequent cyst formation and death of the patient. Gliomas of the Cord. — Gliomas of the cord show two tendencies which require mention: (i) They tend to infiltrate the cord and form solid tumors which reach for considerable distances; (2) these long growths tend to cavit}^ formation and resemble very closely the process called syringomyelia. In the cavities thus formed ependyma cells are frequently, found. Oppenheim does not believe that this condition should be considered with gliomas. Gliomas of the Retina. ^ — Rapidly growing tumors of the retina are observed particularly in children, and are regarded as gliomatous in character. They are soft yellowish or grayish tumors which develop in the interior of the ej'e, perforate the sclera, and invade the orbit. 1 Emanuel, Vircliow's Arch. f. path. Anat., 1900, clxi, 339; Wintersteiner, Das Neuro- epitheHoma retinae, Leipzig, 1897. A/./ koCI.IOM I s~ 1 lu\ tiiciiunt 1\ iii\;uli- ilu- M()()(l\ essils aiul itacli ciisraiu ()r;;ans in this iiianiui. I his nuans ot chssiniinarioii, tont-rhir with their vci\' rapid growth, oaiisis thtni m this sitiiation to iiscmhic- sarcomas even more than those aHeetmj; the- hiam. I he small size ot the cells and their perivasculai aiiangement sometimes makes a distinction impos- sible. Iheir nature is su^^ested h\- the resemblance to the retinal cells, but a cknionstration of ^lia Hbers is necessar\' lor a j-xjsitive identification. Neuroglioma. -Some writers' believe that neural and ^lia tissue ma\ develop simultaneously. Ganglion cells, however, probably do not develop, and those cells which resemble or represent them are probably embryonal cells. Treatment. Diffuse gliomas are inoperable. When encapsulated they can be shelled out. Thomas and Hamilton, Jour. Exper. Med., 1897, ii, 635. CHAPTER VIII NEUROMAS Conception. — Neuromas are rare tumors derived from the essential elements of the nervous system. Any of the various nerve elements may at times be the seat of tumor formation. Much confusion has arisen from failure to distinguish between those tumors dependent upon an actual increase of nerve elements and those in which the nerve elements are purel}^ an inclusion. Chief of these spurious tumors are the fibromas arising from the fibrous tissue in and about the nerve sheaths. Strictl}^ speaking, such tumors should be discussed with fibromas; but because the exact differentiation is frequently the task of the laboratory it seems best, on practical grounds, to discuss them in this connection under the general head of pseudoneuromas, expressing thereby their association and topography rather than their histological structure. True Neuromas. — In this group are included those tumors in the development of which nerve tissue itself takes an active part. Each element of the nervous system may form tumors, so that it is neces- sary to recognize ganglion-cell neuromas and neuromas derived from the nerve fibers. The latter are again divided into the myelinic neuromas derived from the medullated nerve fibers, and the amyelinic neuromas derived from the non-medullated nerve fibers. Some pathologists^ question if tumors made up of nerve fibers alone ever occur, but it seems certain that such true neuromas do occur, though they are very rare. Knauss^ has recently attempted to include many tumors generally classed as pseudoneuromas, but his opinions have not received general acceptance. Neuromas are found in the central and peripheral nervous systems. In the central nervous system they occur most frequently near the lateral ventricles. It is believed that they are dependent upon developmental errors, though it is difficult to distinguish between origin from developmental errors and from masses of misplaced tissue. Better defined are the neuromas of the peripheral nerves, particularly of the sympathetic system.^ Here the ganglion cells are particularly numerous and are often of large size, and may contain many nuclei. ^ Ribbert, Geschwiilstlehre, Cohen, Bonn, 1904, p. 320. 2 Zur Kenntniss der achten Neurome, Virchow's Arch. f. path. Anat., 1898, cliii, 29. ^ Schmidt, Virchow's Arch. f. path. Anat., 1899, civ, 557. psi:ri)OM:rR(j.\i IS ^\) Neiironuis art- bt-m^ii tiiniots, luit a tVw casts have been reported in which iiK-rasrasis occunecl. 1 hesc arc hchc\c(l to have been formed from undcxclopctl 'iaii'ilioii cells. Borst has pointed out that the cnor of mistaking (jther conditions for true neuromas depends upon two causes: (i) Ihe nerve cells ma}' appear as inclusions in other tumors, for example, in fibromas, gliomas, or sarcomas, and thus play only a passive role. This error may be avoided by remembering the topographical relationship of the tumor in question. Ihus if a sarcoma is developing into a region normally occupied by nerve tissue, a discovery of nerve fibers should cause no confusion. Confusion is particularly likely to occur in tumors develop- ing in teratomas where no normal location for nerve tissue exists and where the structure deviates from the normal. (2) Developmental dis- placements without any newgrowth of tissue may be mistaken for tumors. Unless such displacements show an active independent development, they should not be regarded as tumors. Artificial displacements of nerve tissue during operations and autopsies have caused confusion. Pseudoneuromas 'fibromas of the Nerve Sheaths, Multiple Neuro- mas). — These tumors are usually derived from the endoneurium and perineurium, less often from the epineunum, and may vary from a barely discernible size to the size of an orange. 1 hey are usually spindle- shaped, but are sometimes globular or irregular and are often bilater- alh" symmetrical. This symmetry, together with their very early or even congenital appearance, suggests a developmental origin. These tumors may occur in great numbers distributed over the entire body. Prudden^ describes a case in which there were eleven hundred and eighty distinct tumors. Their hereditary occurrence has been noted. They may involve either the sympathetic or cerebrospinal nerves and ganglia, both spinal and sympathetic being not rarely involved; when they occur in the ganglia, ganglion cells are, of course, present. Askanazy- has studied these tumors in the gastro-intestinal tract, where they were located in Auerbach's plexus at the mesenteric junction. Their inde- pendence of nerve fibers is attested by the absence of disturbance in the conductivity- of the nerve even in extensive tumors. This is easily explained anatomically by the fact that the nerve fibers are pushed aside by the growth of fibrous tissue and onh- late or not at all, do the fibers exhibit evidence of degeneration. When this does occur, pressure is most likel\- to be the cause. The myelin sheaths may disappear, but the axis cylinders remain intact. Types. — The clinical significance of some of the types has led to various subdivisions. The following ma\ be distinguished: ' Amer. Jour. Med. Sci., 1880, l.xxx, 136. - Festschr. f. Neumann, vol. ii. 90 NEUROMAS Plexiform Neuromas TFig. 27) Rankenneurom, Bruns). — These are characterized by a more or less diffuse thickening of the nerve sheaths of a region interrupted by locaHzed thickenings or nodes. In addition to the increase in thickness of the nerve sheath it is hkewise increased in length beyond the requirements of the region, and since these tumors have a predilection for the head and neck they have been compared to the cirsoid aneurysms. They are often accompanied by an elephan- tiatic thickening of the fibrous tissue. The fact that the nerve, as a whole, IS both thickened and lengthened has led some authors TZiegler, Birch-Hirschfeld, Klebs) to believe that the nerve elements take an active part m the process. Ziegler believes this fact is likewise indicated b}^ the presence of an excessive number of nerve fibers. Borst believes that connective-tissue fibrils may be mistaken on section for nerve fibers, a mistake which, indeed, might easil}^ happen. No attempt has been made to compare numerically nerve trunks in normal regions with areas involved in such tumors. The lengthening may be onl}" inci- dental, the nerve fibers being merely stretched by the fibrous-tissue growth. Fig. 27 Plexiform neuroma. fPrudden.) Traumatic (Amputation) Neuromas (Fig. 28). — The attempts of nerves at regeneration after amputation form one of the most common examples of false neuromas. The fact that these growths are purposive justifies their classification as pseudoneuromas, otherwise their proper place would be in the group of true neuromas. As experimental transplan- tation of entire limbs becomes possible, these regenerative efforts on the part of the nerve may appear in a new light. The development of fibrous tissue seems merely to keep pace with the ineffectual though purposive efforts at regeneration by the nerve fibers, and it might be questioned whether these growths have a place in the pathology of true tumors at all. Borst now classes them with the hypertrophies. Identical processes occur in wounds of nerves. Macroscopic Appearance. — To the naked eye the false neuromas present the appearance of fibromas in general. The smaller ones appear dense and glistening, the densit}^ depending largely on the absence of TKEITMi.Sr Oi A A/ KO.M.IS 91 stc()iul;ii\ (.luiiiKes. Maii\' of tin hii^ir of tlusi- tumors arc sciiu- HiKruatinr()mas. In ilu- pli xitorm auA traumatic t\pcs tin- direction ot tlu- hln I huiullcs can usuall\ hi- made out, and they present a noduhir and <:n;irKHl suitace. Fig. 28 v^% ,^^ Microscopic Appearance.— In all the foregoing t^pes the structure is that of fibroma. The tissue is dense; the fibers are large and less closely arranged than in fibroma, and the cells more numerous and rounder. In the solitary or multiple type found on the peripheral nerves the fibers lie in interlacing bands containing in the meshes a varying number of oval or globular cells. The fibers are often pushed apart by a myxoid material and present the appearance of myxofibromas, but give the chemical reaction of mucin imperfectly or not at all. It is difficult to distinguish between connective-tissue fibers and nerves. Knauss has suggested teasing the fresh tumors in a solution ot osmic acid as a means of diflferentiation. When ganglia are the seat of these tumors ganglion cells are found. They may not differ from the normal, but may sometimes show the effect of compression and often have numerous processes. Treatment.— No general rules can be given. Painful tumors occupy- ing nerves functionally unimportant may be excised, while those affectmg important motor nerves should he undisturbed. Amputation neuromas should be excised. CHAPTER IX SARCOMA General Conception. — While fibromas represent a type of newly formed connective tissue which goes on to the formation of purpose- less adult tissue, sarcomas, starting from morphologically the same elements, differ from the fibromas in that they do not reach the adult stage of completely developed fibrous tissue. They are made up of primordial fibrous tissue, that is, of cells and intercellular substance. A sarcoma may be defined, therefore, as a tumor formed from the connective-tissue elements, cells, and intercellular substance, which fails more or less completely to reach the state of mature development. Fig. 29 0»^ ^ ^V i ) Young connective tissue from a healing wound. Imperfect as are the attempts at formation of adult tissue in some cases, in others the attempt is more or less successful, giving rise to admixtures of tissue which approach the adult nearly enough to be recognized and identified. Such combinations are described by the terms fibrosarcoma, chondrosarcoma, etc. The clinical course of these tumors is modified usually by the degree of maturity of the tissues composing them. Therefore, from the degree of perfection in develop- ment the life history of the tumor may be inferred. STRlCri Rl. Ol S.IRCOM.I 93 Structure. Since sarcomas arc foriiutl of immaiinc connective tissue, we may ex|Hcr to HncI tluii proror\ jx- in emhrvonal or develop- ing connective tissiu-. In the latter particular!)' do we find the nearest I'iG. 30 ' ' o ' <• - : ' «••'*■ ® *- '' • "♦•*.■- •' •-»'.•«>,>?•> ft Young granulation tissue, corresponding to a round-celled sarcoma. approach to sarcoma structure. More than this, it is highly probable that sarcoma develops from such tissue which, failing for some reason to attain maturity, tends to unlimited growth. The primal elements, cells, intercellular fibrillae, and bloodvessels, are observed in all develop- FiG. 31 ^ Old granulation tissue, corresponding to a fibrosarcoma. ing connective tissue (Fig. 29). A preponderance of cells indicates a primitive state, while a preponderance of fihrdlae indicates a more mature state, of development. In the immature states the bloodvessels are 94 SARCOMA mere channels lined by endothelium. In the very early states even the endothelial cells are absent and the bloodvessels are mereh^ channels among the cells (Fig. 30). As the tissue attains maturity the vessels acquire a connective-tissue sheath (Fig. 31). The life history of develop- ing connective tissue is reflected in the biology of sarcomas. The more primitive of these tumors correspond closel}^ to the earlier stages of developing fibrous tissue, while the more highly organized types of tumors correspond to the more full}^ developed fibrous tissue. There is no sharp dividing Hne between the more perfectly developed sarcomas and the fibromas. Sarcomas of all types, therefore, are but variations of the same structural plan, and it remains to study the various combinations which the mature and immature tissues present. In order to do this more clearly it is convenient first to study separately the various elements which go to make up the tumors. These elements are fibrous tissue (stroma) and cells (parenchyma). Connective Tissue (Stroma). — With the connective tissue of sarcoma it is convenient to include the bloodvessels, and since they are usually very conspicuous they may be considered first. The bloodvessels are composed of a very delicate tube of connective-tissue fibrils lined with endothelium. This endothelium is usually higher than that lining normal bloodvessels, and the nuclei project into the lumen. In the smaller vessels the cells frequently form a nucleated syncytium, which may surround the vessel for a considerable distance. The amount of connective tissue in the vessel walls is usualh^ in proportion to that existing among the sarcoma cells. When this is demonstrable with difficulty it is usually hard to detect about the bloodvessels, and may require special methods. Van Gieson's or Mallory's stain demon- strates the larger of the fibrils, but a complete demonstration can be made only by digesting out the cells according to Mall's method. The more rapidly growing tumors may be composed in some areas entirely of cells. In such regions the bloodvessels are merely channels among the cells, and it may be impossible to detect fibrils by any method. Everywhere, except in certain areas of the most rapidly developing tumors, fibrils are to be found among the cells. They are usually most prominent in the region of the bloodvessels and extend from their walls. Not all connective tissue found within a sarcoma is to be regarded as newly formed; stroma from tissue of the part invaded may come to He in the interior of the growth. In this way bloodvessels with thick walls may be found within such tumors. In expansile tumors all stroma is newly formed and thick-walled vessels are not found. Parenchyma. — In the various types of sarcomas cells of every grade of development may be found, from the normal cells in the more highly rorocR.ii'iiic.iL Ri:i..iri<)\s or s.ircom.i 95 developed vaiit-tics to iiiulifferentiarcd tMihr\()nal cells in rlu- rapitll\' ^rowin^, siiiall-cclk-d r\pcs. (jc'iu'iall\ , howc-Ntr, the- ctlls arc miniature i\fn in tin- nioit- iHiftct t\ pes of tumor tissue, and it is to this tact that the tiiulcnc\ to unhmitid ^lowth is to In- ascnhcd. Usuall)' the cells are larji;ei" and contain more protoplasm than normal cells. I'hus in the most fully developed fibrosarcoma as contrasted with hbrous tissue a great variety of cell forms is observed. In osteosarcomas the bone cells a le uregulaily arranged, and show \anations m size not ob- served in normal bone. In man\' sarcomas there is a prevailing kind of cell, which ma\' all be small or large, round or spindle-shaped; bur usualh there are xanations in size and toiin. Different varieties ma\' be mingled, or one t3'pe may persist in one part of the tumor while an entirely different type prevails elsewhere. This is especially likely to occur when one portion of a tumor suddenly begins to enlarge rapidh'. On this account study of the forms and varieties of the cells often indi- cates the life histoi"\' of the tumor. The cells multipl}' for. the most part b\' mitosis, direct division being more rare, but the process is often varied from the normal, and aberrant divisions are frequent. These irregularities may furnish a clew to the abnormal development of the tumor. The number of mitotic figures observed in a microscopic field is an index to the rate of growth. The cells of a tumor invanabl}- originate from other tumor cells. Degeneration of a fulh" developed cell into a tumor cell probably does not take place. Fibromas said to degenerate into sarcomas may have been sarcomatous from the beginning, or the malignant growth ma}' have sprung up within the higher t}'pe of tumor. In the latter class of cases \oung connective-tissue cells ma}' fail to develop into adult fibroma cells and may go on to unlimited growth, just as granulation tissue following an injury may retain its embryonal state together with the proliferative capacity peculiar to that condition. Certain it is that an individual cell w^hich has once attained full development does not itself become a sarcoma cell, but its descendants ma\' do so. In this sense, therefore, a tumor which has previousl\' developed into an adult t}'pe of connective tissue ma}' become malignant. The phrase, if thus understood, may serve a useful clinical purpose without giving offence to biological law^s. Topographical Relations. — Sarcomas ma}' develop wherever connec- tive tissue is found, but certain regions are especially predisposed to their development. These are the subcutaneous tissue, the intermuscular septa, and the periosteum. The very earl}' stages of sarcoma, unlike epithelioma, have not been studied, and we are unable to describe their mode of origin; but for reasons ahead}' given we may infer that the}' begin as a single point and develop from this centre. The relation 96 SJRCOMA which the tumor focus bears to the surrounding tissue depends on the tj'pe of tumor. The higher types of sarcoma resemble in their mode of growth tumors composed of the adult state of the type represented. Thus a fibrosarcoma resembles a fibroma in its mode of growth, and usually somewhat in proportion to the degree of perfection of develop- FiG. 32 J'- ft ^^i.^ i\ ^ A X t- '■^*^ .^ ' 3' 1 V \ ^ 1 ^'^. ^ >^"- Spindle cells invading muscle. ment of the malignant tumor. The higher types, therefore, are expansile in development, the neighboring tissues being pushed aside to form a capsule which is usually infiltrated by tumor cells. In the more primi- tive types of sarcoma the disposition to expansile growth is less marked, and in many of the more cellular types the surrounding tissue may be more or less diflPusely infiltrated Fig. 33 (Fig. 32), though in even the most _^___^ .^ ^ __ _ __ malignant a pseudocapsule may be ivi-^., ...^^ '" '/^.rr.. ;:;_"''~^ -2 well marked (Fig. 33). -^--^-^-^'V^ ,-—:-,. Classification. — For the purpose of -^^'Xvv--r-f-:"--?%;^r-^^^^ mmute study it is convenient to v:': ;^';:v:;. ■;:-.^-..v .•£;;,•; ;;.vi;v aiviae the sarcomas into a number '\y''^l^:ll)\'}li^^.:1^i:-\ii^}:^^;}\';\'^'-^: of groups, keeping constantly in mmd, •''^■':'-)^^:'^':\^/i:^fr'^>'^^'\:y/\y.'. however, that there are no sharp '° ■''■\^'y.::^y'^y^':<°-^°i'^t^/^''} '']■'■ ■;:.[' dividing lines. Some of the tumors ' ■ " •*■ -''^•■- " ■ ' '■••-Vr?. '-<''. are made up entirel)^ or principally of Pseudocapsule formed by connec- one kind of cells, and are everywhere tive tissue compressed before the similar in Structure; these are the advancing tumor. simple sarcomas. In contradistinction to this is the compound class in which the sarcoma tissue is combined with one or more of the elemental tissues, or with tj'pes which at least closely imitate them. Simple Sarcoma. — Under this head may be included those types in which no part reaches a fully developed state of connective tissue, and in which a particular variety of cell predominates. cL.issiin:.iTi(j.\ oi s./kcoM.is 97 R()rNi)-ci;i.i.i:i) Sarcoma. Tlusf may be divided into two ^roii|)s small round-celled and large roimd-celled sarcoma. (a) Small Round-celled (Fig. 34). Iliis type le.sembles most clo.sely rhf c-mhi\()nal connective tissue, riuretori- the name ' 'emhi \()plastic" appliitl In kohin. The)' are made up ot small round cells resembling rhf small mononuclear leukocytes. I he nucKi are round and stain ticiph , while the cytoplasm is very small in amount, so as to be scarcely demonstrable. It is ot a very delicate nature and easil\' degenerates and leaves the nuclei lying free; or this disappearance may be due to a failure of the cytoplasm to take the stain. The intercellular substance is very slight m amoimt or may be entnely absent o\er large areas, or ma\' present itself merely as granular debris. 1 he stroma is ver\' scanty, being represented largely b\' bloodvessels which are often merely endothelial tubes without an}' connective tissue, or the endothelium Small round-celled sarcoma. may be lacking and tumor cells themselves ma\' form the vessel wall. Large areas may be imperfectly supplied with bloodvessels, and thus degenerative processes readih' occur. Usually stroma ma\' be demon- strated by special methods, particularly about the vessels, but in the very rapidly growing tumors it ma\- be absent over large areas. {h) Large Round-celled (Fig. 35). — This t\pe resembles closel}' in general structure the small round-celled variety, but the cells are larger. The cytoplasm is more abundant and is finel}' granular on account of the presence of gl3cogen granules. The cjtoplasm is less delicate than in the small round cells and the tumor in general undergoes second- ary degeneration less readily. Man\' of the large-celled sarcomas as classified by older writers are descendants from mixed tumors and from melanomas. Spindle-ceiled Sarcoma. — This t\pe also has been divided into small-celled and large-celled t3pes. 7 SARCOMA {a) Small Spindle-celled (Fig. 36). — These tumors are more frequent than the preceding, and usuall}^ less mahgnant. The cells are oblong. Fig. 35 ,c h' , , ^ 'I . » , 0[ - i ' «<■ . Large round-celled sarcoma. fusiform, and usually stain clearly. The cytoplasm is usually abundant, a fact difficult to demonstrate in cut sections, but easily seen in fresh teased specimens. It may be shown by this means often to extend Fig. 36 ^;< - /;^-" ^'r ^ ^ ; Small spindle-celled sarcoma. Some bundles of cells are cut longitudinally and some transversely. several times the length of the nucleus. Sometimes two or more nuclei may be seen. The stroma is more abundant than in the round-celled cL.issiric.iTios or s.ircom.is '.)'.) sarcomas. Often rlu- ili\ iclin<; line- between stroma and spindle cells is not sharp (see Filirosarcoma). These tumors are often well encapsulated. Sometimes round cells are also present, and in such cases the two varieties are inrerminJ Large spindle-celled sarcoma of ovary. Une bundle cut longitudinally, others are cut transversel}'. The tumor was edematous and of low maHgnancy. Mixed-celled S.arcoma. — In none of the preceding classes are the cells of one shape, but usualh" one kind predominates. Sometimes, how^ever, the intermixture of cell forms is so great that the selection of a predominating t\pe is not possible. Cell forms may be inter- mingled (Fig. 38) or the form may be similar but the sizes vary ( Hg. 39). To such tumors the term "mixed cell" has been applied. Every possible combination may be observed and the character of the tumor varies in different parts as one or another kind of cell predominates. Melanosarcom.a.. — This is usually composed of large round cells plus melanin (Fig. 40), which ma}- occur either in the cells or in the 100 SARCOMA intercellular substance. The pigment is not distributed regularly, but occurs in patches throughout the primary tumor and in the metastatic Fig 38 >'^>'' / ,f:^,.-rj.%^(^. Mixed-celled sarcoma, containing spindle cells and round cells. nodules. The color of the tumor may be less intensely black than the metastatic nodules or the reverse may be true. They may arise m any region where pigment cells normally exist. Fig. 39 *- ^3/ ) ^ ^ ?) m e^ & s& (* 53 ^ H 4^ iS> p. ^*'. ^\ ^ Mixed-celled sarcoma, containing large round cells and small round cells. Lymphosarcoma. — The relationships of these tumors are still in- volved in much uncertainty. It is often impossible to distinguish between them and certain diseases of the l3^mph glands and blood. They are characterized by small round cells with a very small rim of CL.issfric.iTiox or s.ircom.is kh protoplasm ( Fi^. 41), ami nsiiiiliK in this respect the small round- c-flKd saiionias. It was lnrnurl\ l)ilu\((l that th(\ diflVr from rhi-m I'k;. 40 Melanosarcoma of li\er: <-/, liver cells; h, sarcoma cells containing melanin. Fig. 41 » •• ■ ' Lymphosarcoma of the neck. 102 SARCOMA by possessing a complete reticular tissue, but the recent researches of Selig,^ with his exact demonstration of the reticulum, have shown that this assumption is not justified. The location of the tumor and the history of its development are of greater value in its diagnosis than are the microscopic findings. The picture often changes when the tumor escapes into the surrounding connective tissue (Fig. 42). The differentiation between lymphosarcoma and Hodgkin's disease is often made possible by detecting the endothelial and eosinophile cells which are peculiar to the latter. In its later stages Hodgkin's disease resembles Fig. 42 "» ./»,,,"'" '»-•''»''!' *•'' -*JJ=^ «■« ilX •' -'.•■*•'. .'x *'o 1''.,. Lymphosarcoma of cervical region infiltrating periglandular tissue. sarcoma more closely, and many writers, as Coley^ and Gibbons,'' have sought to escape the difficulty by placing this disease in the group of sarcomas. Many German writers class them together as malignant lymphomas (see Tumors of the Neck). Giant-celled (Fig. 43). — In some of the preceding types large cells with several nuclei may be seen, but the term "giant-celled" is reserved for those tumors, which are composed for the most part of round or spindle cells, but which contain also large cells with many nuclei, up to a hundred or more, the nuclei being situated uniformly throughout the cell. These large cells are not always evenly distributed throughout the tumor, but may occur in irregular groups. Their number 1 Surg., Gynec, and Obstet., 1907, iv, 319. 2 New York Med. Jour., 1907, Ixxxv, 577. ^ Amer. Jour. Med. Sci., 1906, cxxxii, 692. CL.issmciriox or s.ircom./s lo:i varies greatly. In sonu- rumors a consiiliiaMc propoi rioii ot rlie area may be occupied U\ rlust- crils, wliiK- in oiluis onK occasional cells Kic. 4} .*. V. ■.;•* ■■■■■! :;;bv^,"-^^^' ^' .*^ /. « ' -^F" / ' V ..." J. s VN y ^. \^l I Giant-celled sarcoma (epulis). Giant cells and spindle cells, the latter making up the bulk of the tumor. Fig. 44 C3; i0>. ^# ^^r^ g' ^'^-^ ^<^^ eS . m ^ c Foreign body giant cell. Higher Types of Sarcoma. — Under this heading may be grouped those sarcomas in which a more or less perfect development of one of the elemental forms of connective tissue has been attained. Biologically, these may be considered a higher form, but clinically this is not true, except in a limited sense, for even tumors formed largely of tissue which approaches closest the normal development may nevertheless be characterized by early metastasis. Fibrosarcoma (Fig. 46). — By fibrosarcoma is meant a connective- tissue tumor which has gone nearly to full development, certain areas, indeed, presenting adult fibrous tissue (Fig. 47). Since these tumors comprise the most perfectly developed sarcomas the cells are usually spindle-shaped, but round cells may be present in rapidly developing areas (Fig. 48). Spindle cells cut transversely may appear as round cells. The nuclei are more prominent than in fully developed, fibrous tissue and the cytoplasm is more abundant. The bloodvessels come IlIiROS.IRCOM.I 105 in dirtcr contact with the- tumor cills, though often where the con- ncctiNc tissue IS abundant this nia\ not he- easily made out. The Hlu-r Vu:. 46 / / / / Small spindle-celled fibrosarcoma: a, cells cut longitudinally; ^, cells cut transversely; c, bloodvessels. bundles sometimes lie parallel to the vessels, but quite as often they diverge at a more or less acute angle. The nuclei var}' as to the direc- FlG. 47 c Fibrosarcoma of popliteal space, showing small spindle cells embedded in bundles of fibrous tissue. tion of their long axis relative to the direction of the hbers between which thev He and also of the vessels near them. It is just this independence 106 SARCOMA of the direction of the nuclei, together with their cytoplasm that gives the chief evidence of undeveloped tissue. In many cases a micro- Rapidly growing area of a fibrosarcoma. scopic differentiation is impossible because no distinct border line between benign and malignant fibrous tumors exists. Mallory has proposed that no attempt at differentiation be made. Fig. 49 C a Chondrosarcoma (mixed tumor) of testicle: a, area of cartilage; b, sarcoma cells; c, connective tissue. Spindle-celled sarcomas are often called fibrosarcoma because the cell processes may appear as fibrils. Often the relation of cell to fibers CIIOXDROK.IRCOM.I 1(17 is not fasil\ iiiadr out iink-ss specMal stains tlu- ht-st of \slin.h is Mallon's are ciiiplo) ctl. I'he more nearh Hhrosarconias iniitatt- tlu- |)ure hbroiiias in structure the more nearly the\ imitate them in ehnical course. Often the clinical course and the location of the tumor ma}^ give evidence of importance. In addition, it must he rememhered that a given tumor ma}' show some areas which are composed of fully developed connective tissue while other areas ma\- he distinctly sarcomatous. In no tumor does the examination of onl)' a single part lead so often to error as in these. Chondrosarcoma. — These tumors are composed of cartilage and sarcomatous tissue (Figs. 49 and 50). I'hey are most commonly observed as a part of mixed tumors, e. g., of parotid and testicle, but thev occur also in sarcomas of bone. Chondrosarcomas are formed b}- the deposit of a h3'aline intercellular substance about sarcoma cells which have originated from the osteogenic membrane. The\' bear Fig. 50 Chondrosarcoma. Sarcoma cells in centre, cartilage at the right. an imperfect resemblance to normal cartilage. The number of cells in the lacunae varies from one to man}', and they are arranged irregu- larh'. The intercellular substance varies much in amount and there are frequently large areas without cells. The cartilaginous parts are usualh' easily detected because of their densit}- and the pale blue glistening surface. Tumors often appear, on superficial examination, to be pure chondromas; but if they grow rapidly one should search persistently for malignant areas. These will be characterized b\- the irregular arrangement of the cartilage cells withm them and b}' their border of groups of cells without intercellular hyaline substance, and with, perhaps, a number of thin-walled vessels. Often, however, the most careful search fails to reveal sarcomatous areas, and }et the tumor IS clinicall}' malignant. Chondrosarcomas frequentl}' have osseous areas in their centre, particularh' in the more slowly growing tA'pes, and are then called osteo- 108 SARCOMA chondrosarcomas (Fig. 51). The disposition to bone production may be regarded as the normal tendency in the higher types of chondro- FiG. 51 ,^ > a y 4. • -d \ Osteochondrosarcoma of rib: a, bone; b, cartilage; c, connective tissue. sarcomas. A calcareous infiltration which resembles ossification on macroscopic examination is sometimes found in the centres of degener- FiG. 52 a- a Periosteal osteosarcoma: a, bone trabeculae; b, sarcoma cells. ating chondrosarcomas. The addition of an acid will differentiate these from ossified areas. osri'.os.ikcoM.i 109 OsTKosARCOMA. 7luse tiniiors may spiinji eitlur from rlu- peri- osteum or from rlu- medulla, and are seen most tVequentlv in the lonj; bones, pai ru iil;irl\ near the joints. When they grow from rlu- periosteum they c-ncircK- rlu- hone aiul i-xrend along the shaft. 'I'hey may be com- posed of small round cells (Figs. 52 and 53), spindle cells, or giant cells ( I'ig. 54), and rend ro form large masses which are fusiform or bos- selated. I lu- nu-(.lulhir\ type begins as a globular mass which graduall\' invades the bone from within until the periosteum is perforated. 1' re- quenth- the periosteum attempts to limit the growth by throwing out a shell of normal bone. After a time, however, the tumor reaches the soft parts and rhe growrh is usuall\- rhen more rajiul. 1 he tumor ma\' extend along the medullary cavity for a considerable distance before perforating the bone. Giant-celled sarcoma of the medulla ma\- undergo Fig. 53 » • ® . .-w ■ 4>, ..- •» •• -f •*/ • e ^ ' ^ .^ " J ^' ■ k "' i, « ./.'■' ^ K ' ' ,* " ■ fl " „ i - %V ' 6 '^ ■ ./S?^ ^ *■; " ■ ' Myelogenous osteosarcoma. secondary changes from hemorrhage into its substance or otherwise, and become c\stic. The sarcomatous element ma\- so far disappear that the true nature of the tumor may be difficult to determine. This is particularly true in the so-called bone c\-sts which Gaylord has recenth' shown to be always giant-celled sarcoma which have under- gone such secondary changes. The bone sarcoma ma}-, like the cartilage sarcoma, be made up largel\' of cellular elements, with but a small amount of the more highly developed tissue; or, on the other hand, the tumor may be composed almost entireh" of bone with few recognizable sarcom- atous areas. In the former case the bone is arranged as spicules which may extend out into the soft tissue and upward from the bony base and lie free in the tumor without such bon\- connection. The usual structure is a homogeneous groundwork containing bone cells. 1 here no SARCOMA are usually no Haversian systems,^ but attempts at the formation of a medullary cavity have been noted (v. Hansemann, Borst). The bony spicules must be differentiated from areas of calcareous degenera- tion of cellular tumors. In the more completely ossified types pro- longed search may fail to show mahgnant areas. Their slow growth likewise may further tend to suggest pure osteoma. Nevertheless, metastasis may occur and give rise to fibrosarcomas and chondro- FiG. 54 -b . K t ) \ A', \ • -a * , I «. X .lacetl m this uroiip, which should uulutlc- onl\ those in which the l-K,. 5: -A 4. r , 6 : ' X ' Ct; \ 1 '^ V )|l / V \ \ \ \ \ V i ^ J, 1 ^ V \ • t \ V \ \ V \ - \ Myxosarcoma. myxoid cells are actively proliferating (Fig. 55). The myxosarcomas may present only slight evidence of malignancy. Frequently, rapidly growing tumors composed of myxoid tissue will recur after removal Fig. 56 Glioma of retina (gliosarcoma). when no definite areas of sarcoma can be demonstrated m them. Here agam the life histor\' and topography must be considered m order to interpret correctly the microscopic findings. 112 SARCOMA Gliosarcoma. — These tumors are characterized by rapid proHfera- tion of glia cells. They contain, in addition, various other kinds of cells — round cells of various sizes, polymorphic epithelioid cells, cells with processes, and cells without (Fig. 56). Ribbert is disposed to class such tumors with the gliomas. This seems justified in those cases in which the cells have attained the normal degree of development and are abnormal onl}^ in the rate of growth. The segregation of the malignant glia tumors from the sarcoma group is justified further by the fact that the glia cells are epiblastic in origin. In their general characteristics these tumors resemble the malignant myomas, which in turn closely resemble malignant tumors of connective-tissue origin. It is convenient, therefore, to include the rapidly growing tumors springing from glia cells in the sarcoma group, notwithstanding the more dignified ancestry of the glia cells. Fig. 57 Liposarcoma. LiPOSARCOMA. — Von Recklinghausen described under this head those sarcomas in which a large number of the cells become filled with fat (Fig. 57). They may be distinguished from the cases in which fat is included in a tumor by infiltration, by the fact that in the true lipo- sarcomas the fat-containing cells vary much in size, while in the latter case the cells are more nearly uniform. They are very rare. Angiosarcoma. — The present disposition is to classify those tumors which were formerly called angiosarcomas with the endotheliomas or more particularly the peritheliomas. The most of the tumors placed under this head by Waldeyer would now, no doubt, be regarded as perivascular endotheliomas by most pathologists. If the term is re- tained it should be limited to those sarcomatous tumors in which the MYOS.IRCOM I \V.\ cells spriiii; out litim \\\c vtsst.1 \\;ill. It is possihli- iliar ctlls spmi^iiii^ fn)m tin- pri i\ ;ist.ul;ii iiulotluliiiin iii;i\' fiivi- nsf to tiiinors which cxccid tin- pn iilulioiiKis in m;ili<;ii;iMc\ . Such a view would be usetui ill clinical chissiticiitioii, ;intl in \ icw ot the uiiceiraui liiiiits ot the LMolojiical i^ossihilitits of the- cndotluhal cell lued j^inc no ottcnce to theoretical conceptions. I In- nuinhi r and si/e ol the \essels in sarcoma and theii arranj;eiiunr m nlation to tin- cells do nor warrant the desi liquation angiosarcoma. MYOSARCOMA ( Fifi. 58).- In m\()nias ln(|uciul\ rlie cells ol ceiram areas undergo rapid proliteiarion. I' or rlu- most |iair rlu\ retain a spindle shape and form Inindks which like the myoma cells are arranued parallel to the vessels. The nuclei are larger in size than the muscle- celled luicki. Init in general the structure and the tinctorial reaction of muscle is retained, though rapid growth and the formation ot Fic. ;8 \h"Osarcoma of the uterus. metastases take place. These are called malignant myomas. Areas are frequently encountered in which nuclei of ver}' different sizes dis- regard all relation to the direction of the bloodvessels. It is this type that is said frequenth^ to undergo a sarcomatous degeneration. Leio- myomas are often found in which the deviation is still greater; these are myosarcomas. Still greater deviation results in tumors which parallel those derived from connective tissue. Since in rare instances so-called malignant myomas are encountered in which the cells retain at first the form of a normal non-striated muscle cell and deviate only in the rapidit\- of their multiplications, it seems quite possible that in this rapid proliferation the cells would fail eventually to attain perfect development. In such cases we might well speak of an actual sarcom- atous degeneration. We mean by this not that a myoma cell once having attained full development reverts to a lower form, but that 8 114 SARCOMA descendants of such cells fail to attain full development. In speak- ing of a degeneration of myoma into a myomasarcoma we must think, therefore, of the tumor as a whole, and not of the individual fiber. In this way we may harmonize clinical experience with pathological findings without doing violence to any biological law. Many path- ologists who are reluctant to admit the possibility of such a degree of Fig. 59 Myosarcoma of the uterus. Simple myoma below, malignant transformation — the upper part of the cut. metaplasia believe that the sarcomas appearing in myomas arise from the connective tissue of the myoma or that the entire tumor was sar- comatous from the beginning. There is much evidence, both clinical and microscopic, which warrants the belief that sarcomas do develop from muscle cells. In structure these tumors present every variation from muscle cells, .njKOl.lR S.IRCOM.I 115 sliowinii, tlu' sli;i,lu th;iii<;is :il>()\'f noti-ci to cells l>c-wiin^ no resenibhinct- to iiuiscif cells sri ucrui;ill\ or riiicrori;ill\ ( Kigs. 59 aiul 60). L suallv Kk;. 60 - (.-/*■- ; r?5 M\'osarc()nia, showing the \ariation in the size and shape of the nuclei. f'iG. 61 ^'Sir-cvv e?-o_^' '...J,'..i. '•••.".»■' ■ r ■'"'■-. v .