COLUMBIA LIBRARIES OFFSITE HEALTH SCIENCES STANDARD HX00058815 Columbia Umberfiittp tntyeCitpof Jftetogorfe ^ j College of ^fjpsicians ano burgeons itibrarp ^r THE fp O LIBRARIES ■£ ^ 3 HEALTH SCIENCES LIBRARY Digitized by the Internet Archive in 2010 with funding from Open Knowledge Commons http://www.archive.org/details/gynecologicalobs1 1 norr GYNECOLOGICAL AND OBSTETRICAL TUBERCU LOSIS GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS BY CHARLES C. NORRIS, M.D. ASSOCIATE IN GYNECOLOGY, UNIVERSITY OF PENNSYLVANIA SCHOOL OF MEDICINE; ASSISTANT PROFESSOR IN GYNECOLOGY- OBSTETRICS, GRADUATE SCHOOL OF MEDICINE, UNIVERSITY OF PENNSYLVANIA; ASSISTANT GYNECOLOGIST, HOSPITAL OF THE UNIVERSITY OF PENNSYLVANIA; GYNECOLOGIST TO THE CHILDREN'S HOSPITAL, PHILADELPHIA; CONSULT- ING GYNECOLOGIST AND OBSTETRICIAN, HENRY PHIPPS INSTITUTE OF THE UNIVERSITY OF PENNSYLVANIA GYNECOLOGICAL AXD OBSTETRICAL MONOGRAPHS D. APPLETON AND COMPANY NEW YORK LONDON 1921 COPYRIGHT, 19 21, BY D. APPLETON AND COMPANY PRINTED IN THE UNITED STATES OF AMERICA . PREFACE The current literature contains exhaustive references to the various forms of tuberculosis which are of especial interest to the gynecologist and obstetrician. However, few monographs dealing exclusively with the subject have been written. Pulmonary tuberculosis is one of the most frequent diseases, and when present in the pregnant woman has a definite bearing on the ultimate outcome of the case. Tuberculosis of the female genital tract and peritoneum is of fre- quent occurrence and is usually secondary to tuberculosis elsewhere in the body. Not only do these cases exhibit characteristics requiring special local treatment but, due to the fact that other foci of tubercu- losis are generally present elsewhere, often in the lungs, particular care should be exercised in the choice of an anesthetic and during con- valescence following any surgical procedure. Furthermore many surgical patients who are suffering from some non-tuberculous pelvic lesion are the incumbents of pulmonary tuberculosis and, therefore, require special safe- guards, both during and following operation. As has been stated, refer- ences to tuberculosis, as it has bearing upon gynecologic and obstetric practices, are rife in the current literature, but the space devoted to this subject in the text books is often extremely brief. For these reasons an attempt has been made to present the entire subject in one volume. No effort has been made to utilize all the literature bearing on the text contents in the present monograph. An endeavor has, however, been made to incorporate references to the more important articles bearing upon the various subjects. Some of the material employed in the Chapter on Pulmonary Tuber- culosis and Pregnancy has been previously utilized in an article which appeared in the American Journal of Obstetrics. A part of the material used in the chapter on Congenital and Placental Tuberculosis has previ- ously appeared in the Transactions of the American Gynecological ociety. In both instances the material has been added to and brought up to date. I wish to acknowledge my indebtedness to the following authorities vi PREFACE for much of the material utilized in the preparation of the historic review : Sir William Osier, Waldenberg, Predohl, Johne, and especially to Dr. J. Whitridge Williams, whose valuable monograph on Tubercu- losis of the Female Generative Organs has been extensively drawn upon. Charles C. Norris PHILADELPHIA CONTENTS I. Historical Sketch I Early recognition of tuberculosis, i — Varieties of the disease, i — Atrophic, cachectic and ulcerative, i — Practical knowledge of certain features of tuberculosis held by pre-Hippocratic writers, i — History of the disease reviewed by Osier, and other illustrious members of the medical pro- fession, i — Connection between tuberculosis nodes and phthisis first indicated by Sylvius (1695), 2 — Baillie (1793) first to recognize tuber- culosis in organs other than lungs, 2 — Laennec, originator of the stethoscope recognized unity of scrofulous nodes and phthisis, 2 — Open air treatment recommended by Samuel Morton (1834), 2 — First successful inoculation by Klencke (1843), 2 — Various views regarding etiology of this condition, 2 — Valuable work of Furnival (1842) and others, 2 — Tubercle bacillus discovered by Koch, 3 — Genital tuberculosis first recognized by Morgagni, 3 — Historical review of genital tubercu- losis, 4. II. The Diagnosis of Tuberculosis of the Female Genital Tract by Lab- oratory Methods 6 Cervix and lower genital tract, 6 — Methods of treatment, 7 — Curettage, 7 — Value of examination of leukorrheal discharge in tuberculous endometritis, 7 — Organisms likely to be mistaken for the tubercle bacillus, 8 — Smegma bacilli, 8 — Study of morphology and of staining by ordinary methods, 9 — Grethe methods, 9— Czaplewski method, 9 — Etiology of smegma bacilli, 9 — Bacillus leprae, 10 — Resemblance to tubercle bacillus, 10 — Differentiation, 11 — Cultural methods and clinical study, 11 — Animal inoculation almost positively diagnostic, 11 — Danger of mistaking malignant neoplasms for certain forms of tuberculosis, 12 — Diagnostic use of tuberculin in gynecological conditions, 12 — Sum- mary of histologic examination, 12. III. Pathology 15 Two distinct forms of genital tuberculosis, ulcerative and hypertrophic, 16 — Histologic examination of ulcerative form, 16 — Hypertrophic variety demonstrated by staining or inoculation, 17 — Tuberculosis of the vagina, 18 — Ulcerative form the most frequent variety, 18 — Histologic examination, 19 — Hypertrophic form in relation to miliary tuberculosis, 19 — Tuberculosis of the cervix, ulcerative, papillary, miliary, and interstitial, 20 — Histologic examination, 22 — Character- istics and differentiation, 23 — Corporeal endometritis, miliary and caseous, 23 — Study of histologic and pathologic characteristics, 25 — Myometritis frequent occurrence in advanced cases of tuberculous endometritis, 26— -Infections of the endometrium, 27 — Intramural ab- scess, 27 — Tuberculous deciduitis, 30 — Histologic examination, 30 — Placental tuberculosis, 31 — Macroscopic appearance, characteristics and forms, 31 — Intravillous tuberculosis, 33 — Intravascular chorionic lesions and chorio-amniotic, 34 — Tuberculosis of fallopian tubes, 34 — Tuberculosis of the ovary, 41 — Peri-oophoritis and oophoritis, 41 — Histologic examination, 42. IV. Congenital and Placental Tuberculosis 44 Placental transmission of tuberculosis, 44— Conflicting reports of find- ings, 44 — Types, acute, chronic, 45 — Errors in technic, 45 — Definition of congenital tuberculosis, 45 — Discrimination between congenital inf^c- vii viii CONTENTS CHAPTER PAGE tion and congenital predisposition, 45 — Etiology, 46 — Germinative in- fection: Spermatozoic, 46— Variety of infection, 47 — Experiments of Waldstein and Ekler, 47 — Observations of medical experts, 47 — Un- fertilized ovum, 49 — Ovarian infection and germinal transmission of disease, 49 — Congenital germinative tuberculosis, 49 — Placental and fetal tuberculosis, 50 — Opinion of Baumgarten and others, 51 — Tubercle bacilli in the blood stream, 51 — Views of Delore and other investi- gators, 51 — Infarcts described by Williams, 53 — Criticism, 60 — Period at which intra-uterine transmission occurs, 61 — Predisposing factors to placental or congenital tuberculosis, 62 — Undoubted cases, 73 — Anatomical changes and presence of tubercle bacilli, 81 — Histologic changes and presence of tubercle bacilli, 82 — Demonstration of bacilli by staining or by inoculation of animals, 83 — Conclusions, 85. V. Routes of Infection in Genital Tuberculosis 95 Primary genital tuberculosis, 95 — Modes of infection, 96 — History of cases, 97 — Relative infrequency in women, 98 — Analysis of literature, 98 — Summary of experiments, 99 — Clinical proofs, 101 — Secondary genital tuberculosis, 101 — Latency of the disease, 102 — Determination of source of infection, 102 — Difference of opinion regarding frequency of primary and secondary infections of female genital tract, 103 — Study of cases, 104 — Summary, 105 — Predisposing causes, 105 — Fre- quency, 105 — Histologic examination, 105. VI. Tuberculosis of the External Genitalia 109 Etiology, 109 — Possibility of hematogenic or lymphogenic infection, 109 — Causes,' 109 — Frequency, 109 — Combined statistics of many in- vestigators, no — Varieties, no — Forms, ulcerative and hypertrophic, no — Symptoms, no — Number of cases; average age, in — Relative infrequency of direct "inoculation in this locality, 112 — Parturition as causative agent, 112 — Trauma a predisposing factor, 112 — History of cases, 112 — Appearance of ulcerative variety, 112 — Hypertrophic variety, 112 — Tabulation of parts involved, 112— Diagnosis, 112 — Prog- nosis, 117 — Method of treatment, 117 — Primary variety, 117 — Secondary, 118 — Doubtful cases, 118 — General treatment, 119 — Tuberculous non- ulcerative hypertrophy of the vulva, 119 — Histologic examination, 120 — Tuberculosis of Bartholin's gland, 121 — Tuberculous ulcers of labia majora and minora, 122 — Histologic examination, 123 — Study of cases, 124 — Primary tuberculosis of vulva with elephantiasis of clitoris, 127 — Secondary hypertrophic tuberculosis of vulva, 128 — Reports of cases, 129. VII. Tuberculosis of the Vagina 140 First authentic case of vaginal tuberculosis recorded, 140 — Anatomic relationship existing between external genitalia and vagina, 140 — Histologic similarity, 140 — Etiology, 140 — Symptoms, 141 — Experimen- tation tending to show that trauma and irritation are important pre- disposing factors in implantation form, 141 — Varieties, 142 — Ulcerative appearance, 142 — Miliary form, 142 — Hypertrophic, 142 — Characteris- tics, 143 — Syphilis, malignant neoplasms, chancroid, gonorrhea, noma and diphtheria differentiated, 143 — Cases cited, 144 — Primary tuberculo- sis of vagina and vulva, 145 — Histologic examination, 145 — Cases col- lected by Chaton and others, 146. VIII. Tuberculosis of the Cervix 149 Cases proved by histologic or bacteriologic examinations, 149 — Primary and secondary, 150 — Cases on record, 150 — Coincident tuberculosis of other parts of genital tract, 150 — Tuberculous salpingitis with or with- out involvement of the corporeal endometrium a common accompani- ment, 150 — Predisposing causes, 151 — Analysis of cases verified by histologic or bacteriologic examination, 152 — Average age arranged in decades, 152 — Classification of cervical lesions, 154 — Ulcerative, papil- CONTENTS ix CHAPTER p AGE lary, miliary, and interstitial, 154 — Analysis of cases, 154 — Hemorrhage, 154 — Pain, 154 — Histologic examination, 156 — Cases, 160 — Tuberculosis of the body of the uterus, 182 — Endometritis, 182. IX. Tuberculosis of the Fallopian Tubes and Ovaries 192 General considerations, 192 — Fallopian tubes and ovaries anatomically and symptomatically considered together, 192 — Predisposition, 193 — Routes of transmission, 193 — Histologic examination, 193 — Factors, 195 — Analysis of cases, 196— Study of acute and chronic stages, 198— Duration of acute stage, 199 — Characteristics of chronic stage, 200 — Other forms of infection, 202 — Tuberculin an aid to diagnosis, 206 — Differential diagnosis between tuberculous, gonococcal, streptococcal and inflammatory disease, 207 — Family history, 210 — Prognosis, 210 — Cases, 211 — Methods of treatment, 214. X. Unusual Manifestations and Remote Complications 224 Tuberculosis and neoplasms, 224 — Ways of occurrence, 224 — Etiologic relation to cancer, 225 — Histologic similarity of certain forms of tuberculous salpingitis to carcinoma of fallopian tube, 225 — Types, 225 — Cases recorded, 225 — Tuberculosis and non-malignant tumors of the genital tract, 226 — Accidental or coincidental combinations, 22~ — Pseudoneoplasms, 227 — Etiology, 228 — Infection of adenomyomata of uterus, 228 — Cases, 228 — Ovarian cysts, 229 — Histology, 229 — Sum- mary, 229 — Tuberculosis of uterus causing pyometra, 230 — Illustration, 230 — Tuberculous tubal lesions, 230 — Torsion of tuberculous pyosal- pinges, 231 — Factors, 232 — Action of diaphragm in cases, 233 — Rupture of tuberculous pyosalpinges, 233 — Collected statistics, 234 — Rupture of pyosalpinx in adjacent hollow viscera, 235 — Necessity for thorough pelvic examination, 236 — Extension of tuberculosis from pelvic lesion to other distinct areas, 237 — Tuberculous lesions in hernial sac, 237 — Histologic study, 238— Cases cited, 239. XI. Pregnancy and Tuberculosis 243 History, 243 — Fertility of the tuberculous, 244 — Frequency, 244 — Phys- iology of pregnancy bearing on course of tuberculosis, 245 — Organs affected, 246 — Puerperium and its bearing upon course of tuberculo- sis, 248— Strain of lactation, 249 — Condition of children of tuberculous mothers, 251 — Infant mortality, 252 — Influence of pulmonary tuber- culosis on course of pregnancy, 253 — Influence of pregnancy on course of pulmonary tuberculosis, 254 — Tubercle bacilli in mother's milk, 259 — Tuberculin as curative and diagnostic agent, 260 — Law regarding marriage of tuberculous persons, 261 — Indication for induction of abortion prior to fifth month, 265 — Results, 266 — Consultation and pre- caution prior to induction of abortion, 269 — Choice of operation, 270 — Sterilization, 270 — Anesthetic, 271 — Technic of operation (during first two months), 271 — Convalescence, 272 — Technic and choice of opera- tion for emptying uterus from second to fifth month, 272 — Preg- nancy after fifth month, 274 — Delivery of tuberculous patients. 275 — Cesarean section, 276 — Puerperium treatment during nursing, 278 — In- fluence of pregnancy upon tuberculous lesions other than the lungs, 278. XII. The Menstrual Disturbances of Pulmonary Tuberculosis . . . 284 General considerations, 284 — Gassification according to types, 284- — Etiology, 285 — Theories advanced. 286 — Later observations, 287 — Chief indications for treatment, 288 — Dysmenorrhea, 289 — Clinical reports, 290 — Use of tuberculin, 291 — Scanty menstruation, 291 — Statistics, 292 — Irregular scanty flow, 293 — Amenorrhea, 293 — Cases studied, 293— Menorrhagia, 293 — Vicarious menstruation, 294 — Periodic hemoptysis, 295 — Cases cited, 295 — Leukorrhea, 295 — Influence of menstruation on temperature in pulmonary tuberculosis, 295 — Causes, 296 — Considera- tion, 297 — Precautions instituted, 297. CONTENTS PAGE CHAPTER XIII. Pulmonary Tuberculosis and Operation 300 Three distinct dangers, 300— Choice of anesthetic, 300— Classification of pulmonary tuberculosis based on physical findings and constitutional symptoms, 300 — Subdivision into groups, 301 — Study of different stages of the disease, 301— Spinal anesthesia, 303— Precautions before opera- tion, 303— Importance of expert anesthetist, 304— Convalescence, 305 — Results, 305 — Condition of pulmonary lesion six or more months after operation performed under general anesthesia, 307 — Statistical reports, 307. XIV. Tuberculosis of the Breast 3°9 Historical, 309 — Histologic study of tuberculosis of the breast, 309 — Frequency, 309 — Primary and secondary infection, 310 — Routes of in- fection, 311 — Additional foci of disease, 312— Predisposing causes, 312 — Age incidence, 312— Statistics, 313— Varieties, 314— Confluent, 314— Disseminated, 315 — Physical manifestations, 315 — Initial symptoms, 316 — History of cases noted, 316 — Tuberculosis of breast in combina- tion with true neoplasm, 318 — Differential diagnosis between tuber- culosis and certain cases of chronic pyogenic mastitis, 318 — Post- operative results, 320. XV. Tuberculosis of the Peritoneum 323 Early history, 323 — First authentic operation performed by Sir Spencer Wells, Z 2 i — Primary and secondary tuberculous peritonitis, 324 — Cases studied with v;ew of determining primary lesion, 325 — Routes of in- fection, 325 — Pathology, 327 — Classification of tuberculous peritonitis, 329 — Varieties, 329 — Acute miliary, ascitic, fibroplastic and suppurative, 329 — Latent cases accidentally discovered, 331 — Frequency; special fre- quency among colored race, 332 — Variety of tubercle bacillus causing tuberculous peritonitis, 333 — Division into groups, 333 — Histologic study, 334 — Difficulties encountered in differentiating malignancy from tuberculosis, 336— Pseudotuberculosis of the peritoneum, 338 — Methods of treatment, 340 — Operative complications, 343 — Tuberculosis in hernia, 344 — Reformation of ascites following operation, 344 — Com- parison of results of medical and surgical treatment, 344. Index ... 349 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS CHAPTER I HISTORICAL SKETCH Early Recognition of Tuberculosis — Varieties of the disease — Atrophic, cachectic, and ulcerative — Practical knowledge of certain features of tuberculosis held by pre-Hippocratic writers — History of the disease reviewed by Osier, and other illustrious members of the medical profession — Connection between tuberculosis nodes and phthisis first indicated by Sylvius (1695) — Baillie (1793) first to recog- nize tuberculosis in organs other than the lungs — Laennec, originator of stetho- scope, recognized unity of scrofulous lymph nodes and phthisis — Open air treat- ment recommended by Samuel Morton (1834) — First successful inoculation, by Klencke (1843) — Various views regarding etiology of this condition — Valuable work of Furnival (1842) and others — Tubercle bacillus discovered by Koch — Genital tuberculosis first recognized by Morgagni — Historical review of genital tuberculosis. Tuberculosis was probably recognized many hundreds of years before Christ. Hippocrates (B. C. 460-376) described phthisis and Colsus (B. C. 30) wrote of three varieties of the disease : atrophic, cachectic, and ulcerative. Hippocrates referred to tuberculosis as "the greatest and most dangerous disease and one that proved fatal to the greatest number." Isocrates believed that tuberculosis was contagious, and Aris- totle mentions that the Greeks were of a similar opinion. Galen con- sidered tuberculosis to be an ulcerative process and recommended that sufferers from this disease should live in a high altitude. Some students believe that the curse pronounced by Moses (about B. C. 1500) for disobedience had reference to tuberculosis (Leviticus, 26 : 16, and Deuter- onomy, 28: 22) and that the laws recorded in the Talmud (Mischna, B. C. 500) indicated the recognition of tuberculosis in cattle. Osier 1 states that the title of one of the lost books of Democritus. "On Those Who Are Attacked with a Cough after Illness," probably indicates that the pre-Hippocratic writers had practical knowledge of certain features of tuberculosis. The history of tuberculosis includes a host of illustrious names, only a few of which are mentioned here, as the historical side of the disease 1 2 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS has already been so ably reviewed by Osier, 1 Waldenburg, 2 Predohl, 3 Johne, 4 and others. Sylvius (1695) was the first to indicate the connection which exists between the tuberculous nodes and phthisis. Morton (1689), in his excellent book, showed the prevalence of tuberculosis and accomplished much towards obtaining recognition for its importance by the medical profession. Morgagni (1682-1771) regarded the disease in the light of an infection and believed it dangerous to perform autopsies upon tuberculous subjects. Stark (1785) accurately described miliary tu- bercles. Kortum (1790), Baume (1795), Huf eland (1796), and Cullen (1800) were of the opinion that scrofulous glands anteceded phthisis. Baillie (1793) was the first to recognize tuberculosis in organs other than the lungs. Portal (1780) and Vetter (1803) coincided with Baillie in his findings. Laennec (1819) recognized the unity of the scrofulous lymph node and phthisis, described the pathology as well as the physical signs present during the various stages of phthisis, and originated the stetho- scope, by means of which accurate auscultatory findings were made possible. Samuel Morton (1834), a student of Laennec' s, in a mono- graph entitled "Pulmonary Consumption," recommended the open air for these patients and gave excellent therapeutic advice regarding the treatment of phthisical patients. Klencke (1843) performed the first successful inoculation, infecting a rabbit with tuberculosis by an intravenous injection. About this time various views were held regarding the etiology of this condition. Dupuy (1817) and Baron (1822) attributed it to hydatids. Furnival (1842) believed the condition to be due to deficient enervation. Engel (1844) thought the disease was similar in general character to typhoid fever, but caused by a different exudate. Alison ( 1824), Glover ( 1847), Simon (1850), and Villemin (1865) were of the opinion that tubercu- losis was the result of a specific infection. Langhans (1868), Schuller (1877), Tappeiner (1878), and others performed more or less suc- cessful experimental inoculations, the results of which were finally settled by the work of Cohnheim and Salmonsen (1879), who positively reproduced the lesion by inoculating the anterior chamber of a rab- bit's eye. Friedlander (1873), Koster (1873), an d Weigert (1879-1882) con- tributed valuable work. Aufrecht (1881) and Baumgarten (1883), in- dependently of Koch, described bacilli in the centers of tubercles, but did not prove that they were the infecting and active agents. The tubercle bacillus is the cause of tuberculosis. For many years, prior to the discovery of the tubercle bacillus by Koch 5 in 1882 and HISTORICAL SKETCH 3 public announcement thereof on March 24, before the Physiological So- ciety of Berlin, the infectious nature of the disease was suspected. In 1843, Klencke successfully accomplished the transmission of the disease, employing tuberculous material, and in 1865 Villemin 6 did likewise. Baumgarten 7 also reported the presence of what were probably tubercle bacilli in tissue, but had not proved the pathogenesis of the organism by inoculation. GENITAL TUBERCULOSIS Morgagni 8 was the first to recognize genital tuberculosis. This observer, upon performing an autopsy upon a girl of twenty years of age who had died of tuberculous peritonitis, found the uterus and adnexa filled with caseous material and believed that these organs were the primary focus of the disease. The importance of Morgagni's ob- servation was apparently not recognized, for no further mention is found of genital tuberculosis until the reports of Reynaud 9 and Senn 10 in 1 83 1. Reynaud described two cases of genital tuberculosis occurring in phthisical patients. Twelve years later Louis u also recorded cases. In 1853 Virchow 12 described genital lesions which were secondary to tuberculosis of the urinary tract. In reviewing the literature of this period, cognizance must be taken of the fact that the etiology of tuberculosis was unknown and even its pathology was not well understood, so that as a result many lesions were attributed to this variety of infection which are now known to have no connection with tuberculosis. Thus we find Waller 13 describing uterine myomata and writing that they were analogous to the "fleshy tubercle of the womb" described by William Hunter. Similar erroneous observations were made by Diintzer 14 and Osiender, 15 while Lis- franc 16 and Thiry 17 believed Nabothian cysts to be of tuberculous origin and remarked upon the ease with which this form of tuberculosis was cured. Boivin and Duges 18 describe tuberculous adnexitis and give an excellent illustration of a specimen. As time went on, a greater number of cases were recorded in the literature and a more accurate comprehension of the pathology and symptomatology of genital tubercu- losis became prevalent. Kiwisch, 19 Giel, 20 and Paulsen 21 contributed valuable information on this subject. Hegar's 22 important work ap- peared in 1886. Williams, 23 in his admirable historical review of genital tuberculosis, remarks that an interesting feature of the history of this condition is that for so long the ovaries were not considered receptive to the infection. Even such close observers as Virchow 12 and 4 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Rokitansky 24 were of this opinion, and as late as 1880 Brissaud 25 stated that there was not a single specimen of tuberculosis of the ovary in the Museum of the College of France. Koch's discovery of the tubercle bacillus added stimulus to the study of genital tuberculosis and made it possible to positively diagnose the condition. In 1883 Babes 26 demonstrated tubercle bacilli in the vaginal dis- charge in a case of rectovaginal fistula. This is the first authentic record of such findings. Hegar's masterly monograph entitled "Die Entstehung, Diagnose und chirurgische Behandlung der Genital Tuberculose des Weibes" (Stuttgart), which appeared in 1886, and the work of Chiari, 27 ' Bierfreund, 28 Baumgarten, 29 Kronig, 30 Bandl, 31 Martin, 32, 33 Schauta, 34 and later the valuable contributions by Kelly 35 and Osier 3G and by many others appeared in succession and have done much towards elucidating the pathology, symptomatology, and treatment of tubercu- losis of the female genitalia. LITERATURE 1. Osler, Sir W. Tuberculosis, edited by A. C. Klebs, 1909. 2. Waldenburg, L. Die Tuberculose, die Lungenschwindsucht, und Scrofulose. Berlin, 1869, Hirshwald. 3. Predohl, A. Die Geschichte der Tuberkulose. Hamburg and Leipzig, 1888, Voss. p. 482. 4. Johne, A. Die Geschichte der Tuberkulose mit Besonderen Berikksichtigung der Tuberkulose des Rindes. Leipzig, 1883, Vogel. p. 88. 5. Koch, R. Berl. Klin. Woch. 1882, 19:221. Mitth. a. d. Kais. Gesundheitsamt, 1884. 6. Villemin. Gaz. Hebd. de Med. 1865. 2s., 5. 7. Baumgarten. Virch. Arch., 82. Also Centrbl. f. d. Med. Wiss. 1882. 20:257. 8. Morgagni. De Sedibus et Causis Morborum Epistolae 38. 34. 9. Reynaud, M Arch. Gen. de Med. 1831. 36:486. 10. Sexn. Arch. Gen. de Med. 1831. 37:282. 11. Louis. Recherches sur la Phthisic Paris, 1843. 12. Virchow. Virch. Arch. 1853. 5 404. 13. Waller. Analekten fur Frauenkrankheiten. 1842. 3 493. 14. Duntzer. Neue Ztschr. f. Geb. 1840. 8:219. 15. Osiender. Hann. Ann. f. Ges. Heilk. 1840. 5:108. 16. Lisfranc. Clin. Chiv. de la Pitie. 1842, 2:661. HISTORICAL SKETCH 5 17. Thiry. Presse Med. Beige. 1852. 4:1. 18. Boivin et Duges. Traite pratique des maladies de l'uterus et de ses annexes (2d ed.). 1834. Plate 16. 19. Kiwisch. Klin. Vort. 1847,2:400. 20. Giel. Inaug. Dissert., Erlangen, 1851. 21. Paulsen. Schmidt's Jahrb. 1853. 80:222. 22. Hegar. Die Entsiehung, Diagnose, und Chirurgische Behandlung der Genital Tuberculose des Weibes. Stuttgart, 1886. 23. Williams, J. W. Johns Hopkins Hospital Reports. 1893. 3 : ^7- 24. Rokitansky. Lehrbuch der Pathologischen Anatomic. 1861. 3 444- 25. Brissaud, E. Arch. Gen. de Med. 1880. 146:129. 26. Babes, V. Orvosi Hetil. Budapest, 1883. 27:163. 27. Chiari, H. Vrtljschr. f. Derm. Vienna, 1886. 13:341. 28. Bierfreund, M. Ztschr. f. Geb. u. Gyn. 1888. 15:425. 29. Baumgarten, P. Ztschr. f. Klin. Med. 1885. 9:93. 30. Kronig. Centrbl. f. Chir. 1884. 11:81. 31. Bandl. Billroth-Liicke Handbuch der Frauenkrankheiten. 1886. b. 2. 32. Martin, A. Cong. Per. Internat. d. Sc. Med., Sec. Obst. and Gyn. Copenhagen, 1886. 2 :56. 33. Martin, A. Pathologie und Therapie der Frauenkrankheiten. Vienna and Leipzig, 1887, Urban and Schwarzenberg. 34. Schauta. Arch. f. Gyn. 1888. 33:27. 35. Kelly, H. A. Johns Hopkins Hospital Reports. 1890. 2:201. 36. Osler, Sir W. Johns Hopkins Hospital Reports. 1890. 2:67. CHAPTER II THE DIAGNOSIS OF TUBERCULOSIS OF THE FEMALE GENITAL TRACT BY LABORATORY METHODS Cervix and lower genital tract — Methods of treatment — Curettage — Value of ex- amination of leukorrheal discharge in tuberculous endometritis — Organisms likely to be mistaken for the tubercle bacillus — Smegma bacilli — Study of morphology, and of staining by ordinary methods — Grethe method — Czaplewski method — Etiology of smegma bacilli — Bacillus leprae — Resemblance to tubercle bacillus — Differentiation — Cultural methods and clinical study — Animal inoculation almost positively diagnostic — Danger of mistaking malignant neoplasms for certain forms 01 tuberculosis — Diagnostic use of tuberculin in gynecological conditions — Sum- mary of histologic examination. CERVIX AND LOWER GENITAL TRACT In lesions of these localities biopsy offers an easy and almost certain means of diagnosis and, if this method is employed, the histologic as well as the bacteriologic examination is available. If the lesions be ulcer- ative or friable, a light curettage may be performed and the material thus obtained similarly employed. An anesthetic is not necessary. Curet- tage is of little value in the hypertrophic varieties of tuberculosis, such as may be encountered in the vagina and external genitalia, unless there be loss of continuity of the surface. For lesions within the cervical canal or for the ulcerative or hypertrophic varieties of tuberculosis of the cervix, curettage is of distinct value. For those cases in which biopsy or curettage is not advisable, or as a preliminary measure, the discharge may be examined. For this pur- pose the parts should be thoroughly cleansed and a dressing or tampon applied. Some hours later the exudate which has collected upon the under surface of the dressing may be examined. The most frequent site for tuberculosis in the female genital tract is the fallopian tubes. Tuberculous salpingitis, except in the early stages, usually occludes the inner portion of the tube, so that the tubal contents do not gain free access to the uterine cavity. Even if the uterine ostium of the tube is patulous, the opening is generally so small that but little 6 DIAGNOSIS OF TUBERCULOSIS BY LABORATORY METHODS 7 of the tubal exudate escapes, and that which does is likely to be mixed with such a relatively large amount of uterine and cervical secretion that, by the time it is obtainable at the external os, the demonstration of the tubercle bacillus is extremely difficult. Although the organism is found, it is impossible to determine whether it is from the tube or from the endometrium. However, the latter is a comparatively unimportant point, inasmuch as when the corporeal endometrium is involved the tubes are nearly always affected. To obtain material for examination, the vagina and portio vaginalis should be thoroughly cleansed and two close fitting tampons applied to the cervix. The secretion thus obtained on the upper surface of the upper tampon after the latter has been in place for twenty- four hours is utilized. Immediately after removal of the tampon addi- tional secretion should be secured from within the cervical canal by means of a sterile platinum loop. Cummins 1 has reported good results with this method. Negative results do not exclude the tuberculous origin of the disease, particularly when the tubes alone are involved. In tuberculous endometritis the examination of the leukorrheal dis- charge secured by this method is of definite value, and, whereas too much stress should not be paid to negative results, the proportion of cases in which it is possible to demonstrate the tubercle bacillus is considerable. In the selection of material for examination, especial attention should be given to cheesy particles, for it is in these that the organisms are most frequently found. In cases of tuberculous endometritis, tubercle bacilli are more numerous in the discharge immediately after the cessation of a menstrual period. Curettage of the endometrial cavity and the examination of the curet- tings is naturally a much more certain method of diagnosis than is the examination of the secretion, and in some cases is a desirable procedure, which may be employed for diagnostic purposes as well as for other reasons. Occasionally the removal of one or two strips of endometrium by means of a small curette for diagnosis is a justifiable operation. As a general rule, however, curettage, unless immediately followed or preceded by an abdominal section, is inadvisable. A salpingitis is nearly always present, and curettage alone is likely to be the means of setting up an exacerbation of the infection. When depending upon the staining of the tubercle bacillus in prepa- rations of secretion obtained from the cervix or lower genital tract, the great difficulty is to differentiate the tubercle bacillus from other organ- isms which may be present and which are morphologically and tincto- rially similar. A number of organisms which possess nearly the same 8 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS size, shape and staining properties may be encountered. Chief among these is the smegma bacillus. Organisms Likely to Be Mistaken for the Tubercle Bacillus. — Smegma Bacilli. — These organisms morphologically and by the ordi- nary staining methods closely resemble tubercle bacilli. The normal habitat of the smegma bacillus is in the external genitalia and it is espe- cially common in the interlabial folds in women and around the corona of the glans in men. It is also found in the skin between the thighs and is not infrequent in the vagina, especially the lower portion. Gottstein, 2 Ritter, 3 Labbs, 4 Alvarez and Taval, 5 Matterstock, 6 Klemperer, 7 and others have demonstrated the organism in various localities on the skin surface of the body, around the natural opening, and even upon the coat- ing of the tongue and teeth. The bacillus is of frequent occurrence in the urine (especially in specimens passed voluntarily) and in the feces. It appears in especially large numbers in any area where normal skin secretions are allowed to collect, and has been frequently observed about the umbilicus, in the cerumen of the ear, about the teeth, etc. In this respect its occurrence is somewhat dependent on the degree of cleanliness of the patient. Regarding the frequency with which the smegma bacillus occurs in the genital tract, Brereton and Smith, 8 in 126 insane or un- cleanly patients, found red staining bacilli in 85 (67.5 per cent) after decolorization by 25 per cent sulphuric acid, while they occurred in only 19 (22 per cent) of these patients after methylene blue had been employed as a counter stain after decolorization. They were present in 13 per cent after decolorization by acid alcohol or Labarraque's solution. In a second series of twenty ordinary cases these authors found smegma bacilli in 13, or 65 per cent, after the use of sulphuric acid only, and 2 or 1 per cent after counter staining. Young and Churchman, 9 in 24 tests, found smegma bacilli present in 46 per cent of cases. As this or- ganism is so frequently present in the neighborhood of the external genitalia, its differentiation from the tubercle bacillus, when studying material from this locality, is of extreme importance. The smegma bacillus is non-pathogenic. In earlier times a number of operations have been performed under a misconception because of a lack of knowledge upon this point (Labbs, 4 Kronig, 10 Bunge and Trentenrath n ). A study of the morphology and of staining by the ordinary methods, employing inorganic acids as decolorizers, is useless as far as differentia- tion between the tubercle bacillus and smegma bacillus is concerned, although it is claimed by some observers that the smegma bacillus is slightly shorter than the tubercle bacillus. Certain special stains are, however, moderately reliable. Smegma bacilli are decolorized somewhat DIAGNOSIS OF TUBERCULOSIS BY LABORATORY METHODS 9 more easily than are tubercle bacilli and may be decolorized with abso- lute alcohol, although Moller 12 believes them not only alcohol, but acid proof, and admits no tinctorial difference from the tubercle bacillus. Brown 13 states careful work to have shown that no staining methods differentiate tubercle from smegma bacilli : he recommends Petroff's medium as a differentiating agent for tubercle and smegma bacilli. Bunge and Trentenroth 11 Method. — 1. Fixation of smears by abso- lute alcohol for three hours. 2. Five, per cent chromic acid for at least fifteen minutes. 3. Wash, in several changes of water. 4. Stain in carbol fuchsin, in usual manner. 5. Decolorize with dilute sulphuric acid for three minutes, or pure nitric acid for one; or two minutes. 6. Secondary decolorization combined with secondary staining in a concentrated alcoholic solution of methylene blue for at least five minutes. Result: Bunge and Trentenroth state that in all cases smegma bacilli were decolorized, and only rarely did they find one or more bacilli reddish violet, but by no means so intense (tubercle bacilli from sputum only were employed as controls ) . Grethe 14 Method. — The preparation is stained in the ordinary man- ner with carbol fuchsin, washed, and, without further decolorization, is treated with a concentrated solution of methylene blue in absolute alcohol. The tubercle bacilli remain red, while the smegma bacilli are blue. Weichselbaum 15 reported excellent results with this method. Czaplewski 16 Method. — Treat and stain in the usual manner with heated carbol fuchsin. The excess of the fluid is drained off and the prep- aration immersed for five minutes in fluorescein methylene blue and then in concentrated alcoholic solution of methylene blue for one* half to one minute and rapidly washed in fresh water and mounted. In staining smear preparations for the purpose of differentiating the smegma bacillus, it is wise to place, a few tubercle bacilli of known au- thenticity upon the cover glass, some distance from the material for ex- amination, as a control. If urine is to be examined, a catheterized speci- men, the external urinary meatus having first been carefully cleansed in order to avoid contamination, should be utilized. As has been stated, biopsy or inoculation methods offer a more certain means of diagnosis than does the microscopic examination and smear preparation from the genital tract when tuberculosis is suspected, and should be resorted to in all doubt fid eases. All investigators agree on the non-pathogenic character of the smegma bacillus, and animal inoculation can thus be safely depended upon in all io GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS doubtful cases. Young and Churchman 9 insist that the presence of the smegma bacillus cannot be excluded from specimens of urine by cleans- ing of the external genitalia, but that urethral irrigation must be fre- quently undertaken. Brereton and Smith 8 believe that no meihod of dif- ferentiating the tubercle bacillus from the smegma bacillus by staining is entirely adequate for all cases, and that the successful cultivation of the latter organisms is open to question. The smegma bacillus was first cultivated by Czaplewski. 17 Novey 18 recommends the following media: Agar-agar is cooled to 50 C. and mixed with a small quantity (about 2 c.cm.) of fresh, sterile human blood. The mixture is poured into Petri dishes and placed in the in- cubator at 37 C. for twenty-four to forty-eight hours. Subcultures, according to Moller, 12 may be grown upon glycerin agar-agar. The organism may also be cultivated in bouillon, potato, or glycerin agar, at 2>7° C. ; upon the latter, colonies appear in minute, whitish or yellowish scale-like dots, which later become somewhat rounded and possess a soft velvety or corrugated surface. The growth is slow. Bacillus Leprae. — This organism morphologically closely resem- bles the tubercle bacillus. The frequent intracellular position and often parallel or package-like arrangement under such circumstances, and the tendency which they possess to occupy lymphatic spaces, are points aiding in their differentiation. They stain somewhat more readily than do the tubercle bacilli. None of these characteristics is, however, sufficiently marked to positively differentiate these two organisms, and cultural methods and a clinical study of the case are required for a positive diag- nosis. The bacillus leprae is cultivated upon artificial media with great difficulty and some doubt exists as to whether it has ever been grown freely upon the media usually employed for the cultivation of the tubercle bacillus ; indeed, Jordan 19 states that saprophytic growth on the part of the leprosy bacillus is entirely unknown. On the other hand, McFar- land 20 states that in cultures there is a delicate filamentous arrangement of the leprae bacilli, especially where they have become accustomed to a saprophytic existence. There is a large number of other acid proof bacteria (about forty) which, however, as a rule, can be easily differentiated from the tubercle bacillus. Such organisms have been isolated from butter (Petri, 21 Rabi- nowitsch, 22 Korn 23 ) and hay (Moller 12 ). In some cases, it is almost impossible to differentiate these organisms from the tubercle bacillus, although the rapid growth of the hay and butter bacilli in artificial media, at about 20 C, is their chief differential point. As a rule, the clinical picture is sufficient to differentiate the organism; however, Frankel, 24 DIAGNOSIS OF TUBERCULOSIS BY LABORATORY METHODS n Rabinowitsch, 22 and Marzinowski 25 have demonstrated these organisms in pathologic conditions in the lungs. The possibility of these bacilli being present in conjunction with the tubercle bacillus must also be taken into consideration. Leprosy is a rare disease in this climate, and the clinical symptoms are usually sufficient upon which to differentiate the disease from tuberculosis of the genital tract. As can be seen from the foregoing, ways of demonstrating the tubercle bacillus in the discharge by staining methods are open to doubt. The finding of characteristic organisms in the tissue may be considered moderately reliable. When, however, sufficient tissue is obtainable to demonstrate the organism in it, a histologic diagnosis which is equally reliable is usually possible. Tubercle bacilli, in certain forms of tuber- culosis and in certain stages of the disease, are present only in small num- bers, and their demonstration consequently is difficult, whereas, on the other hand, diagnosis by histologic examination is usually easy. Animal inoculation offers an almost positive method of diagnosis. It is more cer- tain when ground up particles of the suspected tissue are employed than when only the exudate is utilized. The disadvantage of animal inocula- tion is the time required. Three or four guinea pigs should be inoculated. The danger of mistaking malignant neoplasms for certain forms of tuberculosis, particularly those occurring in the cervix and external geni- talia, should be taken into consideration when obtaining tissue for diag- nosis. For fear of disseminating a malignant tumor, it is therefore ad- visable, when performing biopsy, to employ the cautery knife heated to a dull red, or the cut edges of the wound may be immediately seared following the removal of the suspected tissue. Rapid diagnosis is an important factor in these cases and is an additional reason for the em- ployment of histologic means, rather than waiting for the slower inocula- tion method. The latter may, however, be utilized with advantage as an auxiliary to the histologic examination, and is especially valuable when the rare hypertrophic variety of tuberculosis of the external geni- talia is suspected. The Diagnostic Use of Tuberculin in Gynecological Conditions. — Pankow 26 states that, in the cases examined by him, he observed focal reactions in the absence of tuberculous foci. In three cases of pelvic inflammatory disease, of non-tuberculous origin, he obtained a focal re- action, but in one of these the local symptoms may have been caused by menstruation. Sahli 27 has emphasized the fact that sensitiveness is in- creased in the premenstrual period. Beer 28 asserts that such focal reac- tions in non-tuberculous cases must be exceptional. Mohr 29 is of the opinion that a negative response excludes tuberculosis; but Beer thinks 12 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS a general, plus a focal, response is practically invariably due to a focal tuberculosis, and that such a response locates the diseased area. A gen- eral, minus a focal, response is of no practical value, as the most careful examination cannot exclude tuberculosis in other parts of the body, which may give the general reaction. Servaes 30 says that Moller made 20,000 injections without any bad effects. In speaking of pulmonary tuberculosis, Brown 31 states that the dan- ger from the proper use of the tuberculin test is slight, but in some un- suitable cases very real. Shattuck 32 believes the subcutaneous test the most reliable and has never seen untoward results beyond a disturbance of two or three days, except in one case of Addison's disease. Jane- way 33 has also observed a fatal issue from the use of this test in Addi- son's disease. Koplik 34 makes an almost routine use of the cutaneous von Pirquet test in children. The author has had no personal experience with the use of tuberculin as a diagnostic agent in patients suffering from gynecologic lesions. It would appear that its use is not entirely free from danger and the results obtained are somewhat uncertain. It is probably of little or no practical value. The fact that tuberculosis of the female genital tract and peri- toneum is generally secondary to tuberculosis elsewhere in the body and that pulmonary or other forms of tuberculosis are frequent and often quiescent greatly nullifies the value of the tuberculin test. The same may be said regarding the von Pirquet and the complement fixation test. Summary of Histologic Examination. — This method offers a rapid and quick method of diagnosis. The various forms of tuberculosis can nearly always be diagnosed by it with certainty and the possibility of malignancy can be easily determined. As a supplement to it, a part of the tissue may be utilized for animal inoculation, and this is a valuable aid, but has the distinct disadvantage of being time consuming, a point which is especially to be avoided in those cases in which the possibility of malignancy cannot be excluded. With light curettage, such as may be performed upon ulcerative lesions of the cervix or lower genital tract, enough tissue may sometimes be obtained for a histologic examination. The employment of such material for animal inoculation is of distinct value. The staining of secretions, often as a preliminary method, is not without value; negative results do not exclude with certainty the possi- bility of the tuberculous character of the lesion, and, owing to the simi- larity of the smegma bacillus to the tubercle bacillus, positive results cannot be absolutely relieA upon. DIAGNOSIS OF TUBERCULOSIS BY LABORATORY METHODS 13 LITERATURE 1. Cummins, H. H. Phys. and Surg. 1912. 34:202. 2. Gottstein. Fortsch. der Med. 1886. 4:252. 3. Bitter. Virch. arch. 1886. 106:209. 4. Laabs. Inaug. dissert, Freiburg, 1894. . 5. Alvarez and Taval. Arch, de Physiol. Norm, et Path. 1885. No. 7. 6. Matterstock. Mitth. a. d. Med. Klin. Wurzburg. 1885. No. 6. 7. Klemperer. Deutsch. Med. Woch. 1885. No. 11. 8. Brereton, C. E., and Smith, K. W. Am. Jr. Med. Sc. 1914. 148 :267. 9. Young, H. H., Churchman, J. W. Am. Jr. Med. Sc. 1905. 130:52. 10. Kronig. Deutsch. Med. Woch. 1894. No. 43. 11. Bunge und Trentenroth. Fortschr. d. Med. 1896. No. 14. 12. Moller. Centrbl. f. Bakt. Par. Inf. 1902. 29 :278. 13. Brown, L. Jr. Am. Med. Assoc. 1915. 64:886. 14. Grethe. Fortschr. d. Med. 1896. No. 9. 15. Weichselbaum. Fortschr. d. Med. 1896. No. 9. 16. Czaplewski, E. Die Untersuchung der Auswarfs auf Tuberkel- bacillen. Jena, 1891, Fischer. 17. Czaplewski, E. Munch. Med. Woch. 1897. 18. Now. Laboratory Work in Bacteriology. 1899. 19. Jordan, E. O. A Text Book of General Bacteriology. Philadel- phia and London. 1908. p. 1358. 20. McFarland. A Textbook Upon the Pathologic Bacteria and Protozoa. Philadelphia and London. 1912. p. 763. 21. Petri. Arb. a. d. Kais. Geshtamt. 1897. 22. Rabinowitsch. Deutsch. Med. Woch. 1900. 26:258. Also. Ztschr. f. Hyg. u. Inf. 1897. 23. Korn. Centrbl. f. Bakt. Par. Inf. 1899. 24. Frankel. Berl. Klin. Woch. 1898. 35:246,880. 25. Marzinowski. Centrbl. f. Bakt. Par. Inf. 1901. 28:39. 26. Pankow. Centrbl. f. Gyn. 1907. 2.y. Sahli. Tuberkulinbehandlung. 1910. 28. Beer, E. N. Y. Med. Rec. 84:652. 29. Mohr. Munch. Med. Woch. 1906. 30. Servaes. Beitr. z. Klin. d. Tuberk. 1904. u. 2. 14 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS 31. Brown, L. Tr. Assoc. Am. Phys. 1911. 26:22. 32. Shattuck, F. C. Tr. Assoc. Am. Phys. 1911. 26:31. 33. Janew ay. Tr. Assoc. Am. Phys. 1911. 26:31. 34. Koplik. Tr. Assoc. Am. Phys. 191 1. 26:31. CHAPTER III PATHOLOGY Two distinct forms of genital tuberculosis, ulcerative and hypertrophic— Histologic examination of ulcerative form — Hypertrophic variety demonstrated by staining or inoculation— Tuberculosis of the vagina ; ulcerative form most frequent variety —Histologic examination— Hypertrophic form in relation to miliary tuberculosis — Tuberculosis of the cervix, ulcerative, papillary, miliary, and interstitial — His- tologic examination— Characteristics, and differentiation— Corporeal endometritis; miliary and caseous — Study of histologic and pathologic characteristics — Myome- tritis, frequent occurrence in advanced cases of tuberculous endometritis — In- fections of the endometrium — Intramural abscess, tuberculous deciduitis — His- tologic examination — Placental tuberculosis — Macroscopic appearance, character- istics and forms — Intravillous tuberculosis — Intravascular chorionic lesions, and chorio-amniotic — Tuberculosis of fallopian tubes — Tuberculosis of the ovary — Peri-oophoritis, and oophoritis — Histologic examination. TUBERCULOSIS OF THE EXTERNAL GENITALIA As has been stated elsewhere, tuberculosis of the external genitalia may be either primary or secondary, the latter being by far the most fre- quent. As the vulva and adjacent structures are covered by modified skin, tuberculosis occurring in this locality is similar in its general macroscopic and histologic characteristics to tuberculosis of the cutane- ous surface as found in other parts of the body. As a result, however, of moisture, heat, friction, and local anatomic conditions, and not infre- quently due to the presence of irritating discharges, certain modifica- tions of the ordinary tuberculous lesions found in other skin areas may occur. As a general rule, the pathologic processes which occur here closely resemble the ordinary forms of cutaneous tuberculosis. Undoubtedly the rarity of tuberculosis of the external genitalia may be largely explained by the protective qualities of the squamous epithe- lium and particularly of the horny layer of the latter. In young chil- dren the development of the outer horny layer is less marked and, as a result, in such subjects this locality is less immune. In the adult con- stant bathing of the parts in leukorrheal discharges, which more or less macerates the protective covering, probably acts as a predisposing factor to infection. The irritating properties of the discharge likely to be 15 16 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS present as a result of lesions in the upper genital tract are also predis- posing causes by producing a vulvitis, which is due to toxins. Experi- mental studies have amply proven the relationship existing between pre- existing inflammation and infection with the tubercle bacillus in this locality. Tuberculosis of the external genitalia may be divided into two dis- tinct forms, the ulcerative and the hypertrophic, the former being much the most common. Combinations of the two are, however, not infre- quent. Ulcerative Form. — The lesions may be single or multiple and vary in size from the microscopic to huge ulcers, involving not only the ex- ternal genitalia, but also the skin perineum, vagina, and adjacent struc- tures. Vulvovaginal ulcers are relatively frequent and may originate either in the vagina or vulva. The clitoris and surrounding parts are frequently involved. Ulcers may occur either on the labia major or minor, or both may be attacked. Contact ulcers on the opposite labium occasionally are present. Fistulae leading to tuberculous foci in the lower alimentary tract, near-by osseous system, or other localities may be present. The vulvar surface adjacent to the ulcer is generally the seat of a more or less well marked chronic inflammation, the skin being red- dened and swollen. Pigmentation is often present. The initial genital lesion is usually a small swelling, papule-like in character, which enlarges, softens and breaks down, leaving in its center an irregular necrotic ulcer. Less frequently the primary pathologic process is a minute shallow ulcer. In either event, the subsequent course is usually slow, but progressive, and more or less swelling and hyper- trophy are likely to occur. The base of the ulcer is usually dark necrotic, but may be reddish, pinkish, or grayish in color. Small yellowish or grayish elevations are not infrequently present, while in some instances more or less typical tubercles may be observed. These are minute, gray- ish or yellowish, semitranslucid homogeneous elevations, and are gen- erally observed on the floor of the ulcer. To the touch the base of the ulcer imparts a soft, somewhat velvety feel. The walls of the lesion are elevated, infiltrated, somewhat edematous and often undermined, and may be the seat of enlarged veins. In some instances the ulcers are friable and tend to bleed easily when traumatized. The older lesions are generally fairly firm and exhibit but little tendency to hemorrhage. One of the chief characteristics of the ulcers is their chronic appearance. Not infrequently, in old chronic cases, the ulcers are to a certain extent serpiginous and leave behind scar tissue as they advance in other di- PATHOLOGY 17 rections. More or less eczema and swelling are usually present in the neighborhood of the ulcer. In some instances, where the ulcer is lim- ited to one side, the opposite vulva is enlarged or hypertrophied. En- larged or varicose veins are frequently present, especially in aged patients. An inguinal adenitis is a frequent accompaniment of the condition. On histologic examination, the ulcers present the usual characteristics of skin tuberculosis. The tissue is infiltrated with chronic inflammatory products, the blood vessels are enlarged and thrombi may be present. The surface may present characteristic caseous structures, and typical tuberculous giant cells are nearly always present. The latter, together with the tubercles, are the chief diagnostic features. Tubercle bacilli are present, but are frequently difficult to demonstrate by staining methods. They are often few in number, but can be discovered if a careful search is made. In some instances the organisms can be demonstrated in the discharge from the ulcer ; in this, however, they are generally sparsely distributed, and consequently difficult to find. Furthermore, the dangers of contamination from tubercle bacilli bearing discharges from other lesions must be borne in mind. For these reasons the examination of the discharges is generally unsatisfactory, and a better method is to care- fully cleanse the surface of the lesion, lightly curette it and examine the material thus obtained. In cases of doubt biopsy combined with animal inoculation offers the surest and most satisfactory means of diagnosis. Hypertrophic Variety. — In this variety of tuberculosis the infec- tion usually results in moderate sized tumor-like masses, the labia being perhaps the most frequent area involved. The lesions are generally fairly firm, somewhat rounded outgrowths, often covered by thickened wrinkled skin. Cases have been mistaken for sarcoma, carcinoma, con- dyloma acuminata, and true elephantiasis. One or both labia may be attacked, but a unilateral involvement is the most frequent. More or less edema and swelling of the opposite side and adjacent structures is, however, the rule in advanced cases. Engorged blood vessels are some- times observed. As will be noted, the macroscopic appearance of this form of tuberculosis presents nothing characteristic. On histologic examination, the skin is usually found to be thickened, the affected area fairly vascular, and a well marked increase of connec- tive tissue is everywhere observed. A considerable infiltration with a chronic inflammatory exudate is present, although this is less marked than is observed in the ulcerative lesion. Here and there tubercles con- taining the characteristic giant cells may be seen. These, however, are not plentiful, and a number of sections should be examined in suspected cases before the possibility of tuberculosis is excluded. Tubercle bacilli 18 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS are generally present only in small numbers and are therefore usually difficult to demonstrate in stained sections. On account of the confusion which has existed in the past regarding this variety of tuberculosis, the author believes that no case should be classified as tuberculosis unless the characteristic lesions, tubercles and giant cells are observed, or tubercle bacilli demonstrated either by stain- ing or inoculation methods. The latter is the best and most certain pro- cedure. TUBERCULOSIS OF THE VAGINA Tuberculosis of the vagina is similar in its general pathologic char- acteristics to tuberculosis of the external genitalia. It is usually of the secondary variety and may result from a direct extension from nearby foci, such as the cervix or intestines, or from a hemogenic or lympho- genic infection. Direct implantation, either from exogenic organisms by means of the fingers, douche nozzle, or coitus, or by tubercle bacilli bearing discharges originating in the upper genital tract, may occur. Three varieties have been observed, the ulcerative, the hypertrophic and the miliary. The Ulcerative Form. — This is the most frequent variety, and in general presents the same characteristics as when present on the external genitalia. Indeed vulvovaginal lesions are relatively frequent. The ulcers may be single or multiple and are perhaps most frequently present in the floor of the vagina. The lesions vary considerably in size; huge ulcers involving almost the entire vagina and extending to the ex- ternal genitalia and adjacent structures have been observed, and on the other hand, almost microscopic lesions have been recorded. The ulcers generally present a chronic appearance and possess infiltrated, elevated hyperemic and particularly undermined edges. The base is necrotic, brownish, blackish, pinkish, or grayish in color, moderately soft and friable to the touch, and is frequently studded with minute rounded grayish or yellowish semitranslucent elevations, which histologic exam- ination shows to be tubercles. The surrounding vaginal mucosa is gen- erally reddened and presents the usual appearance of a vaginitis. In the early stages the ulcers are shallow. The lesions are to be differentiated from malignant neoplasm, syphilis, chancroids, and in children from gonorrhea, noma of the vulva, and other ulcerative lesions. The presence of tuberculosis elsewhere in the body, the chronic char- PATHOLOGY 19 acter of the lesions, and the presence of tubercle-like elevations on the base of the ulcer are points suggestive of this variety. On histologic examination, any doubts which may exist are easily cleared up. The vaginal mucosa surrounding the ulcer is thickened; hyperemic, and more or less infiltrated with chronic inflammatory prod- ucts, and may contain an isolated tubercle. Sections through the ulcers show the absence of the normal superficial tissues, the floor of the ulcer being densely infiltrated with a chronic inflammatory exudate, and the surface consisting of necrotic granulation tissue. The presence of tubercles and the characteristic giant cells makes the diagnosis positive. Tubercle bacilli can generally be demonstrated in stained sections, if care- ful search for them is instigated. Hypertrophic Form. — Cases of this kind in which the lesions have originated in the vagina itself are too few to draw from them definite conclusions regarding their characteristic appearance. By far the greater number of cases in which the hypertrophic form of tuberculosis has been present have resulted from a direct extension from a similar outgrowth in the cervix. The hypertrophic variety of tuberculosis is usually papil- lary in appearance, and more or less friable and necrotic. The tumor- like masses are bathed in an irritating, foul smelling discharge. The outgrowths are pinkish or grayish in color and frequently present ex- tensive areas of necrosis. In the interstices on the surface collections of clotted blood and discharge are frequently present. Histologically, the lesions present the usual characteristics of tuber- culosis, although tubercles are less plentiful than in the former variety. In some specimens there is marked increase in the number of blood ves- sels, and a correlation with the macroscopic and clinical findings will usually show that these specimens are more friable, rapid in growth, and productive of more easily excited small hemorrhages, especially fol- lowing trauma, than in the less vascular pathologic processes. Tubercle bacilli are present, but as a rule less numerously than in the ulcerative variety. Miliary Tuberculosis. — In this variety of the disease, the vaginal mucosa is thickened, reddened, swollen, more or less edematous, and hyperemic. The entire vaginal lining is usually involved, although the inflammation is apt to be most marked in the floor of the canal. Scat- tered more or less profusely throughout the lining of the vagina are small grayish or yellowish semitranslucent discrete elevations — the tubercles. Occasionally one of these breaks down and a small ulcer results. Con- siderable discharge is usually present. Histologic examination shows a well marked inflammation and the 20 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS usual characteristics of tuberculosis. Tuberculous giant cells and tuber- cle bacilli are generally present in considerable numbers. This variety of the disease is generally the result of a hemogenic infection, and is usually associated with a general miliary infection. Combinations of the ulcerative, hypertrophic, and miliary forms may be present. TUBERCULOSIS OF THE CERVIX Cervical tuberculosis occurs in four distinct forms : the ulcerative, the papillary, the miliary, and interstitial. An analysis of 106 cases showed that 52 were of the ulcerative variety, 41 papillary, 7 miliary, and 6 interstitial. Combinations of these varieties are not infrequent. These statistics may to some extent be misleading, owing to the fact that in the late stages, when many of these cases were first observed, ulcerative lesions are prone to develop. Thus, a lesion which began as an inter- stitial type in the later stage may break through into the portio or cervical canal and present a condition which would probably be classified under the ulcerative variety. Cora x believes the papillary variety the most fre- quent, and bases this opinion upon the fact that, in the later stages, this variety frequently undergoes ulcerative changes. Cervical tuberculosis is usually secondary; a few authentic primary cases have, however, been recorded. Beyea 2 analyzed 61 cases, with a view to ascertaining what portion of the cervix was most frequently attacked. In this series the portio alone was involved in 11, the supra- vaginal cervix alone in 6, and both in 44. The disease usually originates in the cervical canal. Ulcerative Variety. — Cervices, the seat of this variety of tubercu- losis, vary widely in appearance. All of the four varieties of cervical tuberculosis are prone to become ulcerative in their end stages. As a general rule, the pathologic process produced by the ulcerative variety of cervical tuberculosis resembles more or less closely the ulcerative ten- sions of carcinoma. Indeed, malignancy of some form has been the clinical diagnosis in a large proportion of cases, not only in the ulcera- tive variety, but also in the papillary forms. The ulcer may begin either in the portio vaginalis or in the cervical canal. The history usually shows that the disease has been slow, but progressive. The margins of the ulcers are often not markedly elevated, less so as a rule than in carci- noma, and are apt to be undermined and fairly soft. The base is covered with necrotic tissue and may be brown, black, yellow or gray in color. It is usually moderately soft and friable, but in some specimens is firm and PATHOLOGY 21 hard. Numerous minute grayish or yellowish semitranslucent discrete tubercles may be scattered over the floor of the ulcer. These may also be present in the walls of the ulcer and on the surface of the adjacent structures, Not infrequently the surface of the ulcer will be found to be partially covered by cheesy particles. The ulcers are usually single, but multiple lesions have been observed. The ulcers vary widely in size ; in some specimens the place of the entire cervix and adjacent vagina is occupied by a large excavated necrotic cavity, while in others almost microscopic lesions have been described. In the advanced specimens inr volvement of the surrounding vagina is frequent. The adjacent covering of the vagina and the portion of the cervix not actually involved by the ulcers are usually reddened, and as has been stated, may contain small tubercles. The ulcers may extend upward into the body of the uterus, but the disease apparently exhibits some tendency toward remaining limited to the area below the internal os, although a tuberculous endo- metritis is frequently present. Papillary Variety. — This variety usually originates from the vaginal surface of the cervix, but in rare instances may spring from the canal. It occurs usually as a cauliflower-like outgrowth, and is generally dark reddish or brownish in color, presents necrotic areas on its sur- face, and is covered with discharge. Firm smooth nodular dome-like elevations may be present, either in conjunction with the cauliflower type of tumor, or less frequently may constitute the chief patho- logic process present. On section, the papillary variety of tuber- culosis presents a smooth, fairly soft, somewhat translucent, mod- erate vascular pinkish, grayish or whitish appearance; small yellowish areas on the cut surface may be observed in some specimens. The tumor-like outgrowth may possess a well defined pedicle, but more fre- quently springs from a broad base. In the early stage this variety is usually single, but later numerous outgrowths are observed. As the dis- ease progresses the tumor-like masses tend to undergo necrosis and ul- ceration. Interstitial Variety. — In this variety the disease begins in the sub- stance of the cervix, primarily causing a slight nodular swelling in one lip of the organ ; this gradually increases in size and finally breaks down, leaving an irregular opening either on the portio or in the canal. The opening becomes larger as the result of a disintegration of tissues, and the final stage is a large crater-like cavity, together with involvement of much of the adjacent tissue. The walls of the cavity are similar in general macroscopic appearance to those of the ulcerative variety of the disease. 22 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Miliary Variety. — In this variety the cervix is usually enlarged, the portio reddened and glistening, and scattered discretely over the surface are small grayish or yellowish semitranslucent elevations, which are the tubercles. These may be seen in various stages of development. Not infrequently one of these may break down, leaving a minute ulcer filled with whitish or yellowish cheesy material. The mucosa of the cervical canal is usually the seat of a tuberculous endocervicitis. The vagina may also be involved. The specimen is likely to be bathed in a moderate thin semipurulent discharge, intermixed with which cheesy particles may be present. As a result of tuberculosis of the cervix, the canal may become occluded and a pyometra or, as it is sometimes termed, uterine phthisis may result. When this occurs the uterus becomes larger and softer, and the amount of leukorrhea is likely to become somewhat lessened as the discharge from the upper genital tract ceases to gain egress. Pyo- metra is most frequent in the ulcerative and papillary forms. Occasionally specimens are observed in which there is marked re- duplication of the columnar epithelium and a tendency towards a con- centric arrangement of the cells, which somewhat resembles carcinoma. A similar arrangement of the squamous epithelium has been described, in which groups of cells are present which somewhat suggest cancer pearls. By a careful histologic examination the differentiations of these two conditions is always rendered feasible. The mitosis, hyperchroma- tosis, rapid proliferation, and penetrating character of the cancer cells and the absence of tubercles should make the differentiation possible in all cases. Tubercle bacilli can frequently be demonstrated in smear specimens and almost always by animal inoculation, although in old cases mixed infections are of frequent occurrence and at times make the demonstra- tion of the original type of infection difficult. As a result of the amenorrhea and uterine enlargement which results, cases in which a pyometra has been present have been mistaken for pregnancy. It is needless to state that a properly conducted examination should easily clear up such an erroneous diagnosis. The pus in the uterine cavity is usually yellowish and creamy in consistency and not infrequently contains cheesy particles and possesses a foul odor. Histologic Examination. — However difficult the clinical diagnosis may be in these cases, the histologic diagnosis is usually easy. The well marked evidence of chronic inflammation, the necrosis, and lastly the characteristic tubercles and the frequent presence of tuberculous giant cells, clear up any doubts which may have existed. Histologically PATHOLOGY 23 the inflammatory exudate is characterized by the presence of small round cells and a few polymorphous leukocytes. In the ulcerative va- riety the inflammation is most marked at the edge of the lesion. In addition, there is loss of surface epithelium and of underlying tissue, the erosion being lined by necrotic material. The cervical glands in some areas may be destroyed or unrecognizable. The blood vessels are engorged. In the papillary type, polypoid or papilla-like projections are present, which are covered by squamous or cylindrical epithelium, according to their point of origin. In either event, the surface epithe- lium is prone to proliferation, although the individual cells may be more or less normal. The stroma of the outgrowths is composed of cervical tissue, and is usually infiltrated with chronic inflammatory products. Tubercles and giant cells in varying numbers are present. Tubercle bacilli may be demonstrated in stained preparations in most cases, if a careful search is instigated. In some instances, how- ever, they are few in number, and in these cases animal inoculation offers a means of positive diagnosis and should be employed in all cases in which doubt exists. CORPOREAL ENDOMETRITIS With the exception of the fallopian tube, the endometrium of the body of the uterus is the structure most frequently involved in tuber- culous infection of the female genital tract. Careful study has shown that tuberculous endometritis is, like a similar infection of the tubes, much more frequent than was formerly thought before routine histo- logic examination of tissues removed at operations was generally prac- ticed. Mayo 3 states that tuberculous endometritis in the menstruating uterus is infrequent. This has not been our experience; nearly 80 per cent of our cases have occurred during active sexual age. The disease exists in two well defined varieties — (1) the miliary and the (2) caseous or ulcerative. The Miliary Variety. — In many specimens, at the seat of this variety of tuberculosis the macroscopic lesions are not marked and un- less a histologic examination is performed, no suspicion of the presence of this type of infection may be aroused. The entire endometrium is by no means always involved. The infection may be blood borne, the primary and perhaps only genital lesion being in the endometrium; or the endometritis may be secondary to a salpingitis, either as a result of a direct infection by continuity from the tubal mucosa, or by contamina- 24 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS tion due to leakage of the tubal contents through the uterine ostia of the tube. It is often difficult to determine which of these routes infection has followed. The tubes are generally involved. Endometritis, the di- rect result of a bacteriemia, is more frequent in the miliary variety than in the caseous, which latter is nearly always secondary to tubal lesions. In advanced cases, the mucosa is reddened and thickened and hyper- emic; the superficial tissue may contain small discrete yellowish or grayish semitranslucent elevations, which are tubercles. In some speci- mens the tubercles can be plainly discerned with the naked eye, while in others they are less conspicuous, and their presence may not be suspected until a histologic examination is made. The uterus is usually slightly enlarged and not infrequently tubercles can be seen upon the peritoneal surface. This is especially likely to be the case, if a tuberculous peri- tonitis has been present. An accompanying salpingitis is the rule. In rare instances, as a result of adhesions at the internal os, occlusion of the canal occurs and results in a pyometra. The pus under such cir- cumstances is usually thick, creamy and yellowish in color, and may contain cheesy particles. Pyometra is rare, unless definite cervical le- sions are present. Histologic Examination. — This presents a somewhat varying picture. In some specimens the infection is chiefly superficial, the deeper portions of the endometrium being comparatively normal. In its very early stages, tuberculous endometritis cannot be distinguished from other forms of inflammation (Schramm 4 ). The inflammation begins upon the surface in the majority of cases (Orthman and Mun- son, 5 Rosenstein 6 ). This type of infection is usually acute, although specimens in the chronic stage are encountered. The mucosa, besides being thickened, is infiltrated with inflammatory products and contains more or less numerous tubercles, in many of which typical giant cells are present. Tubercles may generally be observed in varying stages of development, and are always insterstitial in location. A more or less superficial involvement of the underlying myometrium is usually pres- ent, and in some uteri may be a marked feature of the specimen. The blood vessels of the mucosa are usually enlarged and the infection tends to spread along them and the lymphatic channels. Tubercle bacilli can frequently be demonstrated in stained sections in the tubercles, espe- cially if the lesions are in the acute stage. Caseous Variety. — In the caseous, or as it is sometimes spoken of, the cheesy or infiltrating variety, the macroscopic evidences of the dis- ease are more marked and more characteristic. The uterus is usually somewhat enlarged, the tubes are likely to be involved, and tubercles PATHOLOGY 25 may be present upon the peritoneal surface. Although tubercles upon the peritoneum are of frequent occurrence, they are perhaps less often observed than in the miliary variety. Upon opening the uterus, the myometrium is often somewhat thickened and presents evidence of a chronic inflammation. The entire endometrial cavity may be filled with yellowish or whitish cheesy material, or part of the mucosa may be thickened and reddened and perhaps contain macroscopic tubercles, and other areas may be covered with caseous material. Tuberculous endometritis does not occur with uniform severity over the entire cor- poreal endometrium, but patches of disease are likely to be present, especially in those areas near the tubal ostia, and in all specimens in this location the lesions are prone to be the most advanced. Some areas are likely to be the seat of an advanced change, and others may con- tinue comparatively normal. This characteristic is common to all forms of endometritis, whether tuberculous or otherwise, and has been empha- sized by Hitschmann and Adler, 7 by Strong, 8 and by other observers. During the late stages of the disease, however, the entire endometrial cavity is, as a rule, involved. The superficial and even the deeper lay- ers of the endometrium become necrotic, the endometrium is destroyed, and a marked involvement of the myometrium occurs. As a result of the destruction of the endometrium, amenorrhea is likely to result, and, viewed in conjunction with the uterine enlargement, has led to a mis- taken diagnosis of pregnancy. One or more ulcers may be present; these are lined by necrotic tissue and often partially or entirely covered with cheesy material. As a rule, in both this and the miliary variety of the disease, a tendency exists for the condition to limit itself to areas above the in- ternal os, and in some of the specimens of the cheesy variety of infection, this characteristic is most striking. Only in comparatively rare in- stances is the cervix invaded. Histologic. — The histologic picture depends largely upon the stage of the disease. This variety is more prone to exhibit chronic changes than is the miliary form. The pathologic process may vary from a partial involvement of some portion of the mucosa to a total destruction of the entire mucosa, and more or less of the underlying myometrium. The surface is, as a rule, necrotic, and the deeper portions of the mucosa exhibit the changes common to a chronic inflammation. In tuberculous endometritis the chief changes are in the interstitial portions of the mucosa. The glandular epithelium apparently possesses a partial im- munity. In some specimens observed by the author, a well marked proportion of the glandular epithelium has been present, and even a 26 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS tendency toward a squamous metaplasia has been observed. The epithelium, as a rule, although exhibiting the above mentioned tendency to withstand desquamation, by no means escapes inflammatory reactions, and in many instances the cells are enlarged, swollen, irregular in shape, lose their cilia, and exhibit well marked nuclear changes; the nuclei often are enlarged, nearly filling the cell, and stain irregularly, fre- quently deeply, and may possess well marked hyperchromatic qualities. The surface epithelium is lost before that of the glands, and in chronic cases may be replaced by granulation tissue. The distinguishing histo- logic characteristics of this, as of all tuberculous changes, is the tubercle. These are interstitial in location, show the usual epithelioid appearance and frequently contain tuberculous giant cells. These possess large dis- tinct nuclei, which are often distributed with a certain regularity. The tubercles and inflammatory changes are by no means limited to the mucosa, the underlying musculature usually being more or less invaded. Indeed, in advanced cases no trace of the mucosa may remain, the surface being covered with a layer of necrotic tissue, beneath which is a zone of degenerating and inflammatory granulation tissue, and finally the inflamed myometrium. In still other specimens, necrotic myometrium actually lines the endometrial cavity. Tubercle bacilli can generally be demonstrated in stained sections, if a careful search is instigated. In the acute stages, tubercle bacilli are usually present in considerable numbers and little difficulty is encountered in their demonstration. In chronic cases, the demonstra- tion of the bacillus by staining methods alone is sometimes difficult, as the organisms are comparatively few in number and may possess atypical forms. Myometritis. — As has been stated, a greater or less involvement of the underlying myometrium is of frequent occurrence in advanced cases of tuberculous endometritis. These uteri may be normal or somewhat enlarged. Beyond the fact that a salpingitis, usually bilateral, is generally present, no macroscopic evidences of infection are neces- sarily present. Adhesions over the peritoneal surface may be observed, and in some instances the serous coat is studded with tubercles. In a large proportion of specimens, however, nothing strongly suggestive of the variety of infection can be observed. In tuberculosis perhaps more frequently than in any other inflam- mation calcareous deposits may be formed. These may occur as small flakes or as well defined bone-like particles. In some instances, as a result of inflammatory occlusion at the cervico-uterine junction, a pyo- metra occurs. In this case the body of the uterus is more or less sym- PATHOLOGY 27 metrically enlarged, and is likely to feel softer and suggest an ill defined sense of fluctuation. On opening the uterine cavity, the appearance of the endometrium may suggest tuberculosis. The myometrium is often thickened, and may be the seat of small caseous areas discernible to the naked eye. Histological Examination. — Upon histological examination, the endometrium almost invariably presents the changes previously de- scribed. The myometrium presents the usual evidence of inflammation either of the acute or the subacute or chronic type. In addition, how- ever, tubercles are found scattered throughout the tissue. Many of these contain the typical tuberculous giant cells. As has been previously stated, the inflammation in the myometrium tends to advance along the course of the blood or lymph spaces or between the interstices of the myometrium. Actual intramural abscesses are occasionally present, but these in nearly all instances communicate with the endometrial cavity. The condition is nearly always secondary to endometritis, and as a re- sult the inner layers of the myometrium are likely to be chiefly in- volved; in old chronic cases, however, even the outermost layer of the myometrium may be invaded. Tubercle bacilli are present, but their demonstration in the depths of the muscle is as a rule much more diffi- cult than from the endometrium. Small cheesy particles should be selected for smear preparations, and in stained sections tubercles should be examined. Intramural Abscess. — Intramural abscesses may be small or large, and single or multiple. They are present more frequently in the inner and central layers of the myometrium and less frequently in the external layer. As a general rule, the abscesses are distinctly secondary to in- fection of the endometrium. The pus is usually thick yellowish and often contains degenerated cheese-like particles. These abscesses are perhaps most frequent in the fundus of the uterus in the neighborhood of the cornua. In rare instances, tuberculous abscesses of the myometrium have been observed in conjunction with normal endometrium. In 1840 Osiander 9 reported a case in which there were nine or ten soft tumor-like masses present in the uterus. These were thought to be of tuberculous origin. In view of the ill defined knowledge of the pathology of genital tuberculosis at that period, and the rather meager description, this case must be looked upon with grave suspicion. Madlener 10 has reported the history of a case of secondary tuber- culous infection of an adenomatous polyp, which he believes resulted from an entirely local caseous focus in the myometrium. Zahn ll has 28 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS related a somewhat similar case of infection of a polyp from a tuber- culous ulcer of the endometrium. Wassmer 12 reports from Runge's clinic in Gottingen six cases of tuberculosis, five of the endometrium and one in which there was an abscess of the myometrium. The latter case occurred in a sterile woman, 39 years of age. Wassmer states his be- lief that in this case the infection began from a diffuse tuberculous peri- tonitis, and from thence spread to the uterine musculature and later to the endometrium. The abscesses were moderately large and tumor-like in appearance. Gottschalk 13 cites a case occurring in a virgin, 32 years of age, in which there was a circumscribed tuberculous process clearly intramuscular in location and separated from the endometrial cavity by healthy myometrium. Tuberculous adnexitis was present. Gottschalk believes this proves that tuberculosis can localize itself in the uterine muscles by way of the lymphatics. Papow 14 cites the case of a multi- para, 39 years of age, who, in addition to cervical and adnexal lesions, had an abscess in the anterior surface of the uterus, near the left cornu. Frome 15 performed a vaginal hysterectomy upon a patient 41 years of age for profuse and repeated hemorrhages. Bilateral adnexitis and macroscopic tubercles in the endometrium were present. In the inner- most layer of the myometrium existed numerous tubercles. An intra- mural abscess was found. Watkins w reported the history of a case of an intramural abscess which occurred in a patient 43 years of age. The family history Was negative for tuberculosis, and she was sterile. Five months before operation the patient fell and sustained an injury to the left side of the abdomen, low down. Pain continued for months. The patient was afebrile, but there was loss of weight and strength, and upon examination the condition simulated a uterine myoma. The uterus was enlarged to twice its normal size, and in the anterior wall there was a myoma-like intramural swelling. The adnexa and endo- metrium were normal. On section, the uterine lesion was found to be about 3 cm. in diameter, and was yellowish in color, friable, moist and caseous. No true capsule was present. The uterine focus was found to consist of confluent tubercles exhibiting the typical characteristics of tuberculosis. There was much advanced caseous necrosis and many miliary tubercles were scattered throughout the myometrium. No tuber- cle bacilli were demonstrated in sections, but the histologic diagnosis was confirmed by animal inoculation. The case was probably one of hemogenous infection; the primary focus was not, however, discov- ered. Roberts 17 has reported an interesting case of diffuse tuberculosis of the uterus with abscess formation, which simulated a myoma. The patient was 49 years of age and single. Some time previously she had PATPIOLOGY 29 been curetted for irregular bleeding - . The curettings were not histo- logically examined. The hemorrhages recurred. The lungs were nor- mal. Operation was decided upon. No tubercles were present in the peritoneum, the right tube had been converted into a pyosalpinx and the left into a hydrosalpinx. The uterus was irregularly enlarged and covered with adhesions. The uterine walls were thickened, and contained numerous areas of suppuration from which cheesy worm-like bodies could be squeezed. All the myometrium was more or less involved, but not markedly in the region of the cornua. The endometrial cavity was enlarged, the walls were necrotic and contained purulent material. Sec- tions showed a diffuse tuberculosis, as instanced by numerous areas of caseous degeneration with epithelioid and giant cells. Apparently the same case has also been reported by Stewart. 18 Alessandrie 19 has re- ported the history of a somewhat similar case. The following example of an intramural uterine abscess not com- municating with adnexa or endometrial cavity has been observed by the author: Pathology, No. 4108; age 25 years; shortly after marriage, four years ago, a profuse purulent leukorrhea and symptoms of urethritis appeared, followed by a labial abscess. One child three years ago. The puerperium was complicated by pelvic peritonitis. Since then, sterility and occasional attacks of pelvic peritonitis. Examination on admission to the hospital showed a small tuberculous lesion at the right apex, and a moderately massive pelvic inflammatory disease. It was the latter condition that brought the patient to the hospital. Gono- cocci were demonstrated in the secretion from the cervix and from one of Bartholin's glands. A supravaginal hysterectomy and bilateral sal- pingo-oophorectomy was performed. Convalescence was somewhat pro- longed, but otherwise normal. The pathologic examination of the uterus and appendages showed them to have the usual appearances of pelvic inflammatory disease. The tubes were converted into pyosalpinges. The abdominal ostia were closed and no fimbria could be distinguished, nor were there any tubercles upon the peritoneal surface. One ovary was the seat of a small abscess, evidently the result of an infection of a corpus luteum; the other was enlarged, covered with adhesions, and contained a number of retention cysts. The uterus was normal in size, and in the left cornu, on the anterior surface, was a semifluctuant swell- ing 2.5 x 2 x 1.5 cm. Histologic examination showed this to be an intramural abscess, not communicating with the tube or endometrial cav- ity. No gonococci were demonstrated in either of the adnexa or in the intramural abscess. Numerous tubercles, many of which contained typical giant cells, were present in the tubes and in the intramural ab- 30 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS scess. This case appeared to have been one in which tuberculosis was implanted by hemogenous infection upon the preexisting gonococcal lesions. Whether the intramural abscess was originally the work of the gonococcus, it is impossible positively to determine. This case has been previously reported by the author. 