■ °J "■•. .' ' ■• i ' 130 ANGIOMAS growth. They usually project from the surface (Fig. 66). The vessel walls are often imperfect, sometimes bemg mere connective-tissue tubules lined with endothelium (Fig. 67), which when the blood spaces are empty becomes contracted and gives the tumor the appearance of adenoma of the sweat glands (Fig. 68). In some instances the endothelium shows such activity of growth as to resemble the endo- FlG. 69 }ifsp "'\- a- b—-< i^-^>i -- >. »«^-. 'W '■:^<:'i::: i'A.'>- ('■- ;*V;.' ■ ■,*i- ^[i^-'-'i:-^-^^' -V"* * yj^i^--''^ .:""- .-:^.. .'--^^ "" __.>■_ •" ■■i'f*.'*^- ■ ■ "vJ^i,^ 'J*-'.. ■■ ^»..' -. -'' -' , ' '. ' ' ^ *v*;']'-'t *>/?fti c •^'^VJH'AJ^ji::^,^--' ^. ■■; '^^...^^i' >■" -, ' ■'•i'l^ ,. ■^'^v^.t;- •v?f .; ■ -f':5vv^*'' -^IJ;^^ /Ssi^^'- -;■ '.„..,„„ ''-;■ -i '":"^-,:.^.^; ■ • • . '■' '"■■''T/"'-V?ii-3'' v.'^^v;''^i'---.*- ^'i'*.'.*7?«S'fti:i^;?.( "-'tS?^'^' -'^^r^' •»f,/ v., ■■''- "• - , ■ ■ •^•y-t^i '•"^%, _ , »;" Hemangioma, with proliferation of its walls: a, blood spaces; b, proliferated endothelium; c, connective tissue. ,theliomas (Fig. 69). On the other hand, the connective tissue may be so abundant and cellular as to resemble the sarcomas (Fig. 70). Cavernous Angiomas, — In this class, as the name indicates, the blood spaces are large compared with angioma simplex. Its most frequent location is the skin and liver, less frequently in other internal organs. The walls of the sinuses are usually thin, devoid of musculature, and c./rr.kXfJi s im.iom.is i:;i liiKtl by a flat tiulot luliiim ' I' i<;. 71). 1" ii-(]iicni l\ (.on^cnital, tlK\- \\\\i\ remain statioiiaiN , l>iit at tiiius rluit- is a ^i<)\stli of the vessels in length wimh tiia\ Infill Mi(iiliiil\ ami cause tlu- tumor to assume im|)oit;mr I'k;. 70 o til '-- J/i.J" ' ^ > 1^ '■ ^ '- t Angioma with cellular connecrne tissue Fig. 71 /yo'^^'.r^i! ;i7?;'Q^- v c avernom a of li 132 . ANGIOMAS proportions. The connective tissue likewise may become active and cause varicosities b}' contracting the lumen of the vessels at certain points. This tA'pe is not attended by a development of endothelium. Aneurysmal Varices. — This condition, which is generalh' excluded from the angiomas, has much in common with the cavernous angiomas. It is formed of dilated vessels, and frequentl}' increases rapidly in size, exhibiting a progressive dilatation as well as an increase in length and a connective-tissue formation. It is therefore a veritable tumor. Course. — Angiomas are distincth' benign tumors. Their impor- tance consists m the resulting deformities and occasional rupture from trauma. They often disappear spontaneously. The aneurysmal varices have a disposition to extend and destroy the surrounding tissue by pressure. Diagnosis. — Macroscopic. — The simple angiomas are easily recog- nized as livid spots in the skin or mucous membrane. The}" can be confused onh' with petechia, from which the}- are easil}' differentiated, b}' noting that the angiomas readil}- disappear on pressure onl}" to return when the pressure is removed. The h}'pertrophic t}'pe appear usuall}' as elevations above the surface, resembling warts, but are softer, redder in color, and are reducible b}' pressure. The cavernous angiomas appear as livid elevated tumors, soft to the touch, disappearing on pressure, and becoming more tense with increased blood pressure. Microscopic. — The h}-pertrophic t}'pe, particularl}' when m course of active development, may be confused with sudorific adenomas. Lsuall}' at some point m the periphery of the angioma a definite lumen containing blood can be discovered. This, together with the fact that in adenomas a basement membrane is present, obviates any error. In the cavernous t}'pe the large endothelial-lined spaces filled with blood are unmistakable. In all t}'pes when there is an active proliferation of endothelium the question of the presence of endothelioma arises. Usuall}', if there is evidence of endothelial proliferation m angiomas, it is in the nature of an h}'pertroph}', and there is no change in cell type, while in vascular endotheliomas there is a change in the form of the cells most remote from the lumen of the vessel or from the intima. In those instances in which there is an active proliferation of connective tissue between the vessels the cells likewise are of uniform size as distinguished from the variation in cell form m the slowly growing sarcomas. In all instances the form, situation, and life history of the tumor must be taken into consideration. The disposition to append the appellation "angio" to every sarcoma possessed of an abundant blood suppl}' has caused unnecessar\' confusion. i.r.\ir//.i\(:/o.\i,is ]:v.', Treatment. Since angiomas ait.- iisuallv ot no more serious conse- (|iifncf than as a ckformit\ , tluir imporraiKc and treatment depend on the location. In general, it may lie said that the cavernous types, it localized and suitahh siruati-d, may he excised m order to pre\ent suhsecjuent rapid growth. If situated on exposed parts, electrolysis or segmental ligation will ha\e preference (see fumors of the Face). LYMPHANGIOMAS In the l\ niphatic tumors there is usually less evidence ot newly formed vessels than in the angiomas. fhey are usuall\' congenital, resulting perhaps from maldevelopment of the lymph channels. In some instances there is probably development ot new sinuses or at least the dilatation of preformed channels. The spaces vary in size and are lined by a thin endothelium surrounded by a thin wall of con- nective tissue (Fig. 72). In rare instances, as in macroglossia, there Fig. 72 — o • '-^ V »' 'KV' I \ Lymphangioma of the face, showmg lymph spaces, interstitial tissue, and section of sweat gland. is an infIammator\" increase of the connective tissue. The relations between connective tissue and lymph spaces is not clear. Attempts have been made to divide the lymphangiomas into classes similar to the hemangiomas. Simple superficial lymphangiomas correspond closeh' to like tumors of the capillaries, but usually the lymph spaces corre- spond to the cavernous angiomas, but often exceed these greatly in size. In some situations (tongue) the new development of lymph spaces unmistakably corresponds with the aneurysmal varices. 134 ANGIOMAS Course. — Lymphangiomas are of importance because of their size and situation and because of the possibihty of rupture in the large cystic type with the ensuing exhaustive drainage, and of gradual exten- sion in the proliferative type. Diagnosis. — Macroscopic. — The soft transparent lobulated tumor is usually sufficient for recognition. They are sometimes simulated by other cysts. When situated so deeply that the color of the contents cannot be made out, they are differentiated from hemangiomas by their failure to be reduced by pressure. Their most frequent site is in the cheek or about the neck. In cases of doubt aspiration may be practised. Microscopic. — The identification of cavities lined by endothelium is strongly suggestive. The nature of the contained fluid is the only certain diagnostic point. Treatment. — Excision should be practised whenever possible. It is frequently the case that because of their situation, their extent, and the age of the patient, excision is impracticable. Drainage and injec- tion of irritating substances must then be resorted to, or operation must be delayed to a more favorable age (see Lymphangioma of the Neck). c w \v r K k X I I KNDOI IIKI.IOMAS CI lOLKS TKA lOMAS ENDOTHELIOMAS' Conception and Dehnition. A conception ot the nature ot the tuniois heretofore considered was easily gained by comparing the pathological anatom\' of the tumor with the normal tissue from which it sprang. This was eas\' because the normal tissue was well characterized and its peculiar features, both histological and pathological, were uncjues- tioned. W ith the endotheliomas the characterization is more difficult, not only because of their great variation in structure but also because the nature of the mother tissue, the endothelial cells, is still a matter of dispute. We are compelled, therefore, to preface the discussion of the pathology of endotheliomas b}- a few remarks on the nature of endothehum. The term was first used b\- His to apply to the fiat cells lining the cavities, including the blood and l\-mph vessels as well as the joint surfaces. Later anatomists have sought to limit the term to those cells which correspond morphologically to the conception of His but have their origin in the mesenchyme. This excludes those lining the pleuro- peritoneal cavity. Other anatomists have sought to abandon the term entirely. More recenth' Borst and Ribbert have declared themselves in favor of retaining the original conception of His. There is no reason in patholog}', Borst declares, to distinguish between the cells lining the vessels and those lining the serous cavities. Histologicall)', the endothelial cells are flat, with large oval nuclei and prominent nucleoli. The protoplasm is abundant and clear. Phy- siologically, the -endothelial cells pla\- a prominent part in nutrition (Heidenhain), not only filtering the passing fluid, but altering its con- sistenc\' also. In pathological conditions the endothelial cells have a ver\- definite action. Under irritation they swell and become cuboidal and even columnar. In function, the\- stand midwa\- between the epithelium and connective tissue, assuming secretory functions like ' For literature see v. Volkmann, Deutsch. Zeits. f. Chir., 1895, xli, i; Marchand, Verh. d. Deutsch. path. Gesellsch., 1900, ii, 38; and Krompecher, Beitr. path. .Anat., 1905, xxxvii, 28. 136 ENDOTHELIOMAS— CHOLESTEATOMAS the former and playing an active part in the production of the latter. Much useless controversy has been waged as to whether endothelium resembles epithehum more than it resembles connective tissue, and the discussion was naturall}^ carried into the classification of endothelial tumors. As a matter of fact, endothelium is as well individualized in structure and function as either of these tissues, and is quite as well worthy of a place among the primary tissues. Accordingly, the term endothelium should suggest a definite type of tissue and not a sub- division of any other tissue. The same may be said of the tumors originating from it. Standing midway between epithelium and con- nective tissue, the tumors arising from it may be expected to maintain a corresponding relationship to epithelial and connective-tissue tumors. The deviation in structure from either of these is less than they are from each other. Naturally, the resemblance to either the carcinomas or the sarcomas becomes closer when one or the other of the normal functions is emphasized. Thus if for any reason the connective-tissue forming properties are brought to the foreground, it will not be sur- prising if the sarcoma should be closely imitated. It is these peculiarities which make questions of differentiation so difficult. The problem has needlessly been made more complex by the addition to the endotheliomas of groups of tumors bearing no genetic relation and but little morphological resemblance to them. These groups are the mixed tumors, the cholesteatomas, and the carotid tumors. Some investigators (Henke) are disposed to class certain nevi with the endotheliomas. This has a certain justification in struc- ture; but the cells of nevi are embryonal and have not undergone differentiation into cells possessed of a higher function, as have endo- thelial cells. When nevi undergo active proliferation they give rise to cells of a very low order, as is manifest by the great malignancy of the tumors they form. It will be an aid to clearness to retain such tumors in a separate class until their nature is understood. Endotheliomas may, therefore, be defined as tumors which develop from the cells lining the blood and lymph vessels and the lymph spaces. It is not necessary to include the cells lining the serous cavities, because they have not given rise to tumors. They are characterized by pro- hferation of the endothehal cells and invasion of the surrounding tissue. Classification. — Much aid in the classification of the endotheliomas could be obtained if the life history of the tumors were considered. The entire problem heretofore has been considered from a purely morphological standpoint. Unquestionably a tumor cell is of interest quite as much because of what it does as because of its origin, form, and structure. Pathologists admit that the nature of a single cell cannot be determined except by its relation to other cells and to the EMxyriii.i.KjM.is \M surrounding tissue. I Ins is an admission that the chief evidence for determining the nature of a cell is ohtamed from its morphogenesis. It would seem that if the entire life histor\' of tumors were considered as re\ealecl in tin- clinical lnst()i\ , much aid vsould he gi\en in the classiH- cations of tumors. In no class ot tumors d(K*s this seem as important as in the eiulotheliomas. All types have in common the characteristic cells, aiul their recognition depends upon the discovery of the situation, usualh at the periphery, where the transition can be made out. I he cells springing from the typical situations extend into the connective- tissue clefts. I heir nature tits them for lining spaces, and narrow columns can often in- made out extending among the connectue-tissue hhrils. I'his close relationship is retained in their subsequent development, and the\' are usually in close contact with the connective tissue, t re- quenth' Hbrils may be demonstrated extending between the cells of endotheliomas, the ease with which they can be shown depending upon the nature of the tumor. The nearer the approach to a sarcoma t\pe the more likely will such demonstration be possible. There are no well-defined groups either from arrangement of cells, as in cancer, or from types of cells, as in sarcoma; but for convenience the\- may be divided according to the situations of the cells from which they spring. These sources are the cells lining the l}-mph vessels, those lining the bloodvessels, and those derived from endothelial cells which surround the bloodvessels in certain situations. \\ e have, therefore, (i) Ivmphangio-endotheliomas, (2) hemangio-endotheliomas, and (3) perivascular endotheliomas, or, more briefl\', peritheliomas. Macroscopic Appearance. — Endotheliomas present a varied picture. The t\"pical form is a more or less globular, sometimes slightl}' lobulated tumor, usuall\" firm in consistency and often dense. Softer varieties, particularlv among the more vascular types, are often encountered. Thev usually grow b\' expansion, and therefore are more or less encapsulated. In many instances, they develop difFuseh' and may spread over large areas. Occasionally CA'Stic dilatations may be observed. On section, they are usually reddish gray. The cell strands can usually- be made out b\' the naked eye. Microscopic Appearance. — Lymphangio-endothelioma. — This class is the most frequent of the endotheliomas, and a network of cell columns corresponding to the network of lymph vessels (Fig. j}) may well be taken as the typical arrangement. This comes about from a prolifera- tion of the endothelial cells which tend to fill the lumen of the vessel without invading the surrounding tissue (Fig. 74). This typical picture is most likeh- to be found in the periphery of the tumor, the more central parts being confused b\' a more advanced development. In the older portions of the tumor the connective tissue surrounding the cell columns 138 ENDOTHELIOMAS—CHOLESTEATOMAS proliferates and the picture closely simulates a carcinoma (Fig. 75). The cells within the columns may undergo certain changes which cause Fig. 73 e" '■r, o « ' 'nUltu^ ,"•.?,••.,.. ' "vl ^ ViW^>' 00 00 '0° at ,' .. \ **'»*-.» ^-> .'^_. a." „ » *3 y - |;f^ i^- .. ^f® ^ ■'- B S > a; \ .«"9.«flVe'.f Lymphangio-endo thelioma. Fig. 74 — r ^ /' -^ ^^ J/ / ^ 1 IS r 1 / ^ ^ r.-, , 'i^ ,S % ^ i '& -t, ^ '^ Endothelioma, showing columns of cuboid cells in tissue spaces. i.Muyriir.i.ioM.is i:i9 rluin to icstnibli' rht- c;iiuii piarls, tli()ui;h kc-ratohyalin is always absent. In otlui- insraiuis tin- i luloihcliomas may rc-st-niMc tin- saicomas, thf cells luin^ nunuioiis ami spindle- loiimd. Tluy tif(|uc-iul\- have a coiuiiit I K' an aniicnu lit . I In- ilisposnion ot i ndotliclionias to resemble carcinomas at oiu- pnioil (»l tluii iK \ clopnunt and sarcomas at another leatls lliiuke to nniaik that in tiiosi' instances in which one part of a tiimoi siii:;jiests carcinoma wliiK- anotlui part suf^f^ests sarcoma the endothelial oiiiiin ot the rnmoi is probable; and it a transition ot the epithelial-like cells ot the one into the sarcoma-hkc- cells ot the other can be made out, the diagnosis is certain. Fig. 75 Lymphangio-endotht-lioma, showing masses of endothelial cells enclosed by con- nective tissue. Hemangio-endothelioma. — This t\pe springs from the lining cells of bloodvessels, which may retain a patent lumen (Fig. 76). The cells are continuous with the intima, and a gradual transition from intimal cells to tumor cells must be made out before diagnosis is possible. This relationship is usualh- most easily discernible in the periphery of the tumor. The vessels may be filled with columns of cells. Hemorrhage into epithelial glandular tumors or epithelial tumors growing into the lumen of the vessels may simulate these tumors. Certain capillary angiomas (Fig. 77), particularly those with hypertrophied walls, may also simulate hemangio-endotheliomas; indeed, the border line between 140 ENDOTHELIOMAS— CHOLESTEA TOM AS the two is morphologically indistinct. The endothelium of vessels in carcinomas may proliferate, simulating endotheliomas (Fig. 78). Fig 76. .-i'?*'^'''.'.^''- Hemangio-endothelioma. Peritheliomas. — These tumors are formed from the proliferation of the perivascular endothehal sheath. Since such sheaths are present in certain situations only, these tumors are, of course, restricted to the Fig. 77 ' "::-^ir ■r^y'■■i ',.*■.■•■•. /<'<%•. - y--^ ',— -/' '^ ! ■■ - - •''' '■ " ' ■'"' '" ' "61 .-v; • <■': /I, .^ >-;;. -^^^/^^^^i^:^ __^. —y .\f^^. '" ' \ J.. ;r.~ ^.-^' ■' ;'■- " ^ '"' .r '^.^/■^J;;;^i'-_:i Capillary angioma with proliferated endothelium. same locations. The tumors present intact vessels surrounded by concentric cell masses having a characteristic radiating arrangement. These tumors even more than the hemangio-endotheliomas are apt ESDorilEl.lOM.IS 141 to Ih- inist;ikin tor i;l;iiuiul;ii tumors, p:ii t Kuhii 1\ ulun rlu- source of tin- tumor is not rakiii into account. I he dixulin^, Imc between rlu- pcri\ascul;ii ami tiidovasculai tumors ma\ hi- ohsiiin-, lucausc- the (.-nriit- \t-ss(.-l ma\' be replaced b\ columns ot cells. Much controversy has arisen about the relati()nshi|) between the perixasculai endotheliomas and the an • C'*^'^, , '0 ^''^f.^ Fibrous papilloma, showing epidermis, connective tissue, and vessels surrounded bv embr\onal cells. 152 PAPILLOMAS thickened in comparison to the normal skm, and sometimes the super- ficial layer of keratin-containing cells is enormously thickened, giving rise to a horny formation. The transition of epithelium in thickness from the normal skin to the highest point is usually gradual, so that there is no sharply defined border, and the impression is therefore of an hypertrophy rather than of a tumor formation; the epithelium may project deeply into the fibrous tissue, but sometimes the tumors are sharply defined, so that the epithelium covering the growth sepa- rates readily from the normal epidermis. The senile wart (Fig. 84) occurs oftenest on the face and generally in persons beyond the fiftieth year. In form and structure they differ little from the adolescent wart. There is often evidence of a disturbed Fig. 84 V-.- ■ ■ , ■ ' ■{ ,:/•• \ / % /^ *^/,^ :"-;■? 2^- Epithelial papilloma (senile wart). relationship between epithelium and connective tissue, manifested by a lessened affinity of the connective tissue for the acid dyes and by areas of round-celled infiltration. That there is some epithelial proHferation is indicated by a stronger reaction to the basic stains. There is frequently evidence of a more active branching of the cell columns than in normal areas and occasionall}^ some irregularity in the size of the cells. All of these observations are in accord with the well-known disposition of these tumors to undergo malignant change. Melanotic Papillomas (Nevus Pigmentosis, Fig. 85). — Under this head may be collected those congenital lesions of the skin which are usualh^ characterized b}^ abnormal pigmentation. They vary in color from pale yellow to black, depending on the amount of pigment present. Occasionally there may be an entire absence of pigment, although the .\ii:i..i.\()Ti{: I'.ii'ii.i.oM.is I5:i tumor in:i\' intscnr tin- mm r;il ih;ir;icrcnsrics ot rlu- >;r<)iip, hut in the niaj()rit\ ot (.iisis a caittul scanli will dist-losc- pij^nunt. 1 ht\' are Vu,. S5 Melanotic iiapillonia (piiiiiuiir wart), showinji epidermis and embryonal cells with areas of pigment. usually flat turiiors (Fig. 86) reaching but slightly above the surface of the skin and without a constricted base, although in some cases -C Pigmented mole: a, pigment cells; b, embnonal cells; r, epidermis they may be pedunculated (Fig. 87). Sometimes there is no e but simpl}- a pigmented area in the skin. Often these tumors levation, are ver\' 154 PAPILLOMAS vascular (Nevus vascularis), and sometimes they are covered bv a growth of hair (Nevus piliaris). An endless number of names has been applied to variations from the common tA'pe. The pigmented nevi are usually congenital, and their growth is in proportion to the growth of the child. Thev are frequentl}' multiple, and are sometimes s}'m- metrically arranged in such a way as to suggest some relation to the nervous system. Their structure is variable, and the}' merge insensibly into the soft papillomas. The nature of these tumors has not yet been fully determined. Borst^ classes them with the fibromas, and v. Reckling- hausen regarded them as l)'mphangiofibromas. Frequentl}' the^' are accompanied b)' a considerable development of fibrous tissue. Bauer- regards them as endothelial m character. The cells are often grouped Fig. 87 ^ /" -•^/ .:■/ :-■;"- W^__ ^ ^ ''% '■' :' '-^ ;-/ ^ ■' ■^''.:'\ V-.V ■^^r 1P- Pedunculated melanotic papilloma. The epidermis contains pigment in the deeper layers. The embryonal cells are free from pigment. about the vessels and often have an alveolar arrangement, which may be retained when they undergo rapid development. Unna,-^ on the other hand, regards the cells which characterize these tumors as of epithelial origin. Ribbert^ classes them under a separate head as melanomas or chromatophoromas. The latter term, except for the lack of euphony, is preferable, since it emphasizes the fact that the essential ^ Lehre von den Geschwiilsten, Bergmann, Wiesbaden, 1902, i, 117. ^ Virchow's Arch. f. path. Anat., 1895, cxlii, 408. ^ Ibid., 1896, cxliii, 224. * Geschwiilstlehre, Cohen, Bonn, 1904, p. 255. MUCOUS I'.II'II.I.OM.IS 155 natiiic ot rlu- tmiiDi lies in tlu- chIIs, whiili show a disposition to pigment foi niarioii. All orlu r factors, papillar\- toniiation, the rtlativc proportion ot tihroiis rissiu- or \tssfls, is unimportant. I In- distiissioii, too, as to the origin of" thtsc cells is irrelevant. There can be no doubt of their orif^in in c()niit in ;;iiuial loilnws one of three types — the iiicciarix c, thi- noduhii, ami the fun^itoini. Tht- Ulcerativt' i l"i^. 'y4). I his i\ |n- arises most freciiicntiv from surfaces coN'ered with scjuamoiis epithehum. A ^ood example is rodent \/f W Carcinoma of lip. Growth of cell columns is more advanced, and there is round-celled mfiltration. ulcer of the face, in which, beginning from one pomt, the normal epithe- lial surface is destroyed as the cancerous process advances. 1 he course of destruction resembles in principle other chronic ulcerative processes, Fig. 94 --:'r* '--w ■•^■iiW-^ ^*'* .^':-;;t ,.; ■ ■■ ...■■■ \ * Schematic representation of ulcerative carcinoma. The covering skin is destroyed as the growth advances. particularly tuberculosis and syphilis. Carcinomas, however, are not ulcers pure and simple, for the peripheral infiltration leads to thicken- ing so that the ulcer has a raised border. When the growth is very active and the deeper tissue is invaded, large masses ma\- be formed underlying the ulcer. When the carcinoma cells infiltrate the surround- II 162 CARCINOMA ing tissue extensivel}', the ulcer forms an insignificant part of the tumor. Superficial ulceration with borders but slightly infiltrated indicats a slow growth, while greater infiltration indicates more rapid extension. Fig. 95 ''J: f; .-■— ^.ffl ;;^ ■ if.. .■!■■■/ ; '*'V \ Jjfi .''■^■^' '^\k?? : '■■ v4/? ■ V"; ■'. ' '/: ■ ■ (:■% F \ ^_jw^ '; & Schematic representation of a carcinoma of the mamma. The tumor forms a nodular mass mdependent of the surface. The Nodular TFig. 95). — This is the form which usualh" arises in the centre of parenchymatous organs. In slowh^ growing carcinomas the formation of fibrous tissue is extensive, and the tumor becomes irregular and nodular because of the contractions of this tissue. If these contracting fibers are attached directh^ or indirectly to the Fig. 96 ■>t^- J.:/J / f^ ' Schematic representation of a deep carcinoma over which the skin has ulcerated secondarily. surface, they produce dimpling, which becomes more noticeable if the centre degenerates. The more rapidly growing types compress the surrounding tissue and form pseudocapsules and thus resemble in their external form the expansile tumors. When such carcinomas approach the surface ulceration may arise secondarily (Fig. 96), either by pressure I'llYSlC.II. i-.ll.lR.ICTERS OF C.IRCISOM.I i(i;; necrosis or b\' an actual extension ot the malignant cells into the surface epithelium. An ini|')eiulinii ulceration of the surface is presa^ec^ b\' intense Muiiiii ot the skin, due to iiuiiasecl \ asculan/arion. Fk;. 97 Mr ■7;l ;(^ U^i.' / ■ J •' ' ,-':,?■;"; ;^^^=#^^F^ J> Schematic representation of an intestinal carcinoma, forming primarily a tumor projecting into the free lumen of the bowel. The Fungiform (Fig. 97). — In this t^pe the mass of the tumor pro- jects outward from the surface from which it arises. In externa! form Fig. /^^%^ Pohp of intestine, showing prolifer.r : ., d cells. The basement membrane is intact. It resembles a papilloma, and in early stages it is only b\" examining the base for signs of infiltration that it is possible to determine its malig- 164 CARCINOMA nant character. The surface may show proHferation (Fig. 98), while the deeper portions are unaffected. This type occurs most frequently on mucous surfaces, notably the bladder and intestinal tract. Fungous tumors may form secondarily by the perforation of the surface by deeply seated tumors of the nodular type and the growth of the tumor bej^ond the surface (Fig. 99). These are nodular tumors primarily, and the fungous development is but an accident of growth. They are often seen in glandular carcinomas of the breast. Fig. 99 : ■'' •-;' . ■■ ■ -■"' . -'"l."^^' JP/ Schematic drawing of a breast carcinoma which has perforated the skm and formed secondarily a firm fungous mass. Macroscopic Appearance on Section. — On section, carcinomas are grayish white or reddish grav, or sometimes translucent. Parts that have under- gone fatty degeneration are deep yellow. One can usually distinguish stroma and parenchyma by their colors, white or pink and grayish white respectivel}^ In the more cellular types the surface sometimes appears moist, grayish white, homogeneous, and, in the scirrhous type, the sparse parenchj^ma ma}^ be distinguished with difficulty, the whole presenting the appearance of a fibroma; but in the majority of cases the contrast is so plain that the character of the tumor can be recog- nized with certainty by inspection. In many superficial carcinomas the cell columns can be seen on the surface about the edge of the growth and can be forced out by gentle pressure. These "plugs" may be regarded as the macroscopic units of carcinoma. By crushing the columns under a cover-glass they are seen to be made up of groups and whorls of epithelial cells. MICROSCOPIC STRl CTlRi: Oi C.I RCI .\()\l ,1 105 Consistency. CaicinoiiKis an- al\va\s hard to tlic- roiicli. 'l"lu-\- can In- lust coinpaird to a puce of luhhir, r. <;., tin- hcc-l of" a iiiI)1kt hoot, with small nocliila t ions hki- thi- siii laic of a }iiicii pc-ar; ot, as tin- (ifnnaiis say, like (.lastic cat tila^c, not without elasticity , hut still, in the main, cpiite finii to tlu- touch. This piciiliar (ieiisitv in cancer is remarkahh' constant. It differs hut little, iiiispecri\e of" the proportion of cells and connective tissue. Tlu- lar^e, celluhii tumors are hut little less firm than those poor in cells. C\st formation ma\ mask the consistency, and dejienerative processes may hrinluiual il'iji. I02j. Diimi^ the (.-aily (.haiims in tlu- c\ roplasin rhi- luulci itsi-nihU- rhost- ot rlu- niorlici' cells, hill quukK ilu\ too show inipoitant dcx la t ions. I he details ot these iail\ eliaimes ha\e not lueii liilh woikeci out, luit it lias iieen noticed rliat tin chioniosoinis \ai\ m luiinlni and character and that tn- Imc. 103 Ik;. 104 ;fP\ 1 npoLir iiiiroses in isolatt-d carcinoma cells. Imperfect division of chromosomes in isolated carcinoma cells. polar or multipolar mitoses sometimes occur (Fig. 103). In certain types of cell changes, apparentl}' degenerative, the chromosomes fail to separate normally (Fig. 104). Hansemann' has made these abnormal mitoses objects of minute study, and has come to the conclusion that Fig. 10: aa - ® « o ^ « » "»* » * © « Cells from the border of a rapidly t;ro\vmg breast carci noma. it is due to them that new races of cells are developed. My own studies in early epitheliomas of the skin, in A.-ray carcinomas, and in artificial epithelial proliferations have convinced me that the primary changes occur in the cytoplasm and that abnormal mitoses appear only after the cytoplasm has attained a certain degree of deviation. In the more ' Ztschr. f. Krebsforsch., 1907, v, 510. 170 CARCINOMA advanced degrees of deviation, the cells lose all resemblance not only to their mother cell, but to all epithelial cells, and it is only by their relation to the connective tissue that their character and origin can be determined (Fig. 105). The Topographical Relation of Cells and Connective Tissue. — A microscopic section of a typical carcinoma shows groups of cells surrounded b}^ connective tissue (Fig. 106). These are the "cancer nests" which have played such a prominent part in the discussion of the histology of car- cinoma. It was the formation of these nests that was formerly regarded as evidence of malignancy, for it was assumed that they resulted from the escape of cells from the central group and the formation of new Fig. 106 a C^-—.'^ ;r^:v-': /-•• ^ '.• - ^ '" 1 ■'j^-'^ • " i-i' "■ ,7;V--' ■ ' y. ".'■. iSv! ■ "^•,- •' J> ■' r-^ , ^ ■'f "} ■, :-^'" : '--V 6 tX^-. :, S Carcinoma simplex: a, carcinoma cells; b, connective tissue; c, round-celled infiltration in connective tissue. groups at some distance from the primary tumor mass. The study of serial sections and reconstructions of small tumors has shown, how- ever, that the formation of "nests" is almost always only apparent. The growth branches out like the limbs of a tree, and a single cross- section shows these as isolated groups of cells, the connection with the rest of the tumor falling outside the plane of the section (Fig. 107). Many tumors, however, support the older view, for cells may escape from the border of the tumor into the spaces of the connective tissue and become independent of the mother tumor, forming, in fact, a regional, metastasis. A section through such a group of cells shows a true cancer nest according to the old view (Fig. 108). Serial sections alone, in most instances, make it possible to determine which type is present. Tor(jc.k.iriiu:.ii. ki:i..iri<).\ ui cells .jmj co.wlct/il t/ssi/-: 171 In soiiK- iiistancts cpirlulial ctlls may Ik- cliff"ustl\ distributed in flu- coniucrnc tissue so that tath cell is more or less surrounded h\ C r k;. 107 /*,. I' ■f.,.. M - y.? •• ^' Schematic drawinji of a carcinoma, showing how a section in Hne a-b would show apparent cell nests, while really the cell columns are cf)nnected with the main portion of the tumor. Fig. 108 a Jfi- Schematic section of carcinoma in which cells have escaped from the periphery- forming isolated nodules. A section a-h made through these nodules shows true nests unconnected with mam rumor. connective-tissue Hbrils (Fig. 109), presenting somewhat the arrange- ment topical for sarcoma. Not infrequently small areas about advanc- 172 CARCINOMA ing carcinomas have this arrangement, and occasionally it is found throughout the major portion of tumors which are unquestionably of epithelial origm. Fig. 109 V ) :.. fey ^^a vJn , te 'o ^ > vi' \ \ . X \ ''.V© t-) i \ '% \'' ''■"' 'w< ^^.r- ft' ,.:\^ ^^ ^ Carcinoma of the mamma cells scattered diffusely throughout the connective tissue. Relation of Cells and Stroma. — The relation of cells and connective tissue has been considered in the section on etiology. Our interest centres here not in a causal but in a developmental problem; whether the entire stimulus to growth lies in the epithehum or whether the Fig. 1 10 c ah Epithelium lining an ovarian cyst: a, h, cells piling up without connective tissue; c, fine strands of connective tissue follow the cells, but no vessels. connective tissue takes an active part. In certain tumors, especially in ovarian cysts, epithelial cells develop without the presence of con- nective tissue (Fig. no). This has been regarded as evidence that the epithelium is the active factor and that the connective tissue is secondary. SKCO.yn.iRV cii.ixcks i\ c.ircimjm./ 173 ( )ii tin- Dtlur IkiiuI, it is ;i laot tli;it m m;in\ taiciiiDiiKis, tin- coniuctive- tissui- |")i()iliKti()n IS far iii c-\ocss ot t he- ipirlulial |)i<)ciuiti()iK :iiul even if stiiiuilarcii h\ tin- (.iMtiu-liuiii ir at least ^aiiis a niomtiuiim winch ir iiiaiiuains. Soim- luu ami intcitsriiif; observations havt- hc-c-ii maclc- in iiiousc carcinomas. In some ol these, after a tew generations, the connective tissue becomes greater and the epithelium becomes less until the entire tumoi- is made uji ot (.-mbr^'onal connective tissue; in other wortls, a sarcoma has de\iloped. I he changes which rake jilace in mouse tumors strengthen the assumption otten made that connec- tive tissue limits the tievelopment of epithelium and in those tumors where rlu comucrne tissue becomes acti\e the epithelial cells dis- appear. I mctorial chemistry which might gi\e some evidence has not yet been applied to this problem. Fig. 1 1 1 -■V .-* '■A M /// Cells in scirrhous carcinoma of the breast becommo; fused; the cell boundaries are indistinct, and the nuclei deiienerated. Secondary Changes. — Atrophy (Fig. iii). — Atrophy from lack ot nutrition occurs especialh' where there is contraction ot the connective tissue with obliteration or partial occlusion of the vessels. It is often accompanied by karyorrhexis and chromatolysis. In tumors made up largely of fibrous tissue it has been assumed that cells in greater number were at one time present, but that the}' perished from lack of nutrition. The presence of fat in the protoplasm which so often accom- panies atrophy, and the changes in the nucleus already mentioned, make this seem probable. Calcareous Infiltration. — Epithelium, especiall\- horny portions, and also stroma ma\- calcif\', following h\aline degeneration. Limited necrotic 174 CARCINOMA areas are usually most prone to calcification. On the whole, it is one of the rarer forms of secondary changes. Fatty Metamorphosis. — This form of degeneration is extremely common, especially in rapidly growing tumors. Such tumors often show on cut surfaces fine lines of yellow material which represent degenerated columns of cells. Frozen sections stained with Sudan III show these changes strikingly. Fig. 112 M i^^y , a C-:^ Necrotic lymph gland in a non-degenerated carcinoma of the breast: a, connective tissue with round-celled infiltration; b, carcinoma cells; c, degenerated carcinomatous area. Necrosis (Fig. 112). — Necrosis as it occurs in granulomas is also seen in carcinomas, especially in the rapidly growing pseudoencapsu- lated type, and it may be due to occlusion of vessels by thrombosis or to suppuration following secondary infection. This accident is most likely to occur in situations normally exposed to bacteria, as in the digestive tract or the cervix. The series of phenomena which usually follows infection occurs here in so far as the tissue will permit; and, because of the relatively defenceless character of the tissue, necrotic sloughs are frequent. The absorption of toxic material from this source SKC(J.\I).IR) CII.IM.ES IS (:,lki:i.\(jM,l 17.') adds nuicli to tin- hiiidi n of tin- |):iti(nt. As the de^cnc'r:ii ion pioj^resses, N'fsscls bttoim- tiotlcd ;ind r.\ttn.si\ c lunioiih.i^cs into ilic rumor or externall\ oiiui. In tins \\:i\- cysts ni;i\ toiin, tlu- loiuciits of which vaiy HI i-oloi lioni daik icd to h^ht stiavv, dc-pciidni^ n|)on thi- dc-^it-c- to which tht' hcmo^lohm is ahsoihcd. C'holcstciin ci\stals usiialK' remain to nuhcatc the ()rihatics, and ir is not uncommon to Hnd growing nodules within the walls of a lymph channel. Sometimes, as in the lungs or peritoneum, the lymphatics ma\' he filled with columns of cells, so rhar rhe vessel nerwork srands our prominenrlw rile rhoracic ducr especiall) may he Hlled in rhis \\a\'. UsualK rhe cells are carneil alonji; in rhe lynijih current unnl a hmph gland is Fig. 1 16 C m./:: ■:' fH": Carcinoma nodule extending into the periosteum: a, carcinoma; h, newly formed bone; c, connective tissue; d, bloodvessel. reached, where the\' develop. Somenmes the first gland escapes and a more distant gland is afFecred. Whether the wandering cells are filtered through the first gland or are detoured rhrough some anasro- mosing channel is not known. Secondar}- growths in general, imitate the mother tumor more or less closeh' (Fig. 117). Thus, pearls may be formed where the primary tumor is an epithelioma (Fig. 118), or the glandular structure may be retained when the primary tumor is an adenocarcinoma (Fig. 119). In general, the cells of a metastatic tumor are of a lower order than those of the primary tumor. The metastatic nodules are usually spherical when occurring in regions where resistance is equal m all directions. This is seen most typically in the liver. There is often a sharp line ISO CARCINOMA of demarcation between the tumor and the parenchyma of the organ (Fig. 120). Fig. 117 -^ ; J> '. V Metastasis from tongue carcinoma in lymph gland, showing islands of squamous epithelium in the lymphoid tissue. The question of metastases is practically the most important one, as it is they that frustrate the majority of attempts at cure. The site Fig. 118 '^pt - \ -1 >. " * ' ^i' 1 1 .. 'V,. V / ..• 'Y- : - y " y. i, ". » '■^ ^ ' .0 _ .- ^•^'^ >'^ ft?-.. .^^=p<^ IcH -n^^^iWS Pearl in lymph gland. Metastasis from a carcinoma of the lip. of the earhest metastases has been carefully worked out for the carci- nomas of the various important regions, and may most profitably be M /■:/-./ ST./ s/s Isl iliscussiil m tlu lonsiilci ;i I Kin ol those itjiKJiis. It is sufliiunr to sa\ in this pi. Ill- th;it (..i i (.imiiii.is \;ii\ in thc-ir dispf)siti()ii to loiiii iiH-r:is- rasfs. In .1 hiii.ul \\.i\ those iii whirh n Ik di \iai( most lioni the niotlu-r !• ic. I \') -^ , ..,r*'t5 '"■■*"'*<>• - ".-Off Oorf> ^ . J> :%".. —i '0» ft Osfl' <•■- — r Columnar-celled metastasis in liver from carcinoma of the stomach: a, metastatic gland cells; b, liver epithelium; c, connective tissue. Fig. I20 Metastatic nodule in liver, showing boundary between glandular caicinoma aiiil liver tissue. 