20 Tuberculous Deciduitis. — This condition may result from a pre- existing tuberculous endometritis, or the infection may occur subse- quently to conception. The histologic picture naturally varies with the advancement of the pregnancy and the stage of the lesion. On macro- scopic examination the decidua may be found thickened, or may be normal in depth. It is red, congested, and the surface is likely to pre- sent more or less evidence of necrosis and may be partially covered by fibrin, lymph, and caseous material. In some specimens described the decidua has appeared normal to the naked eye, and in others, although evidently the seat of an inflammation, no changes characteristic or even suggestive of tuberculosis have been observed. Histologic Examination. — Histologic examination shows that the chief changes produced in the decidua by infection with tubercle bacilli are necrosis of the tissue and thrombi in the venous sinuses. Typical tubercles and the formation of giant cells, so characteristic of tubercu- losis in other parts of the female genital tract, do not occur in the de- cidua, although they may be present in the myometrium underlying the basal decidua, or less frequently a tendency toward the formation of ill defined giant cells may be observed in the deeper layers of the decidua. Tubercles are not formed from the decidua cells. Runge, 21 in the first recorded case of tuberculosis of the decidua, commented upon the ab- sence of tubercles and giant cells, and explained it upon the basis of the transient character of the decidua and its slight capacity for prolifera- tion, the latter being proved by the rarity of tumors in the decidua. YVarthin 22 explains the phenomena by stating that the stroma cells in their transformation into decidua cells have already passed into an epithelioid form and are incapable of further proliferation under the action of such stimuli as tubercle bacilli. In the nine cases reported by Schmorl and Geipel 23 tubercles and giant cells were absent from the decidua in every case. Similar findings are observed by Wollstein, 24 Westenhoffer, 25 the author, and others. The degree of necrosis varies markedly in different cases, but this change is usually pronounced and, combined with the aforementioned thrombi, should in all cases put the pathologist upon his guard for this type of infection. The necrotic areas resulting from tuberculosis must be distinguished from the necrosis which is normally present in the placenta at times. In the lesions pro- PATHOLOGY 31 duced by tube-culosis, there is usually marked karyorrhexis of the lymphocytes and of the polymorphonuclear leukocytes, and in other speci- mens a well marked caseous degeneration, all of which points are ab- sent in the normal placenta. In addition to necrosis and thrombi, the usual evidences of a deciduitis are present and are characterized by the production of an inflammatory exudate, composed of serum, small round cells, plasma cells and polymorphonuclear leukocytes, varying in inten- sity according to the stage of the disease. The stroma cells are often edematous and take the stains poorly. The cell outlines are indistinct and the nucleus stain moderately deep. The blood vessels are markedly congested. Even in the same specimen variations in degree of inflam- matory reaction are often marked and are of frequent occurrence. PLACENTAL TUBERCULOSIS Macroscopic Appearance. — The presence of tubercle bacilli in a placenta does not by any means necessitate macroscopic or even histologic changes being present. Tuberculosis apparently exerts no influence on the size of the organ. Nearly all the tuberculous placentas which have been described have corresponded closely in this respect to the normal, and this has been our own experience in a fairly large series of cases. The area in which changes are most likely to occur is at the base near the insertion of the cord. In Wollstein's 24 case, a triangular area 5x7 cm. with the apex near the insertion of the cord was present. This area was yellow, soft and somewhat cheesy. This area of degeneration extended to and involved the membranous surface. In some reported cases a number of cheese-like areas have been observed. These vary in size. The maternal and fetal surfaces are often somewhat rougher than normal. In some specimens small elevations resembling tubercles have been observed. Smears from the degenerated cheesy areas show tubercle bacilli. The cord is as a rule normal. The uterus in these cases is apt to be slightly enlarged, flabby, and the peritoneal surface may show tubercles. Schmorl and Geipel 23 have described four varieties of placental tuberculosis: (1) On the periphery of the villi; (2) in the stroma of the villi; (3) in the basal decidua and (4) in the chorion in- volving also the amnion. Warthin 22 classifies tuberculosis of the pla- centa as follows: (1) Decidual, (2) intervillous, (3) intravillous, (4) intravascular chorionic, and (5) chorioamniotic. This is practically the same as that of Schmorl and Geipel with the addition of the intra- vascular chorionic variety, which is due to the development of tubercle 32 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS bacilli in the blood vessels of the villi or chorionic stems, resulting in a primary lesion of the endothelium of the vessel, followed by secondary thrombosis. The later organization of the thrombus by epithelioid cells derived from the connective tissue of the vessel wall gives an intra- vascular tubercle. Of these five forms of tuberculosis of the placenta, the first is the most common and is especially apt to be observed in full term placentas. The second or intervillous lesions are frequent. Warthin — described this variety as follows : Throughout the inter- villous spaces there are small, round, deeply stained areas composed of firmly granular or hyaline substance, containing lymphocytes and poly- morphonuclear leukocytes in varying stages of disintegration. The ma- jority of these areas are about the size of a pin head, or somewhat smaller. They take the eosin stain more deeply than the hyaline fibrous masses which are formed normally in the intervillous spaces; but the fragmentation and diffusion of the nuclei of the leukocytes give to many of them a bluish tinge. In their general characteristics they re- semble the hyaline thrombi of the decidual vessels. Varying numbers of tubercle bacilli are present, some in the thrombi as well as a few in the intervillous blood spaces. In single sections the tuberculous thrombi often appear lying between or adjacent to villi covered with syncytium, showing apparently no pathologic changes. In other instances they ap- pear to be lying free in the blood spaces. Serial sections, however, show that in old cases the thrombi are attached to a villus at some point where the syncytium has either vanished or was present as a swollen hyaline layer devoid of nuclei. In many cases the necrosis of the syncytium presents a firmly granular appearance, suggestive of a beginning casea- tion. A similar change may be seen in some of the thrombi. In those cases in which the syncytium is absent and the thrombi are resting di- rectly upon the stroma of the villi, the latter in many instances present evidence of epithelioid proliferation at the point of contact. In some instances epithelioid cells and typical Langhans' giant cells are present, extending from the stroma of the villus into the thrombi. Thrombi may be demonstrated, which are being organized by epithelioid tissue arising from the stroma of the villi and are thus changed into typical tubercles. Giant cells are, as a rule, numerous and large. In many thrombi the only evidence of epithelioid changes is found in solitary giant cells, and these, although apparently occupying the center of the thrombus, possess long protoplastic processes continuous with the stroma of the villus. In those thrombi which rest upon necrotic syncytium no tuber- cles or giant cells are found. This tends to prove their origin from the stroma of the villus and not from the syncytium. The primary lesion PATHOLOGY 33 in the production of intervillous tuberculosis appears to be a degenera- tion or necrosis of the syncytium. Here an agglutinative thrombus forms, composed of leukocytes, red blood corpuscles, or blood plaques from the maternal blood. Epithelioid organization from the stroma of the villi next occurs and the tubercles thus formed later undergo casea- tion. Schmorl and Geipel 23 are of the opinion that the epithelioid cells originate either from lymphocytes or from the fixed cells of the stroma of the villi. Warthin 22 emphasizes his opinion that the placenta has no especial protection against tuberculosis. In the event of tubercle ba- cilli gaining access to the maternal blood stream, the chances in favor of placental localization are, he thinks, as great as those of any other organ. This does not entirely coincide with the author's experiences, in that, in the large proportion of cases in which tubercle bacilli are present in the placenta, histologic changes were present only in a small minority. Intravillous Tuberculosis. — Schmorl and Geipel 23 regard this type of tuberculosis of the placenta as very rare, having observed it but once in their series of specimens ; in Warthin's 22 cases, however, although not so numerous as the intervillous tubercles, the intravillous lesions were common. In this variety tubercles are present in the stroma of villi whose syncytium is normal and independent of intervillous thrombi, as shown in serial sections. The lesions present all the characteristics of tubercles, and giant cells may be present in various stages from the first localized necrosis to advanced caseation. The syncytium remains normal until the caseation reaches the subsyncytium layer of the stroma, after which the villus covering becomes necrotic and a thrombus forms at the site of the injured syncytium. Tubercle bacilli can usually be demonstrated without difficulty in the caseous lesions. Warthin very properly points out that the presence of intravillous tubercles in the absence of syncytial lesions must be considered as strong evidence that the bacilli have passed through the syncytium without damaging it, and have produced their characteristic changes first in the stroma of the villus. Further evidence pointing to this conclusion is the absence of thrombi upon the syncytium in the early stages. In addition, the fact that intravillous tubercles are present in cases in which there are no tubercles in the fetus would seem further proof of the above assertion, as it is not probable that the chorionic villi would alone show tubercles, if the dissemination occurred by metastasis directly through the fetal blood stream, while a retrograde metamorphosis seems still less likely. Schmorl and Geipel 23 offer the explanation that the entrance of tubercle bacilli may occur through a defect in the syncytium, or an infection from 34 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS the fetal blood stream through a direct metastasis after passing through the fetal body. Intravascular Chorionic Lesions. — This is a rare lesion, but is probably similar to the foregoing in the method of formation. In this variety tubercles form in the same manner as previously described. These lesions occur in the vessels of the chorion. The thrombi may entirely obliterate the lumen of the vessel, or may partially occlude it. They are similar in appearance to the intervillous thrombi and are deeply staining hyaline or granular masses composed of broken up chromatin. The vessel walls at the site of the thrombi often show beginning necrosis. The epithelioid cells of the tubercles develop from the vessel walls. Warthin demonstrated tubercle bacilli in these thrombi. Chorio-Amniotic Variety. — Warthin 22 states that secondary in- volvement of the amnion by large caseating or epithelioid tubercles of the chorion was observed by him a few times. The portion of the amnion in the neighborhood of the chorionic tubercles was thickened, infiltrated with leukocytes, and showed a beginning caseation. Tubercle bacilli were demonstrated in the caseous area. Schmorl and Geipel 23 have described similar lesions. One or all of these varieties of tubercu- losis may be present in a single specimen. In one of the cases examined by the author some of the tubercles exhibited a well marked tendency toward healing. Similar changes were observed by Warthin 26 in a recently described case. Many of the tubercles in his case showed no caseation, or only slight central caseous changes. Tubercle bacilli could not be found in the healing tubercles, but were demonstrated in those which were caseous. Healing tubercles must be differentiated from infarcts, which can easily be accomplished by noting their circumscribed shape and by the fibroplastic proliferation of the villi induced in the primary intervillous thrombus, which forms a condensed mass of epithelioid cells. Warthin further calls attention to the need for differentiation from small localized areas of syphilitic chorionitis and small infarcts showing reparative changes. He states that in the former the syphilitic process involves only the stroma of the villi and the latter are not fused with the solid fibroplastic or fibroid mass; and in the healing infarcts the villi may be fused, but there is an absence of fibroplastic tissue, or only a small amount present. The heal- ing tubercles may show the outlines of some villi fused into an intervil- lous epithelioid or fibroplastic proliferation. In all cases of doubt, the presence of tubercle bacilli in the smears or sections or, as a final step, animal inoculation will prove the character of the lesion. Tuberculosis of the Fallopian Tubes. — The fallopian tube is the PATHOLOGY 35 most frequent area infected by the tubercle bacilli in the female genital tract. Jellett 27 states that tuberculosis of the fallopian tubes is the commonest form of tuberculosis in women, with the exception of the pulmonary variety. The susceptibility of the fallopian tubes to tuber- culous infection is explained by Pozzi 28 on the ground that the mucosa of these organs offers a favorable nidus for infection in conjunction with the changes which occur at menstruation and is easily accessible to organisms from tuberculous peritonitis. It has been shown experi- mentally in animals that, if fine granules are injected into the peritoneal cavity, some of the material finds its way into the fallopian tubes and can ultimately be demonstrated in the discharge in the vagina. The infection may result from blood carried organisms and by direct exten- sion, or infection may occur by way of the lymphatics, the most fre- quent form probably being a secondary infection from the lungs, although some investigators think infection by way of the peritoneum the most common. This certainly is not infrequent. On the other hand, a hemogenic or other form of infection of the tubes may be followed by a general peritonitis. Direct extension from the endometrium may occur, but the converse is more common. Direct extension may also result from adherent foci, such as tuberculous lesions in the intestines, but here again the converse may occur. For the purpose of pathologic study tuberculous lesions of the fal- lopian tubes may be classified under the heading of perisalpingitis, sal- pingitis, pyosalpingitis and hydrosalpingitis. Perisalpingitis. — This variety of lesion is not infrequently the re- sult of a secondary infection from the peritoneum. Tuberculous infec- tion of the serosa of the fallopian tubes without involvement of the deeper coats is by no means frequent, and, although many specimens are observed which, upon macroscopic examination, present no pathologic changes except adhesions, histologic examination will usually reveal definite involvement of the muscularis or mucosa or of both. In peri- salpingitis the tubes are as a rule normal in size, the abdominal ostium open, and the surface shows adhesions. The adhesions may cause con- siderable distortion of the tube while in situ, but after removal the lesions are less pronounced. Macroscopic tubercles may or may not be present in the serosa. Their presence is, however, usually an indication of an involvement of the deeper coats of the tube. Histologically these tubes present no lesions beyond the above men- tioned adhesions, and characteristic tubercles are comparatively infre- quent, a number of sections not infrequently having to be examined before the etiology of the condition can be determined. 36 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Salpingitis. — As a rule the mucosa is the first portion of the tube attacked, and from thence the infection spreads until finally all the coats are involved, the ampulla being generally the first part of the tube to be invaded. A study of the specimens in the gynecological laboratory of the University of Pennsylvania shows that slightly less than 50 per cent of cases of tuberculosis of the tubes were suspected prior to the histologic examination, and this despite the fact that all specimens are sub- jected to a macroscopic as well as a histologic examination. In Williams's 29 report only 25 per cent were of the suspected variety. The presence of tubercles on the peritoneal surface of the tube, the fact that in this form of infection the abdominal ostium is more prone to remain patulous than in any other variety of infection, the presence of cheesy material within the lumen or adherent to the fimbria at the abdominal ostium, and the fact that these lesions are seldom seen in their early stages, the usual bilateral characters of the infection, are all points that should make the examiner suspicious of tuberculosis. The tendency toward patency of the external abdominal ostium in tubes, the seat of this variety of infection, is most marked. Although contractures at this point are frequent, actual occlusion, as compared with other varieties of tubal inflammation, is unusual, and even when the external end of the tube is entirely closed, the fimbria can usually be seen plastered down over the closed off end of the tube, a condition that is rarely present in lesions the result of organisms other than the tubercle bacilli. In one form of tuberculosis small nodules are present, especially in the isthmus of the tube, somewhat resembling at first glance a small fibroma. This variety is spoken of as a salpingitis isthmiae nodosa. In the early stages tuberculous salpingitis does not as a rule present very acute symptoms, and indeed is usually prone to run a somewhat chronic course, so that in this form of infection subjective symptoms are apt to be less marked than in the gonococcic or streptococcic varieties, and as a result speci- mens are rarely seen in the early stages. An exception to this is some- times observed in autopsy specimens and in late infection from miliary tuberculosis. The above comprise the chief diagnostic features of tuber- culosis of the fallopian tubes; although none are positive they are ex- tremely suggestive of this form of infection. In about one half the specimens nothing even suggestive of tuberculosis can be detected by macroscopic examination alone, the tubes in these instances resembling organs the seat of ordinary inflammatory lesions. For this reason statistics regarding the frequency of tuberculosis of the fallopian tubes are likely to be misleading, unless based upon histologic as well as macro- scopic examination. PATHOLOGY 37 Tubes the seat of tuberculous infection vary widely in appearance. Except in salpingitis isthmiae nodosa, the ampulla is the portion of the tube in which the pathologic process is most marked. The tubes may be normal in size or greatly enlarged. Violet 30 has especially called atten- tion to the hypertrophy of the tubes which may result from tuberculous infection. The surface is usually the seat of numerous adhesions, and as the disease is prone to be chronic, these are likely to be dense. In the late stages or in those cases that are secondary to tuberculous peritonitis, macroscopic tubercles are often discernible. Caseous material is sometimes present at the external abdominal ostium, and is often ob- served adherent to the fimbriae ; in advanced cases the entire peritoneal surface may be coated with yellowish gray cheesy material. The peri- toneum covering the tube is red and inflamed. Similar changes may be observed on the surface of the uterus and ovaries. On section, the walls of the tube may be found much thickened ; in other specimens the walls may be normal in depth. The lumen, unless the abdominal ostium is closed, is, as a rule, not greatly enlarged. The mucosa is generally thick- ened and congested, but as a result of necrosis may be entirely absent. In some specimens, as a result of thickening of the mucosa, the lumen is greatly reduced in size, and on section presents somewhat the appear- ance of the ordinary pseudo follicular hydrosalpinx. The lumen may be macroscopically empty, or may contain creamy pus, cheesy material, or watery fluid. The muscularis is generally thickened and edematous. Occasionally the tuberculous process has apparently been somewhat checked and retrogressive changes are observed. Restoration to the normal is, however, less frequent in tuberculosis than in other forms of infection. In this, as in all forms of adnexal tuberculosis, calcareous deposits are sometimes present, perhaps more frequently in tuberculosis than in any other variety of infection. In salpingitis isthmiae nodosa the tubes may be normal in size or may be somewhat elongated and slightly enlarged in diameter. This form of tuberculosis does not, however, usually result in massive lesions, the chief feature being that small firm fibrous nodules are present, chiefly in the inner and middle third of the tube; these vary from slight enlarge- ments to small tumor-like masses one or two> or even more centimeters in diameter. On section through one of these nodules they are found to be firm and fibrous in consistency; the lumen of the tube, which may pass through the center or eccentrically, is reduced in size, often being no larger than a pin hole. What appear to be multiple lumina are often observed; histologic examination of these, however, shows that they are pseudoglands. 38 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Bell 31 recognizes two forms of tuberculosis of the fallopian tubes, the miliary and the caseous; but while well defined instances of these forms are often observed, more frequently this is not the case. Pyosalpinx. — This as a rule represents the end stage of a salpingitis. In those cases in which the abdominal ostium finally becomes closed, resulting in a pyosalpinx, the fimbria can usually be observed plastered over the closed end of the tube. The actual method of closure of the external abdominal ostium is still somewhat in doubt. Doran, 32 Klein- haus, 33 Opitz, 34 Reis, 35 and Young 36 have devoted papers to a descrip- tion of the manner of closure of pyosalpinges in general.. The last named observer summarizes the various theories as follows, dividing them into two classes : Class I includes those theories based upon the increase in the total length of the, tube wall, which, by expanding in an outward direction, becomes projected beyond the tubal fimbria. According to the theory of Doran and Kleinhaus, the increase in length is dependent on the swelling and increase in the substance of the tube wall, associated with the inflammation. Reis believes the gliding outward of the "peri- toneal ring" over the fimbria is rendered possible by the fact that the walls are loose and redundant subsequent to the collapse of the distended tube. In Class II are included the theories of Opitz and Young. The first explains the process as due to retraction of the muscularis and mucosa of the tube within the serous coat; and the latter claims that the gliding process involves only the inner coat of the muscularis. The so- called perimetritic closure of Doran is explained by the matting together of the fimbria by inflammatory adhesions without preliminary recession. In many instances the intramural portion of the tube probably becomes occluded somewhat earlier than does the external abdominal ostium; This occlusion is the result of agglutination of the mucosa. In some cases this becomes permanent, whereas in others leakage occurs at irreg- ular intervals. In some specimens the occlusion at the inner portion of the tube is largely mechanical, as a result of a kink or bend. The above applies to pyosalpinges in general. Serial sections have been made by the author through the occluded outer end of a number of tuberculous fallopian tubes. From this study it would appear that the fimbria of the tube is attacked early in the dis- ease and that, as a result of infection, it becomes first swollen and then often adherent to the peritoneal coat of the tube, and that, as subsequent closure occurs, the swollen and adherent fimbria being attached outside the tube, cannot be withdrawn inside the lumen, thus accounting for the frequency with which the fimbrias are observed plastered externally on the occluded ends of tuberculous pyosalpinges. The facts that tuber- PATHOLOGY 39 culosis usually attacks the outer end of the tube primarily, and that the onset is often chronic, probably account for the greater frequency with which the fimbria are visible in tuberculous than in other varieties of pyosalpinges. Pyosalpinges of tuberculous origin vary markedly in size, but in some instances grow to enormous dimensions. Some of the largest tubal abscesses which the author has seen have been of this variety of infec- tion. The surface is usually more or less covejed with adhesions, and this is apt to be especially pronounced in those cases which are secondary to a tuberculous peritonitis. The walls vary much in thickness, but as a rule in very large specimens they are moderately thin. This tendency for the walls to be thin in large pyosalpinges is perhaps more marked in the tuberculous than in other types of infection. Indeed, in gonorrheal pyosalpinges it may be stated that the thickness of the walls has practi- cally no relation to the size of the lumen. Not infrequently a pyosalpinx may be present on one side and a salpingitis on the other. The lumen in advanced cases is usually necrotic, covered with cheesy material, and what mucosa remains is red and in- flamed. The contents of the lumen varies ; it is often caseous material, or may be thick creamy pus, sometimes blood streaked ; more rarely the pus is moderately thin and dark in color. Hydrosalpinx. — As the result of a tuberculous infection of the tube, hydrosalpinx occasionally occurs. In these specimens the walls tend to thin out, and the infection is not as a rule active. The usual type of hydrosalpinx is the pseudofollicular variety. The mucosa is generally thickened, the actual lumen often being small, and the tube presents on cut section a honeycombed appearance, the compartments varying con- siderably in size, but generally being small. The contents are thin, watery material, sometimes colorless, but more often presenting a slightly yel- lowish or amber tint. Tubercle bacilli can rarely be demonstrated in these specimens, whereas in the tube, the seat of a purulent collection, the specific organism can often be found with no great difficulty. Histologically, tubes the seat of this variety of infection present the usual evidence of an inflammatory infiltration, generally chronic in char- acter, plus the characteristic tubercles, some of which will be found to contain giant cells. The tubercles are not limited to the mucosa, but may also be present in the muscularis. In the absence of bacteriologic proof, the presence of tubercles is the only characteristic of this variety of infection upon which a positive diagnosis can be based. Certain other characteristics exhibited by tuberculosis are extremely suggestive. Necro- sis of the mucosa, sometimes amounting to an entire absence of this 4 o GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS layer of the tube, is of frequent occurrence, and is found perhaps more often in this infection than in tubes the seat of an ordinary pyogenic invasion. A form of tubal tuberculosis which is not infrequent is that in which the tips of the mucosa folds are agglutinated, forming numerous pseudo glands. These vary in size and shape, but are usually moderately small. The epithelium is not desquamated, but on the contrary exhibits a well marked tendency toward reduplication, sometimes as many as four or five layers being presept in some areas. Together with the reduplica- tion, the individual cells are themselves altered and present more or less irregularity, alike as to size, shape, and staining properties. Often the nuclei are hyperchromatic and occasionally exhibit karyokinetic changes. At first glance, especially in the absence of tubercles or giant cells, such an appearance is strongly suggestive of carcinoma. A more careful ex- amination, however, dispels this theory, whereas the epithelial cells are reduplicated and somewhat irregular and possess deeply staining nuclei; they are not of the cancer type, nor is there any penetration of the basement membrane. To dissipate any further doubt, a search through a number of sections is almost sure to reveal one or more char- acteristic tubercles. A few instances (Lipschitz, 37 Saulman 38 ) have been recorded in which tuberculosis and cancer have been present coinci- dently in the same tube. The previously described lesions can, however, be easily differentiated by the experienced pathologist from carcinoma. The tubes in which this condition has been observed by the author are usually of moderate size and the external abdominal ostium may or may not be closed. Barbour and Watson 39 and others have reported cases of this type, and have observed penetration of the muscularis as a result of proliferation of the epithelium, as well as the formation of strands and masses of epithelium in the substance of the mucosa. Evidence of destruction of the surface epithelium is also present. In forms other than the caseous variety, and excepting the presence of tubercles and tuberculous giant cells and bacteriologic evidence, proliferation of the epithelium of the mucosa is one of the chief characteristics of tubercu- losis in this area. In the pseudocancerous variety there is little or no tendency toward desquamation, even when the disease is advanced. The epithelium cells may enlarge or lose their cilia. The nuclei swell and often occupy almost the entire cell. In the chronic stage of tuberculous salpingitis, there is often an excessive formation of connective tissue, and calcareous formation is not uncommon. PATHOLOGY 41 TUBERCULOSIS OF THE OVARY The ovary possesses a well defined resistance to infection by tubercle bacilli. Indeed, until comparatively recent years tuberculosis of this organ was looked upon as a gynecologic rarity. As late as 1880 Bris- sand 40 stated that there was not a single example of this condition in the museum of the College of France. When it is remembered how frequent is tuberculosis of the fallopian tubes, and that this infection is a hematogenous one in the large majority of cases, and the close anatomic relationship between the tube and ovary, it is surprising that the latter is not more frequently attacked, especially as it is the ampulla of the tube which is usually primarily invaded. Whereas a true infection of the substance of the ovary is not the rule when the tubes are involved, peri- oophoritis of tuberculous origin is quite frequent. Peri-oorphoritis. — This tuberculous affection of the ovary by no means indicates an actual invasion of the ovary by tubercle bacilli. These lesions are generally secondary to tubal tuberculosis, but may be the result of a general tuberculous peritonitis. In either event, they are caused by a deposit upon the surface of the ovary of a tuberculous exudate, which results in more or less thickening of the tunica albuginea and in adhesions to adjacent structures, usually the posterior layer of the broad ligament, the external end of the tube, the omentum, or intes- tine. Upon section, the ovarian substance is usually found normal and developing follicles are generally present. As a result of the thickening of the capsule of the ovary, a tendency toward the formation of retention cysts occurs and one, or more, of these is likely to be present, if the condition has been of long standing. The cysts are usually not large, and the ovary itself is generally nearly normal in size. This disposition of the ovarian structure to remain free from infection, even after pro- longed contamination of the surface, is of importance in deciding upon the type of operation to be performed upon patients the incumbents of tuberculous salpingitis. Oophoritis. — In comparison with the preceding condition, this is a comparatively rare lesion. It is probably usually a hematogenous in- fection, although the possibility of invasion from the surface of a pre- viously contaminated ovary at the time of rupture of a graafian follicle must be considered. As seen in the laboratory or upon postmortem, the infection is usually in the chronic stage, and the ovaries are, as a rule, but little enlarged. Retention cysts are, however, of frequent occurrence, and are usually of the follicular variety. The surface of the organ is 42 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS generally more or less covered with adhesions, and the capsule somewhat thickened. The substance of the ovary may be slightly firmer than nor- mal and somewhat congested, but otherwise no marked macroscopic lesions are usually present. Less frequently, the ovary is enlarged and may contain one or more abscesses, the contents of which are cheesy material or creamy pus. These abscesses may be interstitial in type or may result from the infection, either hematogenous or from without, of either a graafian follicle or corpus luteum. Histologically, the surface of the ovaries presents dense adhesions, which are generally quite avascular. The capsule is more or less thick- ened, not always uniformly. The ovarian stroma is infiltrated with chronic inflammatory products, small round cells, plasma cells, a few polymorphous nuclei, leukocytes and exudate. Often the stroma is edematous. The blood vessels are as a rule congested. Tubercles and tubercular giant cells are here and there present. These are usually sparsely distributed, and a number of sections may have to be studied before the characteristic lesions of this type of infection are detected. As will be observed, ovaries the seat of tuberculosis present no diagnostic characteristics, with the exception of tubercles. When abscesses are present, these possess a lining of caseous or necrotic material. Typical tubercles and giant cells are generally present in moderate numbers in such specimens. Tubercle bacilli can, as a rule, be demonstrated only with difficulty, except in acute or very advanced lesions. LITERATURE i 2 3 4 5 6 7 8 9 io ii 12 13 14 Cora, E. Gyn. Rundsch. 1910. 4:318. Beyea, H. D. Ann. de Gyn. et d'Obst. 1900. 54:169. Mayo, W. J. Mayo Clin. 1918. 10:146. Schramm. Arch. f. Gyn. v. 19. Orthman und Munson. Arch. f. Gyn. 39 :gy. Rosenstein. Monschr. f. Gebh. u. Gyn. 1907. 20:366,966. Hilschmann, von F., und Adler, L. Arch. f. Gyn. 1913. 233. Strong, L. W. Am. Jr. Obst. 19 19. 80:139. Osiander. Hann. Ann. f. d. Ges. Heilk. 1840. 5 :pt. 1. Madlener. Centrbl. f. Gyn. 1894. p. 529. Zahn. Virch. Arch. 115 :66. Wassmer. Arch. f. Gyn. 1899. 57:301. Gottschalk. Int. Cong. Obst. Gyn. Rome, 1902. Papow. Russi Wratch. 1906. 3:12. PATHOLOGY 43 Frome. Centralbl. f. Gyn. 1909. 81 11093. Watkins, T. J. Surg. Gyn. Obst. 1907. 5 :6o3. Roberts, C. H. Proc. Roy. Soc. Med., Sec. Gyn. 191 1. p. 57. Stewart, M. J. Jr. Path. Bact. 191 1. 16:385. Alessandri. La gynecologia moderna. 19 13. Norris, C. C. Gonorrhea in Women. Philadelphia and London, I9I3- Runge. Arch. f. Gyn. 1903. 68:388. Warthin, A. S. Jr. Inf. Dis. 1907. 4:347. Schmorl und Geipel. Munch. Med. Woch. 1904. 2:1676. Wollstein, M. Arch. Ped. 1905. 22:321. Westenhoeffer. Deuts. Med. Woch. 1903. 29:221. Warthin, A. S. Jr. Am. Med. A. 1913. 61 :i95i. Jellett, S. W. A Short Treatise on Gynecology. London, 1908. Pozzi, S. A Treatise on Gynecology. New York, 1897. Williams, J. W. J. Hopk. Hosp. Rep. 1894. 3:114. Violet. Lyon Med. 1912. 119:279. Bell, W. B. The Principles of Gynecology. London, &c, 1910. Doran, A. Tr. Obst. Soc. Lond. Dec. 4, 1889. Kleinhaus. Veit's Handb. 3 :69c Opitz. Ztschr. f. Gebh. u. Gyn. 3 485. Reis, E. Am. Jr. Obst. Aug., 1909. Young, J. Jr. Obst. Gyn. Brit. Emp. 1910. 16:307. Lipschitz, K. Monschr. f. Gebh. u. Gyn. 1914. 39:11. Saulman. Centrbl. f. Gyn. 1892. 16:533. Barbour, A. H. F., and Watson, B. P. Jr. Obst. Gyn. Brit. Emp. 1911. 21 1105. Brissand, E. Arch. Gen. de Med. 1880. 146:129. CHAPTER IV CONGENITAL AND PLACENTAL TUBERCULOSIS Placental transmission of tuberculosis — Conflicting reports of findings — Types, acute, chronic ; errors in technic — Definition of congenital tuberculosis — Discrimination between congenital infection and congenital predisposition — Etiology — Germina- tive infection : Spermatozoic — Variety of infection — Experiments of Waldstein and Ekler — Observations of medical experts — Unfertilized ovum — Ovarian infection and germinal transmission of disease — Congenital germinative tuberculosis — Placental and fetal tuberculosis — Susceptibility — Opinion of Baumgarten and others — Tubercle bacilli in the blood stream — Histology and physiology of the placenta in relation to routes of transmission of tubercle bacilli — Views of Delore and other investigators — Infarcts described by Williams — Results demonstrating congenital or placental tuberculosis — Distinction between placental infection and fetal involvement — Criticism — Period at which intra-uterine transmission occurs — Predisposing factors to placental or congenital tuberculosis — Undoubted cases — Anatomical changes and presence of tubercle bacilli — Histologic changes and presence of tubercle bacilli — Demonstration of bacilli by staining or by inocula- tion of animals — Conclusions. The subject of the placental transmission of tuberculosis and of placental pathologic lesions, the result of this infection, is of especial in- terest. Quite contradictory findings have been reported regarding the frequency of placental tuberculosis. Thus, Schlimpert x reported having found placental tuberculosis in 80 per cent of a series of cases ; Novak and Ranzel, 2 in 70 per cent ; Schmorl and Geipel, 3 in 40 per cent. Pankow, 4 on the other hand, in a series of 20 placentas, failed to demonstrate a single case, and Bossi had a similar experience. As a result of these and other equally conflicting reports, it has seemed advisable to gather and study the results secured in an extensive series of cases, with the hope of throwing some light upon the actual frequency of placental and congenital tuberculosis. At the outset it became apparent that the diver- gent results obtained by various investigators were dependent chiefly upon three factors : ( 1 ) The standard set for the tubercle bacillus — whether the staining of acid fast bodies morphologically similar to the tubercle bacillus was to be accepted, or whether inoculation, culture, or histologic changes were to be demanded before determining the exciting cause; (2) the different types of cases from which material was ob- tained — acute and chronic; (3) errors in technic. 44 CONGENITAL AND PLACENTAL TUBERCULOSIS 45 CONGENITAL TUBERCULOSIS Tuberculosis was probably recognized many hundreds of years prior to the Christian era (Williams, 5 Osier, 6 Waldenburg, 7 Predohl 8 and Johne 9 ). Hippocrates (460-376 B. C), Galen (200-131 B. C), and Celsus (30 B. C.) described the disease, and to-day the mortality statistics show that, of all deaths, from nine to twelve per cent are due to this affection (Rosenau 10 ). It is not strange, therefore, that the etiology of so ancient and wide spread a scourge as tuberculosis should have received careful study. It was early observed that the children of tuberculous parents were much more frequently attacked by the disease than were the offspring of healthy progenitors. Prior to the discovery of the tubercle bacillus by Koch in 1882, the theory that tuberculosis was of congenital origin re- ceived much consideration. Subsequently, however, the belief began to lose ground, and the frequency with which tuberculous offspring were born of infected parents was explained by the doctrine of postnatal in- fection, aided, perhaps, by a hereditary predisposition. That the major- ity of cases are thus caused has been proved beyond doubt. Recent in- vestigations, however, by Schmorl and Geipel, 3 Novak and Ranzel, 2 Sitzenfrey, 11 Warthin and Cowie, 12 and others tend to show that not only does congenital tuberculosis occur, but that it may be relatively more frequent than is generally assumed. Definition. — The name "congenital tuberculosis" should be reserved for those cases in which tubercle bacilli are present in the fetus at or prior to birth. A sharp discrimination must be made between congenital infection and congenital predisposition. Numerous attempts have been made to classify infection of the embryo or fetus. Martius 13 has strongly emphasized the distinction between the terms "congenital" and "inherited." He applies the term "congenital" to any condition that may be present in the child at the time of its birth, and the term "inherited" only to that condition which develops as the direct result of the conjugation of the two sex cells ; in other words, anything that is given to the new organism from the germinative plasma. The terms "congenital" and "inherited" are somewhat confusing. It would seem advisable to use the term "congenital" to cover both varieties of infection, and, in these cases in which it may be necessary to differentiate between the two forms, to use the term "germinative" as descriptive of an infection caused by the spermatozoon or ovum — the germinative cell — applying the denomina- 46 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS tive intra-uterine or placental infection to those cases having hematog- enous or other intra-uterine or placental origin. Etiology. — Congenital tuberculous infection may occur in a number of ways ; it may be due to the spermatozoon or the ovum — the germina- tive infection; or the products of conception may subsequently be in- fected as the result of a maternal bacillemia ; or it may be the result of a direct extension from surrounding structures, either by continuity or through adjacent lymph channels. Most authorities agree that tubercle bacilli are present in the blood stream under certain conditions, especially in the acute miliary form of the disease and in the terminal stages; that they are probably not con- stantly present, are frequently few in number, are generally difficult to demonstrate, and that slight errors of technic may lead to erroneous con- clusions (Rump, 58 Liebermeister, 59 Giirner, 60 Dressen, 61 Gobel, 62 Klemperer, 63 Kahn, 64 Kessler, 65 Bacmeister, 66 Vinogradoff, 67 and others). The writer believes, with Fraenkel, 68 that the microscopic ex- amination of the blood for tubercle bacilli is likely to prove misleading, and that the inoculation of animals is the only possible means of arriving at correct conclusions. Of 22 persons examined by Fraenkel, only two gave positive results. Elsasser 69 tested 41 cases of advanced tuberculosis, and was able to demonstrate the microorganism in J. 2, per cent of cases. Bogason 70 recovered the organism in only two of 41 patients, although he employed 10 c.cm. of blood. The work of Massel and Breton recently reported by Calmette 71 is of especial interest in this connection. These investigators found that tuberculosis could be produced with relative frequency by the direct transfusion of blood from a tuberculous to a healthy guinea pig. By this method it was possible to transmit tuber- culosis quite frequently, even from animals in whom the disease was chronic or the lesions comparatively small. Pregnancy is prone to light up a latent or chronic tuberculosis, and thus produce a condition in which a bacillemia is likely to be present. Secondary infection and metastasis occur in the placenta in the same manner in which they affect other portions of the body. Dardeleben goes so far as to assert that the placenta is the locus minoris resistentiae of the gravid woman. GERMINATIVE INFECTION Spermatozoic. — Tubercle bacilli have never been demonstrated within the spermatozoon. In order to produce infection, however, it is not necessary for the bacilli to invade the spermatozoa, for a tubercle CONGENITAL AND PLACENTAL TUBERCULOSIS 47 bacillus adherent to the outer surface of the cell may effect a similar result. A tubercle bacillus may become attached to a spermatozoon at any point along its course — testicle, vas deferens, prostatic fluid, urethra, external surface of penis, vulva, vagina, cervix, uterus, or even the fallopian tube. It is, therefore, theoretically possible for an ovum to become infected by a spermatozoon, the tubercle bacillus having orig- inated in the woman or been derived from an exanthropic source. That spermatozoa may be the germ carriers in diseases other than tuberculosis, has been demonstrated (Bab, 14 Sakurane, 15 Fouquet, 10 Feuillee, 17 and others), and although the likelihood of such an event occurring varies markedly in the different diseases, the possibility of their being the carriers of tuberculous infection must be considered. The presence of organisms other than the tubercle bacillus in or at- tached to the spermatozoon does not always inhibit the activity of the latter. In considering the variety of infection, the experiments of Wald- stein and Ekler 18 are of interest. These authors report that in normal rabbits the biologic tests appeared to show that in the female organism absorption of the spermatic fluid occurs. This observation will, how- ever, require further verification. Tuberculosis of the male genito-urinary tract is by no means in- frequent. Viet 19 and Martin 20 assert that involvement of the genital tract occurs in three per cent of tuberculous males. Guiteras 21 reports that, next to the gonorrheal, the tuberculous variety is the most fre- quent form of epididymitis. When tuberculosis of the genital or urinary tract is present, the semen frequently contains tubercle bacilli; on the other hand, in some cases, notably in that of D'Aubeau, 22 the discovery of the tubercle bacillus in the semen and the absence of lesions in the genito-urinary tract were the first evidences of the existence of pul- monary phthisis. Jani 23 and others have reported the finding of tuber- cle bacilli in the testes of phthisical men in whom no demonstrable genital lesions were present. Jani found the bacillus present in 5 of 8 cases examined. Sirenae 24 injected the semen from a tuberculous pa- tient into dogs, which then developed tuberculosis, Solles 25 and Foa 26 report similar results. Spano, 27 in six cases of phthisis, demonstrated the presence of tubercle bacilli in the seminal vesicles in five. Jackh 28 likewise demonstrated the presence of tubercle bacilli in the testicular secretion in two cases of acute miliary tuberculosis. Somewhat similar results were also obtained by Lowenstein. 29 Theoretically, tubercle bacilli free in the blood should not gain access to the testicular or prostatic fluid, but should become enmeshed in the 48 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS fine capillaries leading to the glandular structures of the testes or pros- tate; Grawitz, 30 however, showed that the mold germs, which have a larger diameter than tubercle bacilli, may under certain conditions reach the testicular secretion by way of the blood stream. Murphy 31 calls attention to the fact that it is extremely difficult, in some cases, to make a diagnosis of tuberculous seminal vesiculitis, and that probably many of the cases in which the genitalia have been considered normal have in reality been instances in which this focus has been over- looked. The work of Rohlff 32 and Westmayer 33 tends to support Murphy's opinion, in that these investigations have demonstrated that tubercle ba- cilli are rarely present in the semen of phthisical men, if genital lesions are absent. Rohlff inoculated goats and rabbits with the spermatic fluid obtained from ten men who had died of pulmonary tuberculosis, with negative results. Westmayer injected the ground up particles of the testicles of similar subjects into the peritoneal cavity of rabbits, with like results. Dobroklonski, 34 by means of smears and inoculations, tested the semen of 25 men who had died of pulmonary phthisis. Twenty-four were negative, the one positive result being obtained from a subject in whom a tuberculous epididymitis was present. Walther 35 examined 161 sections made from the testes, epididymes, and prostates of 12 phthisical subjects, without finding a single tubercle bacillus. Gartner 36 injected a pure culture of tubercle bacilli into the testes of 22 rabbits and 21 guinea pigs; he then mated these animals with 65 females. In none of the 29 rabbits or 45 guinea pigs which were born did tuberculosis develop, except in one, the infection in this case prob- ably being caused by food. Cornet 37 was unable to demonstrate the presence of tuberculosis either microscopically or by culture methods in 32 fetuses and young animals bred from guinea pigs the male parents of which had been inoculated in the testes, prior to breeding. Numerous instances are on record in which fetal tuberculosis has been produced experimentally in animals by injecting cultures of tubercle bacilli into the vagina just before or immediately after coitus (Friedman, 38 Varaldo, 39 and others). These results are, however, valueless, for a maternal infection followed by a hemogenic infection of the products of conception was probably the etiologic factor. A tuberculous ulceration of the penis may also be the means of introducing tubercle bacilli into the vagina with the semen. Cornet mentions the possibility of tubercle laden sputum being used as a lubri- cant during coitus, with resulting infection. From what has been said it CONGENITAL AND PLACENTAL TUBERCULOSIS 49 would seem fair to assume that although germinative congenital tuber- culosis of spermatozoic origin may occur, it is extremely rare. Unfertilized Ovum. — Infection of the ovum may take place in the ovary, either before or after rupture of the graafian follicle, in the peri- toneal cavity, fallopian tube, or even in the uterus, although it is generally accepted that fertilization of the ovum usually takes place in the fallopian tube. As in the spermatozoic infection, the tubercle bacilli may be either in or attached to the ovum; in the latter event it may subsequently gain entrance with the fertilizing spermatozoon or very shortly afterward. The deeidua reflexa is probably formed almost immediately after the entrance of the fertilized ovum into the uterus, so that the event last intimated is extremely unlikely. Sitzenfrey, 11 in one case, found tubercle bacilli situated in a primor- dial follicle of a human ovum. The patient was eighteen years old, and four years previously had had a peritonitis, presumably of tuberculous origin. At operation both adnexa were found to be tuberculous. Schott- lander 40 has produced tubercles and giant cells experimentally within developing graafian follicles in rabbits. Landouzy 41 believes that in rare instances infection of the ovum from a tuberculous oophoritis or salpin- gitis may occur. That intra-ovarian infection of the ovum does take place has been definitely proved, but that extra-ovarian infection occurs rests only upon a theoretic basis. It is doubtful if an ovum in- fected within the ovary, if fertilized, could develop. Ova infected out- side the ovary would naturally possess a slightly greater chance of developing. Ovarian infection and germinal transmission of disease have been demonstrated by Rettger's 42 investigation of bacillary white diarrhea in the common domestic fowl. Chicks which survive frequently become permanent bacillus carriers, the ovary being the important seat of infec- tion. The eggs from such carriers often harbor the organism of the disease in the yolk, and chicks from these eggs are congenitally infected. The fact that when intra-ovarian infection does occur, the fallopian tubes are usually involved, and are often occluded, may to a certain extent prevent the more frequent fertilization of such ova. That the ovum may be infected by microorganisms other than the tubercle bacillus has been amply proved (Hoffmann and Wolters, 43 Levaditi and Saur- age, 44 Bab, 45 Magalhaes, 46 Koch, 47 Simmonds, 48 and others). Congenital Germinative Tuberculosis. — From what has been stated it may be seen that germinative tuberculosis may take place in three ways : ( 1 ) the tubercle bacilli may enter the ovum with the fertiliz- ing spermatozoon, either attached to the surface of, or actually within, 50 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS the male germinative cell; (2) the tubercle bacilli may be lodged within the ovum, which is later fertilized; (3) or the tubercle bacilli may have been attached to either the spermatozoon or the ovum and gain access to the latter shortly after fertilization. The question as to whether an infected fertilized ovum would develop is open to grave doubt. We are willing to admit that, theoretically, germinative infection may occur; but viewed from a practical standpoint, this form of tuberculous infection is probably too rare to be seriously considered as a factor in congenital tuberculosis. Further investigation of this subject is neces- sary, and a careful study of the embryos and early gestation sacs of tuberculous parents would doubtless yield much information. The dem- onstration of a germinative infection is obviously extremely difficult, and most authors agree with Cornet that its existence has not yet been proved. PLACENTAL AND FETAL TUBERCULOSIS It is a generally accepted fact that fertilization of the ovum takes place within the fallopian tube — probably in its outer portion — and that from this point it is carried along by the action of the cilia of the surface tubal epithelium to the uterine cavity, where it becomes implanted. It is possible, therefore, for tubercle bacilli to enter the fertilized ovum at any point along its course. Obviously, the uterus is the most likely point for infection to take place. This may occur by four different routes, ( 1 ) hematogenic; (2) lymphogenic; (3) by direct extension through con- tinuity; and (4) by tubercle bacilli gaining access from without. Susceptibility to Tuberculosis. — It would seem advisable, at this point, to digress and to discuss briefly the action exercised by the ma- ternal toxins and antibodies upon the presumably hitherto uninfected products of conception. This subject has received much attention and been widely discussed by Hollos, 49 Rosenau and Anderson, 50 Huppe, 51 Bartel, 52 Klebs, 53 and many others, and the question as to whether the embryo, fetus, or child of a tuberculous mother is hyposusceptible or hypersusceptible to the action of the tubercle bacilli is still in doubt. When we consider how susceptible the fetus is to the maternal tuber- culous toxins, it would seem that the general nutrition must become impaired. Carriere 54 showed experimentally that in animals tuberculous toxins influenced pregnancy by reducing the number of the offspring, and that in many instances these died in utcro or shortly after birth, or that those that survived were weak. The effects were most marked when toxins CONGENITAL AND PLACENTAL TUBERCULOSIS 51 from both parents were injected. This investigator believed that the surviving young animals were more susceptible to tuberculous infection than were the control animals. Ballantyne states that, once the tubercle bacilli have gained access to the fetal tissues, they find there an excellent soil for development. Pankow 4 inoculated a number of guinea pigs with portions of placentas obtained from twenty cases of suspected or congenital tuberculosis. Three of these pigs died within a few days, presumably from a toxemia, for in none of them was it possible to dem- onstrate the presence of tubercle bacilli. Bossi 55 injected the ground up particles of placentas of tuberculous women into guinea pigs, and found that marked evidences of toxemia resulted. The effects were more lethal when placentas from women far advanced in tuberculosis or in poor general health were used, and in those from patients who showed large numbers of tubercle bacilli in the sputum. Control experiments with eight placentas coming from healthy women gave negative results. Bossi, therefore, concludes that there are in the placenta of tuberculous women toxins that are transmitted to the fetus and that may cause death or miscarriage, or result in the birth of weaklings. Cornet's 37 views agree with those of Bossi, and he believes that the toxins result from a process of osmosis. He does not, however, consider that such toxins increase the susceptibility to tuberculosis, but, on the contrary, he believes that the fetus in utero acquires a certain amount of immunity. With this latter view, Sitzenfrey, 11 Hollos, 49 Warthin, 12 and the author are in accord. 