182 CARCINOMA type are most apt to form early metastases, and, on the other hand, the less the deviation the less the disposition to metastatic formation. For example, the glandular tumors are the least apt to form metastases. This is not universally true, however, for the basal-celled epitheliomas, which bear little resemblance to the epidermal cells, are not prone to form metastases. Those which show a proneness to secondary changes are less apt to form metastases. This is particularly true of those which undergo colloid degeneration. If the abundant formation of fibrous tissue is a conservative process, as some pathologists affirm, tumors characterized by the formation of much stroma should show less tend- ency to the formation of secondary tumors. Systemic. — The bloodvessels are sometimes the avenues of dissemina- tion. The cells may reach them through the thoracic duct, hence into the venous circulation, or may penetrate directly into a vein, or more rarely into an artery. The carcinoma may gain access to the bloodvessels by way of the lymphatics of the vessel wall, thence, by extension, into the lumen of the vessel. When extension takes place by way of the bloodvessels any part of the body may be invaded, but especially the lung, liver, bones, nerve centres, and the skin. The reason for this predilection of secondary growths is not understood. It is likely that suitability of soil is a more important factor than accessibility, for certain kinds of carcinomas have regions of predilection of their own. Probably those cells which lodge in un- suitable soil are destroyed, while those which are deposited in a more suitable medium develop. As an example, multiple nodules ma}^ occur in the skin and the remainder of the body remain free. When migrating cells reach favorable soil other than the lymph glands (liver, lungs), they usually grow rapidly, with a disposition to form globular expansile tumors. It is exceptional for such metastases to retain with any degree of perfection the mother type of cell, though they sometimes do so in a striking manner. Types of Carcinoma. — For the purpose of stud}^ it is sufficient to classify carcinomas according to the cells from which they spring. Thus they may be divided into the squamous and columnar-celled types. Variations of the cells of these groups give rise to modifica- tions in tumor types which some writers express by more complicated classification, but this simpler classification, adopted by Borst, has much to commend it. Squamous-celled Carcinoma. — Carcinomas derived from the squamous cells have different characteristics according to the situation from which they spring. It is convenient to separate them into groups. Epidermoid Type (Fig. 121). — The squamous epithelium of the skin gives rise to tumors which have a disposition to slow growth and SQUAMOUS CELLED CIRCI \()M ,1 183 relativtlv late iiutastasis. riu'se biolofiical characteristics express themselves histoloj^ically hy an approach to the normal development of P"lG. 121 at* {^■^ '/ v^ ..-V /^.■;. / ^ i \ ■;.^- ^ V Carcinoma, epidermoid type. Fig. 122 . i 5« .* i »i 'tiJ^'V- 5P^ Epithelial pearls. 184 CARCINOMA epidermal epithelium resulting in the so-called pearly formation (Fig. 122). These pearls are groups of cells arranged concentrically, which present progressive changes from the periphery toward the centre, corresponding to the changes in the normal epidermis from the deep layers to the surface. They are striking objects when seen in stained sections, because of their affinity for acid d^'es. They obtain their name from their bluish-white, glistening appearance in unstained sec- tions. In teased specimens they are seen to be made up of irregular spindle-form cells with large ovoid nuclei surrounded bv a structure- less protoplasm (Fig. 123) which has a marked affinity for acid dyes. The disposition to pearl formation is often seen in metastases in the Fig. 123 Ci, '■**> z •%' '9 ^,,V ^ ^^4 & Teased pearl, showing forms of individual cells. lymph glands. Carcinomas which develop elsewhere form metastases in the skin relatively late.^ Epidermal tissue in abnormal situations, as m dermoids and teratomas, occasionally gives rise to carcinoma. Not mfrequently persistent gill clefts are the starting point of such tumors. Tumors of this t^'pe are sometimes found where the presence of squamous epithelium cannot be accounted for even on the theory of congenital displacements. That they result from metaplasia from columnar cells seems then to be the only explanation. This explana- tion is accepted by all pathologists for such tumors in the gall-bladder, and those in the breast and stomach may belong to this class. ^ Daus, Virchow's Arch. f. path. Anat., 1907, cxc, 196. S-*-'*^ i^ ' i - 0l^ , -•' ■■■■■■ , - ■;-frk"l Z' 'fl» .c'^^ ? i _ ../. ^^^ " y > r %^ ^ ^ / »» -- ^ ^ t* r ^ ^' ■ ^ ' ^'-'^^-i-^y-:: Carcinoma of the tongue, showing nests of cells, the centre of which can be identified as of epithelial origin, but show evidence of rapid growth. The Mucous Membrane Type. — The epitheliomas going out from squa- mous cells which line mucous membrane differ from those developing from epidermis in having less disposition to form pearls, since they deviate more rapidly from the mother tj^pe and are correspondingly more malignant. This may be because mucous surfaces are more subject to maceration, which necessitates a more frequent replace- ment of cells and, therefore, more active proliferation. In addition to this, the abutting connective tissue is less dense and undergoes reactive changes more quickly. They are found most frequenth^ on the cervix, lips, tongue (Fig. 125), and gums, less often in the esophagus, pharynx, larynx, vagina, and penis, and urinary system. Ulceration is common GLAXDUL.IR CIRC I SO M.I ^<^7 and the formation of fungiform tumors is more usual than with skin carcinomas. In harmon\ with the- greater clt\iation of these cells tioni the iioiiiial they u;rovv more lapulK aiul form early metastases. I hey make up a larj^e proportion of the surface carcinomas. I heir location in regu)ns where irritation is fre(juent has ^i\en prominence to the theory of irritation in etiolofiy. Glandular Carcinoma. This t\ pe is developed from glands. Formerly it was the custom to include the more highly differentiated tumors under the head of mahj^nant adenomas, which was a suitahle designa- tion, inasmuch as it indicated that the normal gland structure was retained and that the sunounding tissue was invaded. 1 he term de- FiG. 126 ^^^'M Glandular carcinoma of mamma. The gland lumina are retained in some places and filled up m others. mands that the tumor be formed of glands normal except for the rate and direction of growth. The glands must be lined by a single layer of epithelium, but so rarely are the glands lined by a single layer of epithelium that a separate classification for such forms has been largely- abandoned. In contradistinction to these the type in which the glandular lumen is filled with cells was designated by the term adenocarcinoma; and when all trace of gland structure was lost the t}pe was called carcinoma without a qualifying prefix. But since all these varieties frequently occur in a single tumor they are at present grouped together under the term glandular carcinoma' (Fig. 126). These tumors are > Kaufmann, \'ircho\v's Arch. f. path. .Anat., 1898, cliv, i. 188 CARCINOMA found most frequently in the gastro-intestinal tract, uterus, mamma, and prostate, but may occur in any glandular organ, but rarely in the kidney (Fig. 127) and liver. Fin. 127 ■■■-■■.iAMi<- Glandular carcinoma in the kidney, secondary nodule Fig. 128 ^ • ■ .,..,;■.•.•"•■ ^^...y§^. Intestinal mucosa showing earliest stage of change from normal; slight elongation of glands, with round-celled infiltration. f,r^ . It is the tumors derived from glands that give point to the observa- tion of Adami that in malignancy the "habit of function is replaced GL.l.\Jjt I..IR (:.ll- ^r.f -^ '^ J Encephaloid carcinoma. Large alveoli with relatively small amount of connective tissue. degeneration in cancer of the stomach. In their manner of growth they do not differ from other cancers except that they are more hkely to give rise to transplantation metastases (cancers of the ovary) and, perhaps, less hkely than squamous-celled carcinoma, to give rise to bloodvessel metastases. The contrary is true of glandular carcinomas which are not connected with the surface. In these metastasis is more apt to occur than in the squamous variety. Because many of the glandular organs possess a natural capsule this disposition to metas- tasis may be lessened and the tendency to invade the surrounding tissue is strikingly Hmited. The importance of this capsule becomes very evident when tumors of the gastro-intestinal tract are compared with those of the thyroid. STRL en M./L CL.lSS/I/C.mo.X Of C.lkCJXO.M.l 1<)1 Structural Classification. The classification of carcinomas accord- \n^ to tin- amoiiiu ol coniuctive tissue or according to the t\pe of clefieneration present, is ol \aliu- in ceiram situations in tktt innnin«; the !''<•■ '32 protinosis. Accorilmir ro this I'l.issi- /e^ ao^ ^-.'^^^ tit posei ■)rot:nosis. Accoiclinn to this classi- fj^f 6 ^ t- Ai'^fS)' (!i K: 2f ',' ication. when a carcinoma is com- '?'* «^-^/v «, &*• '^•Ji' «.^V; ... ,*> DOsed of hhrous tissue and cells ^ <\*^><^,S'^' 'i^^-^'^iJ' 4 ',>!>! k^/? they ma}- be very maHgnant (tig. '^y^<^^'i-^.0®MS^^^i:h-y-.^ 132). It is the type of the cell and Encephaloid carcinoma, showing the not the number that indicates the very small amount of stroma, rate of growth. If the connective tissue is in excess it is a scirrhous cancer. The cells may appear as small nests (Fig. 133) or singl}' or in narrow columns (Fig. 134) in large areas of connective tissue. Here again the proportion of cells slow growth; but, on the other hand, ^.^^'^^3>|''i';'*si^^<-C'^lC{'?.'^^^^ -^.- •-.. ^ Fig. 133 tUt^' Scirrhus of breast, with well-dehned cell nests in certain areas Individual cells at left. to connective tissue ma\' give little evidence of the character of the tumor. Those in which the cells are sparsest may give rise to early 192 CARCINOMA metastasis. Any of the elementan^ t3^pes (squamous, glandular, etc.) may form any structural type at some period of their development, and, therefore, the consideration of the latter confuses rather than clarifies the classification of carcinomas. In like manner, the desig- nation according to the type of degeneration, as colloid carcinoma, while it suggests the most common source of the tumor, yet, on the whole, emphasizes too much the importance of secondar}^ characters and tends to confuse rather than elucidate the real nature of carcinomas. Fig. 134 . # •■ * Scirrhus of breast, with cells in single columns. Diagnosis. — Clinical Diagnosis. — After the detailed consideration of the nature and pathogenesis of carcinoma a brief consideration of their clinical manifestations will be useful from a practical point of view. Jge. — Age is a contributory datum in the diagnosis of carcinoma, smce the large majority of cases occur after the thirty-fifth year. While carcinoma is comparatively rare in young persons, it is at the same time the most frequent epithelial tumor; hence a chronic ulcer in a young person, if a specific ulcerative disease can be excluded, should give rise to suspicion of carcinoma. In the young the growth of these tumors is rapid, the age of the patient being, in fact, more useful in prognosis than in diagnosis, for the local signs should always make the diagnosis possible. i)/.i(:.\(js/s or c.iRcisoM.i I9:i Cachexia. — "1 his is alvva\s a late sij;n ot carcinoma and si^niHcs an extensive in\asi<)n or tlie existence of complications. Much mischief has aiisin lioni consiihi in<; the |-)resence ot cachexia necessan' tor diagnosis, the nature ot a tumor, especially a deep-seated one, being too often unrecognized because cachexia is not present. Cachexia is distinctiveh' a terminal manifestation, and is more properl\' placed in the |iosrm()irem protocol than in rlu- clinical history. Pain. Pain in cancer is due to the invasion of sensitive areas by the growth. It is a late symptom and is usuall\' preceded by extensive infiltration. It is usually lancinating and intermittent but later may become constant, in which case other factors than invading cells usuall\' pla\ a part. InHammatorx- complications or pressure on other organs ma\ produce pain. Macroscopic Appearance. — A carcinoma manifests itself as a tumor or an ulcer or both combined. If an ulcer appears primaril\' it indicates that the disease originated in the epithelium ot the ulcerated surface. In some cases superficial carcinomas may for a time present no ulcera- tion, but sooner or later the surface breaks down. On the other hand, a neoplasm which originates in deeper tissue may involve the surface secondarily. Usually ulceration is one of the earliest clinical manifesta- tions of carcinoma of the epithelial surface. Ulcerative Type. — Since carcinoma is an infiltrating growth, ulcera- tion is obviouslv certain w^hen surface epithelium is affected. Con- verseh", ever\' chronic ulcer should suggest the possibility of carcinoma. The contour of an infiltrative fibro-epithelial ulcerating tumor is, in general, circular, though it does not possess a distinctive outline. 1 he ulcer which it produces is characterized by its dense border, which is one of the chief signs of carcinoma, and cannot be too carefully- noticed. The border is usually vertical, but may be sloping or under- mined. If it is extensively undermined the growth is probabl}' of deep origin and involves the surface b}' an ulcerative process due to a second- ary infection. The floor of the ulcer is granular and friable and bleeds easily. Frequently, in the granulations, whitish areas, the "cancer plugs" interspersed b}' pinkish connective tissue, may be seen. In the border of the ulcer wherever the epithelial nests approach the surface theA' ma\' be seen as small whitish points or "plugs," which may be pressed out like comedones. In mucous membranes, likewise, the edges of the ulcer are dense, though less so than when surface epithelium is invaded, and the "plugs" may not be so readily discernible. Tumorous Type. — The nodular carcinomas present little that is char- acteristic except their location. Pitting of the skin over them is often vet}- noticeable, and the increase of the superficial vessels indicates somewhat the nature of the tumor. 13 194 CARCINOMA These points taken together with the location in an epithehal tissue often give a presumptive diagnosis. The density of these tumors is the characteristic feature and even when deeply seated, is at once apparent to the palpating fingers. There is usually a lack of mobility owing to the infiltrative character of the tumor, which does not permit it to glide about within the capsule as in true encapsulation. There may be a pseudocapsule, due not to true encapsulation brought about by formation of new fibrous tissue, but to the central growth which pushes aside and compresses a shell of the surrounding tissue. Microscopic Diagnosis. — It cannot be too strongly emphasized that, no matter how certain the clinical or macroscopic diagnosis of carcinoma may be, a careful microscopic examination of each tumor should be made since only in this way can the clinician keep his bedside diag- nostic skill on a high plane of pathological advancement. The variety of carcinoma often can be determined only by microscopic examination, and in rare instances even a general diagnosis of carcinoma is impos- sible without this procedure. For this purpose numerous sections of the tumor must be removed from such regions of the tumor and treated in such manner that the following points can be determined With the greatest facility. Invasion of Surrounding Tissue. — The only certain criterion of car- cinoma is the invasion of the surrounding tissue by the tumor cells. In microscopic examination, the border-line, between the advancing epithelium and the connective tissue surrounding it, must be constantly kept in mind. This border-line may be sharply defined where the basement membrane is easily determined, as in the intestine, but less easily where it is absent, as in the uterus. In the absence of definite evidence of invasion of the surrounding tissue a positive diagnosis is not permissible notwithstanding the appearance of the cells. In deter- mining this point certain conditions which are likely to cause confusion must be kept in mind. Among them are displacements resulting from the healing of wounds, particularly where two types of epithelium join, as in the cervix; mechanical displacement of tissue in the prep- aration of the section; and reactive inflammatory processes which may cause proliferation simulating an extension of the epithelium into the surrounding tissue. When epithelial cells are discovered out of their normal relations where such possibilities exist other evidence of malignancy must be considered. Important as is the occurrence of epithelium beyond its normal boundaries for recognition of the mahg- nant growth, it is in itself not conclusive. When epithelium is discovered without its normal habitat, if there is no change in cell type point- ing to changed biological characters, the probability is that it has gained this position by means other than its own autonomous growth, and MICROSCOPIC DI.ICXOSIR OF C.I Rcl .\/j.\l .1 1'/>'•■. I hf pro^rcssixc change in ct-ll t \ pc tiom thi- noiiiial, liill\ ditKiintiatc-d t\ |h- to the less diflereiitiared, usualU' in some, thouj^h variable degree, parallels the invasion of the surroundiiif^ tissue. I his manifests itself m the lorni of the cells, and if the\' have specific characters, r. i^., prickle cells m skin, goblet cells in digestive tract, the\' lose them and aiijiroach more closel\' the undifferentiated embi\()nal cell. 1 he change in cell typt- is always present in some degree in carcinoma. In malignant adenomas and other more highly developed of the glandular carcinomas the cell type is very slightly altered, so slightl\ , m tact, that the function of the normal cell may be imitated, l^sually in these tumors, a gradation into adenocarcinomas with a corresponding change in cell t^pe ma\' be observed m other parts of the tumor. Invasion of surrounding tissue and metastatic formation ma}' occur with little or no change in cell t\'pe. Ihis has been called, with doubtful propriety-, metastasis of benign tumors. It is observed among epithelial organs most frequently' in the thyroid. These should be regarded as malignant tumors with the minimum of cell deviation. Usually the degree of malignanc\' is indicated by the degree of alteration m the cell t^'pe; the greater the deviation from the normal the more malignant the tumor. Atypical Mitosis. — The presence of involutional cell forms heightens the probabilit\' of malignant growth in some instances, but the}' may be very numerous in tumors of low^ malignanc}' or even in reactne proliferations. Atypical mitosis has been advanced, particularly by Hansemann' as a valuable diagnostic sign of malignancy, and while not of universal applicability it is useful in man\' instances. The mere presence of mitotic figures is indicative of rapid proliferation, and in itself is a presumptive sign of abnormal growth; but it is the irregular and unequal development of the chromosomes at the nuclear poles that IS particularly suggestive of malignancy. Care must be taken that in a dividing cell which has passed the metaphase, an oblique section, which cuts the centre of one aster and but the edge of another, is not interpreted as an irregular mitosis. Hencke believes only serial sections can obviate such an error, but usualh xhey are unnecessary if the sections are not cut too thin. The unequal division of the chromo- somes IS evidence of a disturbed proliferation, but as a diagnostic sign of carcinoma its value is lessened hy the fact that it is b\' no means always present and that it ma}' be present in reactive processes. It IS of value, however, in the presence of other signs. The number of ' Die mikroskop;sch. Diagnostik bcisartige Geschwiilste, Hirscliwald, Berlin, 1897. 196 CARCINOMA mitotic figures is sometimes regarded as a measure of the degree of malignancy, but it is an unreliable sign. Direction of the Mitotic Axis. — The direction of the axis of the mitotic spindle has been considered by Amann^ as a valuable sign in diagnosis. This axis normally is parallel to the basement membrane, so that the. cells resulting from mitosis lie in the same plane. ^ In malignant growths the axis of division deviates from the parallel and consequently the cells become piled upon each other. This sign is mentioned only to warn the beginner against its uncertaint}^ Mitosis is seen but seldom in routine examinations in tumors of this character, and artefacts are often mistaken for it, with the result that a diagnosis of malignancy is frequently made on these premises. The mere piling up of cells should not be mistaken for malignancy, since it may be simulated by contraction of the gland in a longitudinal direction. It is never more than a suspicious sign and never, in the absence of infiltrative growth, is it alone evidence of malignancy. Round-celled Infiltration. — Ribbert emphasized the importance in diag- nosis of round-celled infiltration about the border of the advancing carcinoma. The sign is of great value in many instances, but I am disposed to agree with Hencke that it is subject to too many errors in interpretation to be of decisive value. Carcinomas, especially those of the glandular type, may be malignant without cellular infiltration. On the other hand, round-celled infiltration may be found associated with other conditions than carcinoma; for example, chronic ulcers, syphilis, and tuberculosis. Chronic ulcerative processes (especially of the cervix and stomach) can often be distinguished from carcinoma by the fact that in the former the infiltration is greater in the older necrosing area, while in carcinoma it is most intense about the advanc- ing border. In syphilitic and tuberculous processes, specific signs of these diseases must be found before they can be distinguished from carcinoma. The presence of round-celled infiltration is most valuable as a diagnostic sign of carcinoma where other causes for the condition are not present, especially in cases of parenchymatous tumors. Relation to Elastic Tissue. — Abel advanced the belief that in beginning epithelioma the cells grew around and enclosed the elastic fibers, while benign proliferations pushed these fibers aside and did not enclose them. Recent investigations tend to confirm the fact that epithelial cells have some influence on the production of elastic tissue, but in a sense diflTerent from that understood by Abel. It has been determined that elastic tissue does not develop about epithelial growths. This seems to be 1 Mikroskopisch. Gynakologischen Diagnostik, Bergmann, Wiesbaden, 1897. ^Abel, Arch. f. Gynak., 1901, Ixiv, 316. TRi:.rr.\fi:\T or circisomi I!»7 clc;iil\ shown iii ;icl\ aiuitl ^lourlis, Inir lor tin- diagnosis of earl\' car- cinonui it has hrtk- value. M icrochc'tnical Rt-cictiuii. I he- niicrochciiiical reactions of cells in carcinoma is too \ariahle to be ot f^reat diagnostic value. Lsualh in beginning malignancy the cells stain more deepl\' with many nuclear d\es, particularly hematoxylin. In many carcinomas in which there is a disturbance in nutrition the nuclei ma\' be large and pale. \\ hen the cells undergo degeneration they react specifically for the t\ pe of degeneration present. Treatment. I reatment may be dnided into palliative and curative. Palliative. It is to Czerny's credit that renewed attention has been directed to the fact that much can be done for the carcinoma patient after hope of cure is past. I he treatment of incurable cases has been too often relegated to the quack. For this the practitioner is not entireK' to blame. A drowning man will grasp at a straw, and when the patient's disease is pronounced incurable by his physician, he too often hastens to place himself in the hands of any charlatan who will promise a cure. It becomes the problem of the reputable practitioner to counteract this tendency. No one should care to retain a patient by holding out false hopes even to save him from falling into the hands of the charlatan. The duty of the practitioner is to state plainly to the patient or to his friends that the disease is incurable, but at the same time he can emphasize the fact that he has the power to relieve suffering and pro- long life. In his personality and his reputation for skill and probity on the one hand, and in the intelligence of the sufferer on the other hand, lies the possibility of retaining these patients. Medical men have been at fault in two directions. Their knowledge of tumors is often obviousl}' limited, as the patient discovers from a faulty diagnosis or prognosis, and too often the patient receives the impression that a fatal prognosis implies a total lack of power for good, an opinion heightened because too often the utmost is not done to relieve his suffering. ihen, too, the charlatan plays on the general ignorance of the public by declaring every tumor malignant, and thus treats as carcinomas innocent conditions with a larger percentage of cures than the skilled surgeon obtains in true carcinoma. In the recog- nition of these facts lies our chief power to prevent these patients from falling into irresponsible hands. Man\' practitioners do not regard it as unfortunate if the patient overestimates the power of palliative treatment to limit the rate of growth and regards such treatment with a faith not warranted by experience. In retaining the patient the physician preserves for himself an opportunity to study the life histor\' of the disease which, aside from perfecting his knowledge, stimulates him in the effort to recognize the disease earlier in future cases. 198 CARCINOMA Palliative treatment includes a great variety of measures, varying from the exhibition of an anodyne to the boldest operation. When the primary tumor can be removed without any considerable risk to life it is frequently desirable to do so. With glandular involvement beyond the reach of the operator, death is certain; but with the chief tumor removed and the patient freed from the toxins produced by it, the renewed hope and mental tranquillity often give the patient a period of comparative health and courage which well repay for the incon- venience of the operation. If death is ultimately produced by visceral metastasis it is usually attended by infinitely less suffering than the primary tumor would have caused. Frequently, the secondary tumor, perchance in lung or liver, kills the patient without his being long aware of the inevitable end. I am firmly of the opinion that much suffering can frequently be saved by judiciously selected radical palliative operations. Unfortunately, the problem is often a very complicated one and demands the highest degree of surgical judgment. Risk of operative death and the immediate suffering from the operation must be taken into account in determin- ing the right path to pursue. These factors depend upon conditions that cannot be treated in an abstract way because the skill of the surgeon, the facilities for good work and the estimation of the power of the patient to withstand the operation must all be considered in each individual case. The efficiency of the surgeon depends largely upon his ability to estimate these factors correctly. The patient is often asked to decide for himself, but obviously the surgeon is in possession of facts which will enable him alone to form an accurate judgment, in which judgment the patient need not in every instance be asked to participate. It certainly is within the province of the surgeon to decide what is best for those who place themselves in his hands, and he should not hesitate to assume a proper share of responsibility. He has, indeed, to answer to the science he represents for his acts, and must exercise care to avoid bringing discredit upon it. A palliative operation may be counted as a failure among the friends of the patient and the value of operations for the cure of malignant disease in general be discredited. This misunderstanding may be avoided somewhat by explaining to them the risk and the reason for assuming the risk and making it clear that a permanent cure is not expected. How much the notion that, even if the operation result fatall)^, the patient will be saved much suffering, has to do with deciding in favor of the operation must be estimated in each case. This phase often appeals to the none too sympathetic friends; but even though the patient is in accord with this view, and even though the surgeon may TRE.ITMEXr OI C.IKCISOM.I I!l!» lean toward a Nur/scluan j^hilosoplu , In- has no n^hr in an\' case t¥ VCV.,,^ ^ 1 ! i; ^,: '■ \ ' ■''^■:y,- -.v. • 1 Dermoid cyst. -jl- may undergo gradual growth, but this is usually due to an increase of the cystic contents. Rarely are they the starting point for malig- nant growth. KMIikYoM.IS 209 Embryomas. I iimois torinccl from ;ill tlntf c iiil)i \ oiial hiytis with- out c\ srir toiin.iiion ait- iioi ml ncjiuni l\ ot oiu'olo^ical iiUcicsr. I hey ail- toiiiul most I H(Hiiiii l\ m flu- ictiioii ot ilic kiihu\', in the region ot the saciiim aiul m thr hiam and |diai\n\. I he- emhryonal hiyc-rs are i ipiisiiiteil m \ aiiahle proportions. I hese tumors iKJt intre(juentl\' takt on lapiil growth, tormin^ tumois lesenihlm^ the sarc(imas in structure. Etiology. The e.xact origin of the emhiyoul tumors has not heen dt hmtil\ istahhshed, hut the followinii e.xjihinations, out ot the man\' which ha\c- hiin atUanced, mac he mentioned: 1. Mu two or three pohir hodies which appear during the matura- tion ot tlie ovum are beheved to become impregnated and give rise to the embryomas. In favor of tliis view is the fact that the location of some of the embryomas, notably epignathous and sacral tumors, corre- sponds to the most fiequent location of the polar cells. Against this has been advanced the fact that as many as five embryomas have been observed in the human subject, and such a large number of polar bodies is unknown. 2. Wilms has advanced the theory that these tumors are derived from the blastomeres of the developing ovum, which are supposed to become displaced at a very early stage of development. As these cells are separated at a time when differentiation has not yet taken place, it ma\' be assumed that wherever they develop they ma}', following their mherent developmental power, produce tissues contaming prod- ucts of all three germinal layers. This theor\- admits of the most universal application, and is now generall}' accepted. Diagnosis. — The teratoid tumors are recognized b)' the presence of cells representing two or more germ layers and by their location in certain regions. The constancy- of their location and their habit ot growth are often required to permit the identification of these tumors, after they have taken on rapid growth. The failure to grasp the inter- relations of the various types and stages accounts for the contusion ot views which prevailed until the ver}' recent past. Treatment. — 1 he removal of the tumor when in the inactive stage IS usually a simple procedure. After the\' have escaped their capsule removal requires the extensive operations of any other malignant tumor. u CHAPTER XVIII GRANULOMAS General Conception. — Under this head are collected a number of conditions which have, in common with true tumors, a structure dis- tinctly cellular in character. The term itself suggests its double char- acter, a tumorous enlargement and a structural resemblance to the newly formed tissue of wound healing. It is desirable to discuss granu- lomas in conjunction with true tumors, as their origin, situation, and form frequently make them objects of interest in differential diagnosis. The fundamental difference between the granulomas and the true tumors is that the former are purposive, the purpose being to combat bacterial invaders and to repair tissue. Differentiation depends upon observing the changes in the cells which tend to effect this dual purpose. On the one hand, we must expect to find the leukocytic and phagocytic activities of acute infections; and on the other hand, the fibrin forma- tion and the polyblastic and fibroblastic activities which end in the production of new tissue. The presence of these activities implies that there is an enemy to combat, and usually we can identif}'^ the offend- ing organism; but sometimes it cannot be found. We are still obliged to class such cases with the granulomas purely on the analogy of the tissue activities. The simplest granulation process is observed in the healing of wounds, particularly when delayed b}^ disturbed nutrition or infection. The pus-producing organisms are usually responsible for the production of luxuriant granulations (proud flesh), but other bacteria, and even chemical and mechanical irritation may be the cause. The initial process is usually inflammation, in which there is a heterogeneous mixture of mononuclear and polynuclear leukocytes with large cells, probably identical with the polyblasts of Maximow. It is this char- acteristic variety of cells that differentiates young inflammatory tissue from neoplasms. In addition to the cells are newly formed vessels which are at first mere clefts in the immature fibrous tissue, but soon they develop fibrous walls. The character of the vessel walls depends upon the age of the process. Certainly, if the process has lasted long enough to warrant a question of neoplasm, the vessel walls will be sufficiently developed to furnish a very definite diagnostic point. About the vessels are SY I'll I LIS 211 numerous hands or granules of Hbnn vvliich may check the development ot the f^ranulatne process at an eari\' immature state. In such cases the mononuclear cills iuconu- iiromnunr, the- polv nuckars are sparse, and tile pol\hlasts tew or ahsent. The vessel walls are conHned to a smj^le layer of cells and are surrounded b\' Hbrin hands. In such con- ditions the resemhiance to sarcoma is very great, and to arrive at a diagnosis the specific cause of the granulation process (tubercle bacilli, spirochetes) must he sought and the life histor\- of the growth be taken into account when the specific causative agent is unknown or cannot be identified. The variation in the structure of the granulomas, despite a general common plan, is so great that it is desirable to present each type sepa- rately, since the histological structure in itself frequently presents sufficient exidence on w Inch to base a diagnosis, and also indicates the specific organism involved. They may be considered in the order of their frequenc\- of occurrence as diagnostic problems. Syphilis. — There is no disease so often confused with neoplasms as is syphilis. Tumors are no exception to the oft-quoted dictum that syphilis may simulate any disease. The syphilitic lesions likely to be mistaken for tumors are the gummas and the ulcerations resulting from them. Fig. 137 "W?P«3 ."V . Gumma of liver: a, spindle cells within degenerated area; b, small round cells. Microscopic Appearance. — The gummas are the syphilitic lesions which may be included with the granulomas (Fig. 137). The}" are formed primarily- by the invasion of a region, usuall}- about a vessel, by small 212 GRANULOMAS spherical cells with a small amount of protoplasm and a relatively large nucleus. The area invaded extends, and when the gumma is fully formed three distinct zones can be discerned. In the centre is a necrotic area composed of granular debris containing nuclei which retain their stain- ing properties. Surrounding this area are spindle cells arranged in a manner more or less tangential to the circumference. These cells are formed from the granuloma cells above mentioned, which follow in a general way the development of the granulation tissue cells into fibrous connective tissue in wound healing. Surrounding this layer, the tissue is infiltrated with unchanged granuloma cells. This region is not sharply circumscribed, but shades gradually into the tissue external to it. Fig. 138 '^■^^'^ ;^ >>'°',.^''--^' '^-C "'-i'''' "»''''-^' J""' ' ..-X"' /'" ■-°' ■■,"''■•' '-\^'" •'"''' :,i^^'.- ''• v"J-, - ^■.-J-.*^»\*i■■'C**-'^• ^ - / ' .■»»'V* ■ ''; ^~~»- -' , '• .■ •■'.'.- *- • ». « " '. .'■ Syphilitic periostitis, perivascular infiltration. Giant cells may be found in any of these regions. Their nuclei are usually diffusely scattered through the cytoplasm, but are sometimes arranged about the centre. In the peripheral layer, bloodvessels are usually to be found showing the typical changes of syphilis. The endo- thehum of the intima shows extensive proliferation. To such an extent may this proliferation extend that the lumen of the vessel may be nearly or quite occluded. The media may likewise be the seat of extensive hyperplasia. The adventitia is often infiltrated with granuloma cells. The centre of gumma is prone to liquefaction, which follows fatty degeneration. Macroscopic Appearance. — The submucous or subcutaneous gummas are nodules varying in size from a pea to a walnut. They are indefinitely circumscribed, are unattached to the skin, and are of firm consistency. As they extend they liquefy, the skin is no longer freely movable upon them and becomes deep red or bluish red. It then breaks down and an ulcer results. The ulcer bears, in general, the form of the gumma. Because the area of liquefaction is spherical, the ulcer has overhanging edges, and since the centre alone becomes softened, the border of the Tl HI.RIA I.OSIS '1\:\ iilcfi" IS liiin. As till- uK( I .i(l\ ;iiuts ilu- process ol mlili lation ;in(l li(|ii(- facrion <.()iitiiuiis. Ilu- llooi ol ilu- nicer is c<)\i-ir(i \Mth ;i \(ll<)\sish (.Ichiis, tin iisiilr ot \\\v li(|iut;Kt ion ol tin- casi-^in-d Cfiuif. \\ lull tlusc iiUiis lual sjioni ancoiislv or as a ic-sult ol t iiat iiu-nt , rlu- \i'llo\\ dihiis (lisappcais and <:,i annia t ion tissnc nplains it. I lie oNii hantini!:, i-d<;cs ap|"»i()ach the luw ii,ranniai ions and hc^in to c(ncr rluiii with a dilu-arc la\(i ol tpitlulinm. Sometimes the processes of hrahnLi, and hlc•akln^ down ait- (.