12a Many excellent observers, however, hold a contrary opinion. Pehu and Charlier 56 believe that the offspring of tuberculous parents are prone to defective development. They think that these children undoubtedly present a receptive soil for all diseases, but not especially to tuberculosis. Undoubtedly, many cases have been recorded in which tubercle bacilli were positively demonstrated in large numbers in the fetus or new born child, no other pathologic changes being present — a finding that requires further study. Tuberculous Bacillemia. — The frequency with which tubercle bacilli occur in the blood stream in infected individuals is a somewhat disputed point. The Histology and Physiology of the Placenta in Relation to the Routes of Transmission of the Tubercle Bacilli. — For a thorough understanding of placental and congenital tuberculosis a knowledge of the pathological processes that occur in these conditions is necessary. An important point that immediately presents itself to the investigator is whether or not the transmission of the tubercle bacilli occurs through the 52 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS normal placenta. A number of cases are on record in which congenital tuberculosis has been observed in the child and no histologic changes were detected in the placenta. Even more numerous are the cases in which tubercle bacilli were demonstrated in the placenta, or fetus and placenta, and in which no histologic changes were found. In accepting these cases as genuine, care must be observed, as the demonstration of tubercle bacilli in the blood, either by staining methods or by animal inoculation, is not reliable unless a strict technic is adhered to. It must also be remembered that, in order positively to prove that the placenta in a given case was normal, the entire structure must be subjected to serial section. So far as can be ascertained, this stupendous task has never been at- tempted, and even if it were, the possibility that healing might have occurred in the placenta after the tubercle bacilli were transmitted could not be entirely excluded. Warthin and others have described the healing of tuberculous lesions in the placenta. The placental transmission of syphilis, leprosy, variola, anthrax, pneumonia, and recurrent fevers has been positively demonstrated in man (Lubarsch, 72 Schaudinn, 73 Paschen, 74 Wallich and Levaditi, 75 Menetrier and Rubeno-Duval, 76 Neuhaus, 77 Freund and Levy, 78 Van der Wittigen, 79 Dorland, 80 Runge, 81 Nattan-Larrier and Brindeau, 82 Delestre, 83 Bar' and Renon, 84 and others). Preyer, 85 Savory, 86 Fournier, 87 and others have demonstrated that toxins injected into the mother may produce the death of the fetus. As early as 1877, Zweifel 88 showed that chloroform administered to the mother also affected the fetus, and more recently the work of Jung 89 has shown the passage of certain drugs through the placenta. Under normal circumstances the blood in the intervillous spaces is entirely maternal in origin (Waldeyer, 90 Bumm, 91 Leopold, 92 etc.), as the fetal blood at no time gains direct access to the intervillous spaces, the two blood supplies being separated from each other by the vessel wall and the two layers of chorionic epithelium. During the latter stages of pregnancy Langhans' layer is, however, absent. It seems probable that, when material is transmitted through the placenta, the process is effected partly by osmosis and partly by the direct action of the syncytial cells, the physiology of the latter being somewhat analogous to that of the renal tubules. Williams, 93 and Cornet 37 are of the opinion that when the placenta is normal and the epithelium covering the villi is intact, transmission of the disease germs cannot occur, but that when lesions of the placenta are present, transmission may take place. It remains to be decided, however, whether the lesions that have been demonstrated in some cases of tuberculosis have antedated the disease, or whether thev CONGENITAL AND PLACENTAL TUBERCULOSIS 53 have been the result of tuberculous toxins produced by the bacteria in the intervillous blood. Delore 94 is of the opinion that the disorganization of the syncytium is not the result of toxins or of inflammation, but is due to a myxomatous degeneration. Warthin 95 believes that the syncy- tium of the chorionic villi is no more immune to the action of the tubercle bacilli than is the vascular endothelium in other parts of the body, and that the theory that tubercle bacilli can pass through this layer of cells without causing injury to it is founded on fact. Sitzenfrey ll is of a similar opinion. In this connection it should be stated that the inter- villous spaces are not lined by endothelium, except for a short distance on the surface of the decidua basalis, into which the endothelium of the maternal vessels extends. The author is of the opinion that tubercle bacilli may be transmitted through the normal placenta. This opinion is based upon the fact that a number of cases are on record in which un- doubted congenital tuberculosis has been present and a careful examina- tion of the placenta has failed to show histologic lesions in the same. In the case of tuberculosis, at least, this question is perhaps of more theoretic than of practical value. It is probable that the toxins, the result of enmeshed tubercle bacilli, may produce injury to the syncytium, and thus by the damage incurred, prepare the way for the entrance of tubercle bacilli. Furthermore, in the latter months of pregnancy, infarcts are frequently present, and doubtless constitute foci by which ingress is secured for the organisms in the intervillous spaces. Williams describes five varieties of infarcts. All placentas contain small infarcts, and Williams states that these attain a diameter of one centimeter or more in 63 per cent of cases. Owing to the histologic structure of the decidua basalis and placenta, these structures offer espe- cial facilities for the enmeshing of microorganisms circulating in the maternal blood stream, a point that has recently been emphasized by Warnekros, 96 It is probable that, in the majority of cases in which congenital infec- tion occurs, the tubercle bacilli travel through the decidual arteries to the covering of the villi and there accumulate, causing thrombi in their own and in adjacent intervillous spaces, with subsequent destruction of the syncytial cells, enter the villous stroma, and finally reach the chorion. To what extent this process is aided by the action of the toxins is not definitely known, but it would seem probable that they act as predispos- ing factors and tend to weaken or even destroy the intervening layers of cells. Schmorl and Kockel, 97 in their carefully prepared report of the path- ology of placental tuberculosis, state that placental villi, even when en- 54 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS tirely embedded in the tuberculous areas, tend to retain their integrity and are easily differentiated from the surrounding tissue. "Even the identical villus on which the primary localization of the tubercle bacilli occurred, and which, in consequence, lacks its cellular sheath in places, remains for a long time unchanged. The tubercle bacilli may be very plentiful in the tuberculous new growth, yet within the villi surrounding this area we find but few." These authors further declare that, even if the villus becomes tuberculous, a thrombosis and partial obstruction to the vessel occur, which may entirely or in part prevent the passage of the bacillus. As has previously been stated, the infarcts that are so frequent in the latter months of pregnancy also probably serve in some instances as channels for the invading microorganisms. The virulence of the infect- ing agent and the resistant power of the host are also probably important factors in the production of the disease. It is a significant fact that a large proportion of the reported cases of congenital tuberculosis have occurred in conjunction with the acute miliary variety of the disease. Undoubtedly the strong uterine contractions incident to labor con- stitute a most important factor in the transmission of tubercle bacilli at the end of pregnancy. Organisms that, prior to the onset of labor, were lodged in the placenta or in the intervillous spaces, may, as the result of these contractions, be forced into the fetal circulation. Schlimpert, 1 Schmorl and Geipel, 3 Warthin and Cowie, 12 Dardeleben, and others are very insistent on this point. Tubercle bacilli are relatively frequently transmitted through macroscopically normal placentas, and may possibly pass through histologically normal organs, although positive proof of the latter is lacking. In addition to the hematogenous infection, tubercle bacilli may reach the decidua by direct extension from the fallopian tubes or cervix, and thence, by continuity, pass into the placenta. A lymphatic infection from an adjacent tuberculous lesion may also occur. In either of these ways a focus of infection is set up in the decidua, and may extend to the chorion, thus reaching the body of the fetus, and infecting it through the respiratory tract, the result of inspiration of the amniotic fluid, through the gastro-intestinal tract, or through the skin. Asch 98 and numerous other observers have recorded instances of supposed intra-uterine suck- ing, and even in adults, whose dermis should be more resistant than that of the fetus, infection without macroscopic loss of continuity has occa- sionally been observed (by Leloir," Baginsky, 100 and experimentally by Wasmuth, 101 Roth, 102 and others). In a case described by Schmorl and Geipel 3 a tuberculous area in the chorion had penetrated the amnion, CONGENITAL AND PLACENTAL TUBERCULOSIS 55 and tubercle bacilli were found on the surface of the membrane. Herr- gott 103 inoculated guinea pigs with the amniotic fluid of a tuberculous woman who died in the sixth month of pregnancy. The animals devel- oped tuberculosis, showing that, in this case, the amniotic fluid contained virulent tubercle bacilli. Aside from the infarcts previously mentioned and the fact that high fever is likely to produce a loss of continuity of the maternal and fetal blood vessels and thus favor transmission of the bacillus, disease of the placenta other than tuberculosis may produce lesions that will facilitate the occurrence of a congenital infection by opening up avenues for the entrance of tubercle bacilli. This is particularly the case in syphilis, of which Hochsinger's 104 case is an example. Trauma may also serve as a predisposing factor in the transmission of the bacillus through the placenta. In examining specimens of suspected congenital tuberculosis, the greatest care must be observed to exclude cases of possible extra-uterine infection. As pointed out by Virchow, syphilis may closely simulate tuberculosis. Henle 105 described a case of pseudotuberculosis in new born twins. It must be remembered that placental tuberculosis does not necessarily imply a transmission of the infecting organism to the fetus, although, of course, the condition strongly favors congenital tuberculosis, for, if advanced, it must produce lesions that facilitate the passage of the bacilli through the placenta. Frequency of Congenital Tuberculosis. — Tuberculosis is the most frequent serious infectious disease that attacks mankind. It has been estimated that from 9 to 12 per cent of all deaths are due to tuberculosis. In Germany, during one year, the mortality statistics show that diph- theria, pertussis, scarlatina, rubeola, and typhoid fever were accountable for 116,705 deaths, whereas during a similar period tuberculosis was responsible for 123,904 deaths. Genital tuberculosis is by no means an uncommon affection. Genital lesions are predisposing factors to congenital tuberculosis, especially if the fallopian tubes are patulous. The frequency of this form of infection is, therefore, of especial interest. In the gynecological laboratory of the University of Pennsylvania it has been found that seven per cent of all cases of pelvic inflammatory diseases are of tuberculous origin. Williams 5 states that eight per cent of all cases of adnexitis are tuber- culous. Merlitti 106 places the proportion at 12.6 per cent. The reports from the University laboratory and from Williams are based upon operat tive material, and are of especial value, as in both clinics all specimens 56 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS are subjected to a routine histologic examination. Martin 20 found twenty-four tuberculous cases among 1 ,600 gynecologic specimens. Bark- ley, 107 in 789 autopsies performed upon women dead of tuberculosis, found the genital tract involved in 7.7 per cent. In 174 cases of similar material from the Henry Phipps Institute, examined by the writer, 6.6 per cent showed macroscopic involvement of the genitalia. In studying the frequency of congenital tuberculosis a number of points must be considered. The great prevalence of tuberculosis and the comparatively small number of authentic cases of congenital infection that are recorded are conclusive proof, it would seem, of the rarity of the variety of the disease in question. On the other hand, it should be remembered that tuberculosis of the placenta does not by any means always present macroscopic lesions. There is, moreover, probably no branch of pathology that has received less attention than the histologic study of the placenta. Baumgarten's theory, although doubtless extreme in some respects, has done much to show that congenital tuberculosis may occur, and that tubercle bacilli may remain latent in the child for quite prolonged periods. It has been shown that the tubercle bacillus may remain latent for some time. Under such circumstances congenital tuber- culosis is probably mistaken for, and classified as, a postnatal infection. The transmission, through the human placenta, of microorganisms of other diseases is a point tending to show that congenital tuberculosis may be more frequent than is generally supposed. Until comparatively recently it was the general belief that congenital tuberculosis rarely, if ever, occurred, and for this reason but few pla- centas were examined in order to determine its existence. As has been stated, even a negative histologic examination does not by any means exclude the presence of tubercle bacilli in the placenta, and it is only by routine histologic and bacteriologic examinations of a large series of placentas and other products of conception from tuberculous women that reliable conclusions can be reached regarding the frequency of placental and congenital tuberculosis. Owing, probably, to the difficulties of secur- ing such material, a sufficiently large number of such examinations have not been made. Sitzenfrey's X1 series of 26 cases is the largest found in the literature. When such studies have been carefully carried out, the results have almost invariably shown that the presence of tubercle bacilli in the placenta is by no means infrequent. It is a significant fact that recent investigators have found both congenital and placental tuberculosis much more frequent than did those of the previous decade, the result, probably, of the more thorough methods of study now employed. Bossi 55 failed to find tubercle bacilli in any of the placentas exam- CONGENITAL AND PLACENTAL TUBERCULOSIS 57 ined by him. Pankow, 4 in a series of 20 placentas from tuberculous women, was unable to demonstrate tuberculosis in a single case. Despite this fact, however, this investigator believes that placental and congenital tuberculosis are not rare. Schmorl and Geipel, 3 on the other hand, found eight cases (40 per cent) of tubercle bacilli in the placenta in a series of 20 tuberculous women. Schlimpert x found tubercle bacilli in the pla- centa in eight of eleven cases. Sitzenfrey, in a series of 26 cases, found the organism in seven, and recovered the infecting bacilli twice from the fetal blood. Novak and Ranzel 2 examined the placentas from ten cases of advanced pulmonary tuberculosis. The placentas were minced, washed in sterile water, digested in soda solution and pancreatin, and then mixed with 40 per cent antiformin solution. The sediment was again washed in alcohol, stained, and examined for tubercle bacilli. A histologic exam- ination of the placentas was also made, and inoculation of guinea pigs was likewise carried out in many cases. As a result, these authors found positive evidence of tuberculosis in seven of the ten specimens examined. They regard the negative findings of other observers, especially those of Bossi 55 and Ascoli, 108 as due to faulty technic ; or that, probably, as in some cases, the specimens were obtained from early pregnancies, in which case the infection would most likely be limited to the decidua. A summary of the cases of Schmorl and Geipel, Novak and Ranzel, Schlim- pert, and Sitzenfrey shows that of 67 cases examined 30 per cent pre- sented positive evidence of tubercle bacilli in the placenta, of placental tuberculosis, or congenital tuberculosis. Many of the earlier results secured regarding the demonstration of congenital or placental tuberculosis are open to criticism because of the methods employed. Von Leyden 109 inoculated animals with por- tions of the liver, spleen, and lungs of a child born of a tuberculous mother, with negative results. Jaquet 110 was unable to find tubercle bacilli microscopically in several human fetuses of tuberculous mothers. Vignal 1X1 inoculated portions of the livers and spleens of eleven human fetuses of tuberculous mothers into guinea pigs, with negative results. Treisser 112 performed similar experiments with the livers and lungs of three still born infants of tuberculous mothers, with similar results. Bernard, Debrer, and Baron, 113 in a series of 36 cases of advanced tuberculosis, found the placenta involved in 12.5 per cent. Bar and Renon 114 found tubercle bacilli in the blood from the umbilical cord in two of five cases in which the mothers were tuberculous. It should be stated, however, that one, at least, of Bar and Renon's cases was not above suspicion. 58 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Armann 115 regards tuberculosis in infants as not infrequently of congenital origin. As a general rule, in hematogenous infections tubercle bacilli attack the lymphatic glands nearest their point of entry. In con- genital tuberculosis, therefore, the liver, being the inlet for the placental blood, the lymphatic glands in this region would exhibit the first changes, and this is frequently the case. That the liver possesses some bacter- icidal properties should, however, be taken into consideration. In those infants in whom the liver is the organ chiefly involved, this fact is at least suggestive of congenital tuberculosis. Leroux, 116 in 214 autopsies performed upon tuberculous infants, found the liver affected in eighty- two. Lannelongue, 117 in 1,005 cases of surgical tuberculosis occurring in young children and infants, observed three that he considered of con- genital origin. Muller, 118 in 150 autopsies performed on tuberculous children, found the liver involved in 33.3 per cent of cases. Haupt 119 was able to demonstrate that, of 617 of his tuberculous patients, 143 had tuberculous mothers. In 1834 Hardy 120 reported the history of a case of tuberculosis of the uterus and placenta occurring in a phthisical woman. The report is, however, somewhat vague, and in view of the general ill defined knowl- edge of the pathology of tuberculosis at that period, this case must be regarded with doubt. The cases of Charrin 121 and Jacobi 122 are, for similar reasons, also open to suspicion. Until 1 89 1 no undoubted case of congenital tuberculosis had been recorded. During that year two cases were reported — one by Sabou- raud 123 and one by Schmorl and Birch-Hirschfeld. 124 The latter authors were the first positively to demonstrate the presence of tubercle bacilli in the human placenta. Runge 125 regards his case as the first in which tubercle bacilli were positively identified in the decidua. As usual, no giant cells were found, but numerous tubercle bacilli were present, chiefly in the decidua basalis. Johne 9 was perhaps the first to report the history of an undoubted case of congenital tuberculosis, his specimen consisting of an unborn calf. Macroscopically, the uterus and placenta were normal. Since the pub- lication of Johne's case, many instances of congenital tuberculosis in animals have been recorded. As early as 1897 Klepp 126 reported that he found numerous calves affected with this form of the disease, and stated that 2.63 per cent of all young born of tuberculous cows were infected in titer 0. Cases of congenital tuberculosis in cattle have been reported by Malrox and Brouwier 127 (2 cases), Czoker, 128 Bank 129 (3 cases), McFayden, 130 Siegen 131 (38 cases), Lungwitz 132 (2 cases), Nocard, 133 Grancher, 134 Kohler, 135 Misselwitz, 136 Bayersdorfer, 137 CONGENITAL AND PLACENTAL TUBERCULOSIS 59 Becker, 138 Ruser, 139 Barland, 140 Galtier, 141 Bucher, 142 and Lohoff. 143 The recent carefully prepared report of Brooks 144 tends to show that the frequency of congenital tuberculosis in cattle has been exaggerated. Of 200 calves born of tuberculous parents, all of which were imme- diately after birth removed from the mother and guarded from postnatal infection, not one became tuberculous. EFFORTS TO PRODUCE CONGENITAL TUBERCULOSIS Animal Experiments. — Gartner, 36 in an extensive series of experi- ments upon white mice, succeeded in producing congenital tuberculosis in from five to ten per cent of cases. This was effected by making intra- peritoneal, intravenous, or intratracheal inoculations. Of nineteen litters in which the mothers were subjected to intraperitoneal injection of 0.00 1 to 0.002 c.cm. of a pure culture of tubercle bacilli, in two cases the young became infected. In an attempt to simulate miliary tuberculosis (bacil- lemia) this investigator injected 0.5 to 2 c.cm. of a pure culture of tubercle bacilli into the vein of the ear of ten healthy rabbits. Of fifty- one fetuses of young born to these animals, five (10 per cent) were tuberculous. In no case were all the young of a litter infected. The method of determining whether or not infection was present in the young was extremely thorough; it consisted of grinding the young or fetuses to a pulp and inoculating this into the peritoneal cavities of guinea pigs. For the purpose of producing conditions analogous to pulmonary tuberculosis, Gartner injected a drop of a pure culture of tubercle bacilli into the trachea of each of sixty-four mice; eighteen litters, consisting of seventy-four young, resulted. These were inoculated into 39 guinea pigs, and tuberculous young were found in 80 per cent of the litters. Another similar series of experiments performed at a later date upon twenty-eight subjects showed only one infected young animal. Lan- douzy 41 and Lodenih 145 performed a similar series of experiments, two of the eighty-six young animals which resulted showing congenital tuberculosis. Numerous other investigators have, however, failed to produce congenital tuberculosis. Sanchez-Toledo 146 performed intra- venous infection on fifteen guinea pigs, and in none of the twenty-five fetuses from these animals was tuberculosis present. Similar results were secured by intrathoracic inoculation. A summary of Sanchez-Toledo's results show that no tuberculosis was demonstrated in sixty-five fetuses from thirty-two tuberculous mothers. Cornet, 37 in an extensive series of experiments, was unable to produce 60 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS congenital tuberculosis. Of 233 fetuses or young examined, not one was tuberculous. Von Nocard 147 inoculated thirty-two fetuses of four tuberculous rabbits into thirty-two guinea pigs, with negative results. Wolff 148 performed a similar test, employing forty-two fetuses of rabbits and guinea pigs, with one positive result. Galtier 141 was unable to demonstrate tuberculosis in nine young from tuberculous guinea pigs and in one calf from a tuberculous cow. Grancher 134 and Straus 149 obtained similar results from the inoculation of suspected organs of fetuses from nine tuberculous female guinea pigs ; fourteen of the progeny were inoculated at birth, with negative results, and the remainder were examined at varying periods up to sixteen months of age. Tuberculosis could not be demonstrated. Vignal 11X performed similar experiments with eleven guinea pig fetuses from tuberculous mothers, with negative results. Carajnis, 150 by inoculating the spleen of a fetus from a tubercu- lous guinea pig, secured a positive result. From the findings just recorded, it can well be seen that congenital tuberculosis is difficult to produce experimentally. In studying congenital tuberculosis, a sharp distinction must be made between placental infection and fetal involvement. It by no means follows that, because a placental infection exists, the child is necessarily contaminated. Criticism and Possible Sources of Error. — Attention has else- where been directed to the small number of cases of congenital or of placental tuberculosis in man that have been reported, and possible reasons for this have been advanced. A review of the literature since 1 89 1 shows that much work has been done on this subject, and that the opinion of most observers is strongly opposed to the view that holds the condition to be frequent. On the other hand, a careful study shows that in most of the investigations in which a moderately large number of specimens were examined and thorough methods of research employed, a definite proportion of positive cases was demonstrated. The microscopic demonstration of tubercle bacilli alone in smear preparations is in some cases to be looked on with suspicion, as the differentiation of other acid fast bodies must be carefully considered — a point that should be emphasized, I believe with Fraenkel and others, that the microscopic demonstration in smear preparations of blood is likely to be misleading. Slight errors in technic, or the presence of acid fast bodies other than tubercle bacilli, are prone to produce very erroneous results. The findings of Novak and Ranzel, 2 which have previously been quoted, and whose experiments were evidently carefully carried out, may be placed partly in this category. CONGENITAL AND PLACENTAL TUBERCULOSIS 61 Sitzenfrey X1 has very properly sounded a note of warning against accepting even inoculation results, unless the technic has been carefully safeguarded. The frequency of tuberculosis among laboratory animals, and the possibility that incipient tuberculosis may have been present before the inoculation was made; the possibility of contamination during, prior, or subsequent to the inoculation, and the marked susceptibility of guinea pigs to this form of infection, are all sources of possible error. Feyerabend 151 has even mentioned the possibility of spontaneous tuber- culosis occurring in guinea pigs. Even the histologic examination is open to misinterpretation. The similarity of the picture produced by certain forms of syphilis and other conditions to that of tuberculosis has previously been pointed out. Experimentally produced congenital tuberculosis in animals is like- wise not beyond criticism. The relatively large amounts of culture material inoculated usually far exceed what could possibly be present in the pregnant woman. Thus Gartner, 36 whose results are perhaps more convincing than those of any other investigator, employed quantities of culture which, if increased proportionately to the weight of an average woman, would amount to 350 gm. introduced into the trachea, 35 to 140 gm. into the circulation, and about 0.5 to 1.5 liters into the peritoneal cavity. These results show, therefore, that while congenital tuberculosis may be produced experimentally in certain animals, it should not be compared to what takes place in the tuberculous pregnant woman. Period at which Intra-uterine Transmission is Most Likely to Occur. — Placental tuberculosis is undoubtedly most frequent in the latter months of pregnancy. As gestation progresses the maternal focus in the lungs or elsewhere is especially prone to exacerbations, and, as a result, organisms are more likely to be present in the maternal blood stream. Furthermore, hyperpyrexia tends to produce a solution of con- tinuity of the fetal and maternal blood vessels. Not only is a bacillemia prone to develop at this time, but a great quantity of blood is passed through the placenta. The larger amount of blood present in the pla- cental sinuses and the relatively slow blood current at this period predis- pose to the enmeshing of tubercle bacilli circulating in the maternal blood stream. The placenta itself is probably more receptive to tuberculosis than during its earlier development. Langhans' layer of cells is absent. As the end of pregnancy approaches, the placenta assumes characteristics that have caused it to be termed a senile organ (Williams, 5 Eden, 152 Warthin, Cowie, 12 and others). Very early hematogenous infection is also unlikely to take place ; indeed, during the first few weeks the chorionic villi are devoid of blood vessels and are nourished entirely by osmosis. 62 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS As has previously been stated, some authorities regard the placenta as a secure filter that prevents the organisms of a maternal bacteriemia from graining access to the fetus. This view is correct to a certain extent. This theory is based upon the assumption that the syncytium of the placenta is everywhere intact. In the latter months of pregnancy the chorion undergoes progressive atrophy. Anemic infarcts are con- stantly present in the placenta, and at these points egress is offered for the maternal microorganisms. It is probable that the fibrinous exudate which forms in these areas is an important factor in safeguarding the fetus. The bactericidal properties of the blood are perhaps also sufficient to destroy or inhibit the growth of a certain proportion of the tubercle bacilli. Indeed, Warthin and Cowie believe that it is only under excep- tional circumstances, or when the organisms are present in large numbers, that transmission is likely to occur. When tubercle bacilli are present in the intervillous spaces, labor itself, with its incident strong and frequent uterine contractions, is espe- cially prone to force the microorganisms into the fetal blood stream. That labor is a powerful agent in the production of congenital tubercu- losis has been recognized by practically all observers ; in fact, Bardeleben goes so far as to perform a cesarean section as a prophylactic operation before the onset of labor in certain cases. PREDISPOSING FACTORS TO PLACENTAL OR CON- GENITAL TUBERCULOSIS As most of these factors have been mentioned in the preceding pages, only a summary will here be given. Excluding the germinative type, the existence of which rests merely upon a theoretic basis, the require- ments for a placental infection to take place are a tuberculous bacillemia or a tuberculous focus in the immediate neighborhood of the placental site. Under predisposing factors, therefore, must be placed all conditions that favor the presence of tubercle bacilli in the maternal blood stream, such as acute miliary tuberculosis, phthisis florida, ulcerative lesions that tend to rupture into blood vessels, acute exacerbations of the disease, and a high temperature, which in itself tends to destroy the continuity of the blood vessels. In this connection, however, it must be remarked that Warthin has recently described a case of placental tuberculosis in which the lung lesion was quiescent, and the attention of the attending physician was called to the tuberculosis only by the finding of miliary tubercles in the placenta. CONGENITAL AND PLACENTAL TUBERCULOSIS 63 Syphilis and other disease of the placenta, as well as trauma, prob- ably not only predispose to the development of placental tuberculosis, but undoubtedly serve as predisposing factors in the production of congenital tuberculosis by forming avenues of egress for the circulating or enmeshed tubercle bacilli. Tuberculosis of the peritoneum or of the female genital tract is also a predisposing factor to infection. The author has, however, seen two cases of advanced tuberculous peritonitis in pregnant women, and in neither was there gross evidence of infection in any of the products of conception. Unfortunately a thorough examination could not be made in either case. Williams has observed a somewhat similar case. The Fate of the Congenitally Tuberculous. — Under this heading Cornet formulates an extremely unfavorable prognosis for children who are the victims of this variety of infection. Embryonic and fetal tissues possess no immunity to tuberculosis. It is probable that, if the fetus were infected during the early months of pregnancy, intra-uterine death or abortion would be likely to follow, whereas if the infection occurred late, it is probable that tuberculosis would manifest itself in the liver or adjacent lymphatic glands. It would appear, therefore, that the prognosis would be decidedly less favorable in the case of a congenitally infected child than in one who acquired the disease postnatally. Furthermore, the virulence of the strain of an infecting organism is of importance in this connection. The mothers of congenitally infected children are often the incumbents of an acute miliary tuberculosis, a form of disease in which the organisms are usually extremely virulent. A few years ago the author reported the results obtained in the examinations of fourteen placentas from tuberculous women. Since these 107 additional specimens have been examined, making a series of 121 placentas. The subjects from which these placentas were obtained were all suffering from pulmonary tuberculosis. They consisted for the most part of ambulatory cases, which were coming to the Henry Phipps Institute for treatment. Almost 15 per cent were in a mod- erately acute stage of the disease at the time of delivery, and two were nearly, moribund. The remainder were mild or quiescent. One hundred and one were at or nearly at term ; the remainder were premature, some as early as the second month. In brief, the following technic was em- ployed: five to ten small pieces were cut from various parts of the placenta and were finely ground up with sand in a mortar, to which was added a little salt solution. This was allowed to stand for a short time, and about a dram of the solution was injected into the peritoneal cavity of a guinea pig, five pigs being used for each placenta. All pigs 64 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS dying after the fifth day were autopsied, and at the end of six weeks all the remaining animals were killed and examined. A case was considered positive for tubercle bacilli where three of the five inoculated pigs showed tuberculosis. There were fourteen such cases. Inoculations from three additional placentas showed tuberculosis in one or two of the five pigs injected. Specimens in which all the inoculated animals died during the week subsequent to inoculation are not included in this series. Thus virulent tubercle bacilli were demonstrated with moderate certainty in n per cent of the specimens, and may have been present in an additional 2 per cent of the placentas. CONGENITAL TUBERCULOSIS (Case Histories) Undoubted Cases. — In the following cases the diagnoses were based upon anatomical changes and the presence of tubercle bacilli. Sabouraud. 123 Child aged eleven days, born of mother in advanced stage of pulmonary tuberculosis, who died shortly after delivery. Autopsy of infant showed the presence of countless miliary tubercles in the liver and spleen, in part showing caseation, and containing tubercle bacilli. Lehmann. 153 Woman forty years of age, suffering from advanced pulmonary tuberculosis, gave birth to a premature, male child in the ninth month. The mother died two days after delivery. Autopsy showed acute miliary tuberculosis of lungs and tuberculous meningitis. The placenta was not examined. The child died twenty-four hours after birth. Autopsy showed miliary tuberculosis of the lungs, liver, spleen, and kidneys. Tubercles were also present in the portal, mediastinal, bronchial, mesenteric, and retroperitoneal glands. The microscopic appearance was that of typical tubercles. Large numbers of tubercle bacilli were found. No giant cells were present in the tubercles. Advanced stage of the process makes the case undoubtedly congenital. Honl. 154 Child, fifteen days old. Autopsy revealed typical caseous miliary nodules in the spleen, liver, lungs, containing numerous tubercle bacilli. These were also found free in the blood vessels. Chronic tuberculous lesions were found in the liver. Mother was brought to the hospital with pulmonary tuberculosis after birth of child. The case is regarded as undoubtedly congenital, as such advanced lesions could not have formed during the short period of extra-uterine life. Ustinow. 155 New born female child weighing 3,060 gms. Died of CONGENITAL AND PLACENTAL TUBERCULOSIS 65 inanition after a few days. Nothing known of the mother. Autopsy of infant showed general tuberculosis, most marked in the liver. The spleen was somewhat enlarged, and contained so* many tubercles that the surface presented a marbled appearance. Lungs contained a smaller number of tubercles. The brain and retina of both eyes were free from tubercles. Large numbers of tubercle bacilli were found in the tubercles and also free in the capillaries. In some sections the bacilli were so numerous that when stained for tubercle bacilli the red areas were visible to the naked eye. Auche and Chambrelente. 156 Mother, in advanced stage of tubercu- losis, died three days after a premature delivery in the seventh month. Autopsy showed advanced tuberculosis of the lungs, liver, spleen, intes- tines, mesenteric glands, and kidneys. The uterus and adnexa were normal. Peritonitis was not present. The placenta showed numerous caseous tubercles with tubercle bacilli. Inoculation of guinea pigs with portions of placenta gave positive results. Child died on the twenty-sixth day. Autopsy showed miliary tubercles in the lungs, liver, spleen, and endocardium of the right heart. Typical tubercle bacilli were present. Inoculation of rabbit with portions of fetal organs gave positive results. The tuberculous endocarditis is of especial interest, as the first case noted in infants. The woman had three other healthy children. Veszpremi 157 reports a case of congenital tuberculosis. Tubercle bacilli were demonstrated from the fetal blood by means of inoculation. The mother was the victim of advanced miliary tuberculosis. Owing to unfortunate circumstances, it was not possible to examine the placenta. Dufour and Thiers 158 report a case of tuberculosis of a fetus. Mother, aged nineteen years, having symptoms of advanced pulmonary and meningeal tuberculosis. The latter was proved by puncture and demonstration of the microorganism. She died twelve days after admis- sion to the hospital. Autopsy showed extensive tuberculosis. The fetus was partially expelled into the vagina. The abdomen was enlarged and ascitic. The placenta was macerated and suggestive of tuberculosis in its appearance. Acitic fluid from the chest of the fetus, by inoculation, was found to contain tubercle bacilli. The placenta showed histologic evidence of tuberculosis, but tubercle bacilli were not demonstrated in it. Brindeau. 159 The child of a tuberculous mother, died on the twelfth day. Autopsy showed very advanced pulmonary lesions, and from these tubercle bacilli were demonstrated. The advanced character of the lesion present makes it extremely probable that this case was one of true congenital infection. Stockel. 160 The mother had advanced tuberculosis. No tuberculosis 66 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS was found in the placenta. The child lived fourteen days. At autopsy the cadaver showed extensive miliary tuberculosis, especially of the lungs, liver, and intestines, and marked caseations, especially of the periportal glands. Zarfl. 161 The infant was born of a tuberculous mother. On the seventeenth day the von Pirquet reaction was markedly positive. The author believes such sensitiveness to tuberculosis could not have developed in seventeen days. On the eighteenth day there was enlargement of the liver and spleen. The swelling of the spleen increased and was the most prominent symptom. Until the last week there was no clinical or Rontgen ray evidence of tuberculosis. This child died on the fifty- second day. Autopsy showed involvement of lymph nodes of the liver region, the most seriously involved. Slight involvement of bronchial lymph nodes and no focus in the lungs. The mother lived for three months after the birth of the child. Jens Bugge. 162 The patient, aged thirty-nine years, died of tubercu- losis four days after delivery. Autopsy showed tuberculosis of the lungs, bronchial glands, kidneys, and intestinal tract. The placenta was not examined. The infant was eight months advanced and died thirty hours after delivery. No tuberculosis was found macroscopically, but tubercle bacilli in the umbilical vein were demonstrated by staining and inoculation. Moller 163 reports a case of tuberculosis in a child which died on the third day. The mother left the hospital well, but returned five months afterward with tuberculosis of the uterus, and died of miliary tubercu- losis in two months. Autopsy of the child showed miliary tuberculosis of the liver and spleen, a tubercle in the pancreas, two typical ulcers in the ileum, miliary tuberculosis of the lungs, massive tuberculosis of the retroperitoneal lymph glands, and a caseous mass in the thymus. Tubercle bacilli were found in the lesions. Recent tuberculosis lesions were found in the decidual membrane and panhysterectomy was per- formed. Old tuberculosis processes were evident in both fallopian tubes, and to these Moller attributes the infection of the uterus and the fetus. When the woman was delivered there was no suspicion of tuberculosis and the placenta was not examined. Grulee and Harms 164 reported a case of miliary tuberculosis in a child which died on the eleventh day. This child showed throughout an irregular temperature. On the fifth day it had a convulsion which con- tinued until death. The liver and spleen were found to be enlarged. At autopsy there were found caseous tubercles of the periportal and mesen- teric lymph glands, miliary tuberculosis of the spleen with caseous CONGENITAL AND PLACENTAL TUBERCULOSIS 67 nodules, and a few scattered nodules in the liver, lungs, and kidneys. The mother of this child had what was apparently only a healed tuber- culosis of the hip. She had, however, a vaginal discharge of unknown etiology. The mother was alive several months after the infant was born. Probable and Doubtful Cases. — In these cases the diagnoses were based upon anatomical appearance only, gross or microscopic, without demonstration of the presence of tubercle bacilli ; or doubtful because of age of child, non-elimination of possible syphilis, extra-uterine in- fection, etc. Delmas. 