■oinhnuil, so that <;\ ions hmiics are pioducitl. I he- ixsnltiiiL;; scars arc iisualK' rcmarkahic lor rluii dchcacw I hc\' all' pearly white, ami oitcn jircscnr a radiatin^i; arran < ,>^eVs>:f^ ft^*' -V^ '^-'^ Mycosis fungoides. lymph glands are sometimes involved, particularly late in the disease. The structure resembles the granulomas in a general way, there being a reticulum within which the cells lie. The cells are round, ovoid, spindle form, or branched. ^ ^ For a complete discussion see Hazen, Jour. Cutan. Dis., 191 1, xxix, 521. ^ Tilden, Boston Med. and Surg. Jour., 1885, cxiii, 386; Unna, Die Histopathologie der Hautkrankheiten, Hirschwald, Berlin, 1894. F A \l V 111 REGIONAL CONSIDERATION OF TUMORS CHAP '1' E 1^ X I X BIOLOGY OF TUMORS AND GP:NERAL CLINICAL CONSIDERATIONS General Diagnosis. — In the foregoing chapters the biology of tumors in geneniL as well as. their individual characteristics, were considered. It is a combination of these factors that make it possible to judge the kind and character of an\' particular tumor. The application of this knowledge demands the same rules as govern any chnical stud\-; but special considerations must be employed here, because of the difference which exists between the nature of tumors and the nature of other processes which are reactive in character. This fact warrants a summar\' of the facts ahead}' presented. The general biology of tumors presents many facts common to all species of tumors, since they are all proliferative in character. It also presents many points of difference in that they have different capacities for growth, dissemination, and degeneration. It is the comparison of these likes and differences that make it possible to identify- the species, that is, to make a diagnosis. In making a diagnosis in a tumor problem, one has the triple question to solve, "Is it a tumor?" if so, "What kind of a tumor r" and "Is it a primary or a secondary tumor?" The first is often the most important clinically. The second is sometimes onh" a matter of scientific interest, but usually it is the answer to this question that determines the treat- ment. The answer to the second carries with it, as a rule, the answer to the third, but sometimes a secondary tumor is regarded as primary, which may lead to error in treatment. The question as to whether a given lesion is a true tumor or not may open up the widest problems in general diagnosis and ma}- require the most diverse means of differential diagnosis. For instance, a mass within the abdomen may be due to scybala, inflammatory exudate, or to a neoplasm. Agents directed to the removal of the scybala may 220 BIOLOGY OF TUMORS AND GENERAL CLINICAL CONSIDERATIONS exclude them as a factor. Inflammatory processes have s3'mptoms of their own — pain, tenderness, rise of temperature, and characteristic changes in the leukocyte count. Each tumor likely to be confused with such conditions has individual characters that count in the diagnosis. To be able to name a tumor correctly is of little use unless the varia- tions within the class can be correctly interpreted. To identify a tumor as a carcinoma, for instance, is of little use if the kind of carcinoma, its rate of growth, duration, extent, and the state of the patient are not fully understood. In every case of tumor, therefore, the most painstaking study should be made. It is in this way only that the clinician learns to understand the slighter points of difference in tumors. The problem of diagnosis can best be studied under two heads, clinical and laboratory. Clinical Diagnosis. — The clinical diagnosis is a judgment based on all data which can be obtained from the history of the case, from the examination of the patient in general, and of the tumor m particular. In general, it should include all those methods applied to the diagnosis of anv other disease. History of Case. — The record of a patient should comprise not only data relative to the tumor, but should take into account the patient as well. Certain points are of particular importance, and the fullest information should always be sought. In all cases a record of the previous course of the tumor is of importance. Sometimes, it is true, the history can be read from the appearance of the tumor; but even the simplest problems are subject to error and every detail should always be taken into consideration. The data should be made a matter of record. The order in which information is obtained is a matter of indifference, but the following is suggested as natural and convenient: Jge of the Patient. — The age of the patient is never in itself decisive, but it is often of great confirmatory value. Certain tumors (teratomas, certain cysts, nevi) are usually present at birth or appear at an early age. Other tumors are most frequently seen in earl}^ Hfe (sarcomas), others in middle life (myomas), and others, finally, at an advanced age (carcinomas). No fixed time, of course, exists for the appearance of any particular tumor, but the general direction of inquiry is frequently indicated by the age of the patient. Furthermore, if a specific tumor is diagnosticated, the general prognosis may be influenced b^' the age of the patient; for example, carcinoma occurring at an earh' age is apt to be especially malignant. General Health. — The state of the patient's health at the time of observation, must be noted. If it is impaired, the cause must be deter- mined; for the fact that the patient has a tumor does not prove that the impaired health is due to it. Thus a patient w^th a fibroma of the CL/MC./L Pl./(;\OS/S OJ ri MORS 221 uterus nia\ have, for example, a concurrent nephritis or pernicious anemia, either one ot which would seriously affect the health indepen- denrl\ ot rhe riiinor. ( )rlu i diseases beinj^ excluded, the tumor itself may cause the decline m health m different ways — b\' its inherent nature, that is, its mahunancy; or by accidents which attend its growth; or horh hictors nuiy be combmed. riuis a Hbroma of rhe uterus ma\' cause hemorrhage and in this wa\ affect the general health. A carcinoma of the uterus ma\ pro- duce cachexia, and at the same time hemorrhage ma\' still further reduce the patient. Kven worry about a recently- discovered tumor may cause deterioration of health such as is sometimes seen when a patient regards a benign tumor of the breast as malignant. Duration. — The most important question which the histor\- must answer is. How long has the tumor lasted .' Upon this depends large!)' not only the determination of its nature, but also the prognosis. Great care must be exercised in accepting the patient's statement, for obvioush- the duration of the tumor will be dated from the time of discovery, and this is often made after some injury which leads the patient to examine the part affected. Thus, frequentl}" in adenoma of the breast an injur\" ma\' cause the patient to discover a growth of considerable size, and if the surgeon accepts the patient's statement he may assume that the tumor is of a rapidly growing variety. Even when the rate of growth has been found to be rapid, a conclusion must not be made at once that the growth is malignant. The sudden increase in size ma\' be due to some change in a benign tumor. Edema in a myoma of the uterus ma\- cause a sudden enlargement which might lead to the erroneous conclusion that a so-called sarcomatous degeneration had taken place; or a hemor- rhage into the interior of a benign tumor (lipoma, fibroma, or th\roid cyst) ma\' cause it to enlarge suddenly. Similar accidents in the course of moderately malignant tumors may lead to a mistaken diagnosis of excessive malignancw Inflammator\' infiltration about a malignant tumor may simulate rapid growth, the nature of which becomes apparent when the exudate is absorbed. On the other hand, a rapid enlarge- ment should always be regarded of grave moment until it is shown bevond doubt to be accidental. Family History. — As previously stated, it is questionable whether heredity has any considerable influence in the production of tumors; however, other diseases in which an hereditary predisposition exists may simulate tumors. This is particularly true of tuberculosis, and the .presence of a famil\" predisposition to this disease should caution us not to mistake a tuberculoma for a sarcoma, or a tuberculous ulcera- tion for a carcinoma. The mistake made in diagnosticating s\'philis 222 BIOLOGY OF TUMORS AND GENERAL CLINICAL CONSIDERATIONS as some other tumor has been on occasion so frightful as to caution us to consider this disease in the vast majority of tumor cases, and the consideration of hereditary influences is often of aid in obscure cases. A complete history of the patient's antecedents, therefore, should be recorded. Pain. — Pain in tumors may be due to the pressure upon a nerve by an expanding tumor or to the inclusion of a nerve by an infiltrating tumor. The pain varies with the location and character of the growth, which are at times factors of importance in diagnosis. Usually, pain in an infiltrating tumor appears late and indicates that a considerable degree of invasion has already taken place. Pain may also be the result of secondary factors. Infection in a tumor may give the usual pain of an inflammatory process. Accidental pain may be referred to a tumor when, in fact, no relation exists. The greatest care is necessary, there- fore, in interpreting the significance of this symptom. Injury. — The history of injury preceding a tumor may be of impor- tance in two ways; it may indicate that the tumor is inflammatory in character, or it ma}^ indicate that the tumor belongs to that class which not infrequently follows trauma, as sarcoma. On the other hand, an injury occurring in the course of development of a tumor may cause it to change its character; a previously slow growing tumor may begin to grow rapidly (thyroid, parotid); a benign growth may be caused to degenerate (myxoma, lipoma); a quiescent tumor may suddenly burst out in malignant development (papillomas, pigmented moles). A tumor may be wrongfull}^ attributed to an injury when the trauma has served merely to attract the patient's attention to a previously unrecognized condition. A patient may wilfully misrepre- sent conditions in order to make a claim for injury. The result of the actual injury may simulate a tumor; for instance, exuberant callus or extensive induration about an infection may raise the question of beginning sarcoma. Physical Examination. — The physical examination of a tumor miay be divided in the same manner as physical examination of the chest, or heart, namely, inspection, palpation, auscultation, and percussion, the last two being, however, of but occasional use. Inspection. — The first point to note is the region of the body occu- pied by the tumor. When we have determined from which tissue a tumor springs we have made the greater part of a diagnosis.^ When we can say that a tumor has grown from such and such tissue we may be certain that it follows the type of that tissue. This is true of all primary tumors. Therefore, by noting the location of the tumor it ^ Liicke, AUg. Chir. Diagnostik des Geschwiilste, Volkmann's Vortrage, No. 97. riiYsicii. i:\.iMi\,iri()\ oi- re mors '1X\ IS tutiiunil) possiMr to phut- II ;it oiuc- uiiliin ;i ri-lati\fly narrow i;i()iip. I his IS |iaiiKulail\ tiiu- ot tuiiiors arising from fetal niis- plactnunts whuh air situatrcl, of course, in the hnes of the fetal folds. This iiuliulcs tkiniouis, hiaiuhiofjenic c\'sts, and tumors derived from aberrant t^lands. ()rlur tumors while less characteristic in situation, still Ikixc- nlati\il\ dcHniti- areas of predilection. 1 his is ]')ai ricularK' true for carcinomas in certain orfi;ans lip, breast, uterus, stomach; for lipomas in the region of the neck and shoulders; and for sarcomas at the joint extremities of the long bones. Relation to the Surrounding Tissue. — Frequently, a tumor displaces anatomical parts which are easily recognized by inspection. 1 his mav give a clue to the nature of the tumor, and the point of origin as well. For instance, a tumor of the neck when growing superficially covers the sternomastoid, but when growing from the deeper tissues it pushes this muscle to one side. Similar phenomena may be observed in almost any region of the bodw Color. — The color of a tumor ma)' give knowledge of the most varied and important significance. A cystic tumor ma\' indicate its nature b\' a bluish or yellowish color, a lipoma b}' a \ellowish color, a malig- nant tumor by a reddened surface, caused b}' vascular activity'. Ihis activity may be due to invasion of the skin by an infiltrative growth, or to vessels which dilate m response to the excessive requirement for nutrition b\- the underlying tumor. A whitish glistening surface mdicates an expansile tumor of relatively rapid growth in which the skin does not partake. This is particularh' to be noted in benign c\sts. The character of vascular tumors is determined b}' the color of the surface; thus, in lymphangiomas, the surface is pale, in venous tumors bluish, and in arterial tumors red. Other tumors characterized by the formation of pigment (melanomas, xanthomas, etc.) disclose their nature by the color of their surface. Surface Conformations. — The surface of tumors which are exposed to direct examination is covered by skin or mucous membrane. Expan- sile tumors cause no change except that brought about by distention. In the infiltrating type the overlying skin becomes affected in one of three ways: (i) If the underlying tumor is made up largely of fibrous tissue, the retraction of this tissue produces a dimpling (carcinoma of the mamma); (2) newgrowths ma}' form in the skin b\' extension from the mother tumor (carcinoma en cuirasse); (3) the most common result is ulceration of the overlying surface caused by invasion of the new growth (fungating tumors). The ulcerated surface varies much in appearance according to the nature of the tumor beneath. Suppurating processes are frequently implanted upon the tumor, and then the 224 BIOLOGY OF TUMORS AND GENERAL CLINICAL CONSIDERATIONS resulting condition depends upon the character of the infecting organism plus the character of the tumor. Forms of Tumors. — The external forms of tumors vary withm wide limits, and if correctly interpreted may lead to a surmise of the nature of the neoplasm. Pedunculated tumors are usually expansile in growth, and therefore usually benign. Tumors protruding prominently from the surface with a well-marked outline are likewise usually benign, and mildly malignant tumors of slow growth often assume this form. Elevated tumors which gradually terminate in the surrounding tissue are most likely malignant. This probability is much heightened if the surface is bosselated. Tumors may present ulcerations as their most prominent feature. Ulceration is usually due to participation or invasion by the tumor, which indicates malignancy. Benign tumors, when the surface is subjected to irritation or traumatism, may ulcerate; but in these the extraneous cause of ulcer- ation is at once apparent. The habit of carefully noting the external form of tumors is valuable chiefly as a means of excluding enlargements, deformities, or diseases which may stimulate one of the general forms of tumors. Palpation. — By the sense of touch the consistency and the more intimate anatomical relation to the surrounding structures may be deter- mined. Certain precautions, however, must be observed in determining these facts. The density of the tumor may vary from soft fluctuation to stony hardness. The fluctuation of cysts may be readily deter- mined if the wall is not too thick or the tension too great. Rapidly growing cellular tumors may simulate fluctuation because of the amount of fluid in the tissue (sarcoma), or because the substance of the tumor is semifluid (lipoma, myxoma). The relation of the tumor to the surrounding parts may be detected by testing its mobility. Thus, expansile tumors permit the skin to be lifted and the tumor to be pushed about upon the underlying tissue. This is indicative of benignancy. The exceptions are those tumors which go out from the skin itself, to which they are then, of course, attached (papillomas and wens). Infiltrating tumors, on the other hand, because of their disposition to invade surrounding tissue, are fixed. Fixation, therefore, is in general a sign of malignancy. It is necessary to exclude benign tumors which have become secondarily fixed by inflammatory processes. Palpation may serve to exclude true tumors by revealing such signs as the expansion of an aneurysm, the crepitus of a collection of gas, and friction rubs about the tendons and bursae. Auscultation. — Auscultation has a very limited application in tumor diagnosis, but is useful at times in excluding conditions which are not Ki:.\i()i .11. or rissii: iok dlkaostk: I'l r/'oses 22.") luophisnis, siuh ;is ()\;iii;in cwsrs, \slmli ^i\i- icitaiii tinrion muiimiis \s lull tlu\ arc- inoxiil. Aiuui \ snis ni;i\ also In- diriitcd l)\ rlu- hiiiir. Laboratory Diagnosis. I In l;il>()!;it()r\ cxamiiKirioii ot rimiors has two ohiicts: ill I lu- uaiiiiiiL;, ot a compuluiisivi- knovvlcdf^c o{ tin- sriiu-tui(.- ami litr liis(()!\ of tumors m ^iiuial; 2) rlu- diagnosis ot rlu- rumor ar liaiul. I lu- oiu- is ruiulamt'iualiy educative in piirpprran\f rriatnunr is not indiiartil until it has lutn ck-nionstrattcl that tin- (.ntirt.- growth can he icniovt'cl. This recjuires an examination of the surroundni^ tissue and of other locations where metastases may form; for instance, the l\niph nt)dcs in carcinoma and distant parenchymatous organs (li\"er, lungs, etc.) in sarcoma. These general principles are modiHed h\ iiian\' factors, which differ not only for each tumor group, hut tor each variety and for the region which they occupy. Benign tumors can usually be shelled out ot their capsule 'lipomas, fibromas, etc.), or their base easily circumscribed b\- the incision (papil- loma). The old delusion as to the fre(|uenc\' of degeneration of benign tumors has given rise to much needless mutilation "in order to be on the safe side," and is an expression of an uncertain comprehension ol the character of the tumor. When the nature of the tumor is exactly determined, the technical requirements can be clearly defined. Often a clear determination of the nature of a tumor is difficult, but the dispo- sition of surgeons to justify a needlessly radical operation under this cloak has done much to retard the clinical side of oncological science. The rules for operation in the different biological t\pes ot tumors may be stated in a broad wa\', which can be narrowed only by a specific knowledge of the concrete problem at hand. The indications for the removal of benign tumors are: (i) For cosmetic reasons; (2) when important structures are .indirectly involved by pressure; (3) when secondar\' degeneration is present or impending (suppuration, necrosis, etc.)- The operative indications for malignant tumors are: (i) The removal of the tumor and its infiltrating borders; (2) the accessible areas of possible metastases. The indications present for benign tumors all obtain in the malignant, but in the latter they are lost in the presence of the one great problem of the removal of the limitless growth of the malignant cells. The extent of tissue to be removed must be decided at the operating table; the decision demands a knowledge of the kind and life history of the tumor. The recognition of the probable situa- tion of metastasis is a problem in anatomy. Presentation of the details of this general plan must be deferred to the section on the different t\pes of tumors when situated in particular regions. Palliative Treatment. — When the operative cure ot a tumor is no longer possible, a number of procedures are available for the relief ot certain symptoms. The removal of the tumor may be justified to save the patient pain or to limit an infection, though inaccessible metas- tases ma}' be present. Channels closed b}' the invasion ot the tumor 230 BIOLOGY OF TUMORS AND GENERAL CLINICAL CONSIDERATIONS may be relieved by artificial openings beyond the point of obstruction (tracheotomy for carcinoma of the larynx, gastrostomy for tumors of the esophagus, colostomy for tumors of the rectum, etc.)- When no form of operative treatment is possible, the resources of the surgeon are still not exhausted. Various methods may be employed for the rehef of the patient's mind and body. Injections of alcohol may accom- plish both results. The x-ray treatment is of use in allaying pain and in many cases in inhibiting the growth of the tumor. When nothing else can be done the suffering of the patient may be relieved by morphine. c II A 1* 1 1-: k x x TUMORS ov riii-: cranium 'I'ni'; c-iaiiial rc{;ion may be calKtl tin- oiicoIo^k al imiscum ot the hoch , since nearly every known tumor ma\ he encoimteied here either as a primar\- «i;r()\vth oi" as a metastasis of a growth in a distant or^an. The Hfe hist()r\- of the neophisms which occur in this region is better known than that of those of other regions because of its exposed posi- tion. The early detection of tlieir presence hy the patient leads him to seek advice early, and he is usually able to give a fairl\- accurate account of the growth. Notwithstanding these favorable factors, the results of the treatment of malignant tumors of the head have not been as good as the conditions would seem to warrant. Often for cos- metic reasons the radical treatment indicated is not carried out with the same freedom as in less exposed parts of the bod\-. Timidity in technique is frequently due to an uncertain diagnosis in the early stages of the disease. The suspicion of the operator that he may be dealing with a benign lesion leads to a more conservative procedure than would be employed if the malignant character of the neoplasm were full\- appreciated. Wens (Atheromas) (Fig*. 143). — These most frequently affect women past middle age, and are very rare in the developmental period of lite. The scalp is the most susceptible area, nearly one-half of all wens occur- ring in this region. The\' are, in fact, the most frequent tumor of the scalp. This frequency is due to the great number of sebaceous glands and to the unsanitary condition of many scalps. Wens are formed by a collection of sebaceous material and desqua- mated epithelium within the sebaceous glands, due to occlusion of the ducts. They vary from the size of a pea to that of an orange or even larger, but are usually not larger than a walnut. The\- are usually globular in form but ma\- be bosselated when several develop in close proximit\- and coalesce; when firm constriction bands pass over a single tumor, it may be pseudolobulated. When small the\- are hard, and as the^• grow they become soft and pultaceous. The skin covering them is usually smooth and shin)-, and ma)- be bluish from venous conges- tion. Hair is generalh- absent on account of atroph\- of the follicles caused by stretching of the skin. Wens are often peculiarl\- striking and repulsive deformities. 1 hey 232 TUMORS OF THE CRANIUM are intimately attached to the skin, and because of the bony base upon which they rest tend to project outward, and in some cases may even become pedunculated. The}^ sometimes extend into the subcutaneous tissue, but never extend below the pericranium. In rare instances they become detached from the skin. They frequently become inflamed, and suppuration with perforation of the skin may follow. After dis- charging for a time the opening closes and they usually refill, only to repeat the process. The suppuration may, however, by destroying the lining epithelium, lead to a spontaneous cure. On the other hand, a fistula may form, or occasionally serious suppuration may supervene by extension into the loose surrounding connective tissue. Prolonged irritation sometimes results in the formation of epithelioma. Fig. 143 Sebaceous cysts. Diagnosis. — Wens must be distinguished from dermoids, meningoceles, and serous cysts. From the dermoids the}" are differentiated by their intimate attachment to the skin, by their freedom from attachment to the pericranium, by their location elsewhere than over the fetal fissures, and by the fact that dermoids are frequently first observed during adolescence, while wens are growths of later periods of life. From meningoceles they are distinguished by their freedom from attachment r ii'ii.i.oM.is 233 to tin jHiiciiinium, ;incl h\ absence of pulsation and of" compressibility. 1 he occasional association of wciis with nuninjicjccles should not be foriiottcn. Treatment. Sinct- they arc caused b\' excessive secretion of the hning cells, the cyst wall must be removed. I his can be done most conveniently under cocaine or quinine anesthesia, by injecting the fluid freely about and beneath the tumor. In this way the surrounding connective tissue becomes edematous and iHiniits an easier enucleation. If the tumor is small, the incision is made through the overlying skin down to but not into the sac, which is then carefull\' dissected out. In moderate- sized and larger tumors an elliptical mcision should be made following the margin of the wen where the skin is less firniK' attached, and the ic. 144 Fibrous papilloma ot the scalp. sac ma\ then be easily dissected awa\-. Hemorrhage may be con- trolled by pressure or b}' ligatures of fine catgut. A few sutures in the edge of the wound complete the operation. Because of the greater danger of infection, v. Bergmann preferred to omit sutures entirely. If the tumor is alread}' the site of suppuration, the sac must be removed piecemeal, which is often difficult to do completely. In this case, at least, sutures should be omitted. With the removal of the sac a definite cure is obtained. The scalp is particularly subject to infection, and therefore careful asepsis is urged in this otherwise trifling operation. Papillomas. — Papillomas of the scalp form small tumors, from the size of a pea to that of a hazelnut. The}- are sharph' circumscribed, sometimes slightly constricted at the base (Fig. 144). Their surface 234 TUMORS OF THE CRAXIUM is roughened by numerous filiform projections. They are intimately united with the skin and move with it. Histologically they are com- posed of fibrous tissue, with a thick overh'ing layer of epithelium. They are essentially epithelial warts, and occasionall}^ form large dense pro- jections on the scalp. They are pigmented with yellowish or brown shades, and often on the temples are almost black. Fibrous papillomas are less common. They are covered with a thin layer of epithehum and are composed of fibrous tissue containing masses of spherical (em- bryonalj cells. These, particularly the pigmented varieties, frequently give rise to melanosarcomas (Fig. 145). They ma}' be confused with Fig. 145 ^-x Melanotic papilloma of the scalp, with metastasis in the lymph glands and neck. endotheliomas, but can be distinguished from them by their papillo- matous surface. These tumors usually cause disturbance only b}' interfering with the ordinary care of the scalp, but are sometimes the starting pomt of epitheliomas. Treatment. — They are best treated by excision. An elliptical incision is made through the entire thickness of the skin because of the deep attachment of the growth. The defect is then closed with sutures. Elephantiasis Nervorum.^ — The scalp is considered to be the seat of this lesion in about one-third of the cases, the temporal region being most commonly affected (Fig. 146;. Most of the cases occur in young 1 Helmholtz and Gushing, Amer. Jour. Med. Sci., 1906, cxxxii, 355. r.l.l:l'll.l\ri.l S IS \ I. Kl OKI M 28.-) adults, hiir tlu- disease is soiiutiims conjit-mral. 1 liere is often a family pieiiisposition.' I'he affecfion usually begins as a tliickeniiifj; (A the skin, Nvlu-nce rlu- term "pachydermatocele" was applied to it hy Mott.- As tile ilisease progresses the skin saj!;s over the side of the head. If in the teiiipoial region, it tc luls to displace the ear, or if" m the frontal refiion, ma\' even cover the face. When the pendent mass is lifted b}' the hand the displaced organs are restored to their normal situations. Mott Fi<;. I \t'< Elephantiasis nervorum of the temporal retiion. A large mass had been removed from the temple before the photograph was taken. described this mass as being to the touch like the dependent breast, the thickened and nodulated nerves suggesting the lobules of the gland. The characters mentioned above will hardly permit this condition to be mis- taken for any other. A hmphangioma ma}' present a slight resemblance but the disposition to sag is absent. ^ Feindel, quoted by Helmholtz and Gushing. -Med. Chir. Trans., 1854, xxxvii, 156. 236 TUMORS OF THE CRANIUM Treatment. — The only treatment is excision. This may be attended by some difficulty because of hemorrhage. Infection is prone to take place in the loose tissue. When the disease is extensive, several sittings may be necessary for its removal.^ Lipomas. — About 2 per cent, of all lipomas occur in the scalp.- They occur most frequently on the forehead (Fig. 147), 37 out of 54 cases according to Chipault.^ They form more or less lobulated tumors, except when in the temporal region. Here they are situated below the aponeurosis of the temporal muscle and are diffuse. Fig. 147 Lipoma of the loiehead. They vary in size from a hazelnut to a hst, and occasionally even form enormous tumors which hang over the face and shoulders. One case is reported in which the tumor reached from the forehead to the knees. They are not attached to the skin, but are frequently fixed at 1 Billroth, Arch. f. klin. Chir., 1869, xi, 232. 2 Grosch, Deutsch. Ztschr. f. Chir , 1887, xxvi, 307. ^ Traivaux de Neurologic chirurgicale, Paris, 1895. DERMOIDS 237 ilu- l>;isi . In ilif ()fii|)it;il legion tins fixation nsiilis tioni pioloii^a- rions winch tlu\ smd out ainoiii; tin- inusilcs ol tin- tuck. Over the sc;ilp, as \. l>iiL;inaini |)oniti(l out, tlu\ ha\c a vei\' intimate relation with the ptiiostcum. I'lhlciscn ic^aidcd thcif periosteal attachment as t\ pical ol li|)onias. ( )ii ilu' loiehead rlu\ lie in the f'at-fiee con- lucriM- tissue hiiuath tiu muscle. I'luie is tVeciuentU- an inhltration into the peiicianiuin ahoui their periiiluiw which makes it appeal" as thoutih the tumor la\ in a (.Icpiession in the hone. Diagnosis. -Lipomas differ from wens in that the\ aie unattached to the skin and ha\e intimate connection with the peiiosteum. 'I'he\' ditter from dermoids in heing located elsewhere than at the sites of fetal clefts. 1 hey are also, as a rule, tumors of advanced \'ears, while the dennouls appeal" in early life, although .Monnie has collected a number of cases of congenital lipoma. He emphasized particular!)- the fact that the\' may be associated with meningoceles and may obscure them, especially when located over cranial fissures. This association probably forms the basis for this author's statement that lipomas ma\- be connected with the dura. The infiltration of the pericranium about their base, as mentioned above, sometimes simulates closely the osseous depressions of dermoids. The subfascial (temporal) lipomas have been mistaken for cold abscesses and also for syphilitic processes. Treatment. — Excision is the only treatment. Their removal is usually a relatively simple matter unless lobulations extend into the surround- ing muscles. In such cases the operation ma)- involve some difficult\-. Their lobulation, their frequent firm attachment to the periosteum, and the possibility of th"eir association with intracranial lesions make them little suited for removal under local anesthesia. Dermoids. — These tumors are situated in the natural lines of junction of the developing cranium, and are common in the temples, orbit, root of the nose, more rarely over the mastoid. They are sometimes observed over the fontanelles, particularly over the great, less often over the glabellar and occipital. Confusion is sometimes caused by their presence at some distance from the natural location of these fetal clefts. Von Bergmann^ explains this discrepancy b)' assuming that they become early attached to the bones, and as the bones develop the tumors are carried with them. They form smooth hemispherical elevations over which the skin is freel)' movable. They are not movable upon their base because of the intimate association with the periosteum. Quite regularl\- the\- are received in a depression of the bone. The depression is not due to atrophy from pressure, but to failure of the bone to develop, since the ' Handb. d. Pracr. Chir., Band i, Enke, Stuttgart. 1900. 238 TUMORS OF THE CRANIUM tumors are present before ossification takes place. The bone ma\' be entirely absent, so that a portion of the tumor ma}' project into the cranial cavit3\ Dermoids ma}" escape notice during the earliest period of life, but the}" are frequently seen in the second and third }"ears, and are particularly likely to develop rapidh" about the time of pubert}\ They contain a whitish fatty fluid composed of fattil}" degenerated epithelium, cholesterm crystals, the secretion of fat and sweat glands, and frequently hair. Sometimes, however, the contents ma}" be serous, due to the fact that the solid elements have become precipitated on the c}^st wall; the cyst contents ma}' become hemorrhagic from traumatic rupture of the bloodvessels.^ Diagnosis. — Dermoids must be differentiated from wens and meningo- celes. From the former they ma}" be distinguished by their non-attach- ment to the skm, b}" their deep attachment to the periosteum, and^ above all, by" their location over the site of the fetal clefts. Tumors which occur before the fifteenth }"ear are dermoids and not sebaceous cysts. 2 Wens project from the surface as spherical tumors, while der- moids appear as hemispherical tumors because their bases are flattened and are set in the periosteum or bone. Dermoids are always single^ while wens are frequently multiple. From meningoceles the}' ma}' be differentiated by their failure to disappear or to cause s}'mptoms of increased intracranial tension when pressed upon. Dermoids when projecting into the skull transmit pulsa- tion from the brain, and aspiration may be necessary for differentiation. It must be remembered that the mere demonstration of the presence of a clear fluid is not suflBcient to exclude dermoid, since a separation of the fluid and solid elements may take place, as has been stated alread}", particularly in dermoids which have remained stationar}" for a long time. Age is important as a diagnostic factor, inasmuch as encephaloceles usuall}" result fatally before dermoids ordinarily become objects of clinical observation. The location of the tumor is of great importance in the diagnosis. Dermoids are very rare in the occipital region, while encephaloceles are particularly common in this region. Encephalo- celes, except in the occipital region, are never encountered in the mid- line of the cranium.^ When the anatomy of this region is remembered, this can occasion no surprise. In order for brain substance or meningo- celes to appear at this point, a failure of development in the falx cerebri would be necessar}". A failure of development of this magnitude occur- ^ Fehleisen, Deutsch. Ztschr. f. Chir., 1881, xiv, 5. ^ Chiari, Ztschr. f. Heilk., 1891, xii, 189. ^ Rathlef, Zur Casuistik und Diagnostik der epicraniellen Dermoidcysten, Dorpat, 1876. .fXCIOM.IS 2.\U v\n^ so t;irl\ in cl(.\ rl()|>nunr would l)f artt-ndcd by cerebral anomalies which could 1)1- (.asiU ditii niiiicd it", indeed, they were compatible with life. Treatment. Removal is usually mdicatc-d tor cosmetic reasons. This m:i\- freeiuenrl)- be accomplished under local anesthesia, but their inrimate attachment to the periosteum sometimes makes the use of a general anesthetic desirable. This is particularly true where a por- tion of the tumor extends into the cranial cavit\' and the use of a peri- osteal elevator or bone-cutting instrument is required for its removal. When the tumor is in direct contact with the meninges there is a line of cleavage which permits an easy separation. In all instances, there- fore, where an exact differential diagnosis cannot be made between a dermoid and a meningocele, a general anesthetic should be empk^ed, and precautions usuall\' observed in brain operations should be taken.' Angiomas. — Simple angiomas are most frequently seen about the temples and forehead, but the}' occur also about the occiput or at the root of the nose. They are usually congenital or appear soon after birth. The cavernous angiomas are likel\' to occur at the occiput or on the forehead. Their rate of growth corresponds to the rate of growth of the child. Sometimes they disappear spontaneously, but rapid devel- opment frequently takes place, particularh' in the cavernous type. Fatal hemorrhage has occurred from their accidental rupture or from incision made under a mistaken diagnosis. Cirsoid aneurysms (Fig. 148J are found most frequently upon the temples. They may become implanted upon nevi or ma\- result from trauma. Thev may cause .erosion of the underl}ing bone or ulceration of the skin covering them. If the}- are in connection with vessels of considerable size pulsation may be perceptible to the examining linger and the patient may complain of the constant whirring noise produced by the blood current. Treatment. — Simple angiomas ma\' be excised when small or ma\' be treated by electrol}'sis. The latter is preferable in regions where scarring w^ould be objectionable. Carbon dioxide snow may also be useful. Cavernous tumors of the scalp ma}' be excised when their size per- mits. Mass ligation may be useful if the base can be easily circum- scribed. Electrolysis is tedious and is seldom justified. Peccioni, however, recommends it in all cases because it may cure; if it does not cure, subsequent operation is made safer. If excision is practised the incision should be at least one centimeter from the base. A narrower margin might endanger large sinuses and give rise to dangerous hemor- rhage on account of the more deepl}^ situated sinuses not showing ^ Fenger, Clinical Lecture, November, 1893. 240 TUMORS OF THE CRANIUM through the skin. In angiomas of very young children it is preferable to defer operation until the child is four or five years old, provided the tumor does not grow rapidly. However, should this happen, opera- tion need not be delayed. In very young children, transfixation and ligation is preferable when possible because of the lessened danger of hemorrhage. Bryant has used injections of boiling water success- fully. Babes recommends surrounding the tumor with a chain stitch, the sutures passing deep enough to include all the vessels. The tumor is then excised and the wound accurately sutured, after which the hemostatic sutures are removed. Acupuncture needles have been used for this purpose. Fig. I. Cirsoid aneurysm of the scalp. In cirsoid aneurysms ligation of as many vessels as possible at a safe distance from the tumor is often curative; and in other cases excision may be made possible by this preliminary step. Bryant has employed a method in aneurysms on the forehead that has much to commend it. The shaved head is incised from ear to ear, as in postmortem section. The anterior portion of the scalp is then deflected over the face, and as many vessels as possible are ligated from the under side. Rubber tissue is placed over the vault of the cranium and the scalp replaced over it. In a few days it is again deflected and the tumor dissected from the under side of the skin. This permits C.IkClXOMJ 241 the rciii()\al of the tiinior wirlioiir the pei {omtioii of the skin and a scar is thus a\'<)ulecl. Carcinoma. L'aieinonias aie nuieh less eotiinion in the hair\' scalp than on the face. In <')75 epitheliomas obser\eci in w Herrmann's clinic, on!)' 36, or 5.3 per cent., were on the scalp. 1 hese j^row ths may in general be di\iclecl into two groups: (l) Ihose arising from \sarts, which are ver\ liable to undergo retrograde changes and give rise to fungiform ulcerated masses; these occur particularly in the occipital l-ic. 149 Superficial epithelioma of the scalp. region. (2) The flat ulcerating type (Fig. 149), which arise from pre- existent lesions, such as tuberculous scars and old seborrheic patches, and form superficial spreading ulcers. Both groups have the hard infiltrating border, characteristic of carcinoma, with the small white comedo-like plugs. These, like the epitheliomas of the face, frequenth' e.xist for many years without involving the underlying tissues or form- ing metastases. Both tA'pes tend to grow into the bone, sometimes even involving the dura and the brain substance itself. Rareh- metas- tatic carcinomas of the diploe have been observed. 16 242 TUMORS OF THE CRANIUM Diagnosis. — The ulcerative type must be differentiated from syphilitic and tuberculous ulcerations. From the former they may be distin- guished by their slow growth, by the fact that they are usually solitary, and by their infiltrating borders with cancer plugs. They are not influenced by treatment with potassium iodide. From tuberculous ulceration they may be distinguished by their more deepl}^ infiltrated border and by the presence of cancer plugs. In rare instances micro- scopic examination of an excised piece is necessary for positive diagnosis. The nodular type must be differentiated from sarcoma and endothe- lioma. From the former they may be distinguished by the increased disposition to ulcerate and lessened disposition to early metastasis. From the latter by their much more rapid growth and the greater dis- position to form metastases. Treatment. — If treated early before the pericranium is involved simple excision of the soft parts wide of the tumor is sufficient. Frequently, however, the bone is involved when the patient is first seen. In such cases the entire thickness of the skull should be removed with chisel and rongeur, not the outer table only, as formerly recommended. In cases where the dura and even the brain are involved, wide excision has been followed by relief for a considerable time.^ Osteoplastic cover- ing of the cranial defects so made need not be undertaken. The cervical lymph glands should be removed at the primary operation. This has generally been omitted in the past, and some operators still omit it. After metastasis has taken place, however, there can be no difference of opinion; they must be completely removed. When lymphatic metas- tasis is present the case was formerly regarded as hopeless, but with the modern method of block dissection, further effort is justified. Endotheliomas (Fig. 150). — These tumors occur as slow growing, papillary, nodular, or ulcerated masses. When small they may be slightly constricted at the base and covered with epidermis somewhat thinner than normal. As they increase in size they tend to become nodular, and when they have attained the size of a split walnut they usually ulcerate. They tend to heal, only to break down again. As the tumor increases in size the ulcerative process becomes more general until the entire surface is affected. The growth acquires an irregular granular base and an elevated, undulating, slightly constricted border. The color is redder than the surrounding skin and is mottled with dark blue patches. The normal skin forms a wall about the base of the tumor and terminates at the edge of the ulcer in a thin, sharply defined margin of epidermis resembling very much the division of normal skin from the vascular area in a spina bifida. ^ Braun, Verhandl. d. deutsch. Gesellsch. f. Chir., 1892, xxi, 439. EMXJTIU.I.KJM.IS 24 :i I lie tumors are slow of growth, and usiiall\- years elapse before tiie ulcerative staj^e is reaelu-d. Seeoiulary nodules occasionalh' form in Fig. 150 Endothelioma of the scalp. c fj ■;>-:c>^ ^■ r—j^-:o Benign qstic epithelioma: a, epidermis; b, tumor mass; c, cysts within the tumor. 