165 The mother was moribund from advanced phthisis at the time of her confinement. The child was delivered by forceps and immediately placed in a sterilized incubator. No further communication between the mother and child occurred. The child died when four months old, the lungs being chiefly affected. Delmas believes the infection to have been a hematogenous one. There were no intestinal lesions. Bourges. 160 The mother died of tuberculosis shortly after having been delivered of a viable child. At autopsy she showed advanced lesions. The child survived but a short time. An autopsy on it showed no macroscopic evidence of tuberculosis, but inoculations from the liver and other areas into guinea pigs gave a positive result. In another case, negative results were obtained. Jacobi. 122 Seven months' fetus of a mother suffering from chronic pulmonary tuberculosis had numerous caseating tubercles in liver, spleen, pleura, and peritoneum. Anatomical evidence only. Demme. 167 Two cases : 1. Boy of five weeks. Sick from birth with fever and cough ; showed on autopsy caseous nodules in both lungs and infiltration of bronchial and tracheal glands. Mother died of chronic pulmonary tuberculosis soon after delivery. 2. Child died on sev- enteenth day after delivery. Similar to the first case. Tuberculosis in mother shown by physical signs. Both cases are doubtful. Charrin. 121 Seven and a half months child of a tuberculous mother ; died three days after birth. Autopsy showed widespread tuberculosis of the peritoneum and abdominal organs. Scattered tubercles were found in the lungs. Anatomical evidence only. Demme. 167 Female child, twelve days old, of tuberculous mother. Autopsy showed caseous bronchial glands and numerous caseating nodules in both lungs. In the right apex and right lower lobe many cavities, the size of a pea to a cherry, are found. Mesenteric glands are unchanged. Doubtful case. Anatomical evidence only. 68 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Merkel. 168 (Not reported until 1884 by Ohlendorff). In January, 1875, patient developed pleuritis, followed by bronchial catarrh and infiltration of apices. In February she conceived; in June showed tuber- culous laryngitis; by October the patient had to be fed with tube; and the child was born on November 4. On the 6th the mother died. Autopsy showed tuberculous cavities in the lungs and miliary tubercu- losis. The child was small and was born with a small yellow tumor on hard palate. On the second day this discharged cheesy material ; abscess then developed in left greater trochanter. Child died of inanition. Autopsy showed caseous nodule in hard palate infiltrating the bone, caseation of cervical glands, caseous nodule behind the left hip joint. Probable case. Anatomical evidence only. Demme. 167 Female child, aged twenty-five days. Mother died of catarrhal pneumonia. Autopsy of child showed in the middle of the right cerebellar hemisphere a caseating tubercle the size of a hazel nut. The microscopic examination showed the appearance of caseating tubercles. No tuberculosis elsewhere. Doubtful case. Baumgarten. 169 States that in autopsies on infants dying during the first months of life he had often found tuberculosis of such advanced stage as to make a congenital origin very probable. Berti (cited by Gartner). 170 Tuberculous mother, aged seventeen years. Child died on the ninth day after birth. Autopsy showed two small cavities, filled with caseous material, in the posterior margin of lower right lobe of the lung. Microscopic examination confirmed the gross diagnosis of tuberculosis. Very probable case. Anatomical evidence. (Berti reports a second very doubtful case, which may be entirely ruled out.) Demme. 171 Two cases: 1. Child dying on twenty-first day. Autopsy showed advanced tuberculous ulceration of intestines. 2. Child dying on twenty-ninth day of pulmonary tuberculosis. Both cases very doubtful. History meager. Evidence, anatomical only. Extra-uterine infection not excluded. Money. 172 Female child of tuberculous mother, dying five weeks after delivery. Child ill for three weeks with cough and attacks of vomiting. No evidence of syphilis. Died in eighth week. Autopsy showed caseous tubercles of lung, liver, spleen, and kidneys. Bronchial and mesenteric glands enlarged, but no caseation. Intestines not ulcerated. One tracheal gland caseated. A probable case. Demme. 167 Female child, aged eleven weeks, of tuberculous mother. Autopsy showed extensive cavity formation in the right lobe of the child's lung. A doubtful case. Extra-uterine infection not excluded. CONGENITAL AND PLACENTAL TUBERCULOSIS 69 Queyrat. 173 Three months old child; died and autopsy revealed extensive caseation and cavity formation in both lungs. Mother healthy; father tuberculous. Very doubtful. Extra-uterine infection not excluded. Flesch. 174 Advanced ulcerative lesions were found in the lungs of eight out of five hundred infants upon whom autopsies were performed. All the subjects had died in the early months of life. Extra-uterine infection not excluded. Statement too inexact. Evidence, anatomical only. Frobelius. 175 Found in 16,581 autopsies of children under two years of age, 616 cases of tuberculosis. One died on the third day, one in the second week, one in the third week, three at about three and one-half weeks, fourteen in the second month, and one hundred and nineteen in the third month. No detailed account of these cases is given. Houtinel. 176 In 996 autopsies upon infants under one year of age, eighteen cases of tuberculosis were observed, two dying in the first fourteen days after birth. No detailed account is given. Lannelongue. 177 Out of 1,005 cases of surgical tuberculosis in chil- dren under fifteen years of age, four were observed which he regards as of undoubted congenital origin ; one child, six weeks old, with classi- cal signs of tuberculosis of knee existing from birth; one, one month old, tuberculous osteoarthritis fourteen days after birth ; one, three weeks old, tuberculous abscesses in the left tarsus and right maleolar regions; one, sixteen days old, tuberculous ostitis. In these cases, Lannelongue believes it possible to exclude extra-uterine infection. In another child of two months, right sided tuberculous epididymitis with fistula was present. A scrotal engorgement some days after birth was noticed. Few details are given concerning the parents. All these cases are doubtful. Huguenin. 178 Two cases, one dying at the age of seven weeks of general tuberculosis ; the second at the age of seven months of a general tuberculosis. Very doubtful cases. Evidence, anatomical only. No details are given. Bosselut. 179 In a large number of autopsies on children dying of tuberculosis, meningitis was found in one subject who had died on the fourteenth day ; in two, aged three weeks ; in one, aged six weeks ; and in four, aged eight weeks. Evidence not conclusive. No details. Doubtful cases. Rindfleisch. 180 Mother in advanced phthisis florida, developing during pregnancy, and died of phthisis shortly after birth of child. The child died on the eighth day of general tuberculosis. Large caseous nodules in the liver. Probable case. Anatomical evidence only. 70 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Sarwey. 181 Monster (cranioschisis) ; prolonged pregnancy. Born in eleventh month. The mother was apparently healthy. Father had cough and tubercle bacilli in the sputum. Child showed caseous and partly calcified nodules in the upper cervical vertebrae. Guinea pigs were inoculated. Three out of six developed tuberculosis after three months. A probable case. Baumgarten. 169 Still born monster (encephalocele). Caseating abscesses in three uppermost cervical vertebrae. No bacilli found. Evidence only anatomical. Details not given. A doubtful case. Leroux. 116 The infant died when eighteen days old. Deep tuber- culosis ulcers were found in the intestine. Caseation of tracheal and bronchial glands was present. Probably a case of congenital infection, as the extensive changes could hardly have occurred from extra-uterine infection. Leroux also gives notes of twenty-two other cases of tuberculosis in children under three months ; one, four weeks old ; one, five weeks (premature birth) ; two, six months; five, two months; eight, two and one half months; five, two and three quarter months. No details given. All these cases are doubtful and based upon clinical observations only. Wassermann. 182 Child when first taken ill was six weeks old, and at that time had bronchial catarrh and osteitis (tuberculous?). Died four and one half weeks later. Autopsy showed extensive tuberculosis of both lungs, diaphragm, liver, and kidneys. Wassermann believed the case to be acquired from a relative of the mother, with whom the latter and child had resided for a short time when the child was ten days old. Correctness of this opinion questioned by Baumgarten and Lebkuchner. Very doubtful case. Hochsinger 104 reports three cases, aged thirty-one days, thirty-eight days, and sixteen weeks, respectively, of combined tuberculosis and syphilis. The mother in the first and third cases was tuberculous. Autopsies of children showed advanced tuberculosis in all three cases. Tubercle bacilli present in all. Age of children and the fact that con- genital syphilis predisposes to rapid development of tuberculosis make these cases doubtful. Straus. 183 Child died when three weeks old. Autopsy showed caseous tubercles in lung, spleen, bronchial, and mesenteric glands. Doubtful case. Full details not given. Kissel. 184 In one thousand autopsies upon children, Kissel observed three cases of tuberculosis which he believed to be of congenital origin. These children were aged four, five, and six weeks, respectively. Small CONGENITAL AND PLACENTAL TUBERCULOSIS 71 advanced lesions of the bronchial glands were present. The cases are not reported in detail and must be considered as doubtful. Holt 185 mentions one case, of a child dying on the twentieth day after a premature birth. The mother, suffering from advanced tuber- culosis, died shortly after the child. On autopsy, the child was found to have caseous bronchial glands, and miliary tuberculosis of the lungs, none in liver, alimentary tract, or spleen. Regarded as probably congenital from advanced nature of lesions. Henoch. 186 Father died of tuberculosis. The child had been ill since the sixth week from multiple tuberculous abscesses in various parts of the body. It died in the fourth month, of inanition. Autopsy showed advanced pulmonary tuberculosis, intestinal tuberculosis, and caseation of lymph glands. A very doubtful case. Extra-uterine infection probable. Bonnet. 187 Mother died of pulmonary tuberculosis two months after delivery. Male child, ill from birth, died at three months. Autopsy showed both lungs studded with caseous tubercles, tuberculous ulcers in ileum, caseous tuberculosis 'in mesenteric glands, kidneys, spleen and liver. Fatty liver. The stage of lesions and the fact of illness from birth, the child having been kept from danger of infection, given as reasons for regarding the case as of congenital origin. Johnson. 188 White female child born of mother suffering from tuberculosis of urinary tract. At birth the infant was very weak, small, and could not nurse. The emaciation increased. The child made efforts at coughing and died during a profuse pulmonary hemorrhage at the age of three months and two days. The father was healthy. Urine of mother contained blood, pus, and tubercle bacilli. Placental tuberculosis was probably present, the organ showing the usual histologic evidence of this infection. Autopsy of child showed extensive tuberculosis of the right lung pleura, and pericardium; miliary tuberculosis of the left lung. Tubercle bacilli could not be found in fetal organs. Evidence not complete, but most probably a case of congenital tuberculosis. Lebktichner. 189 Two cases: 1. The mother was of a tuberculous family ; suspicious symptoms, but case not certain. The child was short of breath and coughed from birth. It eventually died, and a postmortem showed advanced tuberculosis in lungs and other organs. 2. Case simi- lar, but child older. The first case may be regarded as probable, though the evidence is incomplete. Friedmann. 190 The mother had advanced tuberculosis. The child died when twenty-six days old ; a postmortem showed a small tuberculous lesion in the apex of the right lung. Evidence not conclusive. 72 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Lyle. 191 Negress, twenty-two years of age, in seventh month of pregnancy, suffering from chronic pulmonary tuberculosis. The mother died two days after premature delivery. Autopsy showed chronic tuber- culosis of both lungs and intestinal ulceration. The placenta appeared normal. The child weighed about three pounds. It was ill from birth ; subnormal temperature for four weeks, followed by fever; died in eighteenth week. Autopsy revealed extensive tuberculosis of lungs, liver, spleen, and kidneys ; tubercle bacilli were present in the caseous areas. The great number of tubercles of same advanced stage, the fact that the child was ill from birth and had been kept under such conditions as to exclude likelihood of extra-uterine infection, are the reasons advanced by Lyle for considering this case as congenital. It is doubtful, however, because of the age of the child. Sitzenfrey. 11 i. Patient aged thirty-eight years and octipara. Three children are living. She was delivered, by forceps, of a female infant, which weighed 2,620 grams. The patient died five days post partum. Autopsy showed pulmonary tuberculosis, tuberculous ulcers of the intes- tines, chronic tuberculosis of the peribronchial, mesenteric, and retro- peritoneal lymph glands, subacute miliary tuberculosis of the liver, spleen, and kidneys, also meningitis basilaris tuberculosa. Tuberculous caries in the tenth and eleventh ribs (left) with fistula; also caries in the second to sixth dorsal vertebrae with prevertebral abscess. The infant was transferred at once to an institution and died six weeks after birth. Autopsy showed chronic pulmonary tuberculosis, chronic tuberculosis of the liver, spleen, thyroid, intestines, peribronchial, mesenteric, retro- peritoneal and portal lymph glands. Case 2. Patient aged twenty-four years and quadripara. Two children died of gastro-intestinal trouble; one child is living and well. She had a spontaneous delivery of an eight months male infant, which weighed 1,800 grams, and which died two days after birth. The mother died the following day. Autopsy showed chronic pulmonary tuberculosis, chronic tuberculosis of the peribronchial, cervical, and axillary lymph glands, tuberculous ulcer of the larynx, tuberculous nodules in the skin of various portions of the body, chronic tuberculosis of the intestines, liver, peritoneum, and kidneys, and tuber- culous meningitis. The inner surface of the lower uterine segment was yellow and caseous, but showed microscopically merely necrosis — no tuberculosis. The child died on the second day. Autopsy showed partial atelectasis of lungs, icterus neonatorum and debilitas vitae congenita. The placenta measured eleven by twelve centimeters. It was macro- scopically normal, but the membranes show at two places, corresponding to the decidua vera, a certain amount of thickening three to four milli- CONGENITAL AND PLACENTAL TUBERCULOSIS 73 meters. The maternal surface of these is uneven, ragged, and grayish yellow. On section through these areas, the tissue shows a caseous condition. On the fetal side, these areas are covered by smooth amnion. Microscopic examination of the placenta and umbilical cord is negative for tuberculosis. Microscopic examination of the thickened portion of the membranes, however, shows extensive caseous foci, often associated with thrombosis in the decidua vera. In these areas giant cells are not found, but enormous quantities of tubercle bacilli are seen. Sitzenfrey believes the entrance of tubercle bacilli into the fetal circulation in these cases may be explained perhaps by the aberrant nutrient vessels of the chorion, which arise from the umbilical cord and which might be invaded by tuberculous foci in the decidua vera. Two cases have been reported by Schrumpf. Case 1. Patient, aged thirty years, died in seventh month of pregnancy, from chronic pulmonary tuberculosis. The decidua vera in the right and posterior uterine walls was found transformed into a caseous sheet, three and a half by five centimeters and four millimeters thick. Microscopically this tissue is full of round cell infiltration with tubercle bacilli and necrosis. There was no extension to the placenta or decidua basalis. Examination of the fetal organs showed these to be histologically normal, but a few tubercle bacilli were found in smear preparations from the fetal heart, blood and liver. Case 2. Patient, aged twenty-three years, died in the seventh month of pregnancy from chronic pulmonary, laryngeal and intestinal tuberculosis. The uterine mucosa showed an opaque, bright yellow area, one centimeter in thickness on the posterior wall and extend- ing downwards from the outer edge of the placenta for about seven centi- meters. Microscopically this proved to be an infiltration of the decidua vera with necrosis in places and a few tubercle bacilli. No miliary tubercles or giant cells. The placenta and decidua basilaris were normal. Examination of the fetus and animal inoculation were negative. Sitzenfrey also reports two other cases which he believes to have been congenital and which have been excluded on account of the age at which the children died, three months and six months, respectively. PLACENTAL TUBERCULOSIS (Case Histories) Undoubted Cases.— Diagnoses rest upon demonstration of histo- logical changes and the presence of tubercle bacilli in the placenta. Runge 81 reports a case of placental tuberculosis occurring in a patient 74 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS suffering from active pulmonary lesions. The tubercle bacilli were dem- onstrated in the placenta, and the characteristic histologic changes produced by this organism were present. Lehmann. 192 Case I. Woman, aged twenty-six years, died of miliary tuberculosis in seventh or eighth month of pregnancy. Male child re- moved by cesarean section five minutes after death of mother, showed no signs of life. Well developed. The only abnormal changes found were two grayish nodules in the right apex. Pieces of liver and spleen were inoculated into guinea pigs. The general appearance of the placenta was normal. In several places small miliary nodules grayish, semi- translucent, and sharply outlined, were found. The microscopical exam- ination of fetal liver and lungs gave no appearance of tuberculosis. No tubercle bacilli were found. The placental nodules presented the appear- ance of typical caseating tubercles, containing a few tubercle bacilli. Case 2. The woman, aged thirty-two years, died in the hospital. No history of the case was obtainable. The autopsy showed tuberculosis of the lungs, endocardium, liver, both kidneys, meninges and the pul- monary veins. The uterus measured fourteen by nine by eight centi- meters. On its anterior and posterior, surf aces were small, grayish pro- tuberances. The uterine wall was one and a half to two centimeters thick and very vascular. The cavity contained an ovum of about four months' development. A loop of the cord had prolapsed into the vagina. The placenta measured one to one and a half centimeters in thickness and was anterior in attachment. There were small hemorrhages in the placenta and a larger one between it and the uterine wall, showing that abortion must have begun before death. The adnexa were normal. Tuberculous granulation tissue was found surrounding the chorionic villi, and in a few tubercle bacilli were demonstrated. Harbitz. 193 The mother was twenty-six years of age and entered the hospital with advanced pulmonary tuberculosis, of which she died twenty-eight days after confinement. Autopsy showed an acute miliary tuberculosis, involving the lungs, kidneys, peritoneal cavity, and fallopian tubes. The uterus, especially in the vicinity of the decidua basilaris, showed well marked tuberculosis. The infant measured forty-nine centi- meters in length and weighed 1,930 grams. It died on the twenty-fifth day, having been previously isolated from the mother. Autopsy upon it showed extensive involvement of the lungs and bronchial lymph glands. Warthin. 194 Patient, aged twenty years, with a previous history of gonorrhea. Tuberculosis was not suspected, but during the routine course of the histologic examination of the placenta, as practiced in Dr. Peterson's clinic, numerous miliary tubercles, many of them healing, CONGENITAL AND PLACENTAL TUBERCULOSIS 75 were found. In some of these areas, tubercle bacilli were demonstrated by staining. Examination of the mother showed a positive tuberculin test, a suspicious right apex, but no evidence of active disease. The child was viable, and no tuberculosis demonstrated in it. Carl. 195 The mother showed advanced pulmonary tuberculosis and the usual clinical symptoms. The child was normal, but in the placenta typical tubercle bacilli were demonstrated, together with the histologic evidence of this infection. Lehmann. 153 Mother died of chronic tuberculosis of lungs and larynx. The child died ten days after birth. It presented no evidence of tuberculosis at autopsy. Typical caseating tubercles containing tubercle bacilli found in the placenta. Schmorl and Kockel. 197 Case 1. Woman, aged twenty-six years, in seventh or eighth month of pregnancy, died of chronic tuberculosis and miliary tuberculosis. Child, removed by cesarean section, lived two hours. No histological changes of tuberculosis or tubercle bacilli found in fetus. Placenta appeared normal to the naked eye. Numerous tubercle bacilli found in smears from the placenta. Animal inoculation was nega- tive. On microscopic examination, placenta showed typical tubercles in all stages. Tubercle bacilli in large numbers were demonstrated. Case 2. Mother, aged twenty-five, died of general miliary tuberculosis. Fetus removed at autopsy. No evidence of tuberculosis in fetal organs. Typical tubercles containing bacilli found in the placenta. Case 3. Woman, aged thirty-three, died in ninth month of pregnancy of chronic pulmonary tuberculosis. The child was removed by cesarean section; dead when uterus was opened. Male child, showed nothing suggestive of tuberculosis. No tubercle bacilli found in fetal tissues. Placenta presented no naked eye appearances of tuberculosis. Animal inoculations with fetal tissue were negative. Microscopically, the placenta showed typical tubercles in varying stages, not so numerous as in cases 1 and 2. Tubercle bacilli were present in large numbers. Jung. 197 Woman showed advanced pulmonary tuberculosis. The placenta presented the usual histologic picture of tuberculosis. Tubercle bacilli were also demonstrated. Tuberculosis was not present in the child. Warthin. 198 Case of tubal gestation with tuberculosis of tubes, placenta, and fetus. Rupture of the tubal sac in fourth month ; operation. Advanced tuberculosis of tubes, and wall of sac. Chorionic villi involved directly by extension from wall of tube. Typical tubercles in chorionic villi. Few tubercle bacilli found. In one of fifty successive placentas examined at the Stadt Kranken- y6 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS haus in Dresden (Schmorl's laboratory) miliary tubercles in all stages of development were found by Warthin. The placenta was full term. A few bacilli were demonstrated in it. Nothing was known of mother or child, as placenta could not be identified. Auche and Chambrelente. 199 Mother, far advanced in tuberculosis, died three days after premature delivery in seventh month. Placenta showed numerous caseous tubercles containing tubercle bacilli. Inocula- tion of guinea pig, positive. Child lived twenty-six days. Autopsy , showed extensive tuberculosis. Wollstein. 200 Mother died of tuberculosis six days after birth of an eighth month child. Placenta measured seventeen centimeters in diam- eter and three millimeters in thickness. It contained grayish, yellowish, or cheesy nodules. Histologically, agglutination, thrombi, and destruction of the syncytium and other evidences of tuberculosis were present. The umbilical cord was normal. The uterine mucosa was the seat of a tuber- culous deciduitis. Tubercle bacilli were demonstrated. The infant lived nineteen days. No tuberculosis demonstrated in it. This. case was one of hematogenous infection. Walther. 201 A patient with a definite family history of tuberculosis died in the seventh month of pregnancy. Macroscopically, the placenta showed yellowish white patches which microscopically proved to be areas of tuberculous caseation involving the decidual portion of the organ. These lesions affected the maternal aspect of the placenta only. Neither the fetus nor the umbilical cord showed any evidence of the disease. Sitzenfrey 1X reports the following cases: Case I. Patient, aged twenty-eight years and primipara, had a spontaneous delivery of a male infant which measured forty-one centimeters in length and which died after four hours. The mother, who had a history of slight lung trouble and hemophesia, rapidly developed symptoms of acute miliary tubercu- losis and died three days later. Autopsy showed chronic tuberculosis of both upper lobes, chronic tuberculosis of the peribronchial lymph glands, tuberculous ulcer of the larynx, universal miliary tuberculosis, and chronic catarrh of the large and small intestines. Autopsy of infant showed total fetal atelectasis of lungs and multiple ecchymoses of pleura and pericardium. No tuberculosis found in the infant. The placenta measured fourteen centimeters in diameter and two centimeters and a half in thickness. It weighed 3,300 grams, and showed no gross abnormalities in the fresh state. After hardening, the outer surface showed numerous gray red to yellow or whitish, opaque nodules, which presented the appearance of the little infarcts often found in nor- mal placentas. In practically all sections made for microscopic examina- CONGENITAL AND PLACENTAL TUBERCULOSIS 77 tion, there were found numerous foci of caseation and round cell infil- tration, in the decidua basalis, in the villi, on the edge of the infarcts, and in the chorionic membrane. In these areas numerous tubercle bacilli, also small nodules containing tubercle bacilli, were found in the lumen of veins in the decidua basalis. These foci of round cell infiltration are not characteristic of tuberculosis alone, but may be found in various other conditions; their definite diagnosis as tuberculous is dependent upon the demonstration of tubercle bacilli. In some areas these in- flammatory foci with tubercle bacilli could be seen to have broken into the intervillous spaces. Numerous tubercles of various types are pres- ent in the villi. The tubercle bacilli apparently work their way into the syncytial covering of the villi and injure it; the syncytium becomes swollen, loses part of its staining properties, and contains vacuoles, in which the tubercle bacilli are found. As a result of this process, the syncytium loses its power of preventing blood coagulation, causing thrombotic deposits to be formed on the surface of the villus, contain- ing numerous leukocytes. Under the influence of the tubercle bacillus these masses caseate; new areas of coagulation are formed which in- volve surrounding villi, and in turn succumb to tuberculous destruction, often earlier than the villus which formed the first nidus for the bacilli. No tuberculosis was found in the umbilical cord in this case. One tubercle bacillus was found in the lumen of a vessel in a villus, hence the author believes it probable that bacilli had reached the fetus, although microscopic examination of various organs was entirely negative. He thinks the toxic effect of the metabolic products of the tubercle bacilli was the underlying cause of the death of the fetus four hours post partum. Case 2. Patient, aged thirty-four years, had induction of labor on account of pulmonary and laryngeal tuberculosis. The male infant weighed 2,150 grams. The child was immediately removed from the mother and placed in an orphanage. Blood from the umbilical cord collected and injected into a guinea pig, which remained healthy and on being killed, showed no signs of tuberculosis. The placenta measured sixteen centimeters in diameter, weighed 480 grams, and macroscopically showed no pathologic changes. The mother did well for a time, but died about four months post partum with symptoms of peritonitis. Autopsy showed chronic pulmonary tuberculosis, chronic tuberculosis of the trachea, larynx, and of cervical and peribronchial lymph nodes, chronic tuberculosis of the intestines with perforated ulcer in the ileum: one half a liter of fecal fluid was found in the peritoneal cavity. Tubercles were found in the liver, kidney, uterus, and tubes. The endometrium 78 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS showed caseous areas and tubercles. The walls of the tubes were thick- ened in places and reduced to caseous masses. The child died at the end of three months. Autopsy showed chronic tuberculosis of both lungs with pleuritis, chronic tuberculosis of the peribronchial lymph glands, extensive tuberculous ulceration of the large and small intestines, tuberculosis of the mesenteric glands, miliary tuber- culosis of the liver, spleen, kidneys; a few whitish nodules in the fossa of Sylvius, suggestive of tubercles. The placenta was cut into slices and thirty blocks studied histolog- ically. In only three of these were tuberculous changes found. Although very few foci were present, some of these were fairly large, involving decidua basalis and villi. No bacilli were found in vessels. Sitzenfrey believes, however, that probably the bacilli got into the fetal circulation. In many cases the peripheral capillaries of the villi were engorged to bursting point, and in some places have actually burst, permitting ma- ternal and fetal blood to mix. This congestion may have been due to pressure on the umbilical cord by the bag which was introduced to induce labor. Microscopic examination of the umbilical cord for tubercle bacilli was negative. There was no possibility of a postpartum infection of the child. Every precaution was taken in the institution under charge of Dr. Ep- stein. It seems justifiable to conclude, therefore, that the tubercle bacilli were introduced into the child in utero or during delivery, probably the former. The case was probably one of congenital tuberculosis. Case 3. Patient, aged thirty-two years, tripara. (Previous children healthy.) She was delivered, by forceps, of a male infant that weighed 2,520 grams. It was sent immediately to Dr. Epstein's institution. The mother died three weeks post partum. Autopsy showed chronic pul- monary tuberculosis, tuberculosis of the peribronchial glands, tuberculous ulcers of the larynx, trachea, intestines, chronic tuberculosis of the liver, spleen, and kidneys. The child was living and well five months after birth. The placenta measured seventeen centimeters by eighteen and weighed 570 grams. Some material from the placenta was injected into the abdomen of a guinea pig, which became very ill ; the right inguinal glands swelled to the size of a pigeon's egg, but the animal gradually recovered and the swelling disappeared. Autopsy showed no pathologic conditions. Only after very careful searching was the first tuberculous focus found histologically; this was an infarct, in whose periphery round cell infiltration and tubercles in villi were found; many tubercle bacilli and Langhans' giant cells were present. Notwithstanding this, positive CONGENITAL AND PLACENTAL TUBERCULOSIS 79 histologic findings and animal inoculations were negative ; this was prob- ably due to the fact that the portions of tissue used happened to be free from tubercle bacilli. This case shows that not only is a negative inocu- lation result no proof of the freedom of the placenta from tuberculosis, but also that a normal, tubercle bacilli free child may be born with a tuberculous placenta. Case 4. Patient, aged twenty-five years and bipara, was spontane- ously delivered of a male infant, which weighed 2,570 grams. The child was not separated from its mother and died one month after birth. Au- topsy showed chronic gastric intestinal catarrh, icterus universalis. The mother, at the time of delivery, had tuberculous ulcers on both tonsils and involvement of both apices. The placenta measured eighteen centi- meters by twenty and weighed 680 grams. A false knot, the size of a hazel nut, was found in the umbilical cord near the placental attachment. Microscopic examination of this showed the swelling to be due chiefly to the presence of inflammatory infiltration in the vessel walls and sur- rounding jelly of Wharton. The veins were more involved than the arteries. In the intima of the veins, in the midst of this inflammatory tissue, a group of three tubercles was found; in one of these was a typical giant cell with two tubercle bacilli. Numerous other scattered giant cells without bacilli were found in the walls of the veins and arteries. Examination of the placenta showed a similar round cell in- filtration in the walls of the larger vessels, but no tubercle bacilli. Only a few pieces of placental tissue were examined, however, as the remainder was lost. Case 5. Patient, aged twenty-nine years and septipara, had four chil- dren who were living and well. Induction of labor in the eighth month on account of the condition of the patient. Male premature infant which weighed 1,160 grams and died in four hours. Autopsy showed debilitas congenita vitae and partial atelectasis. The mother died three weeks post partum. Autopsy showed chronic pulmonary tuberculosis, chronic sero- fibrinous pleuritis, chronic tuberculosis of the peribronchial and cervical lymph glands, chronic tuberculosis of the larynx, tuberculous ulcers of the small intestine, and chronic tuberculosis of the mesenteric glands. Three guinea pigs were injected; one with blood from the umbilicus, negative; one with material from fetal organs, negative ; one with material from a yellow white nodule from the maternal surface of the placenta, a smear preparation from which had shown a few tubercle bacilli. This animal developed general tuberculosis. No tubercle bacilli were found upon histologic examination of the placenta, but one bacillus was found against the internal surface of 'the wall of a vein in the decidua basalis. No tu- 8o GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS bercle bacilli were found upon histologic examination of the fetal organs, bone, marrow, or lymph glands, although numerous acid fast bacilli were present in these organs; however, their morphology did not cor- respond to that of the tubercle bacillus. (Probable contamination.) There were, nevertheless, marked changes found in the lymphatic sys- tem — enormous dilatation of sinuses of the lymph glands; enlargement of afferent and efferent lymph vessels, This condition suggests the pos- sibility of circulatory disturbances of the placenta, or may perhaps be due to the presence of toxic substances — bacterial toxins which have been transmitted from mother to child. Case 6. Patient, aged twenty-eight years and bipara, had had two miscarriages. Induction of labor as in preceding case. Dead female child weighed 1,370 grams. The patient showed advanced laryngeal tuber- culosis, involvement of the left pulmonary apex. Autopsy of child showed atelectasis of lungs of premature infant. Several cubic centi- meters of umbilical cord blood were injected into two guinea pigs at time of delivery, and several cubic centimeters of salt solution extract of the placenta into two others. In about three months the animals be- came ill; killed after four months and all showed extensive tuberculous lesions. Histologic examination of the placenta and umbilical cord was negative; the same result with the fetal lungs, thymus, spleen, kidneys, adrenals, stomach, intestines, inguinal and peribronchial lymph glands. A tubercle bacillus was found in the adenoid tissue ; in one retroperito- neal gland, a bacillus was found in the lumen of a dilated lymph sinus. No tubercles or giant cells. The infection of the lymph nodes must have occurred by one of two routes — either directly by means of emboli, or indirectly through the lymph. The latter is probably the more important. It is quite possible, as numerous authors have demonstrated, for tubercle bacilli to lie dor- mant in the lymph glands for a considerable time without causing typical lesions. In this case, the fact that, notwithstanding the presence of tubercle bacilli in the circulation and in the lymph glands, there was no evidence of tuberculous changes in the organs of the fetus, may be ex- plained on the theory that infection had occurred just before birth, and that bacteria had not had time to cause lesions ; but Sitzenf rey is inclined to apply here Bail's aggressin theory — invasion of the fetus occurred only after it had acquired from the previously received tuberculosis ag- gressins an immunity, whereby it was in a position to resist infection and either destroy completely the tubercle bacilli, or force them into a long period of latency. CONGENITAL AND PLACENTAL TUBERCULOSIS 81 TUBERCLE BACILLI IN FETUS AND PLACENTA WITH- OUT HISTOLOGICAL CHANGES Undoubted Cases. — Diagnoses rest upon the demonstration of the bacilli by staining or by inoculation of animals. Warthin and Cowie. 12 Woman in fifth month of pregnancy, with chronic tuberculosis of kidney and general miliary tuberculosis; abortion; death; tuberculous thrombosis of placental sinus and intervillous spaces; tuberculosis of placenta; tuberculous thrombi in fetal blood; presence of free tubercle bacilli in fetal circulation, with histologic changes. Leuenberger 202 reports two cases of acute miliary tuberculosis of the mother, in which miliary tubercles were found in the placenta. Tu- bercle bacilli were present in the fetal circulation. There were evidences of injury to the placental blood vessels. Landouzy and Martin. 203 Mother died of tuberculosis in the fifth month of pregnancy. Portions of the placenta and twenty-five drops of blood from the fetal heart, inoculated into three guinea pigs, produced tuberculosis in the latter in four months. Three other guinea pigs inocu- lated with portions of the fetal liver, lung, and brain, respectively, were negative for tuberculosis. Landouzy and Martin. 203 A portion of the lung of the six and a half months fetus of a tuberculous mother, dying a few days after de- livery, was inoculated into the peritoneal cavities of guinea pigs. The animals died of general tuberculosis four months afterward. Huguenin. 204 A woman of thirty-six years of age died of advanced phthisis florida during the sixth month of pregnancy. Tubercle bacilli had been present in the sputum. An autopsy showed the usual characteristics of this disease and tubercle bacilli were recovered from the blood. The uterus was enlarged, rising as high in the abdomen as the umbilicus. The placenta and child were macroscopically normal, but tubercle bacilli were recovered from the fetal blood. Charron and Karth. 205 Guinea pig inoculated with portions of placenta from tuberculous mother. The result was positive. No tuber- culosis was demonstrated in the child. Herrgott. 103 Woman dying of chronic pulmonary tuberculosis in sixth month of pregnancy. Inoculation of guinea pigs with amniotic fluid was positive. This is the first case of the kind recorded. Schmorl and Birch-Hirschfeld. 124 Seven months fetus removed by cesarean section from mother, twenty-three years of age, dying of miliary tuberculosis. Tubercle bacilli found in the fetal liver, in intervillous 82 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS spaces, and in chorionic villi. No histologic changes of tuberculosis were observed. Inoculation of guinea pigs with portions of fetal liver were positive. Landouzy. 206 Case i. The seven months' fetus of a tuberculous mother showed no histologic changes. Inoculation of guinea pig with portions of fetal organs gave positive results. Case 2. The five months fetus of mother dying of tuberculosis presented no histologic changes. Placenta apparently normal. Inoculations with portions of placenta and heart blood of fetus were positive; inoculations with fetal liver, doubt- ful; inoculations with portions of lung and brain were negative. Aviragnet. 207 The seven months fetus of mother dying of acute miliary tuberculosis showed no histologic changes. Tubercle bacilli were demonstrated in the fetal blood. Inoculation of guinea pig with por- tions of placenta and fetus gave positive results. This case is somewhat similar to the one of Schmorl and Birch-Hirschfeld. Thiercelin and Londe. 208 Mother died of pulmonary and intestinal tuberculosis fourteen days after delivery. Child died on fourth day. Numerous tubercle bacilli were found in the fetal liver, spleen, and kid- neys. No histologic changes found. Inoculations of guinea pigs with blood from umbilical cord were positive. Londe. 209 Case 1. Mother died of acute miliary tuberculosis eight days after abortion. No tuberculous changes of tubercle bacilli were found in the fetus. Inoculations of guinea pigs with portions of liver and placenta, and with fetal blood, were positive. Case 2. Mother died with advanced tuberculosis. Infant died ten days after birth. No macroscopic or histologic evidence of tuberculosis was found in the fetus. Guinea pig inoculations with venous blood, portions of fetal organs, and placenta were positive. Schmorl and Kockel 196 report three cases of abortion in mothers suffering from general tuberculosis. No histologic changes were found in the fetus in any one of the cases. In one case tubercle bacilli were found in the fetal liver, periportal tissue, and lymph glands by staining methods. The placenta in each case contained bacilli and tubercles. The inoculation with portions of fetal organs was negative in all three cases. Bugge. 162 Eight months fetus of mother with miliary tuberculosis lived but thirty hours after birth. Mother died shortly afterward. Tu- bercle bacilli were found in the blood from the umbilical vein and in the liver vessels. No histologic changes found. Three guinea pigs in- oculated with blood from the umbilical vein. Portions of fetal liver and lung gave positive results. CONGENITAL AND PLACENTAL TUBERCULOSIS 83 Londe. 209 The mother died of tuberculosis ten days after delivery. No autopsy was performed, but the diagnosis was beyond doubt. Por- tions of the placenta were inoculated into guinea pig. The animal sub- sequently died of tuberculosis. The infant died, but no autopsy was obtained. Doleris and Bourges. 210 Mother died of acute miliary tuberculosis three weeks after delivery of a seven months child. The marasmic infant died five weeks after birth. No tubercle bacilli or histologic changes of tuberculosis found in child by staining methods. The inoculation of a guinea pig with the heart blood of the child gave a positive result. The fact that the child's blood should contain bacilli for five weeks without occurrence of tissue changes is most remarkable. The case is doubted by Hauser and Cornet. Kynoch. 211 Patient aged twenty-eight years and primipara. She had symptoms of rapidly advancing phthisis for six weeks. Fever of 102 F. She was three months pregnant. Death occurred from tuber- culosis. Postmortem showed macroscopic lesions resembling tuberculosis (nodules) in the lungs, liver, and peritoneum. The adnexa were ad- herent, but the tubes were patulous. The placenta was studied with gray, non-caseous tubercles. The fetus was macroscopically normal. No histologic or bacteriologic examination was reported. Armanni. 212 Mother died of tuberculosis in seventh or eighth month of pregnancy. Fetus showed no histologic changes. Portions of spleen, liver, and brain were inoculated into two guinea pigs. One died four months afterward of general tuberculosis; the other not affected. Sec- ondary infection of pig not excluded. Thiercelin and Londe. 208 Mother died fourteen days after delivery. Had pulmonary and intestinal tuberculosis. Child died a few days after birth. No autopsy. A portion of the placenta was placed in the peri- toneal cavity of guinea pig and gave positive result. Actual condition of child not known. Bar and Renon. 84 Five cases of tuberculous mothers. The blood of umbilical vein was injected immediately into guinea pigs. Three cases gave negative results; two were positive. Of the latter, one case showed no apparent lesions in placenta and fetus, and no tubercle bacilli could be demonstrated by staining methods. Mother was in the last stage of pulmonary tuberculosis. Three animals were inoculated with pieces of liver and lung and with peritoneal fluid. The ones inoculated with portions of liver and lung tissue had general tuberculosis; the one inocu- lated with peritoneal fluid had tuberculosis of spleen. Case 2. This case is open to grave doubt, as secondary infection 84 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS was not excluded. Mother had tuberculous cavities in lungs. Child died of bronchopneumonia on the fortieth day after birth. Placenta appeared normal. Two guinea pigs were inoculated with blood from umbilical vein ; one was negative ; the other developed local and gen- eral tuberculosis. Bolognesi. 213 Portions of 13 placentas from tuberculous women, and in a few cases portions of the fetus also, were examined for tubercle bacilli. In eight cases in which the fetus was examined histologically and animals inoculated, no histologic changes were found. In only one case the inoculation with placenta was positive. The report is inexact and contradictory. Henke. 214 Mother had chronic tuberculosis. Child died four days after delivery. Autopsy showed pneumonia with fresh, fibrinous pleuri- tis. No microscopic changes of tuberculosis. Portions of an appar- ently healthy bronchial gland were inoculated into guinea pigs. The pig showed general tuberculosis on the thirty-seventh day. Neither histologic changes nor tubercle bacilli could be found in serial sections of another gland. Henke excludes accidental infection of inoculated animal and regards case as a typical tuberculosis inoculation. Kurbitz 215 reports a case of tuberculosis of the decidua basalis from the Marburg Pathological Institute. The patient suffered from chronic pulmonary and laryngeal tuberculosis and died three days after a confine- ment in the eighth month. The child weighed only 1,880 grams and was delicate. It showed no signs of tuberculosis and was negative to the tuberculin reaction. It died at three months of age from volvulus, and autopsy showed no signs of tuberculosis. The autopsy in the mother showed the lungs and larynx involved and several small ulcers near the ileocecal valve and tubercles in the liver. At the placental site was a dark red blood clot to which numerous red thrombi were attached. Macro- scopically this did not show disease, but under the microscope typical tuberculosis was present. The superficial zone of the placental site was thickly crowded with miliary tubercles. These did not appear to have penetrated any vessels, although in many areas they were actually approximated to the vessel wall. The thrombi, which were numerous, did not in themselves show tuberculosis, nor was there any marked in- flammation in the basal decidua. Tubercle bacilli were demonstrated in the membranes. The placenta proper could not be investigated. Rielander and Mayers. 