244 TUMORS OF THE CRANIUM the region of the primary tumor, but there is httle tendency to the formation of metastases. However, endotheliomas have been known to undergo sudden exacerbations in growth, accompanied by rapid invasion of the surrounding tissue and the formation of metastases. Tumors identical with these in external form and clinical course are sometimes formed of cystic epithelial spaces (Fig. 151). Others, again, approach a sarcomatous structure. Diagnosis. — Their slow growth and their ulcerated, thinned epidermal covering are usually sufficient to distinguish them. The ulceration lacks the dirty incrustation of epithelioma, and the firm constricted base differs markedly from carcinoma, which is more fungoid, dense, and fades gradually into the surrounding tissue. Tuberculous processes are less elevated and the outlines are less regular. Syphilis and sarcoma are more rapid processes. On cross-section the structure resembles carcinoma very closely, and careful microscopic study is sometimes required to differentiate them. Their clinical features furnish a safer guide to diagnosis than does the gross inspection of the cut surface. This latter point should be remembered, for no doubt many of these tumors are classed as carcinoma. Treatment. — Simple excision results in a cure. If sufficient tissue is not removed local recurrence may take place, but persistence in the removal of the local recurrences will result in a permanent cure. When the period of rapid growth has been reached the prognosis is much less favorable, and even wide excision is usually followed by rapid recur- rence and metastases. Sarcoma. — Sarcomas of the cranium may be divided into two groups — those beginning in the soft parts, and those originating in the skull proper. Sarcoma of the soft parts may be derived from the skin, from the fascia, or from scars. Those from the skin frequently develop from pigmented moles which have lain dormant, and which because of some irritation begin to proliferate. They often form early lymphatic and visceral metastases. The papillary tumors described by older writers as sarcomas, and as characterized by slow growth and tendency to recur, I have found to correspond to the endotheliomas as described by Mulert.^ The fascial sarcomas form lobulated, bosselated tumors, which rapidly invade the surrounding parts and tend to form visceral metastases, but in a less degree than the pigmented type. Their growth is frequently so rapid and the invasion so extensive that the exact point of origin is difficult to determine. The sarcomas of the cranium may go out from either surface, peri- 1 Arch. f. klin. Chir., 1897, liv, 658. syr/ii/.is -M.) osteal (»i (.lural, oi iVom tin- iliplor in\ iloj^fnic). Vhty usiiall)' torni ht-rnisplurical masses, winch may appear on rlu- siirtace ot the skull or ma\' project into the cranial cavity and manifest themselves b\' svmptoms of hiain pressure. I In- location of these tumors is usuall\' on tin- lateral pait of the skull oi ahout the mastoid process. Ot the timiors arising in the diploe, Neufeld made the observation that the\' extend as far into the cranial cavit\- as the\' project outside the surface of the skull. .Accordinii to I lei/ini:ifid with int rcurial oinrnuiu and dressfd with on i i lap|Mn in many cilliilar tumors j^lia cells nia\ \->v discovered in some portion of the tumor, which lends weight to the belief that the sarcoma cells are dewloped fiom the ^lia cells. Nevertheless, some of these tumors show no tiace of \i}\:i tissue ( 1' i^. 159), and tiuir development from gha tissue is merely assumption. ! he (juestion could apply onl\' to tumors which de\eloped from j;lia tissue. (Jther distinctly sarcoma- Vir.. 161 ^^?^-^^~ "•i-'-o •-• -"'=■ - 'V \ S>^<. %.t2^(€''-'" '^>■^-•■--.§^^■a'^- Endothelioma of meninges. tous tumors springing from the meninges at least cannot have this origin. The rate of growth of these tumors aside from their structure would seem to exclude them w-ith certainty from the endotheliomas. Endotheliomas.— The usual tumors arising in the meninges are endo- thelial in origin. They are dense circumscribed tumors (Figs. 160 and 161) of slow growth. The\' are usually t3"pical in structure 260 TUMORS OF THE BRAIN AND SPINAL CORD (Fig. 162), and may develop C3^sts in their interior (Fig. 163). Endo- theliomas containing psammoma bodies are not rare (Fig. 164). Fig. 163 ^■^Pl ^^■^H ^^^^^^■■1 fv>^! ^^^^^^^H IHii^^^^^^^^^l ,y ' ' ' ■ '^^" ^^m '^ %» ^^H ^B «f ^^1 'A';, ^ ;.i ^A ^ ' ' ^1 ^^^x ^ * " 4 ^H ^m i^^ ^^1 ^B 7 ',^ - « at' ^^^^^^H .^ /■ ^^^^H ^m ■'■ <• • i" 'A t- ^^^H ^^^■k '-'- '^^^ - 'fl^^^^B ^^^^^A^ "* ' ..^ ^^^ ^^.^^^^^^B Endothelioma of the temporal lobe with multiple cyst formation. Fig. 164 ,'.^ -? -' <; 5P0 Endothelioma of the dura, containing psammoma bodies. Carcinomas. — These occur as metastatic tumors (Fig. 165), and when there is evidence of brain tumor the possibility of a primary tumor elsewhere should be kept in mind. riMfjks or the hk.iis 2r,i Granulomas. I his ^loiip is fouiul usuall\' at the base aiul may be sii^m-sttcl h\ tvicleiui' ot disease- clscwluic. I uIhicuIosis is particu- lail\ siiji;/.\,/l cord 2(;o may be vertebral, external to flu- iiunin{i;es, or meningeal, inrradural, and inrraniediillary. All except the last produce injiir\' In pressure. 1 hn r\ |Hr cent, are medullary, 40 per cent, nieninj^eal but intradural, and less than 20 pel cent, are ineinn^eal and extradural. 1 unior.s of the vertebra, nearl\ always nutaslafic, are twice as common as all other forms. Fic. 16S \\^ Fibrom\-xoma of the dorsal cord. (Redrawn from a sketch b\- the author; Plummer-Grinker's case.) S3miptoms. — The intramedullary variet}' presents s\mptoms of a cord lesion, often those of syringomyelia. The extra- and intradural give signs of cord compression, pain and sensory disturbances, and, less constantly, motor disturbance. The pain is intense, and is referred to the distribution of the nerves which are compressed; it is often 266 TUMORS OF THE BRAIN AND SPINAL CORD constant, and ceases only with the destruction of the nerve. A zone of hyperesthesia indicates the height of the lesion. Extradural. — These tumors may begin from any tissue of the spinal column external to the dura or from the external surface of the dura itself. The most frequent extradural tumors are, as above indicated, metastatic carcinomas and sarcomas either within the bone or in the soft tissues surrounding the dura. Tumors originating primarily in the extradural space are usually benign. Cartilaginous or osseous tumors are rare; fibromas, myxofibromas, and Hpomas are more common. Lymphoid tumors and primary malignant tumors have been noted. Intradural. — Tumors which arise within the dura and external to the cord are of the same varieties as the primary extradural tumors; the most frequent being fibromas or myxofibromas (Fig. i68). Other forms are more rarely seen. Tuberculomas and gummas are of relatively frequent occurrence. Neuromas are sometimes observed about the nerve roots. Carcinoma, whether arising primarily or by extension, is rare in this situation. Fig. 169 Glioma of the cervical cord. Intramedullary. — These occur usually about the central canal, but occasionally from the substance of the cord or from the pia. Gliomas (Fig. 169) are the most frequent of the true tumors. Fibromas and myxofibromas and sarcomas in various combinations with connective tissue occur. Tuberculous and syphilitic processes are likewise found. Metastatic tumors within the cord are pathological curiosities. Diagnosis. — The presence of a localized cord lesion with severe pain, when other cord lesions can be excluded, furnishes the main point in diagnosis. The kind of tumor is more difficult to determine. Gliomas often give signs of syringomyelia. The nature of the secondary tumors may be surmised when primary tumors are known. Mammary tumors are particularly apt to form vertebral metastases. Treatment. — Fibromas and exostoses are capable of removal. Glio- mas are inoperable. C II A 1' T J', k \ X I I TUMORS OK rilK ORHIT' AM) I.ACm'.MAI. CI.ANDS Fig. 170 I iiiiiois ot tlu' oiim an- oiclinanl)- ciiNiclccl into th(jse which arise within the- oihir and rhosc which arise without and invade the (jrbit secondarilw The chnical manifestations of a tumor, however, vary more with its nature than with its point of origin, and growths of the same kind ma\' arise either within or without the orhit. It seems more satisfactor\', therefore, to disregard the point of orgin of these tumors and classif)" them pathologically. Sarcoma. — Sarcomas are the most malignant tumors of the e\'e, and are especially frequent in young persons. The\' are of all cellular types, and ma}' develop from the episcleral tissue, Tenon's capsule, and the periosteum of the orbit. Higher types of sarcoma have been reported, such as chondrosar- coma, myxosarcoma, and frequently fibrosarcoma. The disposition of some authors to classify tumors of the last-named group as endotheliomas is warranted both from their structure and their clinical course. Melanotic sarcomas are some- times primary (Fig. 170) in the orbit, arising independent of pigment-bearing areas. The more common type is the melanotic gliosarcoma which begins in the pigmented cells of the retina, where It can be seen with the ophthalmoscope as a nodule bulging into the cavit)'. Later, the pigmented mass presents itself m the anterior chamber and may perforate the cornea. More often it first perforates the eyeball farther back and fills the orbit (Fig. 171). The entire e3'eball may be destroyed, and the orbit then presents a pigmented ulcerating surface (Fig. 172). The opposite eye may later become affected. Instead of destroying the e\eball, extension ma}' take place in other directions, notabl}" into the tissue of the temple (Fig. 173). Lymphosarcoma. — These tumors are sometimes reported. They develop usuall}' in the region of the lacrymal gland, and are not infre- quently symmetrical. Melanoma in the or- bital fatty tissue above the eve. ^ For the literature of tumors of the orbit see J. H. Parsons, The Pathology of the Eye, Putnam, New York, 1905, and Lagrange, Tumeurs de I'Orbite, Steinheil, Paris, 1904. 268 TUMORS OF THE ORBIT AND LACRYMAL GLANDS Fig. 171 Gliosarcoma of the eyeball, which has perforated the ball and fills the orbit. Fig. 172 Gliosarcoma of the right orbit, secondary to the left eye. c.ikciMj.M.i or rill: okhit 2()9 Myosarcoma liiniois lepoited as rhabdomyomas are probably embryomas, alrhoimb such nniiors mijihr bt- dcriNecl from the ocular muscles. ( )iu i;isi' ot Uumn om;i has bit ii described. iMidothehomas ma\ reseinbli' Icioiin onias, and (hfKi tut lal stains ma\ be necessar\' bifoif the iiuestioii can be decided. I he oibiral sarcomas tend to early metastasis, and \\h(n uniovcd recuiieiice iisuall\ occurs in one or two years. Symptoms. The e\ e is displaced directly forward d the tumor is situatetl within the muscle cone ( Fi^. 171), or in the opposite dirt^ction frt)m the location of the tumor if the latter is situated elsewhere in the orbit ( Fi- between the orbit (Fig. 1.S2) aiul the superior labial fold (Figs. 1S3 and 184) extending backward to the ear (Fig. 185), and is occasionall)- seen on the temples ;iiui thr niik. Musi- tumors arc- characreri/.ed by their slow growth, which caused tin ni tormerlv to be classed apart from carcinomas under the term cancroid. Thev begin most fre(}uentl\' as superficial scaling lesions (Fig. 186), which resemble closel\' and are perhaps identical with seborrheic patches. The scale after a time is replaced b\- a scab; that is to say, the first covering is made up of desquamated epithelium ^. Basal-celled epithtlioma of tlu- ala of the nost. which has undergone imperfect involution, while the subsequent cover- mg (the scabj is made up of serum which has escaped and coagulated, being a consequence, therefore, of the loss of the epidermis. The removal of the scab is attended by more extensive bleeding than occurs in the scaling stages. These two conditions are important to bear in mind, for the\- represent the line between the benign and the malignant. Man}' Aears are frequently- consumed in the transition. Small warts (Fig. 187) may precede the development of facial carci- noma. In these instances there is an increasing hardening about the Fig. I Epithelioma of the bridge of the nose. Fig. 185 Basal-celled epithelioma of the ear. ]■{('.. |K^ J ^^ Earl\- basal-celled epitlielionia of rlie temple. Fig. 187 Papilloma of the cheek which has remained unchanged. Behind it is a well-defined carcinoma which developed from a similar wart. 280 TUMORS OF THE FACE base of the wart, which continues until the physical characters of carci- noma, namely, induration and cancer plugs, are to be made out. Years may be required for the development, but usually the time is shorter than in the preceding type and the disease runs a correspondmgh' more rapid course. Each of these varieties extends gradually in extent, with but little tendenc}^ to invade the depth. Usually the surface of the Fig. i88 Epithelioma of the face covered with a dense brown scab hiding the granular surface beneath. ulcer is covered by a thick brownish-gray crust (Fig. i88) which may quite hide the ulcer. When this is removed a bleeding granular surface remains surrounded by a slighth^ raised border (Fig. 189) which is dense on palpation. The deeper tissues are usually not affected, but deep ulceration may result (Figs. 190, 191, and 192). The ulcers are usually irregularl}^ circular, but may assume any form in the later stages. The area occupied by them may after many years come to be as large sL rr.Riici.ii. T) />/■: oi n mors m rui- rici: 281 :is tin- IkiiuI ..i hnger (Fi^. 193J. Occasioiuilh , iiisicd of ,iir( r lomia- tiDii, |-»i()ni.iinu minors ni:i\- form ( l'"i^. \'-)\). Epithelioma of the face, showing a clean granular base with a well-defined border. Fig. 190 Basal-ccllcd cpirhclioma of the face, forniinu a deep ulct Fig. 191 Deep epithelial ulceration from a basal-celled epithelioma of the lower lid. Fig. 192 Deep ulceration of the face from a basal-celled epithelioma. Diii.r '/)/'/■: fji TtMuks or the i'.ice 2K\ .\Krast;isis ilois nor occur, if ;it all, until a considerable surface has been cltsri()\ td. iiu- pcripluiai (.xrtnsion ni one direction is some- times accompanieil b\ cicarn/.ation at another part of the border 'Fig 195). riu- ulcer nia\ actually heal o\cr ni |>laces at the border or in the centre {V\^. 196). 1 hey are often multiple, a dozen or more being scattered over the face and neck. When multiple, they usually repre- sent different stages of development and are usually confined to persons of sandy complexion much exposed to the elements. I his type has been called sun cancer} Sooner or later the regional glands become involved and these growths then follow the course of the deeper t\pe. Fic. 10; Extensive rapidly growing epitfielioma of the ear. Sometimes, from invasion or infection of the brain, death by meningitis may supervene without the formation of metastasis; or hemorrhage may occur from invasion of some of the deeper vessels. Still more frequently, by local extension the soft tissues of the mouth are destro}ed, nutrition is interfered with, and the patient dies from exhaustion, or a septic aspiration pneumonia supervenes. Deep Type. — This form ma}' occur on the neck or the temples, but is most frequent by far on the lips, 241 out of 490 cancers of the face, ^ Hyde, Amer. Jour. Med. Sci., 1906, cxxxi, I. 284 TUMORS OF THE FACE according to Fredenberg, being in this region. Tliey occur twenty times on the lower lip to once on the upper, and hence cancers of the lower lip may be taken as the type of the deep cancers. The greater frequency in males, 409 in 473 cases according to Neiman, has led to the supposi- tion that the proclivity of men to outdoor life and the habit of smoking are important etiological factors. Outdoor life does, in fact, produce drying and subsequent cracking of the lip, and this condition is fre- quently the starting point of the tumor. Similarly, irritation produced Fig. 194 Epithelioma of the ear, forming a polypoid mass. by a broken tooth or by a pipestem, either because it is rough or because it may wear the teeth sharp, may act in like manner. Seborrheic patches, fissures that refuse to heal, indolent ulcers (Fig. 197), and papillomas (Figs. 198 and 199) may likewise serve as the starting point. Whatever may be the lesion of origin, the subsequent course in an}^ case is marked by an increased induration which is harder to the touch than an inflammatory swelling. With the loss of the protecting epi- thelium there is a tendency to bleed readily when the incrustation is removed. The pain at this stage is due to the excoriation rather than DEEP 7) /'A Ol ri MORS Ol Till: I .ICE 2X5 to invasion by rlie urowtli. l'suall\ , as v\\\\\ as rhc excoriation is noted, oanccT plu^s ina\- In- seen at the ((i^c of" the iiUci. These are Hne white points like mmiiie eonudoiies, whuli rie(]iienil\ ina\ he pressed out by grasping ih( hp tuinl\ between ili. ihimib and Im-.i. I'hese, with the shi/■;/■;/' TYi'i. or tumors oi tiii: i .u.l Kic. 203 2S0 Deep ulceration of a rapidh' growing lip carcinoma in a ho\-, aged sixteen \ears. Fig. 204 Epithelioma ot the lip, showing ulceration and fungous formation simultaneous!}- 19 290 TUMORS OF THE FACE tion, and pain which arises from invasion of the nerves, particularly the mental. The extension along the mucous membrane may reach far back into the mouth and make deglutition painful. The exposure of the respira- tory organs to this source of infection frequently leads to pneumonia; or the extension of the growth to one of the large vessels ma}^ result in a fatal hemorrhage. In the most slowl}^ growing types the chief field of activity is frequently the lymphatic glands. The glands suc- FiG. 205 Epithelioma of the Hp without ulceration. cessively become enlarged until massive tumors are formed which may ulcerate. The lip in rare instances may be invaded by these glandular ulcerations, making it appear as though the lip ulceration were secondary to the disease in the neck. In these cases the glandular disease over- shadows the original tumor? and together with the diffuse invasion of the tissues of the mouth and neck may conceal it. The history as given in these cases may be that the disease began in the gland. This may raise the question of a Ij^mphosarcoma, but the density of the tumor and the nodulation prove the growth to be carcinomatous. The DEEP TYPE or rr.MOKs or riir r.iCE •I'M enlarging {ihiiuis iii;i\ i;iusc- tlirticiiltv in icspiratioii and deglutition, either troni the pressure the\ exert in the neck or in tlie niediastinuni, or hy in\ asion or compression ol the ner\'c-. I he slowly mowinj^ t\'pes characten/ed h\' late <;landiilai nutastasis usually run their course in from two to five years. I'hose m which the ju;lands are affected earl\' ami Income the prominent factor usualh run rheir course in six months to a year and one-halt. When this t\"pe of ^row th occurs ahout the orbit the e\e is quickly invaded il'ii^. 206), the bone is attacked, and cerebral invasion may take place. Invasion of the lymphatic glands of the neck is usuall\' 1... 20'. Deep carcinoma ot the inner canthus, involving tlie nasal cavitv secondarily. early and extensive and invasion of the vessels, trachea, and esophagus soon follows. Large fungiform, ulcerating bleeding masses frequently reduce the patient quickly b\' the disturbance of sleep and hv pain, disordered nutrition, and intoxication, both metabolic and bacterial. The superficial t}pe develops so slowly that it is frequently difficult to determine when the papilloma or seborrheic patch has become malig- nant. Even the microscopic examination may not give positive informa- tion. Whenever the removal of the scales gives rise to bleeding, par- ticularly when the scaling has been replaced by incrustation, it is safest to regard the benign stage as past. If positive induration about the border is present the diagnosis is beyond question. 292 TUMORS OF THE FACE In the deeper type the problem of determining when the lesion has become malignant is, unfortunately, usually easy. The early stages are nearly always past and the positive, evidence of carcinoma is at hand when cases come under observation. Albert's statement that there is never any difficulty in the diagnosis of carcinoma of the lip unfortunately meets with few exceptions. Sometimes, induration about the fissured lip may raise the question of epithelioma, but the tenderness, the evident subepithelial induration, and the greater elasticity distinguish such cases from epithelioma. The hard, horny papillomas occurring on the lips of males beyond middle life, even when not painful or indurated, should be regarded with grave suspicion. It has happened twice in my experience that such lesions which did not give evidence of malignancy on microscopic examination were, nevertheless, followed by speedy recurrence when excised. Fig. 207 Chancre of the upper Hp. Diagnosis. — Lupus, though relatively uncommon in this country, is sometimes seen upon the face, and bears some resemblance to super- ficial epithelioma. It usually begins in young persons; the borders are brownish, softer, and more rounded than in epithelioma. A differentiation must be made particularly from tuberculosis and syphilis. Tuberculous lesions seldom invade the lower lip. When they attack the upper, they occur about the alae of the nose away from the vermilion border. The age of the patient, the chronicity, and the greater density of carcinoma make the chance of error in distinguishing between the two conditions very slight. /'R(j(:.\(js/s or ri \U)i Sup. thyroid artery. 7. Com. carotid artery. 8. Com. hypoglossi nerve. 9. Phrenic nerve. 10. Int. jugular vein 11. Masseter. 12. Submaxillary gland 13. Digastric. 14. Ant. belly omohyoid. 15. Thyrohyoid. 10. Platv.«ma myoides ikj:./T.\n:\r or ri mors oi i he i .ice 2'.i:> I In- ilii-piT t\ |n- ii-tiiiins iii all iiisr:inv-(s rlic snmc r:i(lic:il rrcatmcnt thai IS apjiluil to caifiiionia in am oilui ic;;i()ii. C Oiisci \ at isin, based on a iKsiii toi tile l)t-st cosmetic usiilis, lias led to a lni;li ultimate iiiortalit\ , ami should al\\a\ s he resisted. I he lessons learned ni the operatiNi' ticatnuni of eaiemoma of the Imast seem to l")e lost when a similar lesion of the lip is eoiucined ; sui;i,eons should he as Hiiii in refusing;, local excision in the one case- as they are in the other. Not alone should the <;ro\\th with a wide mar^;in he removed, but the reiiional hniiih <:lands as well (Plate I), whether obviously diseased or not.' I'hese <;lands are located m the trianf:;le formed by the anterior bellies of the digastric muscles and the hyoid bone. I'hey are usualh' from tw^o to six in number. From these the lymphatic vessels pass (a) downward across the hyoid bone and terminate in a gland on the anterior surface of the internal jugular vein at a point where this vessel is crossed by the omohyoid muscle; and (b) a number of channels lead outward and downward to the submaxillary glands. At the external end of the lip channels lead to the inframaxillary gland, which is situated over the inferior maxillary bone just anterior to the point of crossing of the facial vein. From this gland channels lead to the submaxillar\- glands. The submaxillar}' glands in turn drain into the deep cervical glands. The mucous membrane of the lip is drained by lymphatics which pass directly to the submaxillary glands. The chain of glands to be considered are the submental and the inframaxillary in the first link; the subh3'oid and the submaxillary in the second link; apd finally, the deep cervical in the third. I believe the first and second links should be removed in all instances. If these are not obviously involved, their removal, together with the gland- bearing tissue in this region, is all that is required. Should they be obviously carcinomatous, then the third link, the deep cervical glands, on the corresponding side, or both sides, as the case may be, should also be removed. The removal of this amount of tissue can best be made b}' an incision passing to one side of the point of the chin, curving over to the inferior border of the inferior maxillary bone, extending back to the anterior border of the masseter muscle, hence downward to the point of cross- ing of the omohyoid and the sternomastoid muscles (tig. 209). It should be noted that the incisions pass on one side of the point of the chin. The object of this is to leave the skin attached at this point; if it is loosened, it is difficult to keep the flaps in position over the point of the chin. When the lesion is extensive this point of attachment ' D. N. Eisendrath, Juur. .Aiikt. Mtd. Assoc, 1906, xKii, 986; Hertzler, Surg., Gynec, and Ohst., 1909, ix, 80. 296 TUMORS OF THE FACE cannot, of course, be preserved. The lymphatics pass on each side of the tissue left attached, and there is no reason for its removal. The submental space may be easily cleaned out by pulhng the skin aside. The character of the incision in the hp itself depends on the size of the lesion. If small or medium in size, the triangle incision, as shown in Fig. 209, is the most convenient. If very extensive, the incision of Fig. 209 I, omohyoid muscle; 2, sternomastoid muscle. Grant is more suitable, to which is added the incision as here illustrated. If it becomes necessary to remove the deep cervical glands, the vertical line of the incision may extend downward to the clavicle and the infra- maxillary incision may extend backward to the tip of the mastoid process. A second transverse incision may be added which runs parallel to the clavicle and just above it. The neck incision then resembles that recommended by Plummer for the removal of tuberculous glands I Hi ROM. IS or Till. I .ICI- 21>7 of rlu' lU'ik. \\ nil siuh an iiuisioii ilu- intiic iliain «>t ti-rvical ^laiuLs ami (In- L:,laiul-l>rai inj; tissiu- can he icnioscil. As m tin- hiiast, tin- ilissi i.t ion sliouKl In-^m at the ponit farthest from tlu- Rhinophyma. infiltrating. The globular form gives a pseudofluctuation which makes them easily mistaken for wens or other cysts. The yellowish color may be apparent through the mucous membrane, while cysts usually A'. /A'/- ri WKJks or riii: i.ice .{o; appt;ir riansparnu or Miiish. Cystic l\ niphaii^ionuis are parricularh' likch ro caiisr (.■ontiiMoii m cliaiiiiosis. i ic;. 221 "^^- '1 ^m0^ \^' ,.-f"'''-?;W^'^iH :^ . <-.Mi^ i ■ y-^: "^ ■:■■■: ':^'' ■yj^t .- >..r.". ^ J^. .^f^^; Adenoma of nose. Fig. 222 ^^r^K^ ^ ■k 1 ■ i J Br X "■ ^ ^ ^^^^^^^^^ ' ^jH ^ ^^^^^^fc «.«««»- ft 4»^st;iiKc of polyps resembles fretjueiul) true iii\ .\»)ul siil>sr;iiu-e. Sonurinus, however, an\- of the t\pes enumerated may r;ike on tlu- ili,miit\ of imiioitaiit tumois ;iiul fioni their si/e cause pressure atroph\ of neighboring structures. Cerebral hernias some- times present in the roof of the nose simulating nasal pol\ps. 1 hev may produce displacmuni of tin- oibiral plates ot the ethmoid ( Hfj. 22;) simulatint; dermouls of the niner canthus. I'heir recofjnition is important, for if remoMcl without precaution, meningitis may develop. Cerebral hernia of the vault of the right nasal cavity. Treatment.— Snare, cutting instruments, or chisels may be required for their removal. After removal, careful microscopic stud}- should be made in order to detect any possible malignant portion. Gliomas of the nose have been reported.' The presence of glia tissue indicates that the growth is continuous with the brain or represents a terato- matous tumor of the neighborhood. Malignant Disease of the Nose.-^ This is fortunately a rare group of tumors. According to M. Schmitt, in 33,000 cases of disease of the nose there were 6 sarcomas and 5 carcinomas. These resemble each other so closelv that thev may be considered together. 1 J. P. Clark, Amer. Jour. Med. Sci., 1905, c.wix, 769. - For a complete account of the malignant tumors of the nose see Harmer and Glas, Deutsch. Ztschr. f. Chir.. 1907. Ixxxix, 433. 312 TUMORS OF THE NASAL FOSSA AND PHARYNX Sarcomas. — The small, round-celled tj^pe is the most frequent, although spindle-celled and even melanotic types have been observed. The point of origin may be the lateral wall, the septum, or the neighboring sinuses, particularly the antrum of Highmore; but this point is usually impossible to determine by clinical examination, and even at opera- tion, with the affected region exposed, it frequently cannot be made out. The advent of the tumor is often preceded by a prolonged inflam- FiG. 224 Sarcoma of the nasal fossa, producing a displacement of the right nasal bone. matory affection. Usually, the symptoms which bring the patient for examination are those of nasal occlusion, though sometimes hemor- rhage ma}^ appear first. Frequently, the first evidence is the expansion of the nasal bones (Fig. 224), producing a marked bulging of the side of the nose. They grow very rapidly and quickly invade neighboring regions, namely, the orbit and antrum, and by penetrating the cribri- form plate gain access to the cranial cavity. Sloughing masses some- times project from the nose or into the pharynx. 77 I/OA'.S or THE flllKYW WV.', Carcinoma.' I lu sr runiois mkin In- cliiixid (lom ciil)()iclal or s(]ii;ini()iis ipitlulium ot tin n;is;il fossa or from tlu |)crfoiatc(l sinuses. Their growth IS rajMil ami tluir cells soon lose- tluir type, so rliat the source of onjiin can no lonti,cr hi- (Kternnniil. Diagnosis. I he rapiilir\ ot ile\elopnienr anil the e\rensi\e nnasion of the surrounilin^ tissue usuall\- chaiactei i/es the tumor as malig- nant. Mil' microscopic ilisrmction hetweeii mali<:nanr anil heni^n tumois is often iliHicult to make. .\lan\' of" the heni}j,n rumors are cellular anil hear a very close resemhlance to malignant tumors. The difference m clinical course between carcinoma and sarcoma some- times permits of a dif^^erennation. Carcinomas nearh al\\a\s occur after the forty-fifth year. Sarcomas are more often developed from the septum and the carcinomas from the lateral walls. Carcinoma tends to break down and form a fetid mass, while sarcoma retains an intact surface and displaces the surrounding tissue, \'et sarcomas are more prone to bleed. The microscopic differentiation is often difficult, and may require a special technique. Treatment. — Operative removal ma}^ be tried. Extensive local e.xcision is important. The lymph glands of the neck need not be removed, since metastasis does not take place in that direction, but toward the base of the skull. It is questionable if a case of either of these tumors has been cured. - This is easily understood, since it is often impossible to determine the extent of the growth or if the growth has already taken place in lymphatics about the base of the cranium and in the deep glands about the vertebral column. TUMORS OF THE PHARYNX Fibromas. — These tumors develop from the vault of the pharynx. They are said to occur with equal frequenc}' in males and females before puberty, but at this age the}' become excessively rare. Curi- ousl\' enough, during the period between pubert}' and full develop- ment the}' are found exclusively in males. They are very vascular and are made up of cellular fibrous tissue; in fact, the border of these tumors is sometimes so cellular as to make microscopic distinction between them and sarcoma almost impossible. The number of cases of sarcoma sometimes reported may be due to a failure to recognize this characteristic. By their growth they hil first the vault of the ' Drev fuss, Wien. mcd. Presse, 1892, xx.xiii, 141 7. - Herzfeld, Berl. klin. Wochenschr., 1900, xxxvil, 796. 314 TUMORS OF THE NASAL FOSSA AXD PHARYNX phaiynx, and extending downward below the soft palate, interfere with deglutition and even respiration. They may encroach upon the nasal fossa, may invade the accessory sinuses, and may, through pressure atrophy, enter the cranial cavity. Symptomatology. — Aside from the pressure s\'mptoms, hemorrhage, sometimes so great as to threaten life, neuralgias, and disturbances of respiration are the most prominent. Diagnosis. — The chief point of diagnosis is the presence of a tumor which occupies or fills the pharynx of an adolescent male. It is a pedunculated mass developing from the vault of the pharynx and fill- ing the retronasal fossa. Its smoothness and density and the relative slowness of its growth distinguish it from sarcoma. The malignant growths usually appear later in life, are of more rapid growth, and are of softer structure. From pharyngeal polyps, fibromas are distin- guished by color, density, and location, the microscopic diagnosis being less reliable. Prognosis. — After the individual has attained full development the tumor frequentl}' regresses and sometimes disappears. When removed, cure is usually permanent, but recurrence may take place, only perhaps to disappear about the twenty-fifth year. Because of the uncertainty in diagnosis, even after microscopic examination, the prognosis should alwaA's be guarded. Treatment.^ — When small in size, pharyngeal fibromas have been removed by means of a snare through the natural passages, but it is a dangerous procedure, for even in the small ones the blood supply ma}' be very abundant and lead to serious after-hemorrhages. The large tumors at least require the temporary- resection of one or both upper jaws, with or without the ligation of one or both external carotids, in order that the hemorrhage may be more certainly controlled. Usually a unilateral operation is sufficient, but if the growth is very large both sides should be exposed. At best, the operation is a bloody and formid- able one, and ever}^ precaution should be exercised to meet every emergency. ^ Custodis, Beitr. klin. Chir., 1905, xlvii, 37. C' II \ V I i; R X \ \' TUMORS OF INK MOLTII AM) loNCill, General Consideration. runiors of rlu- buccal ca\ir\ and its con- ttius nia\ tor ct)n\cnicncc he considered tojierlui". (jlands sinidar in character and subject to like diseases are distributed throughout the cavit}', congenital anomalies affect alike various portions of it, and malignant disease beginning in one part extends quickly to neighboring structures. They are similar in structure and run a like course. CYSTS IN THE FLOOR OF THE MOUTH Ranula.- Ihese c\sts in their typical form are derived from the outlet ducts of the sublingual gland. ^ They may also develop from Xuhn's glands, which lie on each side of the frenulum toward the tip of the tongue, and from the incisive glands which are situated at the base of the incisor teeth. Neumann- has classed with the ranulas cystic dilatations of Bochdalek's glands. True cysts occurring in these glands, when ver}' large, ma}' be present m the floor of the mouth, but usually the\' are found at the base of the tongue alone, and for this reason are best classed with the latter. Ranulas usually occur in adults, rarelx" in children. They have been reported as occurring congenitalh, but confusion with cysts of the th^roglossal duct in these instances is possible. The}' arise from the occlusion of the outlet duct of the glands by chronic inflammation. Acute ranula ma\' be produced by sudden occlusion, as by stone. The closure of the duct is followed b\" the retention of the secretion, and, in addition to this, exfoliation of the lining epithelium and exuda- tion from the surrounding vessels are thought to compose the C3'stic contents. With the growth of the c}st the surrounding tissues are displaced and portions of the cyst ma\" be forced between the fibers of the surrounding muscles. The cystic contents rarel\' contain digestive ferments, but are usualh' clear and ropy, seldom brownish or blood}". When the tumors are small the}' are situated lateral to the frenulum, ' \'on Hippie, Arch. f. klin. Chir., 1897, h", 164. -Arch. f. klin. Chir., 1877, xx, 825. 316 TUMORS OF THE MOUTH AND TONGUE corresponding to the situation of the subhngual gland, and bulge the mucous membrane of the floor of the mouth (Fig. 225). Wharton's duct can usuall}^ be demonstrated passing directly over the summit. When they become larger the}^ pass the median hne, and the frenulum above them may cause them to present a bilobed appearance. Occa- sionall}^ the}^ are multilocular, double, or even multiple. Accessory loculi may form within the surrounding muscle, and when large may extend into the myohyoid muscle and cause a prominence beneath Fig. 22: Ranula forming a lobulated mass under the tongue. the chin, or in the region of the submaxillary gland (Fig. 226). Those forming in the glands of Nuhn are situated under the tip of the tongue, and those forming in the incisor glands are beneath the frenulum immediately beneath the bone. The last two are rarer than those of the sublingual gland. The cysts usually have a grayish-blue or bluish-red appearance, but when, as rarely happens, acute inflammation takes place, they may be intensely reddened. They are soft, fluctuating to the touch, and usually relatively movable. CYSTS l\ Tin. I LOOK or I III: MOl I'll wv Diagnosis. Tluii sitiKition l)ciu:itli rlu- toii^iif, |)r(jjecrinj; into rlu- suhlin^ual spaii . iluii Muisli loloi, ;in(l i\ st ic trtl arc usually surticic-nt to distiuj^^uisli tlu-m. W'luii tin- muious iiunihiaiu- c-()\criii^ riu-ui is vasculai, auuioina iiuist In- thought of, Inir a globular iiiconipicssihic tumor is distinctive of the- cyst. rh\ i()<;lossal c\ sts in the suprali\<)id ritiion, wlun projcctinu; into tin- llooi of tin- mouth, mav simulate laiuila. I suall\ in laiuila tin re is a history of a hist appearance heneath the tongue, ami there is no attachment to the h\()id hone. Fig. 226 Ranula bulging beneath the angle of the jaw. as in the thyroglossal cysts. Ranula simulates tumors of the duct only when of such size as to project below the jaw. This condition has seldom been observed. An inflamed ranula is likeh' to be mistaken for inflammation due to a sialohth, and indeed the dividing line between the two conditions ma}' not be a sharp one. Treatment. — When a cyst is small, excision of that portion of the wall which projects into the mouth and cauterization of the remam- ing portion with silver nitrate is followed by a cure. Stitching of the 318 TUMORS OF THE MOUTH AND TONGUE cyst wall to the mucous membrane of the tongue after excision of the redundant portion and cauterization of the remaining tissue is often followed by success. Repeated cauterization during the healing gives additional security. When the cyst is large, excision of the entire sac should be practised. This may be done through the mouth or, as V. Hippie^ recommends, by an incision beneath and parallel with the jaw. This gives a better view of the field of operation and permits a more effective hemostasis, but does not in the majority of instances avoid an opening into the mouth. Loose packing with gauze for a number of days must be done after the use of either method. Thyroglossal Cysts. — The imperfect obliteration of the fetal thyro- glossal duct may lead to the formation of cystic tumors along its tract. The part most commonly affected is the foramen cecum at the base of the tongue and immediately above the hyoid bone. These tumors are ovoid, bluish, glistening cysts which project above the tongue. Usually, when first observed, they are as large as a hazelnut or a hickoiy- nut. Occasionall}^, when very large, or arising in the substance of the tongue, they ma}' be visible beneath the tongue, and have been classed with the ranulas. A second tumor may exist in the tract of the duct; these are particularly likely to be mistaken for ranula. Often the wall of the cyst contains well-developed thyroid tissue,- which is frequently very vascular. When this complication exists, the color is bluish red or even deep red. The tumors are usually soft, elastic, and painless. When small, they cause no inconvenience unless from hemorrhage. When they become larger, they may exert pressure upon the epiglottis and may cause cough and difficulty in swallowing. Occa- sionally, repeated hemorrhages may occur before the tumor is dis- covered. The blood, because of its unaltered color and the attending cough, ma}^ be ascribed to pulmonary hemorrhage. The next place of predilection is immediately above the hyoid bone. Tumors which occur here have the same characteristics as those situated at the base of the tongue, but the overlying skin obscures the color. Diagnosis. — When situated at the base of the tongue their globular form, bluish color, and situation in the median line are distinctive. When situated more deepl}^, so that their bluish color is obscured, they may be confused with gummas. The latter spring up more rapidl}^, and are situated laterally to the median line. When situated below the tongue they simulate ranulas, but these are more superficial and bulge to one side of the frenulum. Thyroglossal cysts below the chin (see Tumors of the Neck) simulate dermoids and lipomas, but they are lArch. f. klin. Chir., 1897, Iv, 164. ' Bernays, St. Louis Med. and Surg. Jour., 1888, iv, 201. os/s i.