216 The patient was twenty years of age and had been suffering from pulmonary phthisis for some years. At the time of admittance to the hospital, the sputum contained numerous typical tubercle bacilli. The disease was progressing rapidly. A vaginal hyster- CONGENITAL AND PLACENTAL TUBERCULOSIS 85 ectomy was performed. Death followed. The decidua presented a well marked tuberculous inflammation, and tuberculosis of placenta was also present. Tubercle bacilli were demonstrated by the Ziehl-Neelson stain. CONCLUSIONS Congenital Tuberculosis. — 1. This is a rare condition; a number of authentic cases have, however, been recorded. That transplacental infection is most likely to occur in the last few weeks of pregnancy and especially as a result of uterine contraction during labor, together with the well known latency of tuberculosis, are facts which are suggestive that this variety of infection may in some instances be mistaken for a postnatal infection. 2. As a result of congenital infection, the liver and adjacent struc- tures, especially the lymph glands, are the localities most frequently attacked. 3. The prognosis in the congenitally infected is unfavorable; first, because of the vital character of the organs usually involved; and sec- ondly, owing to the virulent type of organism usually present, maternal bacillemias rarely being found, except in a virulent type of infection. 4. For a congenital hemogenic infection to occur, a maternal bacil- lemia and a permeability of the placenta must precede the condition. 5. Whether tubercle bacilli can be transmitted through the normal placenta is still undetermined ; certainly, when lesions are present the placenta cannot be regarded as a secure filter. 6. Preexisting lesions in the placenta, especially those produced by syphilis, are predisposing factors to the transmission of tubercle bacilli. 7. Lesions in the placenta may be produced by the tubercle bacilli themselves, which may result in conditions favoring the transmission of organisms to the fetal circulation. 8. The presence of tubercle bacilli in the placenta is undoubtedly of more frequent occurrence than was formerly believed, and, in view of the results obtained by recent investigations, a more thorough study of the question of the frequency of congenital tuberculosis is desirable. 9. Until a fairly large series of fetuses and newly born infants can be thoroughly studied by carefully performed histologic and inoculation methods the relative frequency of this type of infection cannot be de- termined. 10. The presence of tubercle bacilli in the placenta by no means in- fers a congenital infection, but is undoubtedly a predisposing agent. 86 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS ii. Congenital tuberculosis can be experimentally produced in ani- mals, but only in a small percentage of cases, and even then as a rule only by the injection into the mother of what are relatively enormous quan- tities of a pure culture of tubercle bacilli. Such results can hardly be compared with what occurs in the pregnant tuberculous woman. 12. Animal inoculation is the most reliable method of testing for the presence of tubercle 'bacilli, either in the placenta or fetus. 13. The animals should be tested with tuberculin before inoculation and carefully guarded from possible extraneous infection subsequently to injection. 14. As an additional safeguard, a second series of animals should be inoculated from those dead of the primary injection. This should be performed for the purpose of positively determining the virulence of the microorganisms, as it is, at least theoretically, possible that dead tubercle bacilli might be present. This precaution was adopited in many of our cases, and in all in which there was the least ground for doubt. It was positive in all cases. 15. Whereas the antiformin method is of value, the acceptance of one or two acid fast bodies morphologically similar to the tubercle bacil- lus demonstrated in a large series of slides is unreliable. 16. In certain types of maternal tuberculosis, tubercle bacilli are not infrequently present in the placenta. Undoubtedly the most frequent period at which transplacental infection occurs is during labor; for this reason the umbilical cord in these cases should be tied as soon as pos- sible, certainly without waiting for the pulsation to cease. 17. The child should be taken away from the mother immediately and carefully guarded against postnatal infection. Placental Tuberculosis. — 1. Placental tuberculosis may result from the infection of the spermatozoon or ovum. This assumption is based upon theoretic grounds only and such an instance is probably ex- tremely rare, too rare to have much practical importance. 2. Placental tuberculosis may result from a direct extension from a nearby focus, such as a preexisting endometritis. This also is prob- ably a comparatively rare variety. 3. Placental tuberculosis may result from a hemogenic infection. This is the most frequent variety. It requires a maternal bacillemia, a condition which in itself is comparatively infrequent. 4. Bacillemias are most frequently present in the acute miliary form of tuberculosis; when ulcerative lesions break into adjacent blood vessels; when hyperpyrexia is present — a condition which tends to impair the integrity of the blood vessels; and during acute exacerbations of the • CONGENITAL AND PLACENTAL TUBERCULOSIS 87 disease. It is worthy of note that, in the tuberculous woman, pregnancy frequently produces an exacerbation of the disease and results in con- ditions favorable for the production of a bacillemia. 5. Tubercle bacilli are frequently present in the placenta without macroscopic lesions, as proven by our own investigations and those of others. This is a much more frequent condition than are actual macro- scopic lesions. 6. Tubercle bacilli, when found in the placenta, are frequently viru- lent. LITERATURE 1. Schlimpert, H. Arch. f. Gyn. 90:121. 2. Novak, J., and Ranzel, F. Ztschr. f. Gebh. u. Gyn. 1910. 67:719. 3. Schmorl und Geipel. Munch. Med. Woch. 1904. 5:1676. 4. Pankow. Monschr. f. Gebh. u. Gyn. 1910. 32:579. 5. Williams, J. W. Obstetrics. New York and London, 1912. p. 136, 606, 6. Osler, Sir W. "Tuberculosis," edited by A. C. Klebs, 1909. 7. Waldenburg, L. Die Tuberculose, die Lungenschwindsucht und Scrofulose. Berlin, 1869. A. Hirshwald. 8. Bredohl, A. Die Geschichte der Tuberculose. Hamburg und Leipzig, 1888. Voss. p. 482. 9. Johne. Deutsche Ztschr. f. Thiermed. 23:207; also Forts, d. Med. 1885. 3:108. 10. Rosenau, M. Preventive Medicine, New York and London, 1913, p. 122, 137. 11. Sitzenfrey, A. Die Lehre von der Kongenitalen Tuberculose, Berlin, 1909. 12. Warthin, A. S., and Cowie, D. M. Jr. Inf. Dis. 1904. 1 :14c 13. Martius, F. Uber die Bedeutung der Verlegung under der Dispo- sition in der Pathologie mit besonderer Berucksichtigung der Tuberkulose. 22te Kong. f. Inn. Med. Wiesbaden, 1905. p. 689. 14. Bab. Berl. Klin. Woch. 1907. No. 12. 15. Sakurane. Arch. f. Derm. u. Syph. 1906. 82:22. 16. Fouquet. Bui. Soc. d'Obst. de Paris, January 17, 1907. 17. Feuillee. Bui. de la Soc. Med. des Hop. Paris, March 9, 1906. 88 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS 18. Waldstein, E., and Ekler, R. Wien. Klin. Woch. 1913, 26: No. 42. 19. Viet, J. Monschr. f. Geburt. u. Gyn. 1902. 16:525. 20. Martin, A. Monschr. f. Geburt. u. Gyn. 1902. 16:555. 21. Guiteras, R. Urology, New York and London. 1912. p. 617. 22. d'Aubeau. Cong. p. l'Etude de la Tuberc. 1893, nach Schlu- ter. 23. v. Jani. Virch. Arch. 1886. 103:522. 24. Sirenae, P. Gaz. d. osp. 1887, nach Walther. 25. Solles. Jour. Med. de Bordeaux. 1892, nach Ziegler. 26. Foa, S. Gaz. d. osp. 1892, p. 208, nach Ziegler. 27. Spano, F. Gaz. d. osp. Nov., 1893, nach Ziegler. 28. Jackh, A. N. Virch. Arch. 1895. 142:101. 29. Lowenstein. Deuts. Med. Woch. 1913. 39:499. 30. Grawitz. Quoted by S. Pozzi : A Treatise on Gynecology. New York, 1897. p. 661. 31. Murphy, J. B. Tuberculosis of the Female Genitalia and Perito- neum. Chicago, 1903. p. 4. 32. Rohlff, E. Beitr. z. Frage von der Erblichkeit der Tuberkulose. Kiel, 1885. Lipsius u. Tischer. 33. Westmayer, E. Inaug. Dis. Erlangen, 1893. 34. Dobroklonski. Rev. de la tuberc. 1895. p. 195. tr. from Wratsch, 1895. No. 19, 20; also Centrbl. f. Bakt. Par. 19 :625. nach Hauser. 35. Walther, H. Ziegler's Beitr. 1894. 16:274. 36 37 38 39 40 4i 42 43 44 45 46 47 48 Gartner, A. Ztschr. f. Hyg. u. Inf. 1893. 13:101. Cornet, G. Tuberculosis. Philadelphia, New York, and London. 1904. Friedman, F. Deuts. Med. Woch. 1901. 27:129. Varaldo, F. R. Clin. obst. Jan. 15, 1906. Schottlander, J. Monschr. f. Geburt. u. Gyn. 1897. v. 5. Landouzy, L. Ann. de Gyn. et d'Obst. 191 1. 8:33. Rettger, L. F. Jr. Exp. Med. 1914. 19: No. 6. Hoffman and Wolters. Die Atiologie der Syphilis. Berlin, 1906. Springer. Levaditi et Saurage. Compt. rend. Acad, des Sci., Paris. 1896. i43:559- Bab. Deuts. Med. Woch. 1906. No. 48. Magalhaes. These de Rio Janeiro. 1906. Koch, R. Deuts. Med. Woch. 1905. No. 47. Simmonds. Munch. Med. Woch. 1906. No. 2j. CONGENITAL AND PLACENTAL TUBERCULOSIS 89 49. Hollos, J. Virch. Arch. 1913. 213: No. 2, 3. 50. Rosenau and Anderson. Jr. Amer. Med. Assoc. 1908. p. 961. 51. Huppe, F. Inaug. Dis. Rostock, 1902. 52. Bartel, S. Wien. Klin. Woch. 1905, No. 34, 41 : 1906, No. 16 : 1907; 1908, No. 22. 53. Klebs, E. Munch. Med. Woch. 1901. p. 129. 54. Carriere, G. Centrbl. f. In. Med. 1901. p. 1017 55. Bossi. Arch. f. Gyn. 1916. 77:21. 56. Pehu, M., et Chalier, J. Arch, de Med. des enf. 1915. 18: No. 1. 57. Rosenberger, R. C. N. Y. Med. Jr. June 15, 1909. 58. Rump. Munch. Med. Woch. 1912. No. 36. 59. Liebermeister. Med. Klin. 1912. No. 25. 60. Gurner, E. Munch. Med. Woch. 1913. 60:401. 61. Dressen. Med. Klin. 1913. No. 13. 62. Gobel. Deuts. Med. Woch. 191 3. No. 24. 63. Klemperer. Ther. d. Gez. 19 12. No. 10. 64. Kahn, E. Munch. Med. Woch. 191 3. 60:345. 65. Kessler. Munch. Med. Woch. 1913. 60:346. 66. Bacmeister. Centrbl. f. d. Grenzg. d. Med. u. Chir. 1913. 16: No. 5, 6. 6y. Vinogradoff. Russky vratch. 1914. 13 : No. 22. 68. Fraenkel. Deutsch. Med. Woch. April 17, 19 13. 69. Elsasser, J. Beitr. z. Klin. d. Tuberk. 19 13. 26: No. 4. 70. Bogason, P. Ugesk. f. Lag. 1913. 75 : No. 18. 71. Calmette, A. Press Med. Feb. 7, 19 14. y2. Lubarsch, P. Ergebnisse der Speciellen Path. Morph. u. Physiol, des Menschen u. d. Tierre. Wiesbaden, 1896, J. F. Bergmann; also Virch. Arch. 124:4. 73. Schaudinn. Arb. a. d. Kais. Geshtamt. 1907. 26:11. 74. Paschen. Munch. Med. Woch. 1906. p. 622. 75. Wallich et Levaditi. Compt. rend. Soc. de Biol. 1906. 60:191. j6. Menetrier et Rubeno-Duval. Presse Med. 1906. p. 7. yy. Neuhaus. Bed. Klin. Woch. 1886. p. 389. 78. Freund und Levy. Berl. Klin. Woch. June 2y, 1895. 79. Van der WiTTiGEN. Ned. Tijdschr. v. Geneesk. 1895. No. 11. 80. Dorland, W. A. N. Am. Jr. Obst. and Gyn. June, 1900. 81. Runge. Arch. f. Gyn. 1903. 68:388. 82. Nattan-Larrier et Brindeau. Compt. rend. Soc. de Biol. 1906. 60:181. go GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS 83. Delestre. Sem. Med. Feb. 9, 1898. 84. Bar et Renon. Ann. de gyn. et d'obst. 1895. 44:217. Also Rev. de la tuberc. 1895. p. 237. Also Compt. rend, de la Soc. de Biol. 1895. 10:505. 85. Preyer, W. Physiologie special de l'embryon ; recherches sur les phenomenes de la tnberculose avant la naissance. Tr. de l'Alle- mand par Wiet. Paris, 1887. F. Allen. 86. Savory. Quoted by Preyer; No. 85. 87. Fournier. Sem. Med. Nov. 30, 1898. 88. Zwiefel. x^rch. f. Gyn. 1877. p. 235. 89. Jung, P. Ther. Montschr. 1914. 28: No. 2. 90. Waldeyer. Arch. f. Mikr. Anat. 1890. 35 :i. 91. Bumm. Arch. f. Gyn. 1893. 43:181. 92. Leopold. Verhl. d. Deutsch. Gesel. f. Gyn. 1890. 3, 257. 93. Williams, J. W. Johns Hopkins Hospital Reports. 1893-94. 3:87; 1 141. 94. Delore. Jr. de Med. de Paris., April 16, 1899. 95. Warthin, S. A. Jr. Inf. Dis. 1907. 4:347. 96. Warnekros. Deutsch. Gesel. f. Gyn. May, 19 13. 97. Schmorl und Kockel. Centrbl. f. Gyn. 1894 18:307 98. Asch. Monschr. f. Gebh. u. Gyn. 1913. 27:701. 99. Leloir. In Verneuils Etudes sur la tuberculose. 1892. 3 .-482. 100. Baginsky, B. Berl. Med. Gesel. Jan. 14, 1891. 101. Wasmuth, B. Centrbl. f. Bakt. 1892. 3:824. 102. Roth. Ztschr. f. Hyg. 4:151. 103. Herrgott, A. Ann. de Gyn. et d'obst. 1891. 36:1. 104. Hochsinger, S. Wien. Med. Blat. 1894. 17:255. 105. Henle, A. Pseudotuberculose bei Neugeborenen Zwillingen, Orth. Festsch. f. Virchow. 1893. p. 143. 106. Merlitti, C. Arch, di obst. e. gin. 1901. p. 512, 649, 714. 107. Barkley, C. Jr. Obst. Gyn. Brit. Emp. 1903. 3:31. 10S. Ascoli. Policlin. 1899. Supp. p. 370. nach Bossi. 109. Leyden, E. von. Ztschr. f. Klin. Med. 1884. 8:375. 110. Jaquet. Quoted by Cornet, No. ^y. in. Vignal, W. Deuxieme cong. pour l'etude de la tuberc. Pans, 1891. p. 334. 112. Treisser. Quoted by Straus; No. 149. 113. Bernard, Debrer, and Baron. Quoted by H. Dufour and J. Thiers in La Gynecologic 19 13. p. 400. 114. Bar and Renon. Rep. univ. d'obst. et de gyn. Sep., 1895. 115. Armann. Fourth Int. Cong. Obst. and Gyn. Rome, 1904. CONGENITAL AND PLACENTAL TUBERCULOSIS 91. 116. Leroux, L. Tuberculose du premier age d'apres les observations indedites du Prof. Parrot. In Verneuil's Etudes sur la tuber- culose. 1892. 117. Lannelongue, O. M. Legons de Clinique Chirurgicale, Paris, 1905 ; also Cong, pour l'etude de la tuberculose, Paris, 1889. 118. Muller, D. Munch. Med. Woch. 1899. p. 875. 119. Haupt. Deutsch. Med. Ztschr. 1891. p. 997. 120. Hardy. Bui. Soc. Anat. de Paris. 1834. p. 115. 121. Charrin. Lyon Med. 1873. p. 295. 122. Jacobi, A. Compt. Rend. 2eme Cong. Tuberc. Paris. 1891. 2 1327. 123. Sabouraud, R. La med. Mod. 189 1. 2:749. 124. Schmorl und Birch-Hirschfeld. Ziegler's Beitr. 1891. 9 1428. 125. Runge. Centrbl. f. Gyn. 1884. No. 48. 126. Klepp. Ztschr. f. Fleisch. u. Milch. Hyg. 1897. 7:67. 127. Malrox et Brouwier. Ann. Inst. Past. 1889. 3:153. 1902. 128. Czoker. Vers. Deutsch. Naturf. u. Arz. Wien. 1894. 129. Bank. Deutsch. Ztschr. f. Thiermed. 1890. 16:409. 130. McFadyen, Jr. Comp. Path. Ther. 1891. 6:353. 131. Siegen, C. Cong. f. Erforsch d. Tuberc. Paris, 3rd Session. 132. Lungwitz. Centrbl. f. d. Med. Wiss. 1894. 32:414. 133. v. Nocard. Ztschr. f. Fleisch. u. Milch. Hyg. 1897. 7:98. 134. Grancher, A. Sem. Med. 1886. p. 297. 135. Kohler. Siedamgrotzk'scher Jhb. 1888-89. 136. Misselwitz. Siedamgrotzk'scher Jhb. 1888-89. 137. Bayersdorfer. Mitt. d. Ver. Bad. Thieraz. 1892. p. 55. 138. Becker. Ztschr. f. Fleisch. u. Milch. Hyg. 1895. 5:115. 139. Ruser. Ziegler's Beitr. 16:294. 140. Barland. Baumgarten's Jahrb. 9 749. 141. Galthier, S. Ann. Inst. Past. 2:492. 142. Bucher. Ztschr. f. Fleisch. u. Milch. Hyg. 1897. 7:217. 143. Lohoff. Ztschr. f. Fleisch. u. Milch. Hyg. 1897. 7:163. 144. Brooks, H. Proe. Soc. Exp. Biol. Med. 1914. 11:50. 145. Lodenih, L. Press, med. 191 1. No. 83. 146. Sanchez-Toledo. Arch, de med. exp. et d'anat. path. 1889. 1 -S03- 147. v. Nocard. Rev. de la Tuberc. 1895. p. 226; also Centrbl. f. Bakt. 1896. 19:625. 148. Wolff, M. Virch. Arch. 1886. 105:192; v. 106: also Vir- chow's Festsschr. v. 3. 92 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS 149. Straus, I. La tuberculose et son bacilli. Paris, 1895, Rueff and Cie. p. 545. 150. Carajnis. Citti del Instit. Veneto. 1885-86. 4:i J 45- 151. Feyerabend, P. Beitr. z. Klin, der Tuberk. 1914. 29: No. 1. 152. Eden, N. T. Jr. Path. Bact. 1897. 5 1265, 282; also A Manual of Midwifery. 153. Lehmann, F. Berl. Klin. Woch. 1894. 31 :6oi. 154. Houl. Bui. Internat. Acad. Sc. Prag. 1894. 155. Ustinow, A. Arch. f. Kindk. 1898. 25:66. 156. Auche et Chambrelente. Press med. Mar. 22, 1898. Also Arch, de med. exp. et d'anat. path. 1899. 11 1521. 157. Veszpremi, D. Centrbl. f. Path. Anat. 1904. 15:483. 158. Dufour, H., et Thiers, J. Bui. Soc. de Ped. de Paris. 1913. 16:274. 159. Brindeau. Arch. ped. July, 1899. 160. Stockel. Beitr. z. Klin. d. Tuberk. 1904. 1 :i29. Also Munch. Med. Woch. 1908. No. 10:535. 161. Zarfl, M. Ztschr. f. Kindk. 1913. 8: No. 5. 162. Bugge, J. Ziegler's Beitr. 1896. 19:433. 163. Moller, W. Arch. mens, d'obst. et de gyn. 1914. 3 : No. 7. 164. Gulee and Harms. Tr. Chicago Ped. Soc. Jan. 20, 1914. Also Am. Jr. Dis. Childn. 1914. 9: No. 4. 165. Delmas. L'Obstetrique. May, 1910. 166. Bourges. Jhrb. f. kindk. 47 : No. 1. 167. Demme, R. Ber. u. d. Thatigk. d. Jenners. Kindsp. in Bern. 1868. No. 6. 1875. No. 13. 1880. No. 17. 1886. No. 24. 168. Merkel. Cited by Ohlendorff, Ztschr. f. Klin. Med. 1884. v. 8 : No. 6. 169. Baumgarten, P. Arb. a. d. Path. Anat. Inst. z. Tub. 1 :322. Also Sam. Klin. Vortr. No. 218. 170. Berti. Bol. delle sc. med. Bologna, 1882. 171. Demme, R. Verhl. d. Vrsml. Deutsch. Naturf. u. Arzte. Frei- burg, 1884. v. j. 172. Money, P. Brit. Med. Jr. 1885. 1 :i247. 173. Gueyrat, L. Contribution a l'etude de la tuberculose du premier age. Paris, 1886. p. 179. 174. Flesch. Jhrb. f. Kindk. 1886. No. 25. Also Pest. Med-chir. Presse. 22 :830. 175. Frobelius. Jhrb. f. Kindk. 1886. No. 24:47. 176. Houtinel. These de Paris. 1886. CONGENITAL AND PLACENTAL TUBERCULOSIS 93 177. Lannelongue, O. M. In Verneuil's Etudes sur la tuberculose. 1887. 1. 178. Huguenin. Cited by Lebkuchner in Gaz. des hop. 1888. 61 : 785- 179. Bosselut, F. Contribution a l'etude de la meningite tuberculeuse chez les jeunes en f ants agea de moins de deux ans. Paris, 1888. 180. Rindfleisch. Ber. d. 63ten. Natf. Vrsml. Bremen, 1890. 181. Sarwey. Arch. f. Gyn. 43 1162. 182. Wassermann. Ztschr. f. Hyg. u. Inf. 17:343. 183. Straus, I. Rev. gen. de l'antiseps. med. et chir. 8:97. 184. Kissel. Arch. f. Kindk. 25 :6y. 185. Holt, L. E. Med. News. 69:656. 186. Henoch, E. Vorlesung iiber Kinderkrankheiten. Berlin, 1897. 187. Bonnet, M. L. Lyon med. 87 :224. 188. Johnson, H. M. Phil. Med. Jr. 3:231. 189. Lebkuchner, F. Arb. a. d. Path-Anat. Inst. z. Tub. 1899, 3- 190. Friedmann, F. F. Deutsch. Med. Woch. 1900. 26:381. 191. Lyle, B. F. Phil. Med. Jr. 1900. 6:219. 192. Lehmann, F. Deutsch. Med. Woch. 19:200. 193. Harbitz, F. Munch. Med. Woch. April 6, 1913. 194. Warthin, A. S. Jr. Am. Med. A. 1913. 61 11951. 195. Carl. Ziegler's Beitr. 1907. v. 41. 196. Schmorl und Kockel. Ziegler's Beitr. 1894. 16:313. 197. Jung. Monschr. f. Gebh. u. Gyn. 23:191. 198. Warthin, A. S. Med. News. 69:319. 199. Auche und Chambrelente. Munch. Med. Woch. 45:616. 200. Wollstein, M. Arch. ped. 22:321. 201. Walther, C. Beitr. f. Path. Anat. v. 41. No. 3. 202. Leuenberger, L. Beitr. z. Gebh. u. Gyn. 15:456. 203. Landouzy, L., et Martin, H. Rev. de med. 3 :ioi4. 204. Huguenin, B. Centrbl. f . Bakt. Par. Inf. 48 ^94. 205. Charron et Karth. Rev. de med. 5 :659. 206. Landouzy, L. Rev. de med. 11:431. 207. Aviragnet, E. C. Gaz. hebd. de med. 29 409. 208. Thiercelin, E., et Londe, P. Med. mod. 4 ^98, also Gaz. des hop. 66:189. 209. Londe, P. Res. de la tuberc. 1893, p. 125. 210. Doleris et Bourges. Sem. med. 16:375. 211. Kynoch, J. A. Scot. Med. Surg. Jr. 20:1018. 212. Armanni. Tr. 10th Int. Med. Con., 1890. 94 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS 213. Bolognesi, A. These de Paris, 1895. 214. Henke. Arb. a. d. Path-Anat. Inst. z. Tub. 1896. No. 2. 215. Kurbitz. Munch. Med. Woch. 1908. p. 535. 216. Rielander, A., und Mayers, K. Arch. f. Gyn. 87:131. CHAPTER V INFECTION IN GENITAL TUBERCULOSIS Routes of infection in genital tuberculosis — Primary genital tuberculosis — Modes of infection — History of cases — Relative infrequency in women — Analysis of litera- ture — Summary of experiments — Clinical proof — Secondary genital tuberculosis — Latency of the disease — Determination of source of infection — Difference of opinion regarding frequency of primary and secondary infections of female genital tract — Study of cases — Summary — Predisposing causes — Frequency — His- tologic examination. ROUTES OF INFECTION Primary Genital Tuberculosis. — Tubercle bacilli may gain access to the genital tract in a number of ways. A phthisical patient may con- taminate a douche nozzle or other article with sputum or other infected material, which may be brought in contact with the genitalia, and thus produce what is to all intents and purposes a primary genital tuberculosis. This auto-infection, endogenous, or primary-secondary infection, as it is termed by Pozzi, 1 is, however, more theoretical than practical, for it is impossible to positively exclude the hematogenic or lymphogenic route under such circumstances, unless an autopsy is performed. Even then it is often difficult. Under the latter conditions more or less well developed areas of tuberculosis are often found along the route of the infection, if the case has been one of hematogenic or lymphogenic type. Sachs 2 has very properly pointed out that the term "primary genital infection" should be reserved for those cases in which no other focus of tubercu- losis exists within the patient's body. The difficulty in positively deter- , mining this point has already been referred to, and makes this classifi- cation faulty, as in many cases this cannot be positively ascertained. The fact, however, that primary genital infection may occur shows that the primary-secondary infection is possible and should lead to prophylactic measures being instigated in tuberculous women. On account of the frequency of tuberculosis, especially pulmonary phthisis, its well known latency, and the fact that small lesions not in- frequently heal, certain writers have doubted the existence of primary genital infection. Primary genital tuberculosis by direct infection from without is extremely rare, and many of the examples of this condition 95 96 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS found in literature are not above suspicion. Aman 3 rightly holds that marriage with a tuberculous man offers more chances of infection through the respiratory or alimentary tract than through the genital apparatus. In proof of this assertion, he points to the fact that the tubercle bacillus is non-motile and can, therefore, only follow the secretion current which in the uterus and vagina is outwards. He states that in rare instances primary genital tuberculosis may occur in young children by direct infec- tion. Even were we to admit all of Aman's conclusions, many of which are not borne out by our knowledge of the action of tubercle bacilli in other parts of the body, we must remember that it is at least possible for the organism to gain access to the endometrial cavity by adhering to spermatozoa, which can readily make headway against the outflowing secretions of the genital tract. Veit, 4 after an instructive dissertation on tuberculosis of the female genital tract, summarizes his remarks as fol- lows : ( i ) Tuberculosis of the female genital tract is more frequent than is generally thought. (2) There is a primary form, but the sec- ondary is the more frequent. (3) The infection is usually a descending rather than an ascending one. (4) Recovery may occasionally occur spontaneously. (5) The best treatment of the primary form is extir- pation. (6) In the secondary forms the treatment should be directed toward improvement of the general health and operative intervention in selected cases. Gutierre 5 at the same meeting gave an account of some original work, the results of which strongly favored the theory of primary infection in certain cases. Von Rosthorn 6 emphasizes the rarity of the primary form and the difficulty of positively excluding other tuberculous lesions. A similar statement is made by Wiener. 7 Blau 8 examined 36 cases from Chrobak's clinic and failed to find a single one which he would accept as primary. That primary genital tuberculosis occurs is now well recognized and is practically proven by the many cases which have been recorded, in which no other lesions, except those observed in the genital system, have been found, even when careful autopsies were performed. Assur- ance of it is strengthened by the fact that frequently, when the genital focus has been removed by operation, perfect health has been maintained for long periods. That wound infection by the tubercle bacillus is not particularly in- frequent is well known. Holt 9 has collected 16 cases of tuberculosis ac- quired through ritual circumcision, infection by tatooing has been re- corded, and numerous other instances of wound infection are on record. When tuberculous lesions are present in the genitalia of a man, the micro-organisms must not infrequently be introduced into the vagina INFECTION IN GENITAL TUBERCULOSIS 97 and on the external genitalia of the wife. It seems also to be proven that in rare instances tubercle bacilli may be found in the spermatic fluid of tuberculous men whose genital tract is healthy. Doubtless the reason that primary genital tuberculosis in women whose husbands have tubercle bacilli in their spermatic fluid is so rare, lies largely in the pro- tective properties of the vagina, which is lined by multiple layers of squamous epithelium that offers an excellent protective barrier against infection of any kind. The bactericidal properties of the vaginal secre- tion have also been amply proven by Dubendorfer, 10 Pankow, 11 Menge, 12 and many others. Numerous animal experiments have been carried out, most of which tend to show that virulent tubercle bacilli may be deposited on the normal vagina without producing infection, but that, if the vaginal mucosa be traumatized or an inflammation be present, a route of ingress is produced and infection may result. In subsequent pages these experi- ments will be reviewed more fully. It is only in extremely rare instances, if ever, that tubercle bacilli deposited in the normal vagina produce lesions. Much has been written upon the question of coitus as a mode of primary genital infection. Infection by coitus may be taken as a type representing all forms of direct genital infection. Cohnheim 13 was the first to suggest this form of infection. Three years later Verneuil 14 stated that tuberculous men with sound genital organs might transmit the infection. As is well known, genital or urinary tuberculosis in men is by no means infrequent, and numerous cases have been recorded in which the husband is supposed to have infected his wife in this way. Veit 4 and Martin 15 state that tuberculosis of the male genital tract occurs in three per cent of all cases. The positive proof that infection has been transmitted by coitus is extremely difficult to obtain. Tubercle bacilli are frequently found in the seminal discharge of tuberculous men. When a tuberculous epididymitis or orchitis is present, the seminal dis- charges almost invariably contain tubercle bacilli, while in some cases, notably those of d'Aubeau, 16 the discovery of the bacilli in the semen, without any lesions in the genitalia, was the first evidence of phthisis. Jani 17 and others have recorded finding tubercle bacilli in the testes of phthisical men in whom no demonstrable genital lesions were present. On theoretic grounds, tubercle bacilli, circulating in the blood, should not gain access to the testicular or prostatic fluids, but should be en- meshed in the fine capillaries leading to the glandular structures of these organs; but Grawitz 18 has demonstrated that corpuscles and mold germs (which are larger than tubercle bacilli) may, under certain circumstances, reach the testicular secretion from the blood stream. Murphy 19 points 98 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS out that it is sometimes extremely difficult to diagnose tuberculosis of the seminal vesicle, and suggests that under conditions such as have just been mentioned, tuberculous genital lesions may have been over- looked. Rohlff 20 and Westmayer 21 have demonstrated that tubercle bacilli are rarely present in the spermatic fluid of tuberculous men, unless genital lesions are present. The former inoculated goats and rabbits with the spermatic fluid obtained from ten men who died of phthisis, with negative results. The latter injected ground up particles of the testicles and ovaries, of patients who had died of tuberculosis, into the peritoneal cavity of rabbits, with similar results in 14 observations. Dobrolonski 22 tested the contents of the seminal vesicles of 25 men who died of phthisis, by means of smears and inoculation. Twenty-four were negative and the one positive result was obtained from a subject in whom a tuberculous epididymitis was present. Walther 23 examined 161 sections from the testes, 48 from the epididymis, 63 from the pros- tate, from the bodies of twelve phthisical men, without finding a single tubercle bacillus. Murphy 19 reports a case of tuberculous salpingitis, in which the in- fection is supposed to have been transmitted by coitus. He remarks, however, upon the infrequency of this mode of infection. Derville 24 reports the histories of 8 cases, all of which are suggestive of this variety of infection. Fernet and Derville 25 and Sheills 26 report similar cases. The relative infrequency of genital tuberculosis, in women whose hus- bands are known to be the incumbents of tuberculous genital or urinary lesions, seems to be a proof that the simple deposition of virulent tubercle bacilli in the vagina is not productive of tuberculosis in the female gen- erative organs under ordinary circumstances. Many experiments have been performed with the view of determining this point, and more or less contradictory results have been obtained. After having studied the somewhat voluminous literature upon this subject and carefully analyzed the results obtained, no one can fail to be impressed with the fact that virulent tubercle bacilli, when deposited within the normal vagina, do not, under ordinary circumstances, produce either local or general lesions, but that some trauma, loss in continuity of the vaginal lining, or special susceptibility of the new host, is necessary before the tubercle bacilli can produce pathologic changes. The vagina and portio vaginalis are invested by multiple layers of stratified squamous epithelium and differ but little in their histologic structure from the skin. Indeed, histological investigation tends to prove that the vaginal lining is a modified skin and is in no sense a mucous membrane. In the course of ordinary, mod- ern life, tubercle bacilli are probably frequently brought in contact with INFECTION IN GENITAL TUBERCULOSIS 99 the exposed surfaces of the body, and in phthisical individuals, unless the strictest prophylaxis is enforced, the patient's skin must very often be contaminated; yet tuberculosis of the skin, resulting- from this form of infection, is extremely infrequent. The outward flow of the vaginal and uterine secretions, the general, downward direction of the genital canal which favors drainage, and the more or less occlusive cervical secretion in the canal with the bactericidal properties of the vaginal se- cretion, all doubtless play a part in preventing the ascent of the non- motile tubercle bacilli which must occasionally be deposited in the vagina of women, the wives of tuberculous men. The relative frequency with which genital tuberculosis develops after abortion or labor, and its in- frequency in wives, the husbands of whom are known to have genital or urinary lesions and who* must frequently be exposed in this manner to the action of the tubercle bacilli, is further clinical proof of this as- sumption. Bull 27 relates an interesting case bearing upon this point. A man contracted tuberculosis in his youth, later married, and at the end of one year a healthy child was born. Two years later it was necessary to remove the right testicle and epididymis for tuberculosis. At this time the left testicle was also diseased, but was spared. One year later the prostate became involved. No further operative treatment was insti- gated. Examination of the spermatic fluid at this time showed the ab- sence of the characteristic odor and of the Florence reaction; an injec- tion into guinea pigs was positive for tuberculosis. During this period of advancing genital infection, the wife had borne two children. These children showed no evidence of tuberculosis and were negative with the Von Pirquet test. The wife presented no symptoms of genital infection. Undoubtedly infection by coitus may occur in a number of ways. As already mentioned, the spermatic fluid may contain virulent tubercle bacilli, either as a result of an internal genital lesion, or from the urinary tract. The penis itself may be the seat of a tuberculous ulcer, or the organism may be upon the external surface of a normal male organ. Numerous animal experiments have from time to time been per- formed in an effort to determine the effects produced by the deposition of tubercle bacilli into the vagina. A summary of these experiments shows that ( 1 ) Tubercle bacilli when deposited in the normal adult vagina rarely if ever produce lesions. (2) When, however, the vaginal lining has been traumatized and there is loss of continuity of the lining mem- brane, infection may occasionally occur. (3) Similar results are likely to take place, if the vaginal lining has been inflamed, either by chemical or bacteriologic means. (4) Pregnancy and the puerperium favor in- ioo GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS fection, especially the latter. (5) When infection occurs it may be local, distant or a general tuberculosis may be set up. The amount or character of the trauma or irritation does not seem to be a guide in this respect. (6) Tuberculosis may result from the spread of the mumicroorganism in the vagina to distant parts of the body without producing local lesions. (7) The route of the infection appears to be through the lymphatics, the pelvic glands usually being involved. (8) Local infection occurs in only the minority of cases. (9) Young animals seem to be slightly more sus- ceptible than old. (10) When tubercle bacilli are deposited within the uterus, the percentage of infections is increased. For further details of research on this subject, the reader is referred to the works of Cornet, Flugge, Dobrolonski, Cornil, Oncarani, Basso, Jung and Bennecke, Blau, Varaldo, Andrews, Sugimura, Hartmann, Williams, Gorovitz and Popov. In view of the above clinical and experimental evidence it appears to be an established fact that in rare instances tuberculosis may be spread by coitus. Precautions are, therefore, indicated, particularly in those cases in which there is a tuberculous lesion in the genito-urinary tract. As has already been stated, tubercle bacilli may be deposited in the genital tract in a number of ways : coitus, septic instrumentation, infected fingers, douche nozzles, dressings, etc. All these should be guarded against. In not a few cases, lesions of the lower genital tract have been attributed to tubercle bacilli bearing endometrial or tubal secretions, A number of cases have been recorded in which the genital lesions were plainly the result of infected discharges from tuberculous intestines, the access of the tubercle bacilli being gained to the genital tract through a rectovaginal, or other form of fistula. The tubercle bacilli may also gain access to the genital tract through contamination of the latter by diarrheal discharges, the result of tuberculous lesions in the alimentary tract. In the same manner genital tuberculosis may result from tuber- culous lesions in the urethra, kidney, ureter, bladder ; or the lower genital tract may be contaminated by discharges originating from a salpingitis or endometritis. It will be noted that a large proportion of the above mentioned lesions produce conditions which result in constant and pro- longed irritation, and that in many cases, as a result of the discharges, a local inflammatory reaction results. This is probably a factor in lessen- ing the resistance of the parts and thus making them more susceptible to the action of the tubercle bacilli. Another method of infection is by direct extension. Thus the genital lesion may be due to a direct exten- sion of a tuberculous focus in the bowel, bladder, or other adjacent struc- ture, either by way of a fistula or through adherent inflammatory struc- tures without actual macroscopic loss of continuity. The infection, espe- INFECTION IN GENITAL TUBERCULOSIS 101 daily of the tubes, may and frequently does follow tuberculous peritonitis. The question, under such circumstances, which is the primary lesion, is sometimes difficult to determine. Tuberculosis of the adnexa may result from a direct extension from a peritonitis, from a deposition by the peri- toneal currents of tubercle bacilli in the peritoneal fluids, or may be purely secondary infection resulting from a lymphogenous or hematogenous origin. These three possibilities are of more theoretic than practical importance. The works of Muscatello, 28 Clark and Norris, 29 and others have amply proven that the general direction of the intraperi- toneal currents is towards the diaphragm and that the chief absorption of the peritoneal fluids occurs in the neighborhood of the central tendon of the latter structure. When, however, the openings of the lymphatics of the diaphragm become blocked with debris, as in the case of peritonitis or ascites, absorption through this structure is greatly diminished, as has been shown by the experimental work of Waterhouse 30 and others. Pinner 31 demonstrated that when powdered cinnabar was introduced into the peritoneal cavity of rabbits, a small proportion of it eventually found its way into the vagina through the tubes and uterus, and it would seem probable, therefore, that in the case of a tuberculous peritonitis, tubercle bacilli might in the same way be swept out through the genital tract and secondarily produce a lesion in the cervix, vagina or external genitalia. Jani demonstrated tubercle bacilli at autopsy in the lumen of a macroscopically normal fallopian tube in a phthisical patient. An ulceration of the intestine was also present. In this case the tubercle bacilli may possibly have been carried to the tube by way of the blood stream. Kaufmann 32 was one of the first to record the history of a case in which tuberculosis of the genital tract was the result of direct infection from a tuberculous intestine and genital fistula. In his case there was a fistulous opening between the small intestine and uterus. Kraus 33 has reported a case of ovarian tuberculosis, which resulted from a similar infection of the vermiform appendix. Secondary Genital Tuberculosis. — Under this head should be classified many of the methods of infection just described. For reasons already stated, it is sometimes extremely difficult, when two separate foci of tuberculosis are present in a patient, to determine which lesion has been the primary one. The well known latency of the disease, the fact that the primary lesion is not necessarily the most advanced and may have become of much less clinical severity or may even have progressed to resolution before the secondary lesion has advanced to sufficient mag- nitude to attract attention, add greatly to this difficulty, and make the determination of the source of primary infection almost impossible to 102 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS positively ascertain, unless a careful postmortem can be performed, and even then is often difficult. The study of tuberculosis in parts of the body other than the genital tract and the preponderance of this infection in certain localities, such as the lungs, may, however, to all intents and purposes, practically clear up this difficult problem. Many forms of tu- berculosis are at some stage bacteriemias. When, therefore, genital lesions are present in conjunction with phthisis, the latter should be considered the primary focus. Authorities differ widely in their opinions regarding the frequency of primary and secondary infections of the female genital tract. Frerich 34 states that, of the genital tuberculoses in women, 6 per cent are primary; Mosler 35 places the proportion at 17.3 per cent; Spath, 36 at 24.5 per cent; Schramm, 37 20.9 per cent; Merlitti, 38 at 18.6 per cent; Berkley, 39 10.8 per cent; Frerichs, 40 at 15.6 per cent; and Horizontow, 41 10 per cent. Williams 42 is of the opinion that blood infections are more frequent than generally supposed. Villard 43 found the lungs diseased in one half of the cases of genital tuberculosis. For reasons already stated, the estimate of the proportion of primary and secondary genital tuberculosis is necessarily extremely difficult and cannot be accurately determined. From a study of the cases which have been treated in the gynecological department of the University of Penn- sylvania, it would seem that secondary infections are by far the most fre- quent, probably not more than 5 or at most 10 per cent being primary. Obviously the question as to whether a given case is a primary or sec- ondary infection is of great importance in governing the prognosis and treatment. To summarize — Genital tuberculosis may arise in one of four ways: 1. By direct infection from without. This is a rare form, but its existence has been definitely proven both clinically and experimentally. The infective organism may come from the patient's own mouth or other lesion by way of the hands, etc., or may originate in another host and be conveyed to the woman's genital tract by coitus, septic examina- tions, etc. 2. Infection of the genitalia may be secondary by way of the blood stream ; the primary focus may be distant or near at hand, the lungs being the most frequent site for the primary infection. This is a frequent form of genital infection. 3. Infection may result from a direct extension from a nearby focus, such as the peritoneum, intestine, bladder, etc. This also is a frequent method of infection. INFECTION IN GENITAL TUBERCULOSIS 103 4. Infection may result from a lymphatic infection, usually from a comparatively nearby focus. Predisposing Causes of Genital Tuberculosis. — As has been stated, genital tuberculosis is more frequent in the female than in the male. This can probably be largely accounted for on an anatomical basis. In cases of tuberculous peritonitis, the tubes are naturally exposed to infection. The lower genital tract in women is also more subject to invasion by tubercle bacilli bearing discharges from the alimentary tract and from external infection in general, than are the corresponding organs in the male. Von Franque 44 and Murphy are of the opinion that tuber- culous salpingitis usually results from an infection via the peritoneum. In many of our cases the reverse has been true. The congestion incident to menstruation and pregnancy and the trauma of the latter are also predisposing factors. Gonorrhea seems in many cases to prepare the soil for the invasion of the tubercle bacilli, and the same may be said of any inflammation, especially chronic ones. Loss of continuity of the surface epithelium appears in many cases to offer an entry way for the tubercle bacilli. Schuchardt, 45 Saulmann, 46 and others have directed attention to the frequency with which tuberculosis follows or occurs concomitantly with venereal diseases. The age is undoubtedly a predisposing factor, but this varies with the character of the lesion and will be considered under the description of the various organs. Hegar, Merlitti, 38 de Rou- ville, 47 Schiffmann 48 and others are of the opinion that hypoplasia of the genital organs is a strong predisposing factor to tuberculosis. A study of our series of cases has not borne out this opinion. Frequency of Genital Tuberculosis. — Genital tuberculosis is more frequent in women than in men. According to Amann, 3 20 per cent of tuberculous lesions involve the genital tract in females and 3 per cent in males. In many cases tuberculosis of the genital tract can only be diag- nosed by the microscope or by culture or inoculation, and the proportion of cases in which the macroscopic lesions are sufficiently characteristic to lead to a positive diagnosis is by no means large. Williams 38 states that, in his series of cases of tuberculosis of the genitalia, 75 per cent were of the "unsuspected variety" and were only diagnosed when the tissue was examined histologically. This difficulty in making macro- scopic diagnoses of genital tuberculosis is probably largely accountable for the divergent results reported by various pathologists and surgeons. Thus Courts 49 found genital tuberculous lesions in 1 per cent of women dying of tuberculosis; Louis, 50 in 2.5 per cent, and Cornil, 51 in 2 per cent; Kiwisch, 52 in 2.5 per cent, and Mosler, 35 in 2.5 per cent; Schramm, 37 in 4.1 per cent; Nimias and Christoforis, 53 in 8.3 per cent. 104 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Owing to the increased knowledge of the pathology produced in the genital tract by the tubercle bacilli, more recent references place the pro- portion of genital infection much higher. In 1901 Merlitti 34 placed the proportion at 12.6 per cent, which is the figure given by Foster. 54 Pure- foy 55 places the proportion at 7 per cent. Martin 15 stated that genital tuberculosis was present in 4 per cent of females who died of tuberculo- sis before puberty, in 12 per cent of those dying during the child bearing period, and in 22 per cent of those dying in later life. Berkley, 35 in statistics including the years between 1880 and 1902, found in 798 necrop- sies, performed in females who died of tuberculosis, the genital organs affected in 62, or y.j per cent. Simonds 56 gives the results of 6,000 portmortem examinations, which lead him to the following conclusions : Tuberculosis of the female genital tract is found in 1.5 per cent of all cases. It is most frequent between the ages of 20 and 30 years. In 87 per cent the tubes are affected, and in 76 per cent, the uterus. Meyer- Riigg 57 states that in 2 per 'cent of all female corpses there is found to be tuberculosis of the genital organs. Taking into account only women dying of tuberculosis, there is genital tuberculosis in 10 per cent. The value of many of the statistics regarding the frequency of genital tuber- culosis is somewhat nullified by the fact that the authors fail to state whether routine histologic examinations have been made, or whether the results refer only to those cases which have presented macroscopic lesions. Probably the latter was the case in many of the statistics. Furthermore, information regarding the location of the lesion in the genital tract is not infrequently lacking. The organs of generation are involved in the following order of frequency : tubes, uterus, ovaries, vagina and vulva. This ratio holds good, whether the infection be primary or secondary. Berkley 35 presents the following statistics. His results are from post- mortem subjects. It is not stated whether routine histologic examina- tion had been performed : Fallopian tubes, 30 ; fallopian tubes and body of uterus, 8; fallopian tubes, body of uterus, and ovaries, 5; fallopian tubes and ovaries, 4 ; ovaries, 4 ; cervix, 3 ; corpus uteri, 3 ; vagina, 2 ; fallopian tubes and vagina, 1 ; fallopian tubes, body of uterus, and cervix, 1 ; fallopian tubes, body of uterus, ovaries and vagina, 1. Thus, the tubes were affected in 80.6 per cent ; body of uterus, 29 per cent ; ovaries, 28.5 per cent; cervix, 6.4 per cent; vagina, 6.4 per cent; vulva, o per cent. Williams, 38 whose material was operative in origin and all of which was submitted to a routine histologic examination, is of especial value. He states that 8 per cent of all adnexitis cases are of tuberculous origin. In nearly all cases the tubes were involved ; the uterus, in 60 to 75 per cent ; and the ovaries, in 40 to 45 per cent of cases. Cummins, 58 in a series of INFECTION IN GENITAL TUBERCULOSIS 105 cases of pelvic inflammatory disease, found 10.5 per cent to be of tuber- culous origin. Hannes 59 places the proportion at 4.5 per cent. Mar- tin, 15 in the routine histologic examination of lesions from the gyne- cological clinic at Greifswald, found 24 tuberculous specimens among 1,600 specimens. Edebohls,' 00 in 157 abdominal sections, found 4 per cent were performed for tuberculosis. Horizontow 61 places the order of in- volvement of the genital organs as follows: Tubes, 87 per cent; uterus, 47 per cent ; ovaries, 1 5 per cent ; the cervix secondarily involved with the body of the uterus, 14 per cent; cervix alone, 2 per cent; vagina or ex- ternal genitalia, 6 per cent. Basing his opinion upon his own and other statistics gathered from postmortems performed upon patients dying of tuberculosis in which involvement was proved, he states that pulmonary lesions were present in 89 per cent; peritoneal lesions, 64 per cent; intes- tinal in 56 per cent; and lesions of the urinary tract in 42 per cent. In the laboratory of gynecological pathology at the University of Pennsyl- vania, where all specimens are subjected to a routine histologic examina- tion, it has been found that 7 per cent of all the inflammatory fallopian tubes are tuberculous. Among 6,557 gynecological specimens examined in our laboratory, there was no case of tuberculosis of the external genitalia, there was 1 case of tuberculosis of the vagina, 1 case of tuber- culosis of the cervix (219 specimens of carcinoma or other malignant neoplasms of the cervix, showing the relative frequency of tuberculosis and malignant tumors of the cervix, a condition for which tuberculosis is often clinically mistaken; this also emphasizes the importance of def- initely excluding malignancy before making a diagnosis of tuberculosis of the region) ; 13 cases of tuberculous endometritis (all associated with tuberculous salpingitis), 4 cases of oophoritis, 2 cases of tuberculosis in- fecting the wall of ovarian neoplasms, 7 cases of tuberculous peri- oophoritis, 1 case of tuberculosis of the breast (among 166 breast tu- mors, 91 of which were malignant and 75 benign). LITERATURE 1. Pozzr, S. A treatise on Gynecology. N. Y., 1897. 661. 2. Sachs, A. Centralbl. f. Gynak. Leipzig, 1893. 17, 249-255. 3. Amann. Monatschr. g. Geburtsch. u. Gynak. Berlin, 1902. 16, 586-630. Fourth Internat. Cong, of Gyn. Rome, 1902. 4. Veit, J. Fourth Internat. Cong, of Gyn. Rome, 1902. Monatschr. f. Geburtsch. u. Gynak. Berlin, 1902. 16, 525-555. 5. Gutierrez. Fourth Internat. Cong, of Gyn. Rome, 1902. 106 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS 6. von Rosthorn. In Kiistner's Lehrbuch der Gynakologie. 1908. 7. Wiener, G. Miinch. Med. Woch. 1909, 1602. 8. Blau, A. Uber die Entstehung und Verbreitung der Tuberculose im Weiblichen Genitaltrake. Berlin, Karger, 1909. 9. Holt, L. E. J. A. M. A., Chicago, 19 13. 61-99. 10. Dubendorfer, E. Bakteriologische Untersuchungen des Vulva und Vaginalsekretes. Inaug. Disc. Bonn, 1901. 11. Pankow, O. Ztschr. f. Geburtsch. u. Gynak. Stuttgart, 1912. 71-449. 12. Menge, K. Handbuch der Geschlechtskrankheiten. Wien, 19 10. 13. Cohnheim, J. Die Tuberculose vom Standpunkte der Infections- lehre. Leipzig, 1880. Edelmann. 14. Verneuil, A. Etudes Experimentales et Clinicales sur la Tuber- culose, Paris, 1888-90. Masson. Also Yaz. Hebd. de Med. Paris, 1883, 2nd s. 225-246. 15. Martin, A. Cong. Obst. and Gyn. Rome, 1892. Monatschr. f. Geb. u. Gyn. Berlin, 1902. 16, 555-576. 16. d'Aubeau. Cong. p. l'Etude de la Tuberculose. Paris, 1893. 17. Jani, C. Virchow's Archiv. Berlin, 1886. 103, 522-544. 18. Grawitz. Quoted in Pozzi's Treatise on Gynecology. N. Y., 1897. 661. 19. Murphy, J. B. Tuberculosis of the Female Genitalia and Peri- toneum. Chicago, 1903. Also J. A. M. A. Chicago, 1912. 58. 137- 20. Rohlff, E. Beitrag zur Frage von der Erblichkeit der Tuber- culose. Kiel, 1885. Lipsius & Tischer. 21. Westmayer, E. Beitrag zur Frage von der Vererbung der Tuber- culose. Inaug. Dis. Erlangen, 1893. 22. Dobrolonski. Cong. p. l'Etude de Tuberculose. Paris, 1889. 265. Rev. de la Tuberc. Paris, 1895, 3, 195. Fro. du Vrach, 1895. 19-20. 23. Walther, H. Ziegler's Beitrag. 1894. 16, 274-284. 24. Derville. These de Paris. 1887. 25. Fernet et Derville. France Med. Paris, 1886, 2, 1673-1685. Courrier Med. Paris, 1886. 36,488-491. 26. Sheills, E. Dublin M. Sc, 191 7. 43, 84-86. 27. Bull, P. Deutsche Med. Woch. Leipzig, 1912. 40, 1882-83. 28. Muscatello, G. Arch. f. Path. Anat. Berlin, 1895. 143, 327-359. 29. Clark, J. G., and Norris, C. C. J. A. M. A. Chicago, 1901. 37, 3°0- J- A. M. A. Chicago, 1904. 43, 281. INFECTION IN GENITAL TUBERCULOSIS 107 30. Waterhouse, H. J. Arch. f. Path. Anat. Berlin, 1890. 119, 342-361. 31. Pinner, O. Arch. f. Physiol. Leipzig, 1880. 241-255. 32. Kaufmann, E. Arch. f. Gyn. Berlin, 1886. 29, 407-408. 33. Kraus, E. Monatschr. f. Geb. u. Gyn. Berlin, 1902. 15, 159-166. 34. Frerich. Quoted by Murphy. 35. Mosler. Inaug. Diss. Breslau, 1883. 36. Spath. Quoted by Murphy. 37. Schramm. Arch. f. Gyn. Berlin, 1882. 19, 416-430. 38. Merletti, C. Arch, di Ostet. et Gynec. Napoli, 1901. 8, 612, 649, 714. 39. Berkley, C. Jour. Obst. & Gyn. Brt. Emp. London, 1903. 3, 31. 40. Frerichs. Quoted by Berkley. 41.' Horizontow. Zeitschr. f. Gyn. 191 1. 52, 1731. 42. Williams, J. W. Johns Hopkins Hospital Reports, 1894. 3, 114. 43. Villard. Quoted by Cornet. 44. von Franque. Pathologic und Therapie der Genital Tuberculose des Weibes. Wurzburg, 1913. 45. Schuchardt, K. Arch. f. Path. Anat. Berlin, 1882. 88, 28-49. 46. Saulmann. Gynak. Gesellschaft in Brussl. Abstracted in Cen- tralbl. f. gyn. Apr., 1892. 47. de Rouville, M. Bull. Soc. d'Obst. et de Gyn. de Paris. 1914.. 559-563. 48. Schiffmann, J. Arch. f. Gyn. Berlin, 1914. 103, 1. 49. Courts. Traite Pratique des Maladies de l'Uterus. 1872. 985. 50. Louis. Recherches sur la Phthysie. Paris, 1843. 51. Cornil, V. Cong, de l'Etude de la Tuberc. Paris, 1889-259. 52. Kiwisch. Klin. Vortrag. 1857. 1-462. 53. Nimias and Christoforis. Schmidt's Jahrb. Leipzig, 1850. 103, 326. 54. Foster, C. A. Amer. Jour. Obst. N. Y., 191 1. 63, 475-481. 55. Purefoy, R. D. Med. Press and Circ. London, 1908. 136, 399. 56. Simonds. Arch. f. Gyn. Berlin, 88-29. 57. Meyer-Rugg, H. Schweiz. Rundschau f. Med. 19 14. 1 4, 5 2 5- 58. Cummine, H. H. Amer. Jour. Obstr. N. Y., 1914. .69, 44-51- 59. Hannos, W. Ergebn. d. chir. u. Orthop. Berlin, 191 3. 6, 609. 60. Edebohls, G. M. Trans. Am. Gyn. Soc. 1891. 16, 514-535. 61. Horizontow. Zeitschr. f. Gyn. 191 1. 52, 1731. CHAPTER VI TUBERCULOSIS OF THE EXTERNAL GENITALIA Etiology — Possibility of hematogenic or lymphogenic infection — Causes — Frequency — Combined statistics of many investigators — Varieties — Forms, ulcerative and hypertrophic — Symptoms — Number of cases ; average age — Relative inf requency of direct infection in this locality — Parturition as causative agent — Trauma a predisposing factor — History of cases — Appearance of ulcerative variety — Hy- pertrophic variety — Tabulation of parts involved — Diagnosis — Prognosis—Method of treatment — Primary variety — Secondary — Doubtful cases — General treatment — Tuberculous non-ulcerative hypertrophy of vulva — Histologic examination — Tuberculosis of Bartholin's gland — Tuberculous ulcers of labia majora and minora — Histologic examination — Study of cases — Primary tuberculosis of vulva with elephantiasis of clitoris — Secondary hypertrophic tuberculosis of vulva— Reports of cases. LESIONS Of all forms of tuberculosis affecting the female genital tract, lesions of the external organs are the least frequent. This is probably largely due to the protective properties of the squamous epithelium with which the parts are covered. Much of the surface of the external genitalia is covered by skin, the outer layer of which possesses a moderately well defined development of horny squamous epithelium, such as is usually found on the surface of the integument. As the skin covering the external genitalia approaches the lining membrane of the vagina this outer horny layer gradually disappears. The tuberculous lesion may be primary or secondary ; the latter being by far the. most common. Of fifty-seven cases, the abstracts of which are appended, 79 per cent occurred in conjunction with well marked tuberculosis of other parts of the body. In 66 per cent of these cases the genital lesions were secondary to distant foci ; 33 per cent from the lungs; 5 per cent from the peritoneum, and the remainder to tuber- culosis in other parts of the body. In 18 per cent of cases lesions in the upper genital tract were present. Winckel * was probably the first to record an authentic case of tuber- culosis of the external genitalia. Cayla's 2 observation appeared a short time later. 108 TUBERCULOSIS OF THE EXTERNAL GENITALIA 109 Etiology. — Lesions of the external genitalia may result from a direct inoculation, or direct extension from the vagina or adjacent struc- tures, or from an hematogenic or lymphogenic infection. In many of the reported cases lesions of the external genitalia have apparently fol- lowed a direct implantation, the result of tubercle bacilli bearing dis- charges originating from the lesions in the intestinal, urinary, or upper genital tracts. In this connection, it is interesting to observe that the in- fection may be transmitted from the tubes to the external genitalia or vagina without the uterus becoming involved. In these cases, the possi- bility of an hematogenic or lymphogenic infection must be considered, although a direct implantation would appear the most likely. The relative frequency of tuberculosis of the uterus and the rarity of infection of the external genitalia is further proof, if such were required, that the simple deposition of tubercle bacilli upon the normal vagina" or external genitalia seldom results in the production of lesions; for in tuberculosis of the uterus tubercle bacilli are frequently discharged through the cervix. Tuberculosis of the external genitalia may also result from a direct extension by continuity from the vagina or adjacent structures. Trauma appears to play an important role in the production of sec- ondary lesions, doubtless by producing an area of lessened resistance. In the primary form it is less frequently a factor, although a loss of continuity, by opening up avenues for direct inoculation, should be considered. Preexisting inflammation is also a predisposing cause. In this manner the more or less constant soaking of the parts in toxin and tubercle bacilli laden discharges probably first produces a maceration of the skin, then a vulvitis, and finally an actual infection by the tubercle bacili. A number of cases have been recorded occurring in conjunction with syphilis; gonorrhea has also been present in some cases. In the young vulvovaginitis has preceded the tuberculosis in some instances. Bulkley 3 believes that, in the primary form, infection frequently occurs either by sputum or coitus. Of the secondary variety, infection may occur by the hematogenous or lymphatic route, or by contiguity of tissue or continuity of the surface. The actual route of infection is often difficult to determine in any given case. As has been stated, trauma or preexisting inflammation apparently acts as a predisposing cause, especially in the secondary form of the disease. Frequency. — As has been stated, this is the rarest form of genital tuberculosis. Of 6,657 gynecologic specimens in the laboratory of Gynecological Pathology of the University of Pennsylvania, but two examples of this variety of infection have been observed. Williams 4 states that at the time of the appearance of his monograph, in 1894, no GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS but three cases of tuberculosis of the external genitalia were found in which the correctness of the diagnosis had been verified by inoculations. The combined statistics of Geil, 5 Mosler, 6 Daurios, 7 Schiller, 8 and Martin 9 show that among 379 cases of genital tuberculosis there was no involvement of the vulva. The statistics of Berkley, 10 Simmons, 11 and Schlimpert 12 show that among 12,114 autopsies upon tuberculous women, genital tuberculosis was present in 215 subjects, but in none was the external genital involved. In 1903 Berkley 10 was able to find in the literature but four which were above suspicion. Varieties. — In general, vulvar lesions closely resemble tuberculosis of the skin in other parts of the body, except that they are often modified as a result of local conditions, such as moisture discharge, heat, friction and the presence of special glands and other anatomic conditions. Bender 13 and Patel 14 recognize two forms of tuberculosis of the external genitalia — the ulcerative and hypertrophic. Of these, the ulcera- tive is by far the most frequent. Of the fifty-four cases, the abstracts of which may be found in the following pages, forty-four were of this type and only ten of the hypertrophic variety. Bender 13 found the ul- cerative variety almost ten times as frequent as the hypertrophic. Occa- sionally the hypertrophic form undergoes ulceration, generally upon the prominence of the tumor, under which circumstances the cases are usually tabulated as ulcerative. The majority of the ulcerative lesions are associated with more or less swelling. To the ulcerative and hypertrophic varieties Combeleran 15 adds a third variety, which he des- ignates as lupus vulvae; this is characterized by thickening of the skin and mucous membrane, occasionally taking on a verrucous aspect, or by the development of ulcerations of limited depth and extent, but sometimes without ulcerative process. This is a doubtful variety, and probably mere- ly slightly atypical form of either the ulcerative or hypertrophic form. Formerly, much confusion existed regarding the hypertrophic variety and many cases of elephantiasis and other forms of enlargement were considered of tuberculous origin. The contrary also probably occurs, and this would seem especially likely in view of the difficulty often encountered in correctly diagnosing the* hypertrophic form, even after a careful histologic examination. Symptoms. — The symptoms resulting from lesions of the external genitalia are in themselves generally not very severe and in the secondary variety are usually subservient to those resulting from the primary condi- tion. Not infrequently there is a history of previous injury, this being particularly likely to be the case in the secondary variety. Thus, a fall from a horse, which resulted in injury to the vulva, occurred in the TUBERCULOSIS OF THE EXTERNAL GENITALIA in Bender and Nandrot 16 case, and the history of a fall resulting in trauma to the vulva was also present in the case recorded by Deschamps. 17 In many of the cases a tuberculosis of the upper genital tract can be dem- onstrated, and not infrequently lesions of the lungs or other portions of the body are present. Perhaps most frequently of all, tuberculosis of the external genitalia is secondary to intestinal lesions. In the author's case the disease was secondary to tuberculosis of the hip joint. Thus, it is seen that the condition may result from a hematogenous infection, from direct implantation through tubercle bearing discharges, or even from exogenous microorganisms, and from a direct extension from adjacent foci. In Schenk's 18 case the child had long associated with two 1 playmates known to be tuberculous. It is probable that direct inoculation from sexual intercourse may occur. The infection almost certainly came from a tuberculous husband in Rieck's 19 case, and probably in Montgomery's. 20 The experiments of Spano, 21 PopO'ff, 22 and Gorovitz 23 bear out this assertion. Cornet 24 suggests that tubercle bacilli bearing saliva may be used as a lubricant by a phthisical husband during coitus and thus result in infection. In a previous chapter the modes of direct inoculation have been more thoroughly considered; it is sufficient here to state that in the case of a woman, the wife of a tuberculous husband, there are other and more probable channels of infection than the genital tract, although the pos- sibility of this occurring must be considered, and should be guarded against. While the number of cases of tuberculosis of the external genitalia tract, recorded in literature, is as yet too small to draw definite con- clusions from regarding many of the symptoms, it would appear that no age is immune. Among 39 cases, the average age was 31.82 years. The extremes are 13 months (Demme 25 ) and 88 years (Dambrin and Clermont 26 ). Arranged in decades, these thirty-nine cases show the following: Years 1 — 10 11 — 20 21—30 .... 31—40 .... 41—50 51—60 .... 61 — 70 . . . . 71 and over Cases Per Cent 7 18.2 4 10.2 8 20.5 12 30.7 1 2-75 4 10.2 1 2-75 2 5-i ii2 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS A family history of tuberculosis is frequently obtainable, and a history of previous or present tuberculous lesions, such as pulmonary phthises, intestinal tuberculosis, adenitis, bone lesions, or pelvic inflam- matory disease, is often present. Both single and married women are attacked, the* disease apparently exhibiting no marked predisposition in this respect, thus bearing out what has already been said regarding the relative infrequency of direct inoculation in this locality. For, if direct implantation by means of coitus often resulted in genital lesions, tuber- culosis of not only the external genitalia, but also of the vagina and cervix, would be more frequent among married women than among spinsters. This, however, is not the case to any marked extent. Parturition, how- ever, appears to play some part as a causative agent. In the secondary variety it is certainly a not unimportant factor. It is accepted that the parturient woman is especially susceptible to acute miliary tuberculosis, a form of infection in which secondary lesions of any sort are not uncommon. Pregnancy and parturition also exert an unfavorable influ- ence on almost any form of tuberculosis, especially the pulmonary vari- eties, frequently leading to exacerbation. It is in acute infections that secondary genital lesions are most common. On the other hand, in the secondary variety of genital tuberculosis, trauma is a decided predis- posing factor and the trauma incident to labor or miscarriage must, therefore, be considered apart from the fact that in the parturient state women are peculiarly susceptible to any form of infection. Mont- gomery, 29 Jorfida, 27 and Davidson 28 have recorded the history of cases which occurred shortly after delivery. The results of animal experi- mentation, which have been previously quoted, show that trauma and inflammation are predisposing factors to direct infection as well as to the secondary or metastatic variety. The onset of tuberculous lesions of the external genitalia is generally slow, but progressive. Local discomfort, pain, discharge, and frequent and more or less marked dysuria are usually the most prominent symp- toms; but even these are quite variable. In some cases the pain is quite marked and in others it is absent. The pain may be sharp and cutting in character or a dull ache. Most frequently, as the disease advances, the pain becomes more pronounced, and if the lesion is of the ulcerative variety and so situated that the urine flows over it, pain at or following micturition is nearly always observed. The rubbing of the clothing against the ulcer, coitus, or other trauma is frequently complained of. Not infrequently there is intense pruritus and more or less itching is generally present, as in the cases of Deschamps, 17 Renaud, 29 and Martin. 30 In the hypertrophic variety the pain is less marked, the TUBERCULOSIS OF THE EXTERNAL GENITALIA 113 enlargement, however, from its very size, may produce discomfort. In the ulcerative variety discharge is nearly always present. This varies, according to the stage and character of the lesions, from a thick, purulent secretion to a thin, more or less irritating leukorrhea. In acute cases or following trauma, it may be blood streaked. As a result of the dis- charge, a more or less general vulvitis usually occurs and sometimes pro- duces distressing symptoms, a certain amount of pruritus being almost always present. Many cases being secondary to tuberculosis of the upper genital tract, it is difficult to determine how much of the discharge comes from above and how much from the vulvar lesion. As a rule, the ulcers do not bleed very readily to the touch and are not markedly tender. Tubercle bacilli can occasionally be demonstrated in the discharge, especially if the lesion be an acute one. In the curettings from the surface of the ulcers they can frequently be found. Occasionally, as a result of exten- sion of the ulcer, fistulas form. In the ulcerative variety, and sometimes in the hypertrophic, inguinal adenitis occurs. Murphy 31 states that inguinal adenitis occurs late. This, however, depends largely upon the character and location of the lesion and upon the amount of suppuration present. Appearance of the Ulcerative Variety. — This is generally pre- ceded and accompanied by more or less enlargement. In the case re- ported by Bender and Nandrot 16 the condition began as a fluctuant swell- ing, which finally broke down, leaving a discharging cavity which was extremely chronic in type and which exhibited little or no tendency towards spontaneous resolution. The areas surrounding the preliminary swelling are usually discolored and edematous. The adjacent tissue is indurated. After a varying length of time, sometimes many months, the swelling softens in one or more areas and breaks down. In some cases the lesion begins as one or more small firm nodules, which subse- quently soften and break down. Thus a number of ulcers may be formed. These may finally coalesce, forming a single large granulating area, usually covered with a layer of necrotic tissue. The ulcer may originate as a superficial loss of tissue, and then gradually enlarges. The ulcer may occur on any part of the external genitalia, but is perhaps most frequently on the labia majora or minora. One or both sides may be involved, and contact ulcers on the opposite side are occasionally observed. The ulcers may extend backwards into the vagina or out- wards over the skin, perineum, or adjacent structures. When the vagina is involved, fistulas connecting with the various adjoining hollow viscera are not infrequent and the symptoms from these are likely to be ii4 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS a marked feature. Occasionally the anus is involved. Ulcers of the external genitalia vary in appearance ; the margins are often elevated and swollen or the edges undermined. The base is usually moderately firm and may be covered by minute, grayish or yellowish elevations (tubercles). In some cases the floor of the ulcer is covered by a dirty, yellowish or brownish crust. An appearance of chronicity is common to the majority of these lesions. The color may be grayish, yellowish, reddish or brownish, and the surrounding skin is often chronically in- flamed, discolored, hyperemic, and may contain enlarged veins. The ulcers vary markedly in size. Thus, in Legane's 32 case the entire vulvar region, including the hymen, was destroyed by a yellowish ulcer. A somewhat similar case is recorded by Brault. 33 The ulcers are often serpiginous in character, healing behind as the advance is made. As a result, cicatrices may be present. In some in- stances, where the urethra has been attacked, the disease has apparently followed the mucosa of that canal, forming finally a funnel shaped ulcer with the small end directed towards the bladder. Reed 34 states that frequently the meatus appears to be torn laterally, somewhat after the manner of the Emmet denudation for trachelorrhaphy, while on the other hand almost microscopic lesions have been described. Hypertrophic Variety. — This is an extremely rare form, and too few cases are recorded to base on them a definite description. In the cases reported by Petit and Bender, and Poverlein the lesions were characterized by moderately large tumor-like masses, which in Pover- lein's case were at first mistaken for a sarcoma of the labia. Specimens in the cases of Petit and Bender, Forgue and Massabuau resembled an elephantiasis. The discharge is not profuse and is never purulent or sanguineous. Tubercle bacilli have never been demonstrated in it. Bulkley 3 gives the below summary regarding the parts involved. In this summary the hypertrophic and ulcerative varieties are included, from which it will be seen that the labia are most frequently involved. Parts Involved Cases Vulva 10 Labia majora 29 Labia minora 30 Clitoris 8 Entire introitus 7 Posterior commissure 6 Anterior commissure 3 Mons veneris 2 TUBERCULOSIS OF THE EXTERNAL GENITALIA 115 Parts Involved Cases Edge of urethra 5 Bartholin's gland 2 Prepuce 1 Diagnosis. — A positive diagnosis without the aid of the microscope in either the ulcerative or hypertrophic varieties is impossible. Malignant tumors and syphilis are the two conditions most likely to cause confusion ; although in children gonorrheal vulvovaginitis, anovulvar diphtheria, and noma vulvae must be differentiated. Chancroids can usually be readily differentiated, as can kraurosis vulvae. The hypertrophic variety usually more or less closely resembles elephantiasis. Bender 35 recommends biopsy in all cases in which there is ulceration, but even this is untrustworthy in the hypertrophic variety. The Wassermann reaction should be applied to all cases, and in chil- dren the von Pirquet reaction will be of value. It should be remembered that malignant neoplasms, especially in the aged, are far more frequent than is tuberculosis, and a thorough histologic examination to exclude this possibility should be made without loss of valuable time in all cases. After excision of the suspected area the diagnosis can usually be readily arrived at. Histologic, bacteriologic, and animal inoculation will clear up all doubtful cases. For the histologic examination, it is advisable to examine slides from a number of sections, as, if only one block is taken, characteristic lesions may be absent. In the hypertrophic form tubercles are sometimes rare and only the bacilli, and these in small numbers, are found. The fact should not be lost sight of that syphilis or malignant tumors, or even both, may accompany tuberculosis. The presence of tuberculosis in other parts of the body, grayish tubercle like elevations at the base of the ulcer, the presence of acid fast bodies morphologically similar to the tubercle bacilli in the discharge, all point to tuberculosis. In staining for tubercle bacilli the smegma bacilli must, however, be excluded. The absence of a syphilitic history and a negative Wassermann reaction will practically exclude syphilis; while the longer duration, more chronic appearance of the lesions, and the lessened tendency to bleeding, and perhaps the age of the patient, are evidence against the condition being a malignant tumor. Prognosis. — This, as in all tuberculous lesions of the female genital tract, depends largely upon whether the lesion be a primary or secondary one. In the latter event, the primary focus will often be the more severe and the prognosis will naturally depend upon its location and character. In some cases the genital lesions are extremely chronic : thus, in Pover- n6 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS lein's 36 case the disease had been present for seventeen years, in Viatte's 37 case, seven years, and in Montgomery's 20 case, five years ; whereas in other reported cases rapid dissemination of the infection and death have occurred. Demme 25 emphasizes the rapid course that the disease may follow, especially in children. A lethal termination is, however, rarely due to genital lesions alone. Even after apparent entire excision, recurrences may occur. On the other hand, spontaneous healing occasionally takes place, but this is unusual. More frequently the course of the disease is chronic, but pro- gressive. Unfortunately the great majority of reports are either too recent or mention is not made of the ultimate outcome of the cases. Statistics are thus apt to be misleading. Bulkley 3 in his excellent review of tuberculosis of the external genitalia, presents the following table, but warns us that the heading "healed" cannot be interpreted as an end result : Method of Treatment Excision Curettage and cauterization Excision with cautery Cauterization Nitric acid Iodoform Tuberculin General hygiene 1_ C/J produce hemorrhage than are the ulcerative and papillary. Pain. — Since the cervix contains few sensory nerves, pain is rarely a marked feature. As the disease advances and absorption takes place a cellulitis of the base of the broad ligament, with its resulting symp- toms, is by no means uncommon. As a result of lesions in the upper genital tract and pelvic peritoneum, pain in the lower abdomen is not infrequently encountered. As a result of the cervical lesion, occlusion of the cervical canal may occur, and result in a pyometra, as in one 154 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS of the cases of Pollosson and Violet. 40 If this occurs, enlargement of the uterus and more or less pain is prone to occur. Spread to the adnexa and pelvic peritoneum, produces the symptoms characteristic of these lesions. The macroscopic appearances of the different varieties of cervical tuberculosis naturally vary widely. Varieties. — Pozzi describes three varieties, the ulcerative, vegetative, and miliary, to which Schutt 32 has added a catarrhal form, examples of which are the cases of Meyer, 33 Giglio, 34 Sippel, 35 and Schutt. This variety is rejected by Chaton. 4 Cotte 36 has described an inflammatory variety which closely simulates an endocervicitis, glandular, periglandu- lar inflammation and changes in the surface epithelium being the chief features. The most satisfactory classification is that which divides the cervical lesions into four groups, the ulcerative, the papillary, the miliary, and the interstitial. Of these, the ulcerative and papillary are the most fre- quent, the miliary and interstitial being comparatively rare varieties. An analysis of 106 cases shows 52 to have been ulcerative, 41 papillary, 7 miliary, and 6 interstitial. These statistics are, however, to some ex- tent misleading, as combinations of the various forms, especially the ulcerative and papillary, have been present frequently; while it is prob- able that, if these cases could have been examined in their incipiency, the interstitial and even perhaps the miliary would have been found more often. The interstitial variety, like the similar form of cervical cancer, does not produce marked symptoms until the disease has broken through to the surface of the portio or the canal, and when examined at this latter time, is doubtless frequently classed as the ulcerative variety. Chaton, in his analysis of cases, found 37 ulcerative, 22 papillary, and 7 miliary. Cova 37 thinks the papillary variety frequently presents ulcerations. Patel 38 states that 50 per cent of the cases are of the ulcerative variety. In the secondary cases, the cervical lesions do not necessarily follow the type of the original foci. Thus a general miliary tuberculosis may result in an ulcerative, papillary, or other form of lesion. An analysis of 14 primary cases showed 8 to be of the ulcerative variety and 6 of the papillary. It is doubtful if all of these cases are primary. Not infrequently specimens are reported as primary upon insufficient evidence. Beyea 14 analyzed 59 cases of cervical tuberculosis with a view to ascertaining the portion of the cervix attacked. In these the portio was involved alone in 11, the supravaginal cervix alone in 6, and both in 42. The primary lesion in the cervix is usually in the canal, regardless of the variety. TUBERCULOSIS OF THE CERVIX 155 Ulcerative Variety. — These lesions vary considerably in size and appearance. In some specimens they are large and the place of the entire vaginal cervix is occupied by the ulcer, as in the case of Bonilly. 41 Not infrequently the adjacent vagina is involved. In other instances the ulcers are small and may resemble a chancroid, as in the case of von Franque. 42 Usually the external os is the starting point, the dis- ease spreading from this location toward the vagina eccentrically. The lesions may be situated upon the portio or in the cervical canal. In the cases of Nanard, 43 and Broucha 8 the ulcers were almost entirely within the cervical canal. Or the lesion may commence on the portio and spread upward, involving the endometrial cavity, as in the case of Lepitit, 18 or the converse may be the case. In some instances the lesions are shallow and surrounded by clean cut, slightly raised margins ; more frequently, and especially in advanced cases, the ulcers are moderately deep and present roughened, swollen, and often undermined edges. The base and edges may be fairly smooth and contain numerous raised concentric elevations, often yellowish and grayish and partially translucent; or the sides and base may be cov- ered with darkened, necrotic material and a general worm-eaten appear- ance be present, or the surface of the ulcer may be granular. Occa- sionally there is attached to the ulcer, yellowish, cheesy material. The ulcers may be multiple, but are more frequently single. The lesions usually bleed moderately, easily, although in a few instances this sign has been absent. As a rule the bleeding is less marked than in carci- noma and the lesions appear more chronic. On palpation the base of the ulcer generally presents a soft velvety feel. The friable character can, however, frequently be detected by the touch. The cervix is gen- erally enlarged. Papillary. — In this form there is an outgrowth from the cervix of more or less cauliflower-like masses; when first examined these are usually dark, reddish or brownish in color and covered by discharge. Not infrequently nodular elevations, sometimes of moderate size, are present. If the latter be removed, or after excision, these are found to be papilloma-like masses, red, yellow, gray, pink, or white, often some- what translucent, and frequently contain areas of necrosis. The papil- lary variety generally affects the portio, but may originate from the cervical canal. As a rule this type is moderately friable, and as a re- sult bleeds easily. In some cases, however, especially when small, and before much breaking down has occurred, the masses are moderately firm and exhibit but little tendency to bleed on touch. The papillomata may spring from a broad base or more rarely be definitely peduncu- 156 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS lated. They may be single or multiple. In some instances there is a papillary endocervicitis, and more or less spreading outwards through the external os of small polypoid masses, as in the cases of Pollosson, 40 Beyea, 14 and Lewers ; 44 or the outgrowths may originate from the portio, as in the cases of Cornil 45 and Giglio. 34 Emanuel, 40 Ferrari, 13 also Ressegna 47 and Vitrac 48 have observed a form that exists as a distinct tumor, more or less pedunculated and friable, the origin of which is variable, but usually from the portio. This has been described by some of the French writers as the "vege- tante neoplastique" variety. In many specimens the fungus-like masses closely resemble carcinoma. The cervix is usually enlarged and the surface of the portio not covered by outgrowths is reddened. Some- times this variety affects the external os, as in the case of Hofbauer, 49 and sometimes the canal (Beyea 14 and Lewers 44 ). Pollosson and Violet 40 especially emphasize the fact that the disease may occur as a localized intracervical polypoid condition. Miliary. — In this variety the cervix is enlarged, reddened, turbid, and small, somewhat pale yellowish, or grayish, partially translucent ele- vations may be seen beneath the surface epithelium. These are usually solid, but may contain turbid fluid or cheesy material. The mucosa at the external os may be normal or may be thickened, swollen and in- flamed. In some instances the tubercles are limited to the mucosa of the canal, but more frequently the portio is also involved. The surface of the portio between the tubercles sometimes presents a granular ap- pearance, and a general tendency towards fibrosis is often observed. Instructive reports on this variety of lesion may be found in the con- tributions of Rigal, 5 Cornil, 50 Zweigbaum, 15 Denville, 51 Vitrac, 48 and Bouffe. 52 Interstitial. — This variety begins in the substance of the cervix, which becomes enlarged, usually asymmetrically. As the disease ad- vances, a localized necrosis occurs, which eventually breaks down into the canal, or more often on to the portio, leaving a ragged, undermined opening leading into the primary cervical focus. In the latter stages, a deep, undermined ulcer is present, which is lined by necrotic tissue, blackish or, in some instances, yellowish in color. Tubercles may be present in the friable floor or walls of the cavity and in the adjacent surface. Combinations of these varieties are frequent, especially of the ulcerative and papillary. Diagnosis. — As has been stated, tuberculosis of the cervix produces no symptoms that are by any means pathognomonic. In no case can a TUBERCULOSIS OF THE CERVIX 157 positive diagnosis be arrived at without the aid of the microscope. The majority of cases have been diagnosed clinically as carcinoma and the true character of the lesion ascertained only by a histologic examination. Tuberculosis of the cervix Carcinoma of the cervix No age is immune. Most fre- quent in active sexual life. Rare in the extremes of life. Most frequent between 35 and 50 years. Such history is infrequent. There is a history of tuberculosis in other parts of the body, in the majority of cases. Nullipara by no means immune. Extremely rare in women who have never been pregnant. Local symptoms may have been present for a prolonged period. Tubercles may often be observed in the lesion or on the adjacent structures. Course of the disease more rapid. Tubercles absent. The margin of the ulcer is usu- ally undermined and fairly soft. Floor of the ulcer is moderately soft and may contain numerous macroscopic grayish or yellowish semitranslucent tubercles. Usually elevated and indurated. Hard and nodular. Tubercles are absent. Usually bleeds readily but not al- ways. Bleeds more readily. The discharge may contain cheesy masses and tubercle bacilli, as shown by staining, inoculation, or culture. Necrotic tissue, which presents the histologic characteristics of cancer. Tubercle bacilli absent. As stated above, tuberculous lesions of the cervix are usually softer and less indurated than carcinoma. The appearance and friability may 158 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS be suggestive of carcinoma, but the less indurated and, indeed, often velvety sensation of the tuberculosis is usually in marked contrast to the cancer. Even if distinct nodules are present, these are usually softer than cancer. To the experienced surgeon this is a valuable sign and should in itself, at least, suggest the possibility of tuberculosis. In ex- amining the literature pertaining to this subject, the reader cannot fail to be impressed with the frequency with which this differential diag- nostic point is mentioned. Despite the above differences, many cases will be encountered, in which a clinical differentiation is impossible, and in all in which doubt exists, biopsy should be resorted to, the excision being performed with a cautery knife heated to a dull red. As carcinoma is so much more frequent than tuberculosis, this fact should be borne in mind, and no time lost in arriving at a diagnosis. In not a few cases, tuberculosis of the cervix has been clinically mistaken for sarcoma (Cornil, 50 Frankel, 16 Kaufmann, 10 Giglio, 34 Vitrac, 48 and Emanuel 46 ) and its true character only recognized after histologic examination. In addition to the differentiation from malignant neoplasms, the ul- cerative and papillary varieties must be distinguished from lacerations, hypertrophies or eversion the result of childbirth, other inflammations such as gonorrhea, chancre, and the papular and ulcerative syphilides, gumma, chancroid, condylomata acuminata, benign polyp, leukoplakia, nbromyomata, and sarcoma. With the exception of the last named, no great difficulty exists in excluding these conditions. The miliary variety must be distinguished from other inflammatory lesions, especially when the latter are associated with laceration and eversion, or nabothian cysts, hypertrophies, and subinvolution. The interstitial variety, if observed in its early stages, may be con- fused with interstitial neoplasms, retention cysts, laceration or hyper- trophies. The differential diagnosis between tuberculosis of the cervix and the above named conditions, with the exception of the malignant neoplasms, usually presents no unusual difficulties. Prognosis. — If tuberculosis of the cervix be primary and localized, the prognosis is favorable, provided the proper treatment be adopted. In determining that a given case is primary, extreme caution should be observed. As the great majority of cases are secondary, the prog- nosis depends to a large extent upon the character of the primary lesion. As a general rule, in secondary cases the prognosis is grave; however, cures have been reported in a number of instances. . In the great majority of the reported cases the ultimate outcome is not stated. Beyea's 14 statistics show that, out of 10 cases subjected TUBERCULOSIS OF THE CERVIX i 59 to panhysterectomy, 3 died soon after the operation — 1 from shock, 1 from tuberculous peritonitis, and 1 from an aggravation of the lung condition : of the 7 remaining, 6 were well some years after the opera- tion; and in 1 four months had elapsed. Statistics of this type are, however, misleading. The chief condition in the large proportion of cases being the primary focus, its extent, character, amenability to treat- ment, the apparent virulence of the infection, the patient's age, social status, etc., are all points which should be considered in rendering the prognosis, as well as the condition of the upper genital tract. Treatment. — This, also, is dependent upon whether the case is primary or secondary. In the former event a panhysterectomy or, if the lesion is small and entirely limited to the. vaginal cervix, a high trachelectomy, is indicated. If the latter operation is selected, a curet- tage should be performed and the curettings from the body of the uterus examined histologically for the purpose of excluding a tuberculous endo- metritis. If curettage is performed, especial precaution should be in- stituted to prevent carrying tubercle bacilli from the cervix to the endometrial cavity. As an additional safeguard, it is advisable, as a final step in the curettage, to apply tincture of iodin to the denuded uterine cavity. In patients past the child bearing period, or who al- ready have a number of children, an abdominal panhysterectomy is pref- erable in most cases, for by this means a thorough examination can be made and the condition of the adnexa ascertained beyond the question of a doubt. The exposure of the peritoneum to the air is also of ad- vantage in cases in which either general or local peritonitis is present. Chaton 4 and Petit-Dutaillis 53 favor the vaginal route in these cases. The former states that in 15 vaginal hysterectomies there were 2 deaths; and among 8 abdominal hysterectomies 2 deaths occurred and 2 local recurrences. The question of which route shall be selected is largely a matter of choice with the individual surgeon. The author prefers the abdominal route. The fact that the corporeal endometrium and the adnexa are involved in the tuberculous process in a large propor- tion of cases should also be borne in mind in selecting the operation. Patel 54 especially recommends excision in the hypertrophic varieties of the disease. In cases in which there is involvement of the upper genital tract or peritoneum, hysterectomy is generally the most satis- factory operation. If pulmonary phthisis or other distinct foci are present, their extent and character should decide the treatment to a large extent. Palliative measures are usually preferable in advanced cases. The amount of discomfort produced by the genital lesion must, however, be considered. i6o GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS In this respect, each case is more or less a law unto itself. The general tendency is very properly to treat these cases surgically when the primary focus is of such a character as to permit operation. This subject of treatment of secondary lesions will be more thoroughly considered under a separate heading in a subsequent chapter. Murphy 39 states that of ii cases treated palliatively, I recovered, 5 were temporarily improved, and in 5 the disease progressed. Petit-Dutaillis 53 has reported the his- tory of one case which recurred six years after a curettage and cauter- ization with the actual cautery. In 1903, Murphy 39 advised against hysterectomy in primary cases, stating that the operation gave a 30 per cent mortality. We feel that, with our present methods of operating in uncomplicated cases, this is far in excess of the actual figures, and that three or four per cent would be the maximum under favorable cir- cumstances. Curettage of the cervix, followed by cauterization, either with a zinc chlorid or preferably with the actual cautery, may be employed as a palliative measure in cases with advanced primary lesions. Radium, or the Rontgen rays have apparently produced good results in some cases. Radium or the X-Rays are positively contraindicated, if a sal- pingitis is present. Under such circumstances either of these methods of treatment is prone to light up the infection and produce serious con- sequences. In these, as in all other cases of genital tuberculosis, par- ticularly in the secondary variety, the after treatment is of the utmost importance. This will be considered in detail in a subsequent chapter. CASE HISTORIES Haultin. 