\ ■/■///■: I LOOK oi I III, \ifji III :u\i mon- ii,l()l"iiil;ii' and Iiss tiiil\ moNahlc th;in t hcsi, and an ai lathintiu to {\\v In Old honr nia\ In- made out. Treatment. I'or those sitnatcd at tin Ikisc ot the tonmic, wlun tlu-ir blue color indicatts an ahs«.iut- ol tinioid tissue, simple excision of rlie redundant tissue with caiitti i/ation ol the reinaiinnu portion of the sac, as ad\ised tor lanulas, nia\ In- all that is ie(|iiii((l. ('omplete excision is to hi' piactisetl wluii this iiKthod tails; hiii this measure is contiaindicated when the thyroid tissue al)o\'e the thyro^lossal duct is all the tunctioninji; thyroid tissue the patient possesses. Its remo\al in sucii cases would he followed hy cachexia th\ ropi i\a.' I sualh , however, rlu' entire growth should be removed, and if there is a dis- position to bleed, remoN'al is imperative. It simple removal ot the surface of the c\sr alone is iiucnded, the operation ma\' be undertaken under local anesthesia, Inir it the urowth is vascular, a general anesthetic should be employed and the usual pre- caution for preventing the flowing of blood into the trachea should be taken. Fhis operation may be done through the mouth unless the tumor is very large, wdien an incision from below ma)' be advisable. The lower route should be selected if there are secondary tumors below the one at the base of the tongue. 1 raclieotomy is not necessar\' unless from the size of the tumor the epiglottis is much pressed uj^on. In that event the method of Butlin may be employed. For cysts situated above the hyoid bone a skin incision is necessary, and sometimes the h\'oid bone may be severed with advantage, particularh' if the c\st extends back of it. In iistulas resulting from ill-ad vised treatment in these cysts, the entire tract must be carefully excised, always under general anesthetic. Dermoids. — Dermoid C3'sts are sometimes observed under the tongue toward the base situated in the median line or laterally. The}' form globular tumors, w'hich are smooth, soft, jelly-like, and sometimes fluctuating to the touch. The skin and mucous membrane remain freeh' movable over them. The\" are painless and cause disturbance only wdien from their size they embarrass deglutition or speech. I hey project into the mouth and limit its space or make complete closure of the mouth impossible, or the}" ma\' project outward below the maxilla (see Tumors of the Neck). The}' frequently have a fibrous attachment on the inner side of the inferior maxilla above the attachment of the genioh\oid muscle or to the bod}' of the h}'oid bone, indicating their origin in the mesobranchial fold. The}' are lined b}" striated epithelium which contains hair follicles and sebaceous glands, and are filled with a granular debris which ma\' be brownish and contain cholestenn ' \ on Chamisso dt- Boucourr, Bt-irr. /. klin. Chir., 1S97, \ix, 281. 320 TUMORS OF THE MOUTH AND TONGUE crystals. These tumors may be observed at birth, or they may not be noticed until an advanced period of life, but they usually manifest themselves between the eighteenth and twenty-sixth years. Diagnosis. — Their yellowish or whitish color, usually discernible through the mucous membrane of the mouth, differentiates them from ranulas. The absence of lobulation distinguishes them from lipomas. If the strand of fibrous attachment to the body of the hyoid or maxillary bone is made out, the diagnosis is certain. From suprahyoid cysts puncture alone will differentiate. Cold abscesses occasionally develop in glands in the submental region, and have been mistaken for dermoids. They are more firmly fixed and project less regularly into the floor of the mouth. In case of doubt an exploratory puncture can be made. Treatment. — Excision is the only method which should be considered. When small, they may be shelled out through the mouth, but when large, they are better attacked through a skin incision. If care is taken not to injure the floor of the mouth, the wound heals primarily and the resulting scar is insignificant. The size of the cavity remaining after the removal of the tumor may be lessened by buried sutures, which also make drainage unnecessary. When the tumor is removed through the floor of the mouth, the wound should be treated as an open one, using a temporary packing to lessen the likelihood of infection of the cellular tissue. Local anesthesia is sufficient for operation by either route. Lymph Cysts. — Lymphatic dilatations, from the size of a pea to that of a walnut, may occur in the floor of the mouth and in the tongue, but usually in the cheek in front of the masseter. These tumors are tense-walled cysts, with endothelial lining and without perfect encapsu- lation. They may simulate adenomas in this region. If they cause inconvenience from their size, they may be extirpated entire or partially removed and the remaining portion of the cyst cauterized. MALIGNANT TUMORS OF THE MOUTH AND TONGUE Carcinomas of the mouth may be divided into those of the tongue, those of the floor of the mouth, and those of the buccal mucous mem- brane. They have much in common and run a similar course, yet their great practical importance makes it desirable to treat them separately. Carcinoma of the Tongue. — The tongue is a relatively frequent site for carcinoma, from 5 to 7 per cent, of all malignant epithelial growths being found here. Males are much more frequently affected than females in the proportion of about 8 or 10 to i. This great frequency M.ii.ic.s.isr rr.MfjRs or the moi rii ./\n ro\(Jii-: :',-2\ in iiKilts has l>itii thought to In- dm- to tluir nujie general addiction to flu- use ot akohol and tobacco. In carcniorna of the tonj^ue the iiiHiunce ot irritation has heen tracetl with greater exactness than in any other rej^ion ot the hoil\ . I hi- ina|oiit\ of rnahj^nant diseases of the tongue occur adjacent to the roughened edge of a tooth or a broken dental plate. 1 his association is so common that attention to roughened objects within the mouth should receive correction as a proph\ lactic measure against carcinoma ot the tongue. A considerable proportion of lingual carcinomas, as many as one in four according to some auth(jrs, are preceded by leukoplakial spots. There are no statistics available, however, relative to the number of leukoplakial areas which become carcinomatous. Fk;. 227 Beginning carcinoma of the tongue, showing ulcer developed beside a papilloma. The disease begins usuall\-, as in case of the lip, as an ulcer f Fig. 227), or a leukoplakial patch (Figs. 228, 229, and 230). The transition into carcinoma is characterized b\' an induration about the base, which is harder than an inflammator\' exudate and which spreads beyond the contines of the epithelial defect. The papillomas of the tongue are especially prone to intiltrate the base. Leukoplakial areas undergo changes much more slowly, \ears being sometimes required for them to reach a malignant stage. Each of these types when full}- developed has edges which are hard to the touch, are usually raised above the sur- rounding surface of the tongue, and have enclosed within their walls an angry granular surface which bleeds easih' on touch. As in most carcinomas, "cancer plugs" can often be seen about the border of the ulcer, and can be forced out by firm pressure. The base 21 322 TUMORS OF THE MOUTH AND TONGUE Fig. 228 Leukoplakia of the tongue. Fig. 229 '<7- ' 't- /- ) / U:, X y\i iiJ J V'-- c^ Leukoplasia, benign stage. \i ii.h.\ i\r 77 MORS or Tin: moi rii /\/> ioma i- .lm (>t the iilrci m.i\ tiiiiL^.iti .iiul |iiii(liii( .1 l.ii^i- mass \slm"li projects a (.■oiisulii .ihic ilisi.iiur .il)i)\( il)( Ml 1 1 ( Hindi iiL! siiil.icc i)t tlic roiijiut*. Mir iii.iss is iisii.ill\ liim .iiul tii.iMc. Molt i.iicK, ilic ca ifinonia iit\rlops within ilu- siihst .1 nrr ot the tontiut-, lta\in^ tin- surface tor a turn unhiokiii. ( iiaciual!\ , lio\\t\ii, rlie suiface is encToached upon and an iiUfi results, diH-eiiii<; tin 11 m no wise from those which befiiii oii!j,inall\ as ulcers. I siialh' the caicmoiiia picsenfs itscH as a ciicum- scrdnd iilcei, Inn occasionalK tlu-re is a dilliisc inhhiaiion iiuoKin;^, hom the luij^innini:, a consuleiahle area ot the tonji;ue. I he edue ot the tongue IS h\ fai the most tie(iuent site because of the exposure ot" this jxirt to irritation. I hose arismu trom leukoplakial spots, however, are seen most tii-tpienth m the dorsum. Viv.. 230 Leukoplakia, beginning malignancy. Pain IS usuall)' an early symptom. Contact with the teeth or \Mrh irritating articles ot food causes pain as soon as the epithelium is lost. Radiating pains begin as soon as infiltration is at all marked. The pain IS referred to the cheek or jaw, and, when the tumor is situated on the side of the tongue, toward the ear or mastoid region. The extension of the disease is usualh' rapid. The cause of this has been sought in the loose arrangement of the structure of the tongue, the abundant blood supply, and the number and size of the l\"mph channels. In some instances the extension is excessiveh' rapid, the entire tongue, the floor of the mouth, and tissues of tiie neck being (juickly imadetl b\' the epithelial cells accompanied b\' an extensive round-celled infiltration. So rapid is this process that in the course of weeks the tongue is converted into a solid, immovable mass, making deglutition difficult or impossible (Fig. 231). This type is most often 324 TUMORS OF THE MOUTH AND TONGUE seen in persons under forty A^ears of age in whom a diagnostic section or a partial operation has been performed. Slowly growing carcinomas are rarely found on the tongue. The lymph gland metastasis is, as a rule, very early and has already taken place when the surgeon is consulted, notwithstanding the exposed situation of the disease and the consequent tendency of the patient to seek advice early. The order in which the glands are involved is Fig. 231 Rapidly growing carcinoma of the tongue, involving also the floor of the mouth. not constant. The submaxillary glands, because of the more frequent location of the growth on the lateral surface of the tongue, are usually the first invaded. When the tip is the seat of the disease the submental glands may be the first affected. When the growth is near the base the deep cervical glands may be the first affected. It must be assumed that all these glands are implicated in any given case of carcmoma of the tongue. Not alone are those on the diseased side affected, but M.n.i(.\ i\T ri MORS (ii Tin moi tii i\h 7n\(,rr. :i2.') because ot the iich l\ inphatic anastomosis at tlu root ot the ton};iie the opposite side is (juite as hkel\ to he invaded. Indeed, the glands of the opposite side may he the Hist or even the only ones to be aftected. When s\ stemic cUssemination takes place, the liver is the most fre- (juent site of metastasis. This may occur before the regional l\ mph j^lands are diseased, and indeed before the primar\- tumor is discovered, though this is ran-. I his is particularly true in those which occur at the base of the tongue or in the Hoor of the mouth. I his should be remembered in obscure hepatic disease. Diagnosis. — It is the diagnosis of tongue carcinoma that v. Bergmann was wont to dilate upon in discussing the differential diagnosis of carci- noma, syphilis, and tuberculosis. Syphilis is sometimes multiple, is situated more often on the dorsum of the tongue, begins deep in the substance of the organ and not near a broken tooth, and the lesion is more nearly spherical, more circumscribed, and, above all, is not as dense to the touch as carcinoma. Lsuall\- there are other signs of svphilis. Tuberculosis of the tongue is rare. It may occur on any part; the ulcer is usually superficial, the edges are soft and under- mined. Sometimes, however, the tuberculous ulcer ma\- form a large fungous mass, which resembles very closely the fungating carcinoma. There is usually tuberculosis elsewhere in the body, particularly in the lungs, which would suggest the nature of the tongue lesion. Lung metastasis from carcinoma may be mistaken for a tuberculous pro- cess. Careful physical examination, together with examination of the sputum, will distinguish the nature of the lung alfection. In the rapidh' growing type of carcinoma, a phlegmonous process mav be thought of, but the former is less acute, and the density and nodulation always discoverable in some part of the tumor are distinc- tive of the neoplasm. Besides, there is usually the history of some primarr lesion which has been subjected to ill-advised treatment. The primary lesion of syphilis with extensive gland disease may simulate this type of carcinoma very closely, but a search for the dense nodula- tions of carcinoma and the evidence of syphilis elsewhere in the body will permit a diagnosis. It is in the mouth that surgeons most often desire the aid ot the microscope for diagnosis. It is here also that exploratory incision is so likely to provoke mischief because of its disposition to increase the likelihood of metastasis. Every means should be employed to arrive at a diagnosis before removing a section for examination. Inoculation tests for tuberculosis and the therapeutic or serum tests for syphilis usually allow us to arrive at a diagnosis by exclusion. Moreover, carci- noma, when present, is usually only too clearly such. \\ hen the diagnosis is in doubt, in the vast majority of cases, the disease is not carcinoma, 32(3 TUMORS OF THE MOUTH AND TONGUE but definite proof of this assumption is best brought to the patient by the cure of the lesion if it is other than carcinoma. Treatment. — Local treatment is never permissible. When the lesion is small, an excision of a wedge, ij/^ centimeters beyond the border of the tumor, ma}^ be permitted. If at all diffuse or extensive, a lateral half of the tongue should be sacrificed. Local return is more common than recurrence at distant lymph glands, and as much tissue as possible in the region of the growth should be removed. When removed with the cautery, recurrence is less likely than when a scalpel is employed. In either case the adjacent lymphatics, together with the salivarj^ glands, should be removed in all instances. The sternomastoid glands receive the lymphatics from the tongue. These glands are situated beneath the sternomastoid muscle upon and behind the internal jugular vein. Their complete removal is best assured by removing the vein, together with the surrounding tissue after Crile's method. The neces- sary exposure can be secured by adding to the Kocher incision one extending down the anterior borders of the sternomastoid muscle to the clavicle, as recommended by Plummer for the removal of tubercu- lous lymph glands. If the growth is well confined to a lateral half of the tongue and is in its incipiency, such an operation may be regarded as a complete one. If the midline is affected, or if the tumor is on one border and has existed for some time, or is of rapid growth, both sides of the neck, at separate sittings if need be, should be subjected to similar treatment. If the glands are palpably involved the entire lymphatic apparatus should be removed. The method of removal is unimportant, but to the incision of Kocher the temporar\' resection of the jaw should be added. When the malignant process has extended to the floor of the mouth, a permanent cure, regardless of the extent of the operation, is hardly to be hoped for. Carcinoma of the Floor of the Mouth. — Primary carcinomas are observed in the floor of the mouth less frequently than on the tongue. They begin as fissures and ulcers with gradually extending borders, and are prone to infiltrate quickly the loose connective tissue of the mouth. Their exposed situation invites secondar}^ bacterial infection. Metastatic processes are likely to occur early in the regional lymphatics, which may comprise any of the Ij^mph gland group of the neck, and the}' are especialh" prone to occur in the viscera. The liver is most frequently involved, and in case of obscure tumors of the liver or hepatic enlargement the mouth should be examined for possible malignanc3^ The growth primarily causing such mischief may be so small as not to have attracted the patient's attention. Diagnosis. — The indurated ulcer, the presence of carcinoma nests, and enlarged hard glands are distinctive. Frequently, however, com- () .1/. //,/(, \./.\y 11 MORS (Ji riii: moi rii i\n iDM.ri-: wr, plications iiulu.iiini; ;i simpK- inHamnKiioi \ aHiiiion iiia\ (thsiint- rlu- pu-tiiu. S\pliilitu' Ksions ina\ siimilaic it ami tin- iapiclir\ of tlu- ina lii:,iiani pi orcss m.i \ ;ultl I o l he i.(Hif iimi )n . ()thti cv uitiuc of s\'phili.s ma\, howiNii, usiialU hi iliscoNtiid. Atiiic pliici^monoiis processes I l.iulw ii^'s anjiinal nia\ Imin a tlcnsc tuiiioi, luit tin i;rcatii jiia\"it\' of tlu- scpnc |")i"()ccss in Muh casi-s (lisnii^iinslus iluni hom taicinonia i'oni|i|ica ti-il li\ cHJiulai intiriion. In ail in()in\ losis, xsjicn the Hoor t rlu- moiitli IS imacltil, t lu- cnla i i^inuni is Jess dense than carcinoma and rlu' siiitaci' is loiindci. In this coimriN' actmonncosis, paitiCu- lail\ in tins location, is ixccssin il\ lair. Treatment. Wlun tin timioi is small, tin- Hooi ot tin month, with hoiderins2; lymph glands, ma\ be excised, pieferahl\- with the caiitei\ ; withont, ho\\e\er, any considerable hope of success. W hen the Hoor ot the month is at all extensively affected, the Ian ean he earned out without detormit) resulting. L iitortunatel\ , however, the diaj^- nosis is usuall\ not made until so large an area is invoK'ed that it would necissitate the lemoxai ot the- major portion ot the tongue. When this state is reached, it is as well to allow the organ to remain until from its si/e it demands iemo\al. Hemangiomas. I In- tongue is not uncommonly the seat ot angicMiias. Racemose aneui\sms ha\e heen obser\ed, hut the usual torm is the arteriovenous. K.xtension ot capillar) ne\i trom the lips upon the tongue is occasionally noted. Diagnosis. I he diagnosis ot none ot these types sh(nild cause ditH- cultw I'he color, the pulsation, or the reduction when compressed with siihse(]uent retilling are characteristic. Treatment. In the racemose type the supplying vessels should be ligated. In the arteriovenous type, if small, excision with (^r without a preliminar\' ligation ot the lingual arteries is the treatment ot choice. In the nevoid t\'pe electrolysis is the method ot election. RARE TUMORS OF THE TONGUE Lipomas. — Occasionalh" small lipomas appear on the surface of the tongue or within its substance, or rarely in the floor ot the mouth. When near the surface, the yellow'ish appearance characteristic of these tumors ma\' be seen. To the touch the}' are sott and non-resistant. When the\' occur in the floor of the mouth their lobulation differentiates them from dermoids, but in some cases, puncture ma}' be necessary tor a decision. When they become large enough to cause annoyance, they ma\' be removed. Those situated on the floor ot the mouth should be removed through a submaxillar\' incision, in order to avoid opening into the mouth cavity. Fibroma. — These tumors, like the preceding, are sometimes tound upon or within the substance of the tongue. They are frequently pediculated. The\' ma\' form dense tumors, and when located m the substance of the tongue may cause confusion with gummas, c^'sts, and even carcinoma. Their dense citcumscribed character should distinguish them. They ma}- be removed if the\" cause annoyance. Gummas. — Gummas ma}' occur in an}' part ot the tongue. 1 he}' are usualh' the result of acquired s}philis, and usuall}' occur a number of \'ears after the beginning of the disease. The\' torm nodules var\"- 332 TUMORS OF THE MOUTH AND TONGUE ing in size from a pea to a walnut, and may be multiple. They are usually smooth, and the mucous membrane over them is usually un- changed, but may be reddened. The consistency may be very dense, but when they approach ulceration ma}^ become soft and fluctuating. They may exist for a period of years without showing any disposition to ulcerate. They differ from carcinoma in their smoothness, their intact surface, their less density, their usual occurrence in early life (twenty-five to thirty-five), and in the fact that no chronic irritant, such as a broken tooth, may be found. In addition, there is evidence of syphiHs else- where. They resemble sarcoma in situation and form. The therapeutic test should be appHed in cases of doubt. C II A V I K k X X \ I llMokS Ol- rilK lAWS, GUMS, TEE'IIL I'KR lOSTEUM, AND BONK General Conception. 1 uniois of the ^ums and jaws and those resuit- ini!;ition. More extensive tuniois, piirtuuhii l\ \slu-n :uToni|);inu-(l In' loostiu-d teerh, demand renio\al ot the teeth, tojic-rlu-i with the aKrolar process ot the jaw. I his nia\ i-asil\' he acconiplislu-d with eiitrin^ torceps. Ileinor- ihat;^- ina\ hi- contiolkil with pressure or hy the eaiirei\. I he latter acts also as a safeguard against recurrence. In neglected cases, where the growth has in\()l\ed a considerahle portion of the jaw, a resection ot the jaw has been doiu. i'.\fn in the \er\' extensive cases, unless the growth has been rapid, conserxatne operation should he gi\en first trial. In rieanng epulides of the upper jaw care must he taken to determine whether or not growth has extended into the antrum. Ihis is likely to occur w Iumi tiu- rumor is situated near the canine teeth. Alveolar Granuloma. Among the most common tumors of the gums are exuberant granulations springing from an area irritated b\' a carious tooth, or a local infection. Ihey \ary in size from a pea to a cherry. They are reddish or bluish in color, vascular, and soft to the touch. I hey bleed readily when injured. Ihex' spring up quickly and remain stationarw Their surface is onl\' in part covered by epi- thelium, the summit, at least, being formed by granulation tissue. 1 hey are formed ot vessels and granulation-tissue cells in close association, lo the novice the picture resembles malignancy, and the diagnosis "angiosarcoma" is frequently returned. This error is particularly likely to occur when a history of recurrence is given. Diagnosis. — Their close association w-ith the teeth, particularly- diseased ones, their vascularit}', and disposition to bleed characterize them. Epuhdes are firmer, often whiter or more brownish red, and develop more slowly. The clinical history is of greater importance than the microscopic examination. From sarcoma the\' are easil\' distinguished b}' their softness and their livid color, and b\' the fact that having attained a certain size the\" become stationary. Sarcoma of like structure is diffuse and the growth is continuous. Treatment. — Because of their disposition to recur, like exuberant granulation tissue elsewhere, the base should be destroyed and diseases ot the teeth corrected. The curette, followed by the electrocautery, is most convenient, though chemical cauteries may be employed. Radical operation is unnecessar}'. Carcinoma. — The gums are nor infrequently the starting point tor carcmoma, which usually begins as an ulcerative process about a decayed tooth, a bridge, or a plate, or rarely as a papilloma. Such an ulceration about a decayed tooth, which is the usual case, differs from an ordinary' ulceration in having an infiltrated border. Usualh' the bone about 336 TUMORS OF THE JAWS, GUMS, TEETH, PERIOSTEUM, AND BONE the root of the teeth is exposed and cancer plugs can be seen. The ulcer spreads along the alveolar process anteroposteriorly (Fig. 235), and may extend over the jaw and involve the cheek, the anterior pillar of the fauces, or the floor of the mouth. If the growth affects the upper jaw it may involve the hard palate for a considerable distance and form a flat ulcer with granular floor and raised indurated border. Sooner or later the lymphatic glands, usually the submaxillary, become in- volved. If the floor of the mouth becomes secondarily invaded, the same disposition to visceral metastasis is manifest as when carcinoma is primary m that region. Fig. 235 Carcinoma of the alveolar process: a, lip; h, toothless gum unaffected by the carcinoma. Diagnosis. — When the bone is exposed the lesion is sometimes diag- nosticated as necrosis of the jaw, but the edge of the ulcer is not soft, as in inflammatory affections, and the bone is not affected except where exposed. The broad carcinomas which spread over the hard palate may suggest tuberculosis, but the hard border and irregular outline will lead to the proper diagnosis. Syphilis, too, may be brought in question, particularly when the ulcer involves the soft palate. The syphilitic ulcer has a soft border, is more regular in outline, and fre- quently gives evidence of healing at some point. Besides, there is frequently evidence of previous ulcers in this region or of active syphilis elsewhere. Treatment. — Nowhere else in the body are carcinomas more frequently technically operable with greater certainty of speedy recurrence. If the floor of the mouth or the cheek is invaded, recurrence is certain. If the lower jaw is invaded, a resection of the bone with a complete removal of the adjacent lymphatics is demanded in every case. If the upper jaw is invaded, the alveolar process, the horizontal ramus, in fact, the entire superior maxilla except the orbital plate, together with the glands, should be removed in every instance. When the tumors are small it is not easy to resist the urgent request of the patient for 77 .\i(Jks /)/:i /■:i.ori\(: i i«)\i iiii. nj/ni :;:;7 coiisii \;it ism, ;iiul loo otitii tin sui^ioii foiittiils lumsilt \sitli tlu- iinu)\;il ot I 111- iiKrolai i^roicss ()iil\ ; such ()|-)ti;it ions ;irc- iii\:ii i;il)ly tollowid In nciii 11 luc. Rare Tumors of the Gums. A ^ri;ir v;iiic'r\' of rumors beloiiKiny; to tlu (.•()niU(.ti\ i-rissuc- oi- \;iscul:ii" j^roup wifcii tin- ji;ums. Hem- ;in!i,u)nKis .iiui hhioiis li|>om;is ;iii- sometMiU'S touiul. I In hhroiiKis arc- I-k;. 2^6 frequenth pol\ poul, ol thf sr/i- ot a jit-a or a hean,oic\tn laiji;!.!' (I'lii. 23M, and causi- anno\ aiUH' h\ coniint; m (.-ontacr with the tonuur or with food, or, when hui^e enou«2;h, by hem*!; cau^lit l')etween the teeth. Not mfreciuenrl)-, small tumors formed of dentine, attached to the soft part of the jaw only, are observed. I'hey are probabl)' due to the sirowtli of a cell nest displaced during the development of the teeth. None of these tumors offers any diffi- cult\- in diagnosis, and if b}' their size the\' cause annoyance, simple removal results in a permanent cure. Hard papillnma of rhu jntlarc. TUMORS DEVELOPING FROM THE TEETH Ihe abnormal development of the teeth or dental sacs gives rise to a variety of conditions within the body of the jaw. When these anomahes approach mature development, they produce bony tumors. When the cellular tumors are produced in the latter by secondary processes, c\sts develop. Odontomas. — This term was introduced b}' Broca for tumors result- ing from oN'erproduction of some of the bon\' elements of the teeth. They are hard nodular tumors which raise up the wall of the jaw (Fig. 237). The\' disappear before the teeth are fully developed; but since the\" frequently grow from wisdom teeth, the\" ma\' occur after the period of adolescence. Diagnosis. -The development of odontomas is slow and symptom- less but for the deformity produced. Their nature ma\" be suspected from the slow- growth and the irregularity of outline and from the absence or rudimentary' state of the teeth at the region affected. The proof must depend upon the recognition of the bone-like tumor after its exposure during the operation. 338 TUMORS OF THE JAWS, GUMS, TEETH, PERIOSTEUM, AND BONE Treatment. — The growth should be exposed by the removal of the bon}^ tissue of the jaw covering it and removed. This is frequentl}^ made easy b}^ the presence of a complete capsule. The lining of the cj^st should be taken out with a sharp spoon or rongeur. When inti- mately associated with the alveolar border, odontomas ma}^ frequently be removed through the mouth by means of the cutting forceps. Fig. 237 Odontoma of lower jaw. (Von Bergmann.) Adamantomas. — These tumors grow from embryonal cells associated with the development of the teeth. They are cellular growths, often gland-Hke in structure, and often intermingled with cystic areas. Diagnosis. — They develop during the adolescent period, producing expansion of the jaw, and when they occur in the upper jaw may fill the antrum. Their slow and painless growth suggests their nature, which can be proved only when the tumor is exposed during operation. The complete encapsulation and freedom from attachment to the bone distinguish them from sarcoma. Treatment. — The local removal results in a cure. The capsule, if one exists, must be removed, and if none exists, a portion of adjacent bone should be removed by means of a curette or cutting forceps. Only rarely in neglected cases is it necessar}^ to resect the bone in continuit}^ Odontoid Cysts. — Under this head have been described cysts of the jaw which appear to originate in some developmental anomal}^ of a tooth, usually a permanent one. The cyst walls are formed of epithelial-like cell masses which in their growth expand the jaw and produce a globular, spindle-formed, or lobulated tumor similar to the soHd tumors. They frequently contain a poorly developed tooth, and seem to be a connecting link between the adamantomas and the cysts. Like the solid tumors, the}^ may be confined to the horizontal 77 MfjRs or Tin: I'I.riostu m .i\i> Hfj\j:s ;;;{•) ramus or ni:i\ iinolvi- rlu- iiulii- jaw. I In iliiiuuss ot tlu hoiu- \aric-s in diHiiciii lasis. !• ir(|ucntl\ ii is so ilim that a (.larklin^ soiind IS inoclintd h\ [mssim-. ( )c-(.-asionall\ tlu- lioin wall is l)i(»kfn cnrircly rliiou^h ami tlu (.•oiutius cscapt- into tin- sin louiuiiiiji tissue. Infec- tion of ilu ».\stu- (.-ontints ma\' take place. I In- eavit\' of the cyst, which contains stia w-coloi cd Huid, is l'i((iu(iitl\ di\ided 1)\ hon\' septa. I'.arh in tlu couisi ot" dt\ c lopnieiit neuial^ia-like jiams nia\ he felt in the teeth, hut asuli- iVoni this the only manifestation is the ^raduall)' eI1la^^in^ tumor. When it Incomes ven" larj^e it ma\ press upon the toniiue, or nia\ inriit'ere with the nio\enu-nrs ol the inferior maxilla. Diagnosis. Tin cystic tumors of" the jaw cannot he differentiated from the solid with certainty before operation. C\sts ma\' be sus- pected when the walls are so thinned that a paichnunr-hke crackling is manifest on pressure. From odontomas, they can he distinguished onl\- at operation. From m\eloid sarcomas, they can be distinguished b}- their slow growth. M'he absence of a tooth or the presence of a primary tooth after a permanent tooth should ha\e displaced it points to this class of affection. Treatment. — The older authors ( Heath j advise opening the cysts into the mouth. Present practice seems to lean toward a more radical treatment. The exsection of the walls w hen possible should be practised. At most, a bridge of bone can be left in order to retain the outline of the alveolar process. If cysts are merely curetted, there ma\- be a recurrence, which then will require a radical operation. Periosteal Cysts. — Resembling somewhat the foregoing are the so- called periosteal cysts which occur about the roots of carious teeth in adults. TheA" do not attain the size of odontoid CAsts, but may involve the antrum. Removal of the diseased teeth and curettage of the cavit\' results in a cure. TUMORS OF THE PERIOSTEUM AND BONES Sarcoma. — The jaw is one of the most frequent sites of sarcoma, the upper and lower being affected in about the same proportion. Nine- tenths of the tumors of the jaws are sarcomas. In the lower jaw endosteal tumors are encountered as well as periosteal, while in the upper jaw the former are infrequent because of the thinness of the plates. The endosteal or myelogenous sarcomas of the lower jaw are usually giant-celled tumors exhibiting the reddish structure common to this variety. Their clinical character is dependent on their predominant cell t\pe. By their gradual growth the\- cause a nodular or fusiform 340 TUMORS OF THE JAWS, GUMS, TEETH, PERIOSTEUM, AND BONE enlargement of the bone. The bony covering becomes gradual!}' thinned, so that crackling is produced upon palpation. The teeth become loosened, and either fall out or are easily removed. With continued growth, the bouA' plates are perforated and the tumor extends into the surrounding soft parts. Fungous masses may protrude from the open- ings left bv the extracted teeth, and be subject to injury during mastica- tion and to infection by the buccal flora. The cheek may become invaded and ulcerating crateriform masses produced, which may give rise to troublesome or even fatal hemorrhage. Fig. 238 Fig. 239 Periosteal sarcoma of the lower jaw, Sarcoma developing from the alveolar periosteum. The periosteal sarcoma (Fig. 238) may be composed of an}- cell tA'pe or a mixture of types, and frequenth' there is a large proportion of fibers. This fact has caused man}^ of them to be reported as fibromas ol the jaw, particularly by the older writers, and for this reason the older reports are quite useless for statistical purposes. Notwithstanding the prominence of fibrous tissue, they are very malignant tumors. They form at first a nodular or a spindle-form enlargement upon any part of the surface of the jaw, and ma}" at this stage resemble a periosteal TlMUkS (J I Till. ri:kl()SThl M IM) li()\ES 341 iiiriaiiiniation; their extension to tlu lutks of the teeth and the attend- ing suppuration add to this lesenihlaiui-. I hey usuall\ tend (juickly to in\ade the siiiroiindin^ tissue and to piifoiate the cheek, with the saini' attiiulant phmoiiuiia as are nofiied m the nnelo^tnous t\ pe. Neifhn class tends to early lymphatic iinasion, hut \isceral metastasis m the periostea! t\ pes is fre(pientl\ an eaiiv complication. I suall\ hemoii hajie, pain, inteifeieiice with nutiition, either hy the mjiestion ot putriHed products ot tlu tumor, or h\ mechanical hindrance to mastication or deglutition, reduce the patient to a point where pneu- monia or some other intercurrent disease finds him an eas\' victim. Vv Sarcoma of the antrum. Sarcoma of the superior maxilla may be periosteal (Fig. 239) or may develop within the antrum ( Fig. 2401, the latter being the more common. Often when the disease is advanced the point of origm cannot be made out with certainty. The invasion of the surrounding tissue is often extensive; the orbit and nasal fossa particularly are often attacked. Diagnosis. — Myelogenous sarcoma of the jaw is simulated by many conditions which cause expansion of the bony plates, most frequently' hy odontomas and dentigerous cysts. These benign tumors are indi- cated by slower growth and displacement of teeth. Epulis may cause 342 TUMORS OF THE JAWS, GUMS, TEETH, PERIOSTEUM, AND BONE loosening of the teeth and may extend into the mouth, but when there is expansion of the jaw or severe pain the graver disease is probably present. There is, however, no sharp distinction between epulis and sarcoma; the character of the tumor is the same, the difference being merely one of location. By extension through the bone, the malignant character of the growth is definitely declared. Necrosis of the jaw, tuberculosis, and actinomycosis may resemble sarcoma very closely, and the clinical characters of the affection must always be considered. The periosteal type, because of its exposed posi- tion and more rapid growth, usually permits of an easy diagnosis. In- flammatory affections usually may be differentiated by their history. The more slowly growing sarcomas may be confused with osteomas, especially since this type of sarcoma is frequenth^ attended by the formation of bone. The osteomas, however, are less frequently spindle- form and are of a slower growth. After all means of diagnosis have been exhausted, the surgeon is sometimes compelled to proceed to operation on a probable diagnosis, expecting to vary his operative pro- cedure according to the findings after exposure of the tumor. This is particularl}^ true of tumors suspected to be myelogenous sarcomas. When exposed to view, the reddish solid or cystic sarcomas can easily be differ- entiated from the inflammatory conditions and the dentigerous anomalies. Sarcomas of the superior maxilla less frequently offer difficulties in diagnosis, because usually when patients present themselves for examination the tumor has advanced to such a stage that the diagnosis is clear (Fig. 241). Tumors which go out from the alveolar process with the characteristics of periosteal sarcoma usuall}^ show their char- acter by their rapid growth and by their primary origin from sites other than the dental periosteum. Those which grow within the antrum ma}' bulge and cause a prominence on the cheek, in the orbit, on the nasal wall, or in the nasal cavity, in the last case resembling nasal polyps. When the tumors have extended beyond these limits they present diagnostically neither interest nor difficulty. The soft parts may be infiltrated and ulceration or hemorrhage follow. More fre- quently the orbit ma}^ be invaded, causing dislocation of the eye and perhaps ulceration and blindness. The cranial cavity may be invaded and symptoms of cerebral compression may arise. Treatment. — Opinions as to the treatment of sarcoma of the jaw have not yet been adjusted to the changes in view which have modified the treatment of sarcoma of the long bones, and radical operation is usually advised in all cases. The medullar}^ sarcomas may be sub- jected to conservative treatment provided the surrounding tissue has not been invaded. When this has occurred, however, radical removal is necessary. The periosteal type, including sarcomas of the antrum, 77 MORS or I III: I'E KKJSTIA M IM) liOMlS 343 cUiiKiiul ilu- ic-m()\al of rlic hone. Siiuc tlu l\ in|ih;itu s .m- i.ncly imaiiiil, ilu- iitrc-nrion ot ilu- sinLitoii should he ;;i\in lo adcciuate local iiiiioxal. So (liNiisi- aif tlu- loiiiscs wliuli iiuiiNuluaJ cast-s iiin that {iciuial iiiKs raiiiiol \n- laid ilowii. I siialK iii ihc ia|Mdl\ ^lo\Miif; ptriostial ( \ [hs all tnainunt lads. Sarcoma of the superior maxilla. Chondromas. — Chondromas are found m both upper and lower jaws, but much more frequently m the latter. \ hey may develop as solitarA' growths, together with similar processes elsewhere. They maA' arise from the surface of bones, including the wall of the antrum, or from the medullar}' cavity. The)' develop slowly and expand the bone when the\' grow from the medulla or in the antrum. They ma}' form a diffuse enlargement, but the base is usually more or less con- stricted. 1 he disturbance they cause m the body is purel\' mechanical. Like chondromas elsewhere, their nature is always open to suspicion, and the more so the more rapidly they grow. Frequently, when no evidence of malignanc\' can be found with the microscope, they recur after removal, and in the course of time the recurrences show definite sarconiatous elements. Diagnosis. Their topography being considered, their character can be suspected from their density and rate of growth, but can actualh' be demonstrated onh' on section. 344 TUMORS OF THE JAWS, GUMS, TEETH, PERIOSTEUM, AND BONE Treatment. — Operative removal should be undertaken as soon as they are discovered, because by incision only can their nature be definitely determined; their early removal, because of their doubtful character, is always desirable. Osteoma. — Bone tumors occur in the same situations and present the same evidences as chondromas. They are of slower growth and of an innocent character, and cause mischief solely by pressure. Diagnosis. — Their density, slowness of growth, and location char- acterize them. Inasmuch as bone tissue is frequently present in malig- nant growths, the possibility of malignancy should always be assumed when the tumors develop rapidly, and in such cases early removal is demanded. Treatment. — Local removal is sufficient, though when they are in the antrum or in the centre of the inferior maxilla it is necessary to remove normal bone in order to gain access to them. Fig. 242 Lobulated fibroma of the gums from irritation of a dental plate. (Von Bergmann.) Fibroma. — As stated in the discussion of sarcoma, the malignant tumors frequently have a distinctly fibrous appearance, but true fibromas which form masses extending about the inner border of the gums are sometimes observed. They may follow ill-fitting plates but may develop in gums when the teeth are still present and still normal. Walls of fibrous tissue sometimes have the appearance of a second toothless alveolar process (Fig. 242). Fibromas, both endosteal and periosteal, have been described, but they are infrequent. Careful examination of every part of the tumor, together with an account of the subsequent history of the patient, should be demanded before the diagnosis of fibroma is accepted. Such critical analysis would probably greatly reduce the frequency of these already rare tumors. C II A I' T I', R X X \ I I TiMORs oi' 'iiii-; PAkorii) ANi:) srmiAxii.i.Ain' ci.anos I HI-, tuinois ot tin- p;ii()ri(l and suhmaxillai \ ;;laiuls lia\t- itukIi in coiiinion, rlu' only difference, aside from topo^iraphical relations, heinji that they occur much more frecjuently in the parotid f^land; the same description, therefore, will appl\' to both. The many classes of tumors in these glands, described by the older writers, have in the light of newer studies come to be regarded largely as merely varieties of the mixed tumors. Mixed Tumors. The vast majority of the parotid and submaxillar\- tumors are composed of a variety of tissues, and have very appropriarel\- been called mixed tumors. These tumors occur usuall\' between the twentieth and thirtieth \-ears, equally' in the two sexes and without known cause. Macroscopic Structure. — At first these tumors lie isolated beneath the capsule of the parotid gland. They are more or less perfectly encapsu- lated and are usually globular, sometimes with slight bosselations. As the\- become larger the}' displace the tissue and the bosselations become more prominent. On section, the mixed character of the tumors is usualh' at once apparent. Bluish areas dense to the touch bespeak the presence of cartilage, pinkish semisolid areas indicate m\'xoid tissue, while whitish punctiform or striated areas mark the more cellular areas (Fig. 243), and intermingled with all are bands of connective tissue. The propor- tion of these tissues is variable. The cartilage ma}' be entirelv absent and the myxoid tissue may not be discernible by gross inspection, so that the general appearance is of fibrous tissue with small punctiform areas of white; or these white cellular areas may be much larger and more prominent. These last two types are more common in the sub- maxillary tumors. As more rapid growth takes place the general char- acters change. The surface of the tumor becomes distincth' nodular (Fig. 244), and in very advanced cases the capsule may be perforated and the surrounding tissue invaded. I'he rapid growth finds expres- sion also in a change in structure, which on section becomes apparent in the proportionate increase of the white cellular areas, and often b\' displacement of the myxoid and cartilaginous tissue by a central area 346 TUMORS OF THE PAROTID AND SUBMAXILLARY GLANDS of necrosis (Fig. 245), which may be of considerable extent. It is 3^1- lowish gray in color, and is usually dry and cheesy, but may liquefy and form a cyst. Fig. 243 Mixed tumor of the parotid. Fig. 244 Mixed tumor of the parotid, showing nodules in active proliferation, but the capsule is intact. MicRoscoi'ic .//'/'/■:. n<./.\(:/: oi mim.d tlmors 347 Microscopic Appearance. I lu- i;ii tilauiiioiis ;iiul myxoid elements are usiKilU oln ions ;ii ;i iihiiui iiiulci the nmroscope e\iii wlu-n not :ip)"):irent Ik;. 245 Mixed tumor of rhc parorul, wirh in\H)l\ cmcin ot tin- l\ mph glands. on gross section. The main interest, however, centres about the cellular elements, on account of the difference of opinion which still exists rela- tive to their source and significance. Some hold that the various tissues 348 TUMORS OF THE PAROTID AND SUBMAXILLARY GLANDS are all of similar origin and result from metamorphosis. The cells have been regarded b}^ some as epithelial,^ but the consensus of opinion is in accord with Wilms, that the tissues are of independent origin and that the cells are endothelial. Many German authors, following Volkmann,- regard the mixed tumors as essentially endotheliomas, and classif}' them with this group; but the presence of the cartilaginous and mj^xoid tissues are more than accidental, since the}' often make up the bulk of the tumor, particularly in early stages. The cellular elements are sometimes gland-like in their arrangement, with similar Fig. 246 Mixed tumor of the parotid, showing endothelial areas undergoing active proliferation. cells interspersed in the interstitial tissue (Fig. 246); in other regions the cells have the arrangement of endotheliomas (Fig. 247), and usually areas typical of this class can be found somewhere in the tumor. When rapid development begins these cells become active, often lose the distinct character of endothelioma, and are diffused throughout the connective tissue. The cells then may be larger and contain more 1 Hinsberg, Deutsch. Ztschr. f. Chir., 1899, li, 281. 2 Deutsch. Ztschr. f. Chir., 1895, xli, I. ci.iMc.ii. (:(f{ rsj: or mixed n mors ;j4<) protoplasm. I Ik- ciulotlulioinaroiis ananiicnu-nr is usuall\ iiuiit'ly lost ill tlu- iiutastasc's. Kic. 247 C"''' -."!5'« AV ••• e.>'' ♦•■?••'.•.•-'.. V. • ;. ; . ;v •> • ..: iA Endotheliomatous area of mixed tumor of parotid. Lymph gland metastasis showed similar structure and distinctly sarcomatous areas. Clinical Course. — The development is usually- ver\" slow, these tumors having often existed for ten to fort\- \ears before the patient seeks advice. The growth is often scarcely perceptible from year to year, and during this stage the tumors are but partl\- Hxed and the skin is freely movable over them. The\- appear to be subcutaneous (Fig. 248), but their location within the parotid imparts a definite limitation of motion. In nearly all cases rapid growth begins sooner or later, and is often accompanied b}" dull pain and annoying salivation. With more exten- sive growth, particularly when the surrounding glands are involved, the entire region of the parotid becomes diffusely enlarged (Fig. 249), and no definite outline to the tumor is perceptible. The lobe of the ear is carried upon the summit of the tumors, and involvement of the 350 TUMORS OF THE PAROTID AND SUBMAXILLARY GLANDS facial nerve may become manifest by facial paralysis. Sometimes the tumor itself may show but little evidence of activity, while the lymph glands of the neck are extensively invaded. Distant metastases, particularly in bones, sometimes take place early. The submaxillary tumors follow the same general course as the parotid tumors and present the same physical findings. They project immediately in front and below the angle of the jaw and just behind the facial artery (Fig. 250). Fig. 248 % m J f^^4: ^^^^^^^S|^^^B w 1 n Mixed tumor of the parotid gland. Diagnosis. — The location of a tumor within the capsule of the parotid or submaxillary gland is usually alone sufficient for diagnosis, and a history of long duration with sudden exacerbation and the appearance of nodulations on the tumor is strongly confirmatory. Adenomas forming smooth circumscribed tumors are sometimes seen, but many of these will be found on close examination to be mixed in character. The same may be said of the chondromas. The lymph glands situated over the parotid are sometimes the seat of secondar}/ tumors, par- ticularly melanomas; the identification of these depends largely on the result of a search for the primar}^ growth. The carcinomas are rare and are distinguished by their irregular outline and great density. Treatment. — When presented for treatment before active prolifera- tion has begun, local excision can easily be done with preservation CI.IMC.II. C.Ol RSI: or Ml\l.l> It MORS iiol oi Sttnson's tliK't and thr facial iui\t, rlu- incision l>cinji prc-tt-rahly above the ncixi- and paiallil wirli it. Ilu- tumor can he rcniovcd hv Miint liisscction. Attn acii\t- ^lowili lias bt^iiii tin- prol)liin is less siinpK-. ()tun the hulk dT tin- tumor lu s in the nt lomaxillaiy tossa, and it ma\ he diHiculr to dislodge it. An txammarion h\ palpation thfouuh till- mouth .shouKl al\\a\ .s In- made hetoie the opeiation i.s he;;uii. Fl(.. 24< Mixed tumor of tlif parotid y,laiul. K.\ten.si\t in\ ol\ eiiicnt ot tlie cervical lymph glands. The entile gland must he sacrificed when the growth has become extensive, when the capsule is invaded, and when the l\-niph glands are involved. In such cases it ma\' still be possible to save the facial nerve; it should at least Hrst be located and an attempt made to set it free, although often the tumor will have surrounded it in such a manner that its liberation is either impossible or beyond the technique of the surgeon. A sharp knife applied under full vision will succeed more often than the dull dissection usually recommended. When the nerve 352 TUMORS OF THE PAROTID AND SUBMAXILLARY GLANDS has been liberated, or inadvertentl}' destroyed, the tumor must be removed. When the neck is extensively involved, a block dissection is required sometimes even to the resection of the artery and pneumo- gastric nerve. The submaxillary gland is usually sacrificed in parotid tumors as a matter of convenience. In extensive operations, where the common carotid is not sacrificed, it is often convenient to ligate the external carotid as soon as it is exposed in the dissection before Fig. 250 Mixed tumor of the submaxillary gland. the tumor is dislodged. This step is particularly helpful where the retromaxillar}' fossa is filled with tumor mass. In tumors of the sub- maxillary gland the problem is usualh' much more simple. The tumors of this gland are less prone than the parotid to rapid growth, and when they are presented for operation they are still encapsulated and can be readily shelled out, not from, but together with, their capsule. Ex- tensive operations may be required when the capsule has been invaded or the h^mph glands involved. R.IRI: ri MORS 353 Radical operations upon inixiil tmnois are worth while, for when completely removed the prospiets ot a cure are j^ood, much hertii than with carcinoma or sarcoma of the same extent. Rare Tumors. I umois ot these glands, aside- from riu- mixed rumors, may well he classed as rare. Sarcomas' have heen descrihed, hut the\' prohahl\ represent the malignant stage of the mixed tumors. Carci- nomas undouhtedl) do occur,- usually as dense slowly growing tunK^rs, or more rapidly growing ones with ulceration of the overlying skin. .Adenomas' and chondromas may occur, hut careful examination must he made to exclude mixed tumors. Myomas have been descrihed, hur without douhr leinesent the mixed tumors. Angiomas^ may occur withm the gland. 1 hey must be distinguished from angiomas of the skin covering the gland which is a common site. Lipomas^ within the gland have been described. None of these tumors occurs with suffi- cient frequency to warrant a description or to permit the formulation of a distinct line of treatment. ' Ivaufmann. Arch. f. klin. Chir., iSSi, xxvi, 672. - Micheau, Cancer de la Parotid, I hese de Paris, 1884. ^ Nasse, Arch. f. khn. Chir., 1892, xliv, 233. ^ Lannelogue and .Achard, I raite des Kystes congeniteaux, Paris. 1886. '" Demarquay, Bull, et mem. de la soc. de Chir., 1873, li, 12. 23 CHAPTER XXVIII TUMORS OF THE NECK, INCLUDING TUMORS OF THE THYROID AND PARATHYROID GLANDS, AND CAROTID TUMORS General Conception. — There is no region of the bod}' so fruitful in tumors as the neck, nor one in which tumors demand so careful differentiation. True primary neoplasms are somewhat rare; but con- genital anomalies, branchiogemc CASts, secondary carcinomas, certam obscure diseases of the Ij^mph glands, and inflammations which are characteristic of the neck provide subjects for the broadest laboratory and clinical judgment. PRIMARY TUMORS OF THE LYMPH GLANDS OF THE NECK All primar\^ tumors of the lA^mph glands have much in common, and it is often hard to distinguish them apart; but there are certain differences, both clinical and microscopic, that make a distinction possible. 1 The importance of this matter lies in the fact that the prog- nosis varies with the t^^pe. These tumors, named in the order of their frequenc}', are Hodgkin's disease, lymphosarcoma, and sarcoma of the l3^mph glands. ^ Hodgkin's Disease ("Malignant Lymphoma; Pseudoleukemia). — By this IS usuall}' understood a disease characterized b}' progressive en- largement of the lymph glands, which begins usually in the cervical group and spreads to others. It is attended by general bodil}^ deteriora- tion, which leads to a fatal termination in from two to five years. Its relationship to the true tumors has not been definitely established. There is a gradual transition from this disease to the lA'mphosarcomas, and some authors would class the two diseases together. In the present description they will be considered separately. Its close relation to the tumors is emphasized b}' the fact that Hodgkin's disease is now treated by operative removal. In fact, German writers- designate It by the term malignaiit lymphoma, which indicates at once the seat ^ W. G. MacCallum, Johns Hopkins Hosp. Bull., 1907, xviii, 337. ^ See F. Fischer, Krankheiten der Lymphgefasse, etc., Enke, Stuttgart, 1901, p. 104. I'RiM./k) r I MORS or Tin: i.ymi'ii (.i..im)s or riu: m-ck ;;.'>:. ot the distasf aiul its i h;ii:KtcJ as a riiiiioi. I lusc considciations justif)' its inclusion anion;:, the tumors, althoujih it has had a prominent place in treatises on hematolo^w Etiology. All attempts to estahhsh a hacterial origin for the disease have so far heen unsuccessful. It is lejiaided still h\ some' as tuher- culous m character, though this \ iew is now j^eneralh' ahandoned.- In some ot tin- fatal cases there has heen e\idence of tuberculosis in other parts of the body, notahly in the lungs, and in some of the lymph glands typical tubercles with giant cells have been found. I'he likeli- hood of a double fallacy is, however, very apparent; in the Hrst place, a malignant lymphoma ma\ he the seat of a secondar\- tuberculous focus; and in the second place, a purely tuberculous glandular disease may be regarded as malignant lymphoma.' Twelve cases were care- fully studied by hischer without any evidence of tuberculosis being dis- covered. In some cases various cocci have been found in the blood and lymph glands, but these also are regarded as accidental complications. White and Proescher^ have reported the discover\- of a spirochete. Pathology.- In the early stages the glands are usually soft though their densit\- varies greatly. On section, the capsule is thin; the paren- chyma is pink or pinkish gray, is semitranslucent, and bulges beyond the cut edge of the capsule. At times, however, the capsule becomes involved and the various glands are more closeh" united. This occurs in the absence of a special excitmg cause, such as secondar\' infection or the v-rays. The denser glands are, on section, pale pink, and glisten- mg, fibrous bands traverse the cut surface dividing the scant\' paren- chyma mto lobules, or in some cases the entire surface ma\' appear to be formed b^" fibrous bands. An attempt has been made'' to divide the disease into two forms on the ground of variations m densitA". In rare instances the glands attain a large size and are sclerotic from the first. Such cases are distinct from the ordinary type of the disease, and ma}' well be emphasized b\' a separate classification. On the other hand, the ordinary type varies so much in the degree of fibrosis that it is difficult to maintain a sharp classification. It is a mistake to assume that the small dense glands are of longer development than the large soft ones. The finer structure of these glands varies as greath" as the macro- scopic appearance would lead us to believe. In the large soft glands ' Sternberg, \ erliand. d. deutsche path, (iesellsch., 1904, p. 128. -Dorothy Reed, Johns Hopkins Hosp. Rep., 1902, x, 13s; W. T. Longcope. Hull. .Aver Clin. Lab., Penna. Hosp., Philadelphia. 1903. '■' F. Delafield, .Med. Rec, New "iork, 1887, .x.x.xi, 425. ■* Jour. Amer. Med. .Assoc, 1907, xlix, 1988. ^ Fischer, loc. cit. 356 TUMORS OF THE NECK AND CAROTID TUMORS the lymphoid elements are more numerous and the endotheHal cells covering the septa are increased to a slight extent. The lymph sinuses are filled with mononuclear, polynuclear, and, especially, eosinophilic cells, ^ while the connective tissue is not increased. Finally, there are present large polynuclear cells containing spheroidal particles analo- gous to Russell's bodies within their protoplasm. ^ The characteristic cellular condition is alreadj^ evident in the early stages, the multipli- cation of the pol3^morphonuclear leukocytes, the proliferation of the endothelial cells, and the presence of eosinophils (Fig. 251). Throughout Fig. 251 - d m * ■■{^ "^ Hodgkin's disease: a, endothelial cells; h, eosinophiles; c, lymphocytes; ^, connective tissue. the progressive stages, as the glands become more and more sclerotic, these same cells remain the chief objects in interest. The lymphocytes become relatively less numerous. The endothelial cells become more numerous and also larger and more irregular, and leave their position at the periphery and mingle with the other cells; their large size and irregular form and the number of their nuclei make them very striking objects. In the denser glands the interstitial tissue predominates. Trabeculae extend from the dense capsule into the interior of the gland, present- ing an arrangement similar to the connective tissue m carcinoma. The cell types remain the same, but the epithelioid cells are relatively more numerous and lie in the meshes of the connective tissue. Glands of different degrees of density often co-exist in the same patient; although there is a general tendency to progression from the softer to the more dense, the relation is not necessarily a sequential one. The glands show but little disposition to undergo secondary 1 Goldmann, Cent. f. allg. Pathologie u. path. Anatomic, 1892, iii, 665. 2 Brigidi and Piccoli, Beitr. z. path. Anat. u. z. allg. Path., 1894, xvi, 388. rki.M.iKY II MORS ()i ■/■///■: DM I'll (,i.i\i)s or Tin. m.ck ;;:.7 Hodgkin's disease, imolving the submaxillary glands. Fig. 253 Hodgkin's disease, beginning in the cervical glancb 358 TUMORS OF THE NECK AND CAROTID TUMORS changes. In fact, such changes are specifically denied b}' manj^, but in some instance certainly small yellow necrotic areas may be observed. These areas differ from the caseated masses in tuberculous glands by their canary yellow color and moist surface. Clinical Course. — Usually the submaxillary (Fig. 252) and cervical glands (Fig. 253) are the first affected. Much less commonly the axillary and very rarely the inguinal glands are the primary seat of the disease. The glands on one side are usually enlarged first, followed by involve- ment of corresponding groups of the other side of the body, and then hy extension to other groups (Fig. 254). This sequence is not constant, however, and the disease may remain localized m one set of glands. Fig. 254 Advanced Hodgkin's disease, involving cervical, supraclavicular, and axillary glands. The enlargement proceeds quickly without pain and is usually un- attended in the beginning by any constitutional disturbance. The glands vary in size from a hazelnut to an orange. It is generalh^ stated that the capsule is not infiltrated, and in any exception to this the pro- cess is believed to be inflamrnatory in character. Some observers have come to question this statement. In a disease which is characterized by an increase of the lymphatic elements it is diflficult to prove that when the capsule is invaded by similar cells it is due to inflammatory complications. In some cases at any rate, the surrounding tissue is invaded. Sooner or later, after the primary group is involved, usually a period of months, I'KiM.iK) ri MORS or rill: I.) Mi'ii (.i..i\i)s or riu. xeck :;.■)!> anorluM- f^ioiip In-^ins to cnlar^i-. I liis jii«)U|) iiins a smiilar l)iir hv no means idtiuiial course with rlu- Hisr moup; jr iiia\ t<|iial ir in si/.c or It ma\ soon siii|i;iss it. Siinilailv, otiui ^^loiips rna\ Ixcoini- iinolvcd. 1 lu- simouiulin^ tissue- rt-niaiiis for the- most part iiiiattVcfc-d aiul sutters only by compnssion. I he skin lowrinji; tlic cnlai^ccl j^lands remains unattectcil l\\ rlu- disease, hut may hecoiiu thinned and ^loss\' ' 1* ij^. 255) tiom stretehm^. and the \essels ma\ hccomc |ii omiiunt . 'I'he jjlands may cause trouble b\ compressing; neighboring parts, especiaiU' the bloodvessels and nerves, and likewise the trachea and esophaj^us. Secondary nockik- in Hodgkin's disease, slio\vin 1)\ sli;;htl\ dioppiiii; tlu- (.hiii, in rlu- suhiiKixilhii \ ;in(l cervical jiioup by passively inclininj; the head to the side exaniuu'd. As the nodule ^rows larger it remains dense and often retains a sharp outinu', hiii it is iisualK not ticeU moNahle, heinj^ asso- ciati'd mtimatilv with lui^hhonni; glands. I his loss ot mol)iIit\ occurs before the niali^nanr cells peiforate the capsule, and results troni a reactive piohtiiarion ol the connecti\'e tissue. When the capsule is intiltratctl, which may not occur until rlu- uland is as larj^e as a hickor\- nur or larger, the gland pockets and surrounding tissue become fused into a solid mass (Figs. 264 and 265). It is often difficult to determine whether the mass is formed by a diffuse growth of the carcinoma oi from the coniuctn e-tissue proliferation. I'sualh' the latter is the case. When malignant cells have escaped from the glands small dense nodula- tions can be palpated upon the surface of the mass. While the carci- noma cells are still confined to the glands a more rounded, though often bosselated, mass of characteristic hardness results. The secondarv tumor ma}' serioush' interfere with deglutition or respiration, and the deep nerves may become invaded. Metastases in this region are hable to central degeneration, which tends to gain the surface and result in crateriform openings, which lead to suppurating cavities. These con- ditions are hastened by secondar}' infection from the mouth. The infec- tion may spread to the deeper tissues of the neck, producing suppuration about the esophagus or in the mediastinum, or the necrotic process ma}' open a vessel and cause hemorrhage. Diagnosis. — Usually' the primar}' growth is apparent and the nature of the neck tumor is evident. Rarel}' the primary tumor is small and inaccessible, as in the laii'nx or esophagus; or the primar\' tumor ma\" have been removed jears before, leaving nothing but a scar. When careful search fails to reveal an}' trace of such a tumor, the carcinoma of the neck is probabh' branchiogenic. Prognosis. — In secondai}- carcinoma of the neck, when the primary tumor has been untreated, radical operation ma}' offer some hope of prolongation of life or even cure. When the primary tumor has been removed the disease has been, in my experience, invariabh' fatal. Radical operation possibh' dela}'s the progress sometimes; quite as often, and certainl}- when carcinoma cells have escaped the gland capsule, it hastens the course of the disease. Treatment. — No other problem confronts the surgeon so often as the recent enlargement of cervical glands following a }ear or two after local excision of a small tumor of the lip or tongue. The famih' ph}- sician urges radical operation, and the patient consents. When the growth is limited to the glands and is not spread diffuseh' throughout the tissue nor attached to the carotid sheath, esophagus or trachea, 24 370 TUMORS OF THE NECK the operation is technicalh^ possible. If the removal of the common carotid artery and the deep nerves would be required, or if the esophagus or trachea was involved, the disease is inoperable. If cachexia or gland suppuration is present, the condition would best be let alone. The local use of acetone ma}" lessen the odor from ulceration and perhaps delay the progress of the disease. When the larynx becomes invaded or the trachea compressed, tracheotom}" may be required. When the esophagus becomes occluded, gastrostomy may add a few days to the existence of the patient, though it is questionable in this case if the better treatment is not to use morphine as a diet. Secondary Sarcoma of the Lymph Glands. — Pigmented moles of the temple, scalp, or face, melanomas of the eyes, and more rarely sar- comas situated elsewhere on the head, give rise to secondary' tumors of the Ij^mph glands. In this respect they vary from the sarcom.a t3"pe, though they are usuall}/^ classed as sarcomas. In their course the}' resemble carcinomas more than sarcomas, which recalls the wisdom of Ribbert^ in grouping these tumors separately under the head of mel- anomas. They form nodules within the glands, which are more or less discrete at first, but which fuse later. Usuall}' metastases elsewhere terminate the life of the patient before the growth in the neck causes much discomfort. Diagnosis. — The pigment can often be seen through the unopened skin, and the discovery of the primar}" tumor establishes the diagnosis. The softness of the glands differentiates them from carcinoma. Explora- tory excision should be avoided. Treatment. — Operation is worse than useless, and no means is known that will sta}' the progress of the disease. CYSTIC TUMORS Lymphatic Cysts (Hydrocele of the Neck). — These are congenital dilatations of the lymph channels and occur in those regions where the l}^mph channels are first formed in the embr}^o. They are usuall}^ situ- ated immediately above the clavicle (Fig. 266), but they may be found on the chest or in the axilla. The}^ are soft and }'ielding, their walls being composed merel}" of a delicate endothelium. They may remain nearly stationary or they may grow with some rapidity. The}' are attached to the surrounding tissue, and can usuall}^ be demonstrated to originate deep under the clavicle. They sometimes disappear spontaneoush', the fluid becomes absorbed, Geschwiilstlehre, Cohen, Bonn, 1904. CYSTIC 77 .UOA'.S :i7i aiul rlu- walls colhijisi-. Hicaiisc- of rluir cka-p ;it tachiiunr ;il ciuiic- CDinsf of tin- caniil ..i to ;.n\ i>oi t.oii of ir. TIk-> arc usually found anterior to tlu- sttrnoniastoid (Figs. z(>7 and 268), but wlu-n larjic they occupy tlie entire lateral n-ion of tin- neck. They are usually sniffle, bur mav be nuilrilocular. Ilu- skin over tluni is usuallv unchantitd :,iul frrcl\ iiiovabK. 1 he \ arc- usually more or less rt.xed on the deep aspect to the vessel sheath, rlu- h\ oid bone, or in the region of the styloid process. As above indicated, the\ ma\ be lined with .s(iuamous epi- Fui. z(>7 Branchiogenic c\ sr. thelium when derived from the lower part of the duct or when a trans- formation of the columnar epithelium to this form has taken place in the upper part; or b\ columnar epithelium when derived from the pharyn- geal end of the duct. The presence of 1\ mph follicles in the walls indicates an origin trom the upper portion, no matter what the character of the lining epithehum may be (Fig. 269). When the cysts are lined with squamous epithelium they may contain hair and skin glands. The contents depend upon 374 TUMORS OF THE NECK Fig. 2( Branchiogenic cyst. Fig. 269 \ "y *-'/" Gill cleft cyst: a, area covered with flat epithelium; h, area of squamous epithelium forming papillary projections. CYSriC TLMORS 375 rlu- oiif^iii; wlun liiud l>\ (.oluinnar r|>irlulium tlu-) ait- usikiIIv clomly ami imuiiious. Inn ina\ In- siious; when liiucl h\- cpiclcrnia! tissue then' arc tlciiiionl. I his variation in tontiius has led to classiHca- tion into serous, arluioniatous. anti ciernioiclal t\pes an unfortunate moupinu, since it confuses iheni with tiniiois of (iitterent etioloji;y. rhe\ are likeh to he confused with cysts of other origin, chieHy hniphangionias and c\ sts of ahenant thyroids. From the former they can usuall\' be distin^;uished by their tenser walls, and from the latter In the ditteienr points of attachnient. Aspnarion nia\' be of service. Slo\\l\ dc\ elopinii; abscesses may smuilate them. I he history 1 h3Toglossal cyst. of the case is frequently of service. Wens in this region are distinguished by a closer association with the overlying skin and by the absence of a deep attachment. The diagnosis may be complicated by the presence of other tumors in their walls; lymphomas or lymphangiomas may develop in the adenoid tissue, or the cyst wall ma\- become carcinomatous. Opening and draining the sac is uniformly followed b\' fistula, and removal is often hazardous because of the intimate connection of the wall with the bloodvessels. The latter course, however, should always be undertaken, and if association with the veins is intimate, these may be resected. If the cyst is inseparably attached to the carotid, a por- 376 TUMORS OF THE NECK tion of the cyst wall should be left and the epithelium destroyed, so as to secure obliteration rather than face the mortalit)^ incident to ligation Fig. 271 Wen of the neck Fig. 272 Retro-aural dermoid in man, aged twenty-six years. of this vessel. Cases in which the cyst wall has been removed have resulted in recovery. CYSTIC -nMORS :Vi Thyroglossal Cysts il'ij;. 2701. Nm mln ijiu m l\ tin- mioMiiriiitcd rc'iiKiins ot the t h\ roj»;l()ss;il dint ^i\f iisr to cyst toiiiKirioii. I lu\ In- ahow tlu- h\()icl hone, at ilu- tlooi ot the nioiirli, 01 at rhc base of tin- tonmic. I lu'\ iiuist In- dist mmiisluil lioin ilciinouls aiul lipomas iii tins ictiioii. Wens. SrhatH-oiis i-\ srs an- not iiif r((|iuiirl\ {omid 111 rhc- neck. I heir !j,l()hiihir toiiii ami intiniatt- ulatioii to the skin arc suHicicnt to idciUity t hi 111 I 1' 1^. zyi ). k;. 27 Dernidid of the suhnia\illar\' region. Dermoids. — Both lateral and anterior regions, particularly above the mastoid (Fig. 272) and above the hyoid bone, are sometimes the seat of dermoids (Fig. 273). They usually begin to develop slowly about the time of puberty. Their freedom from attachment to the skin differentiates them from wens, and the lack of attachment to deeper structures prevents confusion with c\ st formation in vestigial ducts. 378 TUMORS OF THE XECK SOLID BENIGN TUMORS OF THE NECK Lipomas. — The neck is one of the chief sites for Hpomas. They appear most frequenth' on the back of the neck (Fig. 274J, not so fre- quenth' on the sides (Figs. 275 and 276), and still less frequently' m front TFig. 277). The}'" appear here, as elsewhere, as soft, lobulated, circumscribed tumors, free both from the skin and from the under- FiG. 274 Fig. 275 Lipoma of the neck. Lipoma of the submaxillary region. lying structures. They are painless unless the}' press upon nerves; when deeph' seated, the}' ma\- cause pain in the area of distribution of the brachial plexus. Their mobility ma}' be limited if the}' are situ- ated under immovable structures. In the lateral regions of the neck they may extend under the sternomastoid and the trapezius, and hence ma}' not be accessible for examination. Diagnosis. — When situated on the side of the neck lipomas may be mistaken for carotid tumors, but their mobilit}' upward serves to differ- entiate them, carotid tumors being immovable in this direction. They SOLID iii:\n;\ ri mors or the ma.k ;i7!» ail- iisualh situaiid luhiiul oi Ixlow rlic (.aiDtuI hitiiicatioii. Their sottiuss IS siirticuiir to clirtciiiuiatc tluni hom tin- various I\ inph ^laiul iii\()I\cnuMrs 111 tins nt^ioii. Deeply seated cysts nia\' |")ieseiit the pseudoHiuruation ot li|)onias, l)ur hponias ate rareh' jilohular, as are c\ sts, ami deeply siatttl lijioiiias iii the ic^ion ot ^dl (.left c\'sts are rare. Lymph c\sts occur in eail\ Idt, and rlu- hhiish loloi ot rlu- conreiirs can otten In- peiceixi-d rliroiiiih rlu- skin. l-'ic. 276 I I.,. Lipoma of the supraclavicular fossa. Lipoma of the submental region. Treatment. — 1 he treatment here, as elsewhere, is removal. The lobulations are likel\- in this region to extend under important struc- tures, and care must be exercised in their removal. Because of the deep cellular tissue open drainage should be employed for a few days. Fibromas.^ — "Rankenneuromas" of the subcutaneous tissue are some- times observed, but do not differ from those situated in the region of the temple. Dense tibrous tumors are sometimes found in the deeper portion ot the neck, most frequently in the lateral and posterior parts. They spring from the deep fascia, and their density and smoothness distinguish them from tumors which they otherwise resemble. Because of their deep attachments extensive operations are required for their removal. Congenital Chondromas. — A number of congenita! tumors composed wholly ot cartilage have been described in the neck.' The\' are with- ' Zahn. \ irchow's Arch. f. path. .Anat., 1889, c.w, 47. 380 TUMORS OF THE NECK out practical significance. Teratomas have been found, but are of theo- retical interest only.^ Tumors of the Hyoid Bone. — Chondromas of the hyoid bone have been reported, though they are very rare.'- A case of this kind operated upon by J. W. Perkins is represented m Fig. 278. Fig. 278 Chondroma of the hyoid bone. TUMORS OF THE THYROID The common enlargements of the thyroid, the goitres, are not true tumors, and an extended consideration of them is not required here. The thyroid gland is, however, subject to true neoplastic formation. Adenomas. — Two classes of adenomas have been described — the fetal and the tubular. The former is congenital, and appears at or soon after birth as a single or multiple nodulation. In structure it resembles the undeveloped gland, being form.ed of tubules with low columnar 1 Pupovac, Arch. f. klin. Chir., 1896, Hii, 59. ^ Spisharny, Deutsch. med. Wochenschr., 1892 xviii, 853. re MORS or riii: rin koid :i8i (.J)!! lulium ami a !aiii,< iiiiinlx i ol MoocK cssi-ls. I lie tiihiihir upi- closeh' itsciiiM( s ilu Ilia! loini in stiiuruic ' 1' in. 279) ami clitters cliiifl\' in its soimw hat ^itarci tciultiu')' to fiiowth.' It forms a solid circunisd ilHcl tumor, iisiiall)' ot slow growth, riic piesc-iicc oi adenomas in tin- th\ loid is iasil\ detected h\ the cxcinsions they make during di«;liititi()li. Ilu\ nia\ lie leadlh enuelcated. Fig. 279 Adenoc\stic thvroid ;iiul luck. oi i\( II hiiakini; down of" tumors, ma\ icsult. '|'lic nerves are ofttii iii\()l\icl i;irl\ ; rlu- iii\:ision of r lie rctiinciis ma\ cause paralysis of tin- cortls, compression of tlie vafii ma\- be expressed as pain in the sromach, and in\()l\ement of the s\inpatheric h\- vascular disturbance ami dilatation of the pupils. As the tunior ^rows larger, compression of the trachea and esophagus or ulceration into them nia\ result. When the Ulterior of the rumor softens there ma\ be fever, though withf)ut an mfection. Metastasis. I lure has been a controversy as to the possibilit\- of a normal gland producing metastasis. It has occurred in glands with slight malignanc\ , and those cases in which metastasis has occurred from supposedly normal glands probably belong to this class. In carcinoma of the thyroid metastasis occurs relativeh" earl\- and in the majority of the cases aftects the lungs first, and ne.xt in frequency- the hver, bones, kidneys, suprarenal bodies, and brain." Metastasis may occur so early that the secondary tumor is regarded as primary. ^ This indicates the importance of suspecting the th3roid whenever obscure epithelial tumors occur in regions where epithelium does not normally occur. Diagnosis.— A struma hitherto quiescent in a person be\ond the age of forty which begins to enlarge or to produce by tension pain which radiates over the neck or to the jaw may, in the absence of inflammatorv condi- tions, be regarded as malignant. Inflammatory conditions may cause similar conditions. Dysphagia from an enlarged thyroid is, according to Liicke, due either to a malignant growth or to thyroiditis. With the further extension of the disease the glands become involved, the various nerve phenomena appear, the skin may become involved, and ulcera- tion occur. Prognosis. — Carcinoma of the thyroid is one of the most malignant forms of the disease. Very few cases have been cured b\- operation (Ewald could quote but foun, but in recent ^ears operative cures seem to be more frequent. When they are left untreated, death occurs in some cases by cachexia and exhaustion, but is more frequentiv due to involvement of the respiratory tract; edema of the glottis, compres- sion or invasion of the trachea, secondary pneumonia or hemorrhage from the erosion of a vessel are also causes of death. Treatment. — Operative removal ma\- be undertaken when the diag- nosis IS made early. The operation at best is a formidable one, and is quite useless when invasion of the surrounding tissue has taken place. \\ hen there is obstruction to respiration, tracheotomy may relieve the ^ Eiselsberg, Arcfi. f klin. Chir., 1893, .\lvi, 430. - Hinterstoi.sser, Billroth's Festschrift, 1892. 25 386 TUMORS OF THE NECK condition. Esophagotomy may be employed when the esophagus is compressed, though rectal feeding is recommended b}'^ Ewald instead of this operation. Gastrostomy would seem a better palliative operation where the esophagus is occluded than any attempt to find the esophagus so near the malignant growth. Sarcoma. — These tumors are uncommon, occurring about one-fourth as frequently as carcinoma (Ewald). Liicke^ states that they occur most frequentl)^ in the j^oung, but this is disputed by Tiffany and Lanier.- They are usuall}/^ of rapid growth and invade early the capsule (Fig. 285) and the surrounding tissue. They produce much the same disturbances in this invasion as were detailed in the discussion of carcinoma. Sar- FiG. 285 Sarcoma of the thyroid gland. comas go out from the interstitial tissue of the gland, and all cell types have been described. The surface is usually smooth, sometimes slightly nodulated; on section, the}' are smooth, white or pink, and ghstening. The treatment is operative removal if the diagnosis can be made early enough. Otherwise, the methods of palliation described under carcinoma may be employed. Rare Tumors. — Fibromas and lipomas have been described. 1 Krankheiten d. Schildruse, Enke, Stuttgart, 1875. 2 Ann. of Surg., 1897, xxvi, 498. C.I Ron I) ri MORS 387 TUMORS OF THE PARATHYROID Bur rlirce cases ot rumors ot rliis i;l;iiul h:i\c- hcen rc-porrt-cL I lu- diajiiiosis IS iiiaclt- ui the hihoi atory.' CAROTID TUMORS riu- (.xisrcnci- of a sj^ccuil r\ pi- ot rumor springing from the carc^tul ghiiul has now been estahhsheti. I he Hrst observation was published b\' Marchand, and was followed the next \ear by a description (jt four cases b\' Paltauf.- Ihe general character of these tumors is constant, but there aie numerous secondary characteristics that van' in the differ- ent cases reported. They have occurred about equalh' between the sexes and in age between the eighteenth and sixtieth ^ears. About three-fourths of the cases have occurred on the left side; which fact has given rise to various suppositions as to etiology, for instance, that the more direct force of the blood stream on this side may have a causal relation. The nature of these tumors is still uncertain. Because of the number and size of the vessels and the nearness of the tumor cells to the vessel walls Marchand classed them as angiosarcomas. He believed that the\' spring from the connective tissue surrounding the capillary loops. Later writers have followed the general disposition of Borst and Ribbert to abandon the designation angiosarcoma, and have classed them with the endotheliomas or peritheliomas. Clinically these tumors, because of their slight disposition to form metastases, correspond very well \vith the endothehomas. Histologicalh' this seems satisfactory, but, as pointed out by Monckeberg, if the cells of these tumors are derived from the cells of the carotid glands, an alignment with the endotheliomas would cause confusion. Pathology. — The tumors reported have varied in size from a pigeon's egg to a good-sized apple, and are irregularh" oblong (hence the name potato tumor). They are firm in consistency, except in a few cases in w^hich the}' are described as soft and elastic. On section they have been described as usually reddish brown, and occasionally grayish yellow. Tneir brow^n color is due to the collection of pigment in the connec- ti\e-tissue spaces. From the relationship of this pigment to extrava- sated blood it has been inferred that it is of hemic origin. As to ' \V. G. MacCallum, Johns Hopkins Hosp. I^ull., 1905, xvi, 87. - Beitr. z. path. Anat., 1892, ii, 260. 388 TUMORS OF THE NECK structure macroscopically, they are homogeneous or show bands of fibers intermingled with granular or spongy areas, and in some of them numerous bloodvessels and spaces are seen on cross-section. Micro- scopically, they are made up of an alveolar arrangement of cells with a small amount of connective tissue, the cells being in close association with the bloodvessels. The cells are large, with from one to six large nuclei, each bearmg one or two nucleoli. The chromatin of the nucleus is grouped about the periphery. The protoplasm of the cells is rather abundant and is sometimes vacuolated, the vacuoles depending, accord- ing to Monckeberg,^ upon the hardening fluid emplo)"ed. He adopts Fig. 286 Carotid tumor: a, internal carotid artery; b, external carotid artery. this view because certain cells in carotid tumors stain yellow in tissue hardened in Miiller's fluid, resembling in this respect cells occurring m the normal gland and in hypernephroma. He also observed that the cells of these tumors react to the various hardening fluids like those of the normal gland. In none of these specimens has an intercellular connective tissue been found. There has been a deviation from this common type in some of the specimens which showed solid columns of cells, and m others in which the connective tissue predominated; some of the specimens presented two or more of these types of arrangement. ^ Die Tumoren der Glandula Carotica, Fischer, Jena, 1905. CIRnrih 77 MORS 389 Symptoms. (. .noiul iuiikhs oihui ;is ili use il.istu' potato-sliapiil tumors III tin hit iiK.it loii (»t tin i.iioikI Iil:,. 2.S()j; that is, at the lc\el ol tlu- h\(»ul hoiu-. Mu\ iisii.ilU .ipiu;ii in front of tht- stfrnoniastoul miisiK ' I' in. 2S7), luit as th(\ hcroiiu hii'^ii th(\ tend to j^row under It. Mu (Atriision IS iipw.iid i.ilh(i ih.m ch i\\ iiw a 1 d . Mn skin co\'cr- injj, thiiii nio\i's tuih and the iiiniois ha\i- a limned hiteral hut no vertical motMht\ ; thi\ ch) not mo\c o\ei tht- underlymfi; tissue. 1 he}' Fig. 2«7 Carond tumor. transmit the pulsation ot the carotid, and m some of the cases reported the tumors moved upward shghtly with each pulsation of the arter\'. They are usualh' dense, but may be so soft as to be mistaken for lipoma, as in Keen's case; they usually grow slowh', resembling in this the mixed tumors of the parotid; at time of operation the duration of the tumors reported had varied between eighteen months and thirt\'-six \-ears. 1 hey are usuall}' painless, but sometimes cause slight pain and slight limitation of the rotation of the head. The}' are alwa^•s isolated and have a more or less distinct capsule. Metastasis was reported in regional ' Keen, Jour. Aiiier. Med. Assoc, 1006, \lvii. 469 and 566. 390 TUMORS OF THE NECK lymph glands in one instance (Monckeberg). Their malignancy is shown by their tendency to invade the surrounding tissue even through the vessel wall. Diagnosis. — Their location, ovoid shape, slow growth, relation to vessels, general consistency, and isolated occurrence are enough to make a presumptive diagnosis. Most of the cases reported were not diagnosticated before operation. In Keen's case the close relation of the tumor with the vessels and the brownish color caused the operator at first to think of an aberrant thyroid, and only the associa- tion with the bloodvessels led to the correct diagnosis. Their rate of growth and isolated occurrence distinguish them from the lymphomas. In addition they seldom involve the Ij^mph glands, but even if thej^ do, their relation to the vessels is quite distinctive. Tumors of the branchial clefts may simulate carotid tumors, but they usually extend nearer the buccal mucous membrane and can be pushed so as to appear to project into the floor of the mouth; however. Keen's case of carotid tumor also projected into the floor of the mouth. They are more globular and less movable than lipomas. Treatment. — Excision should be practised only in the presence of special indications. The deformity is usually what causes the patient to seek relief, less often the slight pain. If the tumor has grown slowly, a con- servative course is to be recommended. Resection of the carotid, the veins, and the vagus, a procedure required in all but a few cases^ makes the operation a formidable one, and it should not be undertaken with- out the patient having full knowledge of the risk incurred. Reclus and Chevassu^ recommend operation only when serious functional trouble or evidence of malignancy is present. Keen concurs heartily with these conclusions. The operation, all are agreed, is facilitated by ligation of the common carotid below the tumor and detaching the tumor from below upward, taking care to avoid the pneumogastric. The superior thyroid and facial arteries require ligation. The tumor mass together with the vessels are then lifted up and the vessels again tied when the tumor has been freed. The mortality from the opera- tion has been 27 per cent., supposedly due to section of the vagus and hemorrhage. When the tumor has been recognized early before the vessels and nerves are extensively involved, immediate operation may be advised. At this time the operative risk is small and the more formidable operation will be avoided.- ^ Rev. de chir., 1903, xxviii, 149, 388. ^ Douglas, Med. Rec, New York, 1909, Ixxv, 397. CM \ r r K R x \ I x IL.MORS OV lllh CUKSr, MKDIAS 1 IM M, llhARl, AND LUNGS TUMORS OF THE CHEST Superficial Tumors of the Chest. I lu- skin ot rlu- chest is siihjccr, with slight niodiHcations, to the same tumors as the skm elsewhere. Carcinoma. In the clavicular region and about the shoulders carcino- mas are observed with relative frequency. 1 hey are slow growmg and tend to spread o\er a considerable area. Growths the size of a palm are not rare, and much larger ones are sometimes observed. 1 hey form ulcers with but slightl\' elevated borders. fhere is some increase m densit}', but it is not so marked as in carcinoma in many other regions. They must be excised and the defect resulting must be covered with grafts. Local recurrences are common. Tuberculous Ulcer. — The tuberculous ulcers are seen more often on the back than on the chest. The\" are superficial spreading ulcers which closel)' resemble carcinoma, but proceed more slowly. They show a tendenc}' to heal in the centre while advancmg about the border. The edge is not infiltrated. Tuberculous lesions of the skin may be secondary to disease of the subjacent bones, especiall}' ot the ribs and sternum, but also ot the spine. Curettage, the cautery, and Finsen ra\s have been used with success. The .v-ra^s ma)' be tried. When the extent of the disease permits, excision and grafting is the method which gives the most rapid results. Syphilis. — About the scapula and clavicle are the regions of predilec- tion for gummas and syphilitic ulcers. The\' give the histor\' ot acute onset, and the}' have the peculiar kidney shape and the soft overhang- ing edges characteristic of syphilis. Signs of the disease elsewhere and the use of potassium iodide soon clear up the diagnosis. Keloids. — The anterior surface of the chest is one of the commonest seats of keloid. They are smooth, elevated, dense tumors, confined to the skin and subcutaneous tissue, are of slow growth, and remain often for years without notable change. They ma\' occur singly or in large numbers. The chief location is about the sternum and between the shoulder-blades, though when multiple they may be scattered over the entire trunk. 392 TUMORS OF THE CHEST, MEDIASTINUM, HEART, AND LUNGS Deep Tumors of the Chest. — Sarcoma. — These tumors occur with relative frequenc}" on the anterior surface of the chest. In the majorit}^ of instances the}^ go out from bone or cartilage, and appear, therefore, as osteo- or chondrosarcomas. They are frequentl}^ preceded by trauma- tism. Their more or less dense bosselated surface, their immobility upon the deeper structures, and the relatively rapid growth are sufficient to characterize them. Lipoma. — Lipomas are relatively infrequent about the chest, but when they occur are found particularly near the clavicle and along the costochondral junction. They are identical m diagnosis and treat- ment with lipomas found elsewhere. Diagnosis. — Tuberculosis and actinomycosis may involve any of the bones in this region and bring up the question of possible sarcoma. They are of slow growth. Actinomycosis may be detected by the discovery of the canary yellow bodies and the demonstration of the ray fungus. Tuberculosis may be detected by the peculiar granulation tissue and by the type of bone involvement. Metastatic tumors from other regions not infrequently find lodge- ment in the ribs and sternum. Syncytiomas and hypernephromas may be discovered in this region before the primary tumor is suspected. Carcinomas of the breast, too, though rarel}^, show early metastasis in these bones where they may present spindle-form enlargements which closely resemble sarcomas. When the primary tumor is hidden, as in small hypernephromas, a diagnosis can be made only by microscopic examination, and even then may be difficult. Aneurysms which appear in this region are most commonly con- nected with the arch of the aorta and exist commonly as distinct tumors after they have eroded the ribs or sternum. They present a smooth surface over which the skin is freely movable, and are fluctuating or semifluctuating. Their location and the transmitted impulse of the heart are ordinarily sufficient for a diagnosis. Sometimes, however, when they are filled with coagulum this may not be noticed. It should be made a rule in observing all tumors in this region to exclude aneurysms by careful physical examination of the heart before instituting treat- ment. Very disastrous consequences have followed the disregard of this rule. Treatment. — Extensive operations are required for the removal of sarcomas, and the prognosis with or without treatment is almost equall}" bad. Extensive resections, even to include a portion of a lung, have been done with at least temporary success. With the use of the newer methods of chest surgery more favorable results may be expected in the future. When metastatic tumors occur in this region little or nothing can be expected from any form of treatment, though if the tumor is the only 77 MORS (II 1111 \! I DI.ISriM M 393 nu't ast asis liimi .i li \ [k i in pli i ( mim, icmox.il nt Imili |)riinai\' aiul sec()iuiai\ Illinois ini^lit ti,i\i- soiiu- piospcii of smccss. l.i|ioiiias ;iie casih slicllcti out. TUMORS OF THE MEDIASTINUM General Conception. I'nmaiv and ,s((.()iulai\ luinois ol the incdias- iiiniin all' amonu llic laiir ntoplasms, hiii ihc h((|ii(iu\ wiili winch iluii possible.' cxistciicc imisr he rciiuiiilHiccl m rlu- consulciation of otiu'i (.liscasfs, notably discast-s of the heart and hloocUessels, ^i\e theni a uieater di<2;nity than tlu\ would coniniami ot their own aecount. Types of Tumor. Sarcoma. Primary sarcoma of the mediastinum may he composed ot any of the cell types, the small-celled variet\' predominatinti;. The point of origin ma\' be connective tissue or, more commonly, the lymphatic l)\i«)iis I thi t ii)iislii|), p;iii(nt.s ;n r f iit|iitiit l\ opiiaticl upon toi (>\;iii;in, list uiihii , and m;iiiim;ii\ tiiniois wlu-n pulmonary coiniilicatioiis ixisr. I In- ic-sult is hut to hastiii rlic- fatal t-nd. 1 lu- lunj; s\ iiiptoms ina\ hv loiiHiud to a sli<;ht coii^h and iiUTcascd it-spiia- r()i\' latc. In nion cxtcnsiNC nnoKt-nuiii, d\ spina and loiali/ici areas ot duliuss nia\ l)i- disioxc ricl. 1 lu- most dcHmtc symptom ot primai\ lun<; niali^nanc\ , rlif dilaicd superficial \'essels, dois not appear ui sicoiulai \ tumors. I he situation is more difficult wlu-n the existeiUH- of a |-»rimai\ tumor is not rt-co^m/A-d. ()\arian sarcomas ma\' produce lunji metastases hefore the existence (^f the pelvic tunu>r is reco • '.V' . K .?.. V.,-.. . .^. ... /ci^'-- • *•••. iN •(• .,•' Fibro-adenoma of the breast, showinji small uland uroups m the mass of connective tissue. 1 lie hbromatous group occurs most fret]ueiitly m young women soon after puberty. The glandular type occurs somewhat latei', usually 400 TUMORS OF THE MAMMARY GLAND between the twenty-fifth and thirty-fifth years. Tumors of this type form globular masses varying from the size of a hazelnut to that of a walnut. They are dense and smooth to the touch, but occasionally are sHghtly nodular. They are fully encapsulated (Fig. 289), and glide about under the examining fingers. The upper quadrant of the breast is the usual location. Patients sometimes complain of a slight burning pain. On section, the fibrous type is pearly or pinkish white and homogeneous, while in the adenomatous type the glandular areas can be distinguished as small pink dots. Microscopic examination reveals large areas of palely stained connective tissue often with myxoid degeneration, interspersed with glandular tissue (Fig. 290). The glands are unchanged or even atrophic in the fibrous type, while in the adenoma- tous group the glands are more numerous and stain more deeply, and often show cA^stic degeneration. In these cases there is a tendency toward papillary formation on the part of the gland epithehum. These changes explain the proneness of the fibro-adenomatous group to undergo malignant change. Occasionally, the fibrous tissue tends to grow into the cysts, giving rise to the so-called intracanalicular fibro-adenoma. These resemble very closely similar formations in the mixed tumors, and may indicate a genetic relation to them. Diagnosis.— Their exquisite encapsulation, smooth surface, and firm- ness suffice to identify them. The slow growth of the tumor and the 3 outh of the patient off'er additional evidence. Often normal gland lobules in the breasts of young women, particularly in those recentl}^ married or subject to sexual excitement, are mistaken for fibro-adenomas. Usually, careful examination will reveal other nodules in the same or the other breast. A diagnostic sign often emphasized consists in placing the flat hand upon the breast. If no tumor is felt in this way the nodule may be regarded as glandular. Occasionally, fibro-adenomas are multiple, in which case they are usually of the intracanalicular variety. They may be difi^erentiated from mixed tumors onl}^ after the latter have undergone rapid enlargement. Their painlessness and definite encapsulation will distinguish them from inflammatory nodules. From carcinoma they are diff'erentiated by their encapsulation. Occasionally, these tumors undergo malignant change. I have examined two such cases in women under thirty, in both of which it was possible to detect the change at the operating table. Treatment. — Removal is advisable in all cases on account of the mental distress they cause the patient and the possibility of malignant change. This is easily accomplished under local anesthesia by splitting the capsule and turning out the tumor. Care should be taken to maintain asepsis, as the deep wound is prone to infection on account of the low resist- ance of the loose fatty tissue. If it is desirable to avoid a scar over the sj:.\ /!./■: I'.iRESciiYM.iroi s ini'Lki koriir 4(H exposed portion of the iiianinia an incision nia\ he made alon^ the told of the lower ed^e of the breast after the method of Warren,' the breast ma\ be elevated and tin rumor excised from behind. This operation is of jireater niaiiiiiriuli-, and sa\i- in txct priDiKil mstanccs wdl nor be st'lecricl. Chronic Interstitial Mastitis. Nor infViMiiu nrl\ intersririal changes in the mamma simulate a tumor. There is an increasing diffuse hard- ness of rlu- cnrire breast, in which inter- stitial tissue is chief!)- concerned, though the «ilandular element shows some in- crease, and c\ st formation is common. This t\pe is most often seen in nervous unmarried women about thirty _\'ears of age. The entire gland is converted into a firm resistant mass (Fig. 291). Usuall\- cysts can be made out with the naked eye. Microscopicalh', the picture closelv resembles fibro-adenoma. Sometimes the lumen of the gland are filled with epithelial cells, which are, how'ever, degenerated forms, and not cells of abnormal development. Histologically, these structures re- semble closeh' some types of senile parenchimatous hypertrophies, but in interstitial mastitis the process is es- sentialh" an interstitial one attended hv cvst formation, though papillarj- proliferation is not unknown. The age of incidence, the extent of the process, and the slight liability to become malignant are the chief reasons for regarding them as a separate group. The breast ma\' require excision, but the removal of the axillary contents and sacrifice of the pectoral muscle are not justified. Senile Parenchymatous Hypertrophy.- — Midway between the pre- viouslv mentioned inflammatory diseases of the breast and carcinoma stands a group of changes which have been designated as senile paren- ch\ matous hypertrophy by Bloodgood.^ This is the most expressive term ^'et suggested for the condition, because the senile changes are observed in the connective tissue and in a part of the glandular tissue, while in other portions epithelial proliferation alone is present. ^ Jour. Amer. Med. Assoc, 1905, xlv, 149. - Schimmelbusch, Arch. f. klin. Cliir., 1892, xliv. 117. ^ Surg., Gynec, and Obst., 1906, iii, 721. 26 Interstitial mastitis involving the entire gland. Numerous cysts can be seen. 402 TUMORS OF THE MAMMARY GLAND The disease occurs most frequently about the menopause, usually between forty and fifty 3^ears of age, but it is sometimes seen in younger women. It is most frequently seen in nulliparas, but not mfrequently in multiparas. It must be regarded as composed of a number of con- ditions; according to Bloodgood, these are phases of a continuous process. Various lesions are present, but their inter-relations are difficult to determine. Bloodgood speaks of an adenomatous (Fig. 292), an ectatic (Fig. 293), and an adenocystic stage. In the first the glandular elements are in- creased so as closely to resemble the adenofibromas. In the ectatic Fig. 292 Adenomatous type of senile parenchymatous hypertrophy: a, gland epithelium; h, basement membrane; c, connective tissue. Stage cysts of varying size (Fig. 294) are formed without marked pro- liferation of the epithelium (Fig. 295), while in the adenocystic stage (Fig. 296) the glandular elements tend to proliferate (Fig. 297). About one-half of these, according to Bloodgood, become malignant. The second stage, it seems to me, represents a benign group with no tendency to malignancy, resembling in this the glandular ovarian tumors; the adenocystic group tends to invade the surrounding tissue rather than to form a secretion, and corresponds to the papillary ovarian cysts. The adenocystic stage of Bloodgood corresponds to Schimmelbusch's type. Fig. 293 Ectatic t3-pe of senile parench}matous h\ [urrrophy. Xmnerous smooth-walled cysts are scattered throughout the eland. Fig. 294 Ectatic stage of senile parenchimatous hypertrophy: c, cyst with slightly proliferated epithelium; b, cyst with unchanged epithelium; a, fibrous papilloma projecting into a cyst. 404 TUMORS OF THE MAMMARY GLAND When malignant changes occur, the epithehal walls of a dilated tubule begin to proliferate and form papillary excrescences, which may fill the entire cyst with newly formed cells. Sometimes the proliferation becomes more active, the surrounding tissue is invaded, and the malig- nant transformation is complete. The crucial point is not the degree of multiplication of the cells, but their relation to the basement membrane; when the malignant stage is ap- ,' proached the basement membrane --' , disappears (Fig. 298). It is impossible to say with what frequency this Fig. 296 •X. \ Fig. 295 fe^«> High-power drawing of cyst, showing the character of the epitheUum. i^ Adenocystic stage of senile parenchy- matous hypertrophy. The larger cyst shows papillary formations in its interior. metamorphosis takes place. In the ectatic type the secretory function, as I have already indicated, is more active than the proliferative, and invasion of the basement membrane rarely occurs. Diagnosis, — Senile parenchymatous hypertrophy differs from adeno- fibroma in being more diffuse and in occurring usually later in life. Interstitial mastitis involves the entire breast and occurs in younger women. On section, these senile breasts are whitish, with pink dots occurring in groups. When the ectatic stage is reached, cysts of vary- ing sizes are scattered throughout a pinkish fibrous-tissue background. Fig. 297 \^!^ '"''"'^cct^-'^i; 7^:='« "^J^ © (?*' ^sge.e.^0 --„-*3^-. s?:^ ^^<^&e Area in senile parenchimatous hypertrophy, in which glands have invaded the sur- rounding connective tissue, indicating that a malignant stage has been reached. 406 TUMORS OF THE MAMMARY GLAND The cysts are usually smooth-walled, but may show papillary excres- cences. Landau believes this type is never associated with carcinoma, in which opinion, as I have already said, I agree. In the adenocystic type the pinkish dots are intermixed with very minute cysts. The cellular element is more abundant proportionately than in the other types. When the malignant stage is reached the pinkish points are replaced by the grayish white of the cancer plugs, and the stroma is much denser on palpation and section. Often an incision into the tumor at the time of operation is necessary in order to determine the diagnosis. In such cases the incision of Warren is to be recommended, since it gives a free access to the entire gland. Many surgeons employ frozen sec- tions at the time of the operation for the purpose of making a diagnosis. It is not uncommon to find nodules of firmer consistency in women approaching the menopause. These are often multiple and frequently bilateral, and are usually well circumscribed; they are not so dense as true senile hypertrophy. These areas disappear if let alone. They seem to correspond to similar states at puberty or from sexual excite- ment, and to be the response to changes incident to the menopause when also other sexual functions are sometimes exaggerated. I have seen this condition present in patients with fibroids of the uterus and disappear when the pelvic tumors were removed. Treatment. — When a nodule is discovered in a senile breast it is a proper object for surgical interference. If the condition is a cystic one the excision of the cyst-bearing area is sufficient. Should the cystic contents be hemorrhagic the complete removal of the breast is the safer procedure. If the adenocystic type is present, the removal of the entire breast is indicated. If areas are discovered which have already under- gone a malignant change, the axilla should be cleared out at once. Tuberculosis. — Tuberculosis of the breast is regarded as a rare disease, but the frequency with which it is found in the laboratory after radical operation makes one believe that the disease deserves more considera- tion than it usually receives. Though occurring most frequently in women between the ages of twenty-five and thirty-five years, it has been observed in older women and rarely in men. The disease may be primary, but is generally secondary to tuberculous foci elsewhere. The danger of unnecessary radical operation make its chnical diagnosis of importance. Forms. — It occurs in several forms: (i) As cold abscesses; (2) as sohtary tubercles; (3) as disseminated nodules; and (4) as conglomer- ated masses. Modes of Infection. — Infection may take place in several ways: (i) By extension from a diseased rib or sternum, or from a tuberculous pleurisy; (2) from lymph glands, particularly those situated at the ri ur.Rci i.ns/s or ////■: hreist MM antciioi ;i\ill;ir\ IxmU i. Korm^' \\;is tin- hist lo cniph;isi/.f this iiiodt- of iiitrction, A\\(\ m.iin icrtiii ohsc-rvcrs sustain his opinion; (jj infec- tion m.i\ lake \A.\cv ihioiii^h the nipple, rhoniih this is rare; '4) it is USiialK t lansinii IihI thioui^h the hhxxl eiiircnt , and main ot the ii-|-)oi'recl c;isi'S ;ire Hist noud tluimLi, hu'iainni. It ahiach |)i(sent, the disease prosiiissis inoie lapulh during tins |>tii()d. Symptoms and Course. I In- course is usuall\ chronic. Dull, aching pain nia\ he the- Hist s\niptoni, or a nodule ina\ he discovered Hist. I' re- quentl\ , a nuinlnr of nodules are jz;rouped in one (juadrant of the ^land, or they may he diffuse. \\\l' nodules are Hrm, not freel\' movable, and moderately tender on pressure. After rlu- initial round-celled iiiHIrration subsides and partial encapsulation takes jilace the mobility is <2;reater. 1 he various nodules may remain discrete, or the\' may coalesce especially in patients of low vitality. After a time the centre of one or more of the nodules may caseate and liquefy. 1 he skin becomes reddened and adherent, and may be perforated so that a permanent fistula is formed. The nodular mass may attain the size of an orange or the breast ma\' be smaller than normal on account of the formation of fibrous tissue. In these cases the nipple is frequenth' retracted, particularly after extensive caseation and fistula formation. In the early stages the fascia is free, but it may become attached later. En- largement of the h'mph glands may precede the mammary disease, and even when not primarily- involved they are soon infected. 1 he}' vary in size from a bean to a hick:or\'-nut. They sometimes become caseated and form fistulas. When the disease has existed for some time there may be an evening rise of temperature. Diagnosis. — When there is evidence of tuberculosis elsewhere he possibility of similar mammary disease should always be borne in mind. It must not be forgotten, however, that malignant disease of the breast nia\' find its chief expression in a slowl}' growing lung metastasis which ma)' be mistaken for a primary tuberculosis. The mammary changes in tuberculosis may resemble carcinoma, but where there is a sinus, its crateriform opening and thin watery discharge are sufiiciently characteristic to declare the tuberculous nature of the affection. In the conglomerate t\'pe there is a uniform moderate hardening of the mammar)' tissue with a varj^ing degree of pain, and the patient is usuall\' a young woman. The mass does not attain the hardness characteristic of carcinoma nor do the axillary glands become so hard. In women of middle life, not the subject of demonstrable tuberculosis elsewhere, a macroscopic section of the tumor during the course of the operation may be necessar\" for diagnosis. The presence of caseated areas and the ' Lehrbuch. d. spec. Cliir., I Iirschwald, Ht-rlin, 1893, '^'"l- !•• 408 TUMORS OF THE MAMMARY GLAND absence of the characteristic unit lesion of carcinoma are sufficient to distinguish the two diseases. A microscopic diagnosis from a frozen section will rarely be needed. The co-existence of the two diseases has been reported. ^ Tuberculosis may be distinguished from actinomycosis by the absence of the characteristic canary yellow granules and of the indurated yellow skin. Fibro-adenomas are usually single and clearly encapsulated, and there is no glandular involvement. Interstitial mastitis occurs in more rugged neurotic young women, and there is no glandular involvement. The entire breast is enlarged and more dense than in tuberculosis, and the cysts are more tender than the softened areas in tuberculosis. Aspiration is to be practised in case of doubt. On section, the tubercu- lous areas are whitish or grayish, with usually cheesy areas in the centre. The newer tests, Moro's and Calmette's, are too recent to permit of a proper estimate of their value. The bacilli may sometimes be demon- strated in the pus, though the search has frequently been fruitless even in the excised gland; or inoculation of guinea-pigs may be necessary to prove the presence of the disease. Prognosis. — When the disease is primary the prognosis is good. When secondary, the prognosis depends upon that of the primary lesion. Treatment. — When an abscess is formed without extensive infiltra- tion, incision, curettage, and drainage (Warren), or incision and tampon- age with iodoform-glycerin or iodoform-ether may be employed. When there is much infiltration, excision will give satisfactory results. Rodman emphasizes the fact that since this disease usually occurs in young women, the gland should be dealt with as conservatively as possible, because of the decreased chances for matrimony after complete removal. When extensively involved, the entire gland may require removal, but the radical operation as employed in carcinoma is never justified. Mixed Tumors.2 — Scattered through the literature are a great variety of terms applied to tumors of the mammary gland, such as cystic sar- coma, adenosarcoma, cystosarcoma, sarcoma myxomatodes, chondro- sarcoma — all indicating combinations of epithelial and connective tissue. The researches of Wilms have done much to simplify the conception and terminology of these tumors, which, it has long been known, have clinically much in common. Wilms groups these heterogeneous tumors together under the term "mixed tumors." He regards their formation as due to some dis- placement of tissue in the anlage of the mammary gland. He calls 1 Warthin, Amer. Jour. Med. Sci., 1899, cxviii, 25. 2 Periductal Myxoma, Warren, Jour. Amer. Med. Assoc, 1905, xlv, 149. MIXED r I MORS or Till: M.I MM. IKY (.L.I SI) 409 iittciition to tin- tacr that tin- mammaiy ^land is not foiiiicd from the skin as such, hut that the skin covering the ^land and the j^land itself are derived from thi- primary epihhistic hiyer, a point which is certainly well taken. In other words, at the time the anla^e of the ^htnd appears, the skin is not yet formed, and a displacement of epihhistic cells is likelv to include cells which would become glandular and also cells which Fig. 299 A- • ^ -,;,?^i^.' *•-.:- ,'- /I 4.. #3^//: ;-.?^^ Section of a mixed tumor of the breast. The fibrous tissue forms papillary projections within the epithelium-lined spaces. would become epidermal. When this displaced tissue begins to pro- liferate, columnar and squamous epithelial cells ma}' develop side by side. These may form cystic spaces if the cells produce a secretion, or fibrous tissue ma}- form papillary masses which fill the cysts and produce an intracanalicular fibro-adenoma or, if the connective tissue is more cellular, a sarcoma phyllodes (Fig. 299) of the older writers. The connective tissue, being also embryonal, continues to produce tissue which is more or less primitive in character. In this way m\xoid 410 TUMORS OF THE MAMMARY GLAND and sarcoma-like tissue is produced, and each of these may approach more or less the structure of adult connective tissue. All of these tissues may be present in varying amounts and with them may be intermingled glands more or less typical, all combining to make a most varied picture. Not infrequently areas typically endotheliomatous fFig. 300) are observed, though not with the same regularit}' nor in the same degree as_in the mixed tumors of the parotid. Fig. 300 e;i^ * ^^' ~^ < /,; ; f " ; .^ ' '• '3.^^ Endotheliomatous area from a mixed tumor of the breast. This hypothesis seems to dispose of most of the difficulties which have heretofore attended the attempt to classify these tumors. They correspond in general to the mixed tumors in being benign, notwith- standing their apparently sarcomatous structure. The fact that they are separated sharply from the surrounding tissue has been thought to indicate that they are derived from structures independent of the gland. In many of these tumors, areas which resemble the intra- canalicular adenofibromas are seen, and indicate a possible relation- ship between the two groups of tumors. Clinical Course. — Mixed tumors of the mammary gland are usually seen in adult life, cases having been reported in patients from twenty- eight to seventy-two j^ears of age; and in some instances small tumors Cl.l.MC.Il. C.Ol KSli Ul n MORS 1)1 TIU. M.IMM.IKY I.L.I \ I) 111 Iki\c- Ihcii ()I)sci\ icI cvtii earlier. \\ Ikii hist noticed rhev are painless nodules, and after a period var\in^ from one to fifteen or more \ears, they take on a more rapiil growth. rsiiall\ when the patient seeks advice the tumor has attained some si/e, is re^uhir ( Fij^. ^OIj or hos- selated ( Ki^. ^02) in outhne, and of varvinji; consistencw Tense, elastic, ami cystic areas are interminuled. The rumors are not painful, are heeiy mo\ahK. and the a.\inar\ l\rnphatics remain unaffected. Like Fk;. 101 Mixed tumor of the breast, showing nodulations of the mass and a cyst immediately below the nipple. the mixed tumors of the parotid, the\' may exist without increase in size for many \ears or may enlarge ver^- slowly. Sooner or later x\\ty begin a sudden growth, infiltrate the surrounding tissue, and assume the cluneal characters of a sarcoma or endothelioma. Diagnosis.— They resemble in form the cellular t\pe of carcinoma, but the varying consistency and particularl\- the complete encapsula- tion are usually sufficient to distinguish the mixed tumors. The absence of lymphatic enlargement and the usual history of slow growth are 412 TUMORS OF THE MAMMARY GLAND other signs which emphasize the diagnosis. If doubt exists an incision into the tumor at the time of operation will clear up the matter. The mixture of pink homogeneous myxoid tissue, with glandular and cystic dilatations, and particularly cartilage, if present, makes a picture never seen in carcmoma. In the light of the studies of Wilms, the relation of these tumors to sarcomas must be considered anew; in fact, nearly all the so-called sarcomas of the breast should be classed among the mixed tumors. This being the case, the number of true sarcomas of the breast becomes very small, too small, in fact, to permit us to form a definite clinical picture. So far is this statement true, that all the tumors previously Fig. 302 Mixed tumor of the breast, showing the nipple carried on the summit. reported as sarcomas, which have been subjected to reexammation, have been found to be mixed tumors. Usually some portion of the rapidly growing tumor shows evidence of an origin within a mixed tumor. Cysts are likely to be present, and even when absent, spaces filled by newgrowth of tumor masses can be made out. These are the sarcoma phyllodes of the older writers. Treatment. — Since mixed tumors do not affect the glands and metas- tasis has not been observed, they are benign tumors, and simple shelling out is sufficient treatment. Like other mixed tumors, they may undergo malignant change, and thorough prophylactic removal should always be practised. After evidence of malignant change appears, the treatment is that of sarcoma and endothelioma. R.I RE TV MORS Of Till: HRE.IST 41 :^ Rare Tumors of the Breast. Lipoma. LiponKi of the breast is an Lincomiiu)!! artccrion. latry tmnors are observed about rbe axilla, about the outer border of the fi,laiul, and below the clavicle, but these should not be classed as niaiiiniar\' lipomas. Ihe gland in man\' instances may be compost-d hirll()\M(l h\ c;in-iii()iii;i ot the hicast. It consists 111 ;in cc/AiiKitous inH;inHii;iti()n about tin- nipi^U- which he compared to an acute hahinitis. I lie surface discharges a stick\- \'iscid Huid, which may dr\' into scales and crusts, not unlike the scales in psoriasis.' This process gradually extends beyond the nipple. The skin soon becomes thickened and hard. Many breasts so diseased are subject to carcinoma, which is most often of the scirrhous type, though other types have been observed. I'he older opinion was that the initial Fig. 3 1 1 Fungating elandular carcinoma of the inainma. skin lesion was an eczema, but later observers- believe that the carci- noma is primary and that the skin affection is the secondary process. The argument pro and con may be summarized as follows: In favor of the primary nature of the skin affection may be mentioned: (i) The disease ma}' exist man}' years before a carcinoma can be demon- strated; (2) similar lesions are found on the penis, scrotum, vulva, ^ Shields, A Clinical Treatise on the Diseases of the Breast, Macmillan, London, 1898. - Jacobaeus, Virchow's Arch. f. path. Anat., 1904, clxxviii, 124; Schambacher, Deutsch. Ztschr. f. Chir., 1905, Ixxx, 332. 424 TUMORS OF THE MAMMARY GLAND and umbilicus;! (-^) ^he presence of peculiar cells in the epidermis; (4) the apparent direct continuation of the epidermal cells into the carci- noma. In opposition to the primary nature of the affection is: (i) The demonstration of early involvement of the cells Hning the ducts (Fig. 313); (2) the tumors when first discovered are deep and may retain a Fig. 312 's^K^'^y' --— ^ „4 !$i'r'^ . -"^§<« ^'-vj':.-'- :. -%>. 'feists „'»- °- __ ■ - . ^°'/y -s.-;- f<=£'" 0^^'i°~'?-° .'/,'/ = «■ ■» §_ <>OOa,'» '^'a 'j t'„"„>f'" Is^^f/""-,?;" ". °%sl- e'^'^W"*' ''V - »»»." '<--■• . »°, "■»• -"«o^"« r '^ .*' « ° • o^.#'^ ^''^ ^ ' <" „° *' f. t? ^^^ ^"-^ o'i- \^ Acute encephaloid carcinoma of the mamma : a, milk duct; b, cancer nests; c, round- celled infiltration; d, normal lactating breast. gland-like structure; (3) pearls are never formed as is usual in skm carcinomas; (4) the Paget's cells ma}^ result from metamorphosis of epidermal cells from the more deeply lying carcinoma; (5) eczema of the nipple may exist without the development of the carcinoma. The ^ Hannemiiller and Landois, Beitr. z. klin. Chir.. 1908, Ix, 296. C.lkCIXOM.I or TIIK BREI^T 425 interest m ilu- lii.st()l()ii\ ot xhv ilisrasc iHiitits iihout rlu- so-calKcl I\ifi;et s ctlls. I lusr il'ii;. 1I4) inc lar^f, |Kilc-sr:iiniii}i, cells sunouiulccl by a sharp honlri. I iuii- sijiniHcancf is not unclcTsto(Kl. Jacobacus believes Fig. 313 '• < ■-.•>■ .'i'»i^**t'T'-i^'^ *-J\.'i' • ■■■■' - .'i^A ■;.*ii ,-*'.;- .. ' . ■: - .... ■;.- ,- ■.-.■.'.p ^ , •t* , »2' ■■ ■■-'■■' . ■- ■■■■-.-- ., • •• '.-v/ »<•'* •• ; V; •••■•.-.■,:-.:•' • .'• . .■ • ■ -■- , .. , • .• ■ -'••:•■»/''.■/' ^/*',i ,'^' rrolitVriirit)!! of rht- tiiirhcliuni in the lacriftmus ducrs in Paget's diseast ot rht- nijiple. that they are carcinoma cells derived from the ducts and extending upward into the epidermis, and this, indeed, seems the reasonable expla- nation. The chief argument against the ductal origin of the disease is that similar processes are observed elsewhere, notably about the scrotum. The disease usually occurs in women beyond middle life tic. 314 who are married or have borne _,.^ -, " children, though it has been ^r^]&^ ^" ■ ^^ seen in virgins and in men. It v? *^ -^^ \^ -'.? %':^ is essentially a chronic process, ; '- M <-> ^ ^^ > ^ > '^ though a rapid course is possi- ^- \, -^ --■=' ^^ ' ^ '^ ^ ^J'^'^'^ ble. 1 he clinical course and '^^ i? s? '^ ■ ^ ^>» v'"?.» general diagnostic signs are ^V V --" /^->- ff ^'^^iP-S^i. those of ordinary carcinoma. ^'^ X-f^^'^ ItA^ ^i^^- "^r^^ It is generally admitted that (^^-■■.rA^--^-'^^^ S^ifh' ^^ y.-^^^' there is a true eczema ot the ^•v*^*'.*^ 'o ^ 0^0 nipple which is not raget s disease andwhich isnot followed ^^^^.^^ ^^ ^^-^^ -^ p^g^^.^ ^-^^^^^ ^f ^,^^ „i,,p,^^ b)- carcinoma, but no observa- showing (a) the so-called Paget's cells, tions are recorded which ma}' guide to a differentiation between those likeh' to be followed b}- carci- noma and those which are not. Chronic disease of the nipple, there- fore, presenting the clinical features of Paget's disease should be sub- jected to microscopic examination. 426 TUMORS OF THE MAMMARY GLAND Secondary Degeneration. — Secondary degenerative changes in mammary carcinomas are not frequent. Colloid degeneration may occur in the epithelium or in the connective tissue (Fig. 315). The tumor is then semisolid and the cut surface is jelly-like. Areas, particularly about the border, retain the carcinomatous structure. Hemorrhage occa- sionally takes place into the colloid area (Fig. 316) and may cause rapid enlargement of the breast and lead to secondary cyst formation. Fig. 315 Fig. 316 ^1 Colloid carcinoma of the mamma. Hemorrhage into a colloid cyst of the mamma. Colloid degeneration is especially likely to take place in metastases. Microscopically, colloid material is seen within or surrounding the cells. These changes within the cells are believed to account for the relative benignancy of these tumors. Fatty degeneration occurs in mammary carcinomas and is manifest macroscopically by small canary yellow areas, and microscopically by detritus and fat globules within the meshes of connective tissue. Calcareous infiltration may take i:,iR(:i\()M.i oi Tiir. nRii.isr 427 placf ill siK-h ureas ( I'i^. 317). CMosil) assoclatici \\\\\\ farr\ ckmiuia- rion is atioplu'. I.ar^e areas of Hlnous tissue are seen in breast carci- noma, and it has Inen assumed that the epithehum has disappeared by the process ol" tatt\ defeneration. Ihis was heheved to lessen the rate of