55 Single, 35 years of age. Menstruation was normal. For several months there had been increasing leukorrhea. Examina- tion of the cervix showed it to possess a rough, irregular outline, not friable, and did not bleed easily. It was purplish in color, and more or less covered with papillomatous outgrowths and bathed in a thick, yellowish discharge. The body of the uterus was normal. A high trachelectomy was performed. The specimens showed the usual histo- logic picture of tuberculosis, and tubercle bacilli were demonstrated in the tissue by staining. The case is of especial interest, as it was ap- parently primary. No history or physical evidence of tuberculosis in any other parts of the body could be demonstrated. Furthermore, the patient was well sixteen years after the operation. The fact that she was single, and that the hymen was intact would tend to exclude the TUBERCULOSIS OF THE CERVIX 161 ordinary routes of direct infection. The most pronounced histologic changes were, however, on the portio vaginalis. Montanelli 17 furnishes brief records of eleven cases of tuberculosis of the cervix uteri, from the Royal obstetricogynecological clinic at Florence, reporting two of these cases in detail. Case I. Patient aged 38 years and nullipara, in whom the onset of menstruation had been delayed. The menses were abundant and fre- quent. She had leukorrhea and papillary tuberculosis of the cervix. Case 2. Woman, aged 41 years and nullipara. The menses were always irregular, had leukorrhea, and bleeding after coitus. Interstitial tuberculosis of the cervix with marked glandular hyperplasia was pres- ent. Case 3. Woman aged 40 years, and had one child. The patient had leukorrhea for eight months, and sometimes bleeding. Papillary tuberculosis of the cervix was diagnosed. Case 4. Patient aged 39 years had tuberculosis of the peritoneum, adnexa, and cervix. Case 5. Woman aged 18 years and nullipara. She had abdominal pains, and amenorrhea had existed for six months. Papillary tubercu- losis of the cervix was diagnosed, with partial ulceration. Case 6. Woman, aged 44 years, and decipara. The patient had leukorrhea, and abdominal pains. Interstitial tuberculosis of the cervix with involvement of the body of the uterus, adnexa, and peritoneum, was present. Case 7. Patient aged 26 years, was a nullipara, and had irregular menses and caseous masses in the uterine cavity. Papillary tuberculosis of the cervix, and tuberculosis of the adnexa and peritoneum were present. Case 8. A woman, aged 25 years, who had been married three years, but had never been pregnant, sought advice for amenorrhea, which had persisted for some months. Examination showed the body of the uterus normal in size and position. On the portio was an erosion which bled easily and a polyp protruded from the external os. Curettage and histologic examination verified the diagnosis and panhysterectomy and bilateral salpingo-oophorectomy was performed. The cervix, the body of the uterus, and the right tube were found to be tuberculous. Case 9. This patient was a sterile married woman of 28 years. Menstruation was irregular for two years. The last period, five months ago, was followed by profuse leukorrhea. The uterus was normal in size. The cervix was the seat of a papillary growth, which bled easily. No tuberculous lesion could be detected in any part of the body, nor had 162 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS there been fever for fourteen days prior. The inguinal glands were slightly enlarged. Panhysterectomy was performed and tuberculosis histologically demonstrated in the cervix, endometrium and tubes. Case 10. Woman, aged 51 years, and nullipara. The menopause occurred at 46 years and had had a bloody discharge for the last seven months. Papillary tuberculosis of the cervix was diagnosed. Case 11. Woman, aged 39 years and nullipara. The patient had abdominal pains and leukorrhea. The condition was diagnosed as papillary tuberculosis of the cervix. Montanelli 17 believes the cervical lesions are often secondary to sal- pingitis and therefore recommends panhysterectomy and bilateral sal- pingo-oophorectomy in all cases. Alterthum. 30 Woman, aged 36 years and married, had abdominal pains and polypoid elevations on the posterior, cervical lip. Extensive pelvic inflammatory disease was present. On microscopic examination, tuberculosis of the polyp was diagnosed. The author makes no claim for the primary occurrence on the cervix in this case. Smith. 56 A nullipara, aged 25 years, believed to have had a two and a half months' abortion a few months ago. This was shortly fol- lowed by pain in the right ovarian region, irregular hemorrhages, of- fensive discharge, and fever. Thus the history simulated one of septic abortion. The cervix was the seat of a soft, friable mass, which bled easily. The fundus and adnexa were normal. A provisional diagnosis of carcinoma was made, biopsy performed, and tuberculosis reported. The lungs were involved, and for this reason curettage and the applica- tion of zinc chlorid were decided upon. Yineberg. 57 Case 1. Nonipara, aged 37. Regular and painful menstruation. Family history negative. Two and a half years ago suffered from a pleurisy with effusion, otherwise well. For last three weeks, pain in lower abdomen and fever, amenorrhea for two months, and had lost flesh and strength. Cervix hypertrophied and presented three ulcers. These were irregular in outline, moderately deep, and covered with a dirty grayish exudate. The remainder of the portio was reddened. There was no marked induration, no friability of the tissues, and no tendency to bleed when slightly traumatized. Tuber- culosis of the body of the uterus and adnexa was present. Hysterec- tomy was followed by death. No autopsy. Histologic verification. Case 2. Single woman, aged 25 years. Suffered from amenorrhea. Rather profuse leukorrhea and occasional attacks of pain in the right groin for two years. Her general health was good, and there was a good family history. Hymen was intact. Uterus small and anteflexed. TUBERCULOSIS OF THE CERVIX 163 Cervix was soft and bled slightly to touch. Inspection showed the portio to be covered with vascular granulations. Case resembled a marked endocervicitis. Attached to the right wall of the cervical canal was a small cyst, the size of a cherry. This contained sebaceous ma- terial. This was not examined histologically. Trachelectomy and dilatation and curettage. Adnexa normal. Histologic examination of the amputated cervix verified the diagnosis. Vineberg considers this a primary case; at least no other focus of tuberculosis is referred to. Martin. 27 Patient, aged 25 years, was married and nullipara. She had amenorrhea, leukorrhea, and pains in the lower abdomen and back. On histological examination, pieces of tissue from the cervix showed that the process was tuberculous and not cancerous. Lorrain and Chaton. 58 Patient, aged 37 years, was married but had no children. She had prolapse of the uterus, hypertrophic elongation of the cervix, and bilateral inguinal adenitis. The cervical tissue was incised and tuberculous products removed by the curet. The histologic examination showed typical tuberculosis, but no tubercle bacilli could be demonstrated by staining. Injection of some of the material into a guinea pig was followed by tuberculosis in the animal. Horrocks. 59 Woman, aged 34 years, who had pulmonary phthisis. The cervix was dotted over with grayish, opaque vesicles, with a red, pulpy substance between, which bled easily when touched, and resem- bled a malignant neoplasm. The ulcer felt rather soft. Hysterectomy showed the genital condition was limited to the cervix. A uterus sep- tus was present. The patient made an uneventful recovery, and was discharged from the hospital cured. The diagnosis of tuberculosis was made on microscopic examination. Garkisch.' 60 Woman, aged 28 years, married but had had no chil- dren or miscarriages. She never menstruated. At external os was a polypoid projection. Biopsy was performed. Microscopic examination of tissue removed for diagnosis showed typical tubercles and giant cells. Hysterectomy was performed, and the corpus uteri and tubes were found to be involved. In spite of the fact that the woman presented no other evidence of tuberculosis, even on a careful examination, and the fact that her husband was healthy, the author hesitates to regard the case as primary. Normal convalescence. Zweigbaum. 15 Tuberculosis developed, apparently primarily, in the cervix, then vagina, and then left labium minus, on which there was a large ulcer. These were cauterized and apparently cured. She died later, however, from a general tuberculosis, so that it would seem at least likely that this was not a primary case of cervical tuberculosis, but 164 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS one in which the primary focus was for a time latent. The cervical lesion was of the miliary variety. Emanuel. 46 A woman, aged 50 years, presented herself, suffering from profuse purulent, often blood streaked leukorrhea. The cervix was found to be enlarged to the size of an apple. The enlargement was due chiefly to a vegetative outgrowth. Deep necrotic ulcerations were also present. Some of the ulcers involved the vagina. The body of the uterus was also enlarged, and the endometrial cavity filled with caseous material. The adnexa were normal. A panhysterectomy was per- formed. The patient died, and autopsy showed that miliary tubercu- losis was present in the liver, spleen, and peritoneum. The lungs were normal. Examination of the cervix showed the usual histologic picture of this condition. Frankel. 16 Woman who died of Pott's disease. The mucosa of the cervix was covered with fungus-like masses. The vagina and body of the uterus were normal, but advanced tuberculous lesions of the tubes were present. The diagnosis was confirmed by histologic examination. Broye. 24 This patient was married, and 24 years of age. Three months after a miscarriage she developed a tuberculous peritonitis, sal- pingitis, and oophoritis. The cervix was the seat of a papilloma-like outgrowth, which, upon histologic examination, presented the usual ap- pearance of tuberculosis in this locality. A tuberculous endometritis was also present, an interesting point being that the histologic changes were most marked at the placental site. Rigal. 25 Patient had a miscarriage at the sixth month and died shortly afterwards. She had general miliary tuberculosis of the lungs, peritoneum, and meninges. The cervix uteri was the seat of an ex- tensive ulcerative lesion, which had involved the adjacent vagina by direct extension. The edges of the ulcer were raised, edematous, and partially undermined. The diagnosis was based upon histologic evi- dence. Death and autopsy showed miliary tuberculosis of the lungs, peritoneal cavity, meninges. Klobb. 9 The specimen was discovered accidentally at autopsy in a woman who had died of an intercurrent disease. The lesion was almost the size of a cherry, and had its origin low down in the cervical canal. The character of the pathologic process was determined only upon his- tologic examination. A careful examination of the body at the post- mortem failed to show any other foci of tuberculosis present. Kaufmann. 10 Patient, aged 72 years. The external os was small, but the supravaginal portion of the cervix was notably thickened and enlarged. A section from this portion showed it to contain a cavity TUBERCULOSIS OF THE CERVIX 165 lined with semitranslucent grayish granulations. The walls were dis- tinctly firm to the touch. The mucosa of the cervical canal presented no marked alterations. On histologic examination tubercles, giant cells, and other evidences of this type of infection were observed. Tubercle bacilli were demonstrated by staining. Michales. 11 Patient, aged 33 years, was married and a nullipara. Her mother had died of tuberculosis. The patient's lungs were normal. A moderate amount of purulent discharge was the only marked symptom referable to the cervix. A moderate sized necrotic ulcer of the cervix was present, which was cured by excision. A tuberculous ulcer and granular hypertrophy of the adjacent mucosa was diagnosed histo- logically. Brouha. 8 A quintipara, aged 41 years, with a family history of tuberculosis. The last child was born fourteen years ago, at which time she suffered from a pelvic peritonitis and pleurisy. A curettement has recently been performed. Now complains of pain in the back and left iliac region. Constipation and leukorrhea. The anterior cer- vical lip was enlarged and reddened. Opening into the cervical canal, and evidently interstitial in origin, is an ulcer, the cavity of which is red and has a worm-eaten appearance. On biopsy histologic evidence of tuberculosis was discovered. The lungs were normal and no extra- genital foci of tuberculosis were found. Chronic pelvic inflammatory disease was present. Panhysterectomy and bilateral salpingo-oophorec- tomy were performed. Although evidence of inflammation in the upper genital tract was present, no tuberculosis was demonstrated. The au- thor believes the condition to have been contracted by direct infection through coitus. Brooks 12 reported the following year that the pa- tient was in good health. Ferrari. 13 Case 1. Patient, aged 30 years. She was a nullipara and had a tuberculous ulceration of both cervical lips. Polypoid ex- crescences were also present. Microscopic examination showed tuber- culosis. Vaginal hysterectomy was performed. The convalescence was normal and the patient was discharged from the hospital cured. No other foci of tuberculosis were demonstrated elsewhere in the body. Case 2. The patient was a multipara, who presented papillary and nodular excrescences on the cervix. She had had irregular menstrua- tion and discharge. A trachelectomy was performed and the diag- nosis verified by histologic examination. Giglio. 34 Patient was 28 years of age. The chief symptom was a profuse purulent and at times blood streaked leukorrhea. Examination showed the cervix to be the seat of a papillary outgrowth. A number 166 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS of ragged, irregular ulcers were also present. A diagnosis of sarcoma of the cervix was made. Vaginal hysterectomy was performed. Death occurred three months later from "poussee de granulie." Histologic ex- amination revealed the true character of the cervical lesion. Tubercle bacilli were demonstrated by staining in the tissue. Glockner. 22 Patient, 20 years of age. She had a papillary tuber- culosis of the cervix, resembling cancer. The husband had tuberculosis of the right testicle and epididymis. The author believes the infection resulted from coitus. A vaginal hysterectomy was performed, and re- sulted in a cure. The diagnosis was made by histologic examination. Zweifel. 61 Patient, 28 years of age, had a family history of tuber- culosis. A ragged, irregular, necrotic ulcer of the cervix was present and diagnosed cancer. The ulcer extended upward and involved the endometrial cavity. A panhysterectomy was performed. Histologic examination showed the lesions to be tuberculous. Mosler. 7 The patient was a child 3 years of age, in whom the cervix, uterus, tubes, lungs, peritoneum, and intestines were all invaded with tuberculosis. Death. Autopsy. Diagnosis was confirmed by a histologic examination. Hamolle. 62 Patient, aged 57 years, had pulmonary and peritoneal tuberculosis, from which she died. The disease manifested itself in the cervix and vagina as small deep ulcerations and here and there papillary masses. Bender. 63 Aged 32 years. Good health until the present illness, always fond of sports. Married. One child; labor normal. Miscar- riage one year ago. Trouble dates from miscarriage. Considerable hemorrhage was present for a time after the miscarriage. This finally ceased for a time but recurred, and was accompanied by purulent leukor- rhea. Examination of the cervix showed it to be the seat of an elliptical ulceration. The tuberculous character of the cervical lesion was not recognized until after curettage of the uterus and amputation of the cervix. Histologic examination of the excised portions showed tuber- culosis. Tubercle bacilli were demonstrated by staining methods. Re- covery after the operation was normal. Apparently no tuberculosis of the other sexual organs or other portions of the body was present. The patient's husband had died of typhoid fever, and had always been healthy. Primary case. Peham/ 34 This patient was a nullipara 30 years of age. The chief symptom was progressively increasing purulent, and at times blood streaked, leukorrhea. Examination showed that the anterior cervical lip was the seat of an ulcerative lesion, which was suggestive of carci- TUBERCULOSIS OF THE CERVIX 167 noma. A piece of the ulcer was excised for microscopic examination, which revealed its true character. Tubercle bacilli were demonstrated. In the discussion of Peham's case, Fabricius stated that he had three cases of tuberculosis of the cervix. Case 1. A young girl who had introduced a hair pin into the rectum. This perforated the rectovaginal septum and punctured the cervix. At the point of puncture on the cervix a tuberculosis developed. The diagnosis was confirmed by histologic examination. The girl died nine months later of tuberculous meningitis. Case 2. The patient was a middle aged corpulent woman, who pre- sented herself, suffering from profuse purulent, and at times blood stained, leukorrhea. Examination showed the cervix to be the seat of a necrotic, sloughing tumor-like mass, which was thought to be a carci- noma. Histologic examination of the tissue, however, proved the con- dition to be tuberculosis. Case 3. The cervix in this case was found to be enlarged, indu- rated, and extremely hard. Histologically the condition was found to be tuberculosis. Details of this case are not given. It is of interest chiefly on account of the hardness of the lesion. Santi. 65 This patient was 23 years of age, married, and had two children. The previous history showed that she had suffered from pleurisy some time prior to her present illness. She had also been op- erated upon for a peritonitis, probably of tuberculous origin. For some time there had been symptoms of Pott's disease. The only symptoms referable to the cervix were discharge and occasional irregular bleed- ing. Examination of the cervix showed it to be the seat of an irregu- lar growth. Trachelectomy was performed and the tuberculous char- acter of the lesion verified by histologic examination, the microscope re- vealing chronic inflammatory changes, numerous tubercles, many of which contained giant cells, and the usual typical appearance of tuber- culosis in this locality. One histologic peculiarity was that, at some points, the squamous epithelium over the papilla had formed into masses more or less suggestive of syncytial cells. No tubercle bacilli were demonstrated. Kromer. 20 A case of cervical tuberculosis, the only other focus of the disease being a patch of lupus on the left buttock. In this instance the chief seat of the disease was the external 03, but there were tubercles in the serosa and muscularis of the tubes and uterus. She had had tuberculous peritonitis some years previously. Deletrez. 66 The patient, aged 21 years. A cauliflower growth was present on the portio, which caused a suspicion of carcinoma, but the 1 68 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS microscope proved it to be tuberculous. The body of the uterus pre- sented numerous small granulations separate from one another. Toward the cervix these became papillary in character. The mucosa here was reddened. The diagnosis was confirmed by histologic examination. The patient was in excellent health six months after a total vaginal hvsterectomv. The author believes this to be a primary case. He calls attention to the fact that the ulcerative and miliary forms are usually secondary, but all recorded cases of primary lesions were of the hyper- trophic variety.* Everling. 67 Patient, aged 25, had apparently primary tuberculosis of the portio. The villous or papillary appearance and friable char- acter, discharges, and irregular bleeding suggested carcinoma and pan- hysterectomy, but biopsy showed the real condition and a high trachelec- tomy was considered sufficient. Recovery. Yon Franque. 42 A negress, aged 21 years, had irregular menstrua- tion. There was a lesion on the cervix, which resembled a chancroid. This had not caused symptoms or discomfort. Biopsy showed tuber- culosis. The uterus and adnexa were apparently normal, therefore only a trachelectomy was performed. Histologic examination of the ampu- tated cervix showed tuberculosis, but no tubercle bacilli or cheesy de- generation. No tubercle bacilli were demonstrated in any other part of the body. Lewers. 44 This patient had "bronchitis for a number of years." She was a nullipara, 36 years of age, had a slight, whitish leukorrhea for years, metrorrhagia for 9 months, and recently the discharge had become effusive. Metrorrhagia usually occurred in the form of "spot- ting," following slight trauma. She had noticed pieces of "skin" from time to time in the blood stained intermenstrual leukorrhea. The periods were regular and there was no increase in duration or amount. The cervix was more patulous than usual, and a soft friable growth was felt in the cervical canal, extending as far up as the fingers could reach. This bled easily on touch. The uterus was normal. A diagnosis was made of carcinoma. Yaginal hysterectomy was performed. Recovery. The diagnosis was made by histologic examination. The patient was well 5 years after the operation. The cervical mucous membrane was involved as well as the portio. Croft. 68 Patient, aged 26 years, had a family history of tubercu- losis. A moderate amount of leukorrhea had been present for some time. The menstruation was always irregular. She had amenorrhea for nine months. The cervix was enlarged, softened, and friable, bled * A careful study of the literature fails to confirm this statement. TUBERCULOSIS OF THE CERVIX 169 easily, and there was a profuse mucopurulent discharge. The cervix felt roughened, the anterior lip was elongated, everted, and the raised por- tion coarsely papillary, the projections being of various sizes, some as large as a pea. The involvement was chiefly around the center of the portio. The uterus and adnexa were normal. Biopsy was performed and a diagnosis of tuberculosis was made. Hysterectomy. Recovery. Histological examination then showed involvement of the corporeal endometrium also. Cullen. 69 Case 1. Autopsy specimen. The patient died of a gen- eral tuberculosis. The uterus measures 7 cm. in length, 4.5 cm. in breadth, and 4.5 cm. anteroposteriorly. In the vaginal fornix is an ulcer 1.5 cm. in diameter and 1 cm. in depth. This has a sharply defined margin and a smooth base, studded with minute yellowish dots, varying from a pin point to 1 mm. in diameter. The cervix measures 3 cm. in length and 2.5 cm. in diameter. There is a slight transverse laceration. The lips are red and congested, but present an intact surface. The outer surface of the cervix, 1 cm. from the os on both sides, presents a raised appearance, the tissue being whitish yellow, and showing an ir- regular, eaten out appearance ; the ulcers vary from 1 to 3 cm. in diam- eter. On opening the cervix a cavity 1.5 cm. in diameter is found, which begins at a point 1.5 cm. above the external os. This contains densely necrotic material. Its walls are ragged, eaten out, and irregu- lar. This tissue is yellowish, soft, and stands out in contrast to the injected uterine wall. The adnexa are also involved. The diagnosis was confirmed by histologic examination. Case 2. Attempted vaginal hysterectomy for condylomata and tu- berculosis of the cervix. The autopsy revealed tuberculosis of the en- dometrium, tubes, and ovaries. Miliary tuberculosis of the lungs and pleura. Tuberculous ulcer of the intestines. Tuberculosis of the spleen and kidneys, and solitary tubercles in the brain. The patient, aged 17 years, colored, had a family history of tuberculosis. Profuse, effusive leukorrhea, fever, etc. No cough. The cervix and surrounding vaginal vault were occupied by firm, smooth, polypoid elevations, lining an ul- cerated cavity. These were pinkish in color. Biopsy was performed. Amputation of the cervix was followed by quite severe hemorrhage. Histologic diagnosis. Driessen. 70 This case occurred in a woman who had been operated upon seven years before for a stricture of the rectum and had complained for some time of menorrhagia and mucopurulent discharge. The cervix was found enlarged and studded with many small ulcers, most numerous about the external os, and growing fewer towards the periphery. In the 170 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS ad de sac were small, red spots with yellowish centers. Vaginal hysterec- tomy. Histologic examination showed characteristic tuberculous changes. Vitrac. 48 This patient was a woman, 21 years of age, who entered the service of Lannelongue and Bordeaux, complaining of pain in the lower abdomen and of leukorrhea. There was a history of tuberculosis and examination showed lesions at the apex of the left lung. There was a history of trauma of the genital organs, which was followed by bleed- ing and dysuria. The cervix was enlarged and the seat of a vegetative outgrowth about the size of a walnut. This was elastic, yielding to the touch. The surface of the portio not involved in the papillomatous growths was reddened and inflamed. The uterus was small and adnexa adherent. Biopsy was performed and the diagnosis of tuberculosis arrived at. A vaginal hysterectomy was performed and the diagnosis confirmed by further histologic examination and also by animal inocula- tion. Operative recovery. Frank. 23 This patient gave a previous history of a tuberculous bone disease, involving the metacarpal bone and one phalanx of the middle finger, which was excised and apparently cured 6 years ago. This patient had never suffered any pain and had sought relief for dysmenorrhea. Examination at that time showed the portio vaginalis to be enlarged and somewhat mushroom shaped. Numerous vesicles and nodules were present, most numerous about the external os. The papillary masses bled readily. The case was diagnosed clinically as a malignant neoplasm. Biopsy showed the true character of the lesion. Following the diagnostic excision there was considerable hemorrhage, which required firm tam- ponage. Hysterectomy was performed ; the tubes were normal. Recov- ery. Frank believes the infection in this case resulted from contamination by the hands or by soiled linen. Beyea. 14 A patient, aged 23 years, with a negative family history. She had irregular menstruation, dysmenorrhea, and more or less leukor- rhea, at times purulent, for 3 years. The portio vaginalis was enlarged to twice its normal size and was the seat of an extensive ulcer, which involved the external os. This ulcer was bright red and bled easily. Trachelectomy and bilateral salpingo-oophorectomy was performed, and the diagnosis made by the microscope. Tubercle bacilli were demon- strated by staining methods in some of the sections. A tuberculous salpingitis was also present. The patient was in good health 16 months after the operation. Baudet. 71 This patient was 51 years of age and presented herself, exhibiting symptoms suggestive of carcinoma of the cervix. Examina- tion showed the cervix to be the seat of an extensive papilla-like growth, TUBERCULOSIS OF THE CERVIX 171 which also involved the adjacent anterior vaginal wall. The tumor-like masses were moderately friable, and were covered with a profuse malodorous discharge. The excised tissue presented the usual histologic picture of tuberculosis in this area. A careful examination of the lungs failed to reveal any evidence of tuberculosis. This was of the type described by the French writers as the pseudoneoplastic. Primary case. Young. 72 This patient's family history was negative for tuberculosis. She was a tripara. The last child was born 2^ years ago. They are all healthy. The patient was healthy until six months ago, when the periods began to become more profuse and of longer duration than usual, and for the last five weeks thick yellowish non-odorous leukorrhea has been present. Constant pain in the lower abdomen and sacral region has been present for a similar period. There was no enlargement of the inguinal glands. The cervix was indurated and greatly enlarged. Its surface was uneven and ulcerated in places, and in other places nodular and papillary, but not friable. The uterus was freely movable. There was a suspicion of malignancy. Vaginal hysterectomy was performed. The patient was examined six months after operation and was found healthy. The diagnosis rests upon the histologic evidence. Nebesky. 73 This patient was a woman aged 33 years. A careful examination failed to reveal any foci of tuberculosis other than those in the genital tract. The cervix was the seat of an advanced tuber- culosis; the endometrium of the body of the uterus and the tubes were also involved, but Nebesky believes these were secondary to the cervical lesion, as the pathologic changes became progressively more pronounced as the cervix was approached. The tubes were but mildly affected. Panhysterectomy and bilateral salpingo-oophorectomy were performed and resulted in a cure. Matthews. 74 This patient was single, 22 years of age, negress, never pregnant. First menstruated at 15 years, one day's duration and scant, regular. Later, every three weeks. Dysmenorrhea for three years, and occasionally colicky pains in the hypogastric region. Recently profuse mucopurulent leukorrhea, and often blood stained, has been present. Ex- amination showed the cervix enlarged to twice its normal size, and the seat of a worm-eaten bleeding ulcer. The right adnexa were enlarged, adherent, and the base of the broad ligament was thickened. The fundus was enlarged and partially adherent. A vaginal hysterectomv, bilateral salpingo-oophorectomy and excision of the upper portion of the vagina were performed. The patient was discharged from the hospital as cured. Histologic examination showed the cervix to be the seat of a diffuse tuberculosis, numerous typical tubercles, many of them cheesy, being 172 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS found. No record was made of the demonstration of tubercle bacilli nor was the vaginal lesion reported upon, but it was evidently a direct extension from the cervix. Buscarlet. 75 26 years of age. Family history of tuberculosis. Pul- monary tuberculosis was present. Complained of pain in the vagina and a profuse, thick, foul leukorrhea. Examination showed the cervix the seat of a friable granular growth, and covered with mucopurulent discharge. Death occurred from the pulmonary lesions. Autopsy showed tuberculosis also present in the tubes, ovaries, and body of the uterus. Chaton. 4 Family history of tuberculosis. Entered the Saint Joseph Hospital with a diagnosis of uterine prolapse. The cervix was hyper- trophied. The suspected area was friable and ulcerated and whitish on section, and here and there caseous areas were present. The adnexa were involved. Diagnosis confirmed by histologic and inoculation methods. The lesion evidently began as an interstitial cervical tuber- culosis. Galabrin. 76 The author merely mentions during the course of a discussion a case of tuberculosis of the cervix, which was mistaken for carcinoma. The correct etiology of the condition was discovered only upon histologic examination. Bouilly. 41 The patient was 26 years of age and gave a family history of tuberculosis. Suffered a cervical laceration during delivery. On the posterior cervical lip, at the seat of the laceration, slowly developing ulcer appeared. This was excised, resulting in recovery. The ultimate out- come of the case is not stated. The diagnosis is founded upon histologic examination. Bouffe. 52 The patient was 26 years of age, and gave a family history of tuberculosis. Married and her husband had suffered from a tuber- culous epididymis. She complained of pain in the vagina and purulent leukorrhea. Examination revealed an ulcer occupying the posterior cervical lip, the base and edges of which were moderately firm and presented a somewhat cicatricial appearance. Palliative treatment was followed by improvement. Tubercle bacilli were demonstrated from the ulcer. Reverdin. 77 Case 1. The patient was 30 years of age and presented a personal and family history of tuberculosis. For three months had been suffering from irregular and moderately profuse hemorrhages, chiefly metrorrhagic in type. Examination showed an ulcer on the vaginal cervix, and the body of the uterus and adnexa also involved. On account of the extensive primary involvement and the poor general condition of TUBERCULOSIS OF THE CERVIX 173 the woman, no radical treatment was employed. Death occurred in 3 months. Case 2. This patient was 23 years of age and gave a family history of tuberculosis. Pulmonary tuberculosis was present. For some time the patient had suffered from pain in the lower abdomen and metror- rhagia. The cervix was enlarged, and on the anterior lip was a reddened ulceration. This was treated with silver nitrate and tincture of iodin, and is said to have disappeared. The body of the uterus was also enlarged and probably involved in the tuberculous process. Nanard. 43 Pulmonary and intestinal tuberculosis was present. Death occurred. At autopsy the anterior cervical lip was found to be the seat of an extensive ulcer, which involved the external os. The tubes and uterus were diseased. The diagnosis was confirmed by histologic examination. Lepitit. 18 This was an autopsy specimen, the subject having died of a tuberculous peritonitis and other complications, the lungs also being involved. The cervix was the seat of an ulcer, which possessed irregular edges and a necrotic base. The diagnosis was confirmed by his- tologic examination. The fallopian tubes and uterus were also in- volved. Cornil. 50, 89 Case 1. The patient presented an ulcer on the cervix. On histologic examination this was found to contain tubercles and giant cells and other evidence of tuberculosis. This is one of the earliest if not the earliest case verified by histologic examination. Case 2. This patient was a middle aged woman, whose chief symp- toms were discharge and irregular bleeding. Examination showed the cervix increased in size and indurated; numerous vegetative outgrowths were present. The diagnosis, sarcoma of the cervix, was made, and a panhysterectomy performed. The correct diagnosis was arrived at by histologic examination. Uhland. 78 This patient was 20 years of age and presented a family history of tuberculosis. The symptoms referrable to the cervical con- dition were leukorrhea and irregular bleeding of the metrorrhagic type. Death occurred from a tuberculous peritonitis. Autopsy showed the cervix to be the seat of a tuberculosis; the corporeal endometrium and myometrium and adnexa were also involved. Laboulbene. 79 A patient, 20 odd years of age, died of pulmonary tuberculosis. At autopsy the cervix was found to be the seat of an irregular ulcer. Parrot. 80 The patient was an infant that had suffered from a gen- eral tuberculosis, pulmonary, intestinal, meningeal, renal, etc. The cervix 174 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS was the seat of numerous small outgrowths which also involved the adjacent vagina. Reclus. 81 The patient was about 30 years of age. She was pale and anemic. There was a fistula in ano, probably tuberculous in origin, present. The cervix was enlarged and the seat of numerous small semi- transparent granulation-like outgrowths. The lesions were small and superficial and were treated by cauterization with the actual cautery and the application of the tincture of iodin. Recover) 7 . Pollosson and Violet. 40 This patient was 45 years of age and gave a family history of tuberculosis. Pulmonary tuberculosis was present. The cervix was the seat of a lesion which resembled carcinoma, involving chiefly the anterior lip. A vagino-abdominal hysterectomy was per- formed, and upon histologic examination of the cervix the true character of the condition was discovered. Adnexal lesions were also present. Operative recovery. Raynaud. 1 Case 1. Death from pulmonary tuberculosis. Autopsy showed a small tumor springing from the posterior cervical lip. A doubtful case. Case 2. The patient was 37 years of age and suffered from pul- monary and meningeal tuberculosis. There was a moderate sized necrotic ulcer on the portio. Vaginitis and adnexal lesions were present. The case is without histologic or bacteriologic verification. Haby. 82 The patient was 21 years of age and had always been deli- cate. No pulmonary lesions were present. Profuse offensive leukorrhea was present. A speculum introduced into the vagina showed the cervix to be the seat of a papillomatous, friable, easily bleeding, tumor-like out- growth, covered with a glairy discharge, which was clinically diagnosed as a sarcoma. Biopsy and curettage, however, showed the true nature of the lesion. Tubercle bacilli were demonstrated. Hofbauer. 49 The patient was 26 years of age and presented a family history of tuberculosis. She was a multipara and the labors had been normal. The lungs and heart were normal. Springing from the cervix was a tumor-like outgrowth. A diagnosis of a cervical neoplasm was made and a vaginal hysterectomy performed. Histologic examination showed, however, that the uterus and cervix were the seat of a tuber- culosis. Thiercelin. 26 The patient was 24 years of age and gave a negative family history of tuberculosis. The chief symptoms referable to the genital tract were menorrhagia and discharge. The temperature was 40 ° C, and there was pain in the lower abdomen. Death resulted from advanced pulmonary tuberculosis, involvements of the fallopian tubes, TUBERCULOSIS OF THE CERVIX 175 pericardium, lungs, and body of the uterus. The cervix was the seat of a deep ulcer, the walls of which were soft, spongy and friable. The lesions had apparently extended from the external os. The adjacent vagina was also involved. The diagnosis was verified by histologic examination. The disease followed a miscarriage, and the pulmonary symptoms developed subsequently. Meyer. 33 The patient was 30 years of age and had suffered from lupus. Metrorrhagia and leukorrhea were the chief symptoms referable to the genital tract. Examination showed that the cervix was consider- ably enlarged, reddened and indurated. A portion of the suspected tissue examined histologically showed giant cells and other evidences of tuberculosis. Godard. 83 The patient was 23 years of age and was admitted to the service of Louis at the Hotel Dieu, where she died of a wide spread tuberculosis, the meninges, lungs, intestines, and other organs being involved. It had been noticed before death that the cervix was reddened; it was subsequently found to be the seat of a caseous ulcer. Histologic examination by Corvisart. Cotte. 36 The patient was an anemic woman 23 years of age. The family history was negative for tuberculosis. There had been irregular bleeding and discharge for some months. The cervix was enlarged, and, surrounding the os, was an area somewhat resembling eversion. The adnexa were also involved, but the uterus was small and sclerotic. His- tologic examination of the suspected cervical lesion showed this to be a tuberculous ulceration. Schutt. 32 The patient died when 33 years of age of a general tuber- culosis. An early pregnancy was found, and the decidua and even, in some areas, the myometrium, was the seat of caseous lesions. The cervical mucosa was also involved. It was in some areas thickened and reddened and, on histologic examination, evidence of a tuberculous cervicitis was found. Schutt states that the inflammation of the cervix was catarrhal in type. The surface and granular epithelium presented characteristic changes, caseation in or near the gland, and tubercle bacilli. Sippel. 35 The patient was 31 years of age and complained of leukor- rhea and irregular bleeding. On inspection, the cervix was found to be reddened and congested and to be the seat of an ulcer, the edges of which were firm to the touch. Biopsy was performed and a typical histologic picture of tuberculosis found. The lungs and fallopian tubes were also tuberculous. Menetrier. 28 The patient was 24 years of age and died of pulmon- ary tuberculosis under the care of Jaccound. The body of the uterus 176 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS was enlarged. The cervix was the seat of an extensive ulceration, the surface of which was vascular and friable. Considerable caseous material was present. Tubercle bacilli were demonstrated. The fallopian tubes were also involved. Mayor. 84 This was a patient suffering from pulmonary tuberculosis in the service of Sireday. A whitish, granular ulceration was present on the anterior lip of the cervix. A pelvic peritonitis, involving the uterus and appendages, was also present. Death occurred and an autopsy was performed. Verification of the etiology of the cervical lesions was obtained by histologic examination. Adenot. 85 The patient was 17 years of age and admitted to the service of Poncet suffering from tuberculous peritonitis. A laparotomy was performed and was followed by death. At autopsy the lungs were found to be involved. The mucosa of the cervix was reddened and a small lenticular shaped ulcer was present. Histologic verification of the diagnosis was made. Boldt. 86 The patient had a tuberculous pleurisy and an ulceration upon the cervix. A curettage was performed and later a panhysterec- tomy. Death occurred six hours later. Histologic verification of the diagnosis. Cheron. 87 The patient was 24 years of age, and entered the Saint-Lazare hospital suffering from pulmonary tuberculosis. An ulcer was found in the cervix, which somewhat resembled an ectropion. This was chronic looking in appearance, and gave no marked symptoms. Chiarabba. 88 Menstruated at 16 years. At 24 years the menstruation disappeared. The chief local symptom was discharge. The body of the uterus was enlarged and the cervix was the seat of an ulcer, the base of which was granular in appearance. The diagnosis was verified by his- tologic examination. The patient also had a tuberculous peritonitis and involvement of the uterus, tubes, and labia majora and minora. Fernet. 90 This patient was 27 years of age and suffered from pul- monary tuberculosis. The cervix was the seat of a small granular erosion, which under local treatment disappeared. Tubercle bacilli were demonstrated in the vaginal discharge. A doubtful case. Frerichs. 91 The patient was 25 years of age and presented an ulceration on the mucosa of the cervix, which extended some distance into the canal. The fallopian tubes, uterus and pericardium were involved and also the kidney, intestines and other areas. Gummert. 92 The patient was a nullipara, 29 years of age, who suf- fered from a purulent leukorrhea and amenorrhea. There was a circular ulcer at the external os, and on the surface of the portio were numerous TUBERCULOSIS OF THE CERVIX 177 small, whitish elevations about the size of millet seeds. Biopsy confirmed the diagnosis of tuberculosis and a vagino-abdominal hysterectomy was performed. Adnexitis was present. Gottschalk. 93 The patient was a virgin, 32 years of age, who pre- sented a previous history of tuberculosis. Pains in the lower abdomen and a profuse, thick, offensive discharge were present. Examination showed a papillary mass originating from the cervix. A vaginal hysterec- tomy was performed with an excellent result. Tuberculosis of the endo- metrium and tubes was present. Histologic verification of the diagnosis. Haidenthaler. 94 The patient was 28 years of age and presented a previous history of tuberculosis. An ulcer was present on the anterior cervical lip. This was curetted, without marked benefit. The patient subsequently died, and autopsy showed pulmonary and renal tuberculosis and a tuberculous salpingitis. This diagnosis of the cervical lesion was verified by histologic examination. Holmes. 05 The subject was a cachectic woman, who died of a general tuberculosis, the lungs, peritoneum, intestines, and adnexa being involved. A miliary tuberculosis is said to have been present in the cervix. A doubtful case. Knauer. 96 The specimen was presented before the Vienna Medical Society. The disease was of the ulcerative type, the vaginal portion of the cervix being the seat of a lesion. A panhysterectomy had been performed. Liouville. 97 The cervix was the seat of a tuberculous lesion and the fallopian tubes were also involved. The case is extremely doubtful, despite the fact that the diagnosis was verified by Lebert. Rivilliod. 98 The subject was an aged woman, who died of pulmon- ary and intestinal tuberculosis. The uterus and adnexa were the seat of inflammatory lesions. An ulcer was present in the cervix. Richelot." Case 1. The patient suffered from pulmonary tuber- culosis. An ulcer was present in the cervix, and a hysterectomy was performed. The diagnosis was verified by histologic examination. This case is not reported by Richelot in detail. Case 2. The patient was a nullipara. Examination showed the cervix enlarged and the seat of an ulcerative lesion. Biopsy was per- formed and the diagnosis of tuberculosis made. Hysterectomy was then performed, and examination of the specimen thus obtained showed in- volvement of the cervical canal. The corporeal endometrium, according to Cornil, was the seat of a non-tuberculous endometritis. He, however, thinks that the cervical tuberculosis was the result of a hematogenic infection. i/8 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Yon Hauschka. 100 The patient was at the cancer age and presented symptoms suggestive of this condition. The cervix was enlarged and covered with partially necrotic papillary outgrowths, which were bathed in an offensive discharge. A vaginal hysterectomy was performed and the diagnosis arrived at by histologic examination. Tuberculosis was also present in the body of the uterus and in the fallopian tubes. Schultze. 101 The patient presented a previous history of tuberculosis. Leukorrhea was the chief local symptom. The mucosa of the cervix was irregular, reddened, and bathed in an offensive discharge. Biopsy was performed and the diagnosis of tuberculosis made. Vaginal hysterectomy was then performed. Bacteriologic and histologic examination confirmed the diagnosis of the cervical condition. Spath. 6 The patient was 26 years of age and gave a previous history of tuberculosis. The lungs, the body of the uterus, were involved. The cervix was the seat of an ulcerative lesion. The base of the ulcer had a granular appearance. Thompson. 102 Hypoplasia of the lower genital tract was present in a young girl. The cervix was the seat of numerous small semitranslucent elevations which were thought to be miliary tubercles. A pelvic peri- tonitis was present. In this case small retention cysts were probably mistaken for miliary tubercles, as has previously been done by Lis- franc 103 and Thiry. 104 Walther. 105 The patient was 26 years of age. Amenorrhea was present. Leukorrhea had been present for some time. The body of the uterus was the seat of a tuberculosis. There was an ulcer on the cervix, which was covered with glairy discharge. The base of the ulcer was moderately soft. Biopsy was performed and the diagnosis of tuberculosis arrived at. Weigert. 106 An aged woman suffering from pulmonary and peri- toneal tuberculosis. An ulceration was present upon the vaginal cervix and had extended to the adjacent vagina. Clinical diagnosis only. Winter. 107 The patient suffered from pulmonary and peritoneal tuberculosis. A tuberculous endometritis and salpingitis was also pres- ent. There was a necrotic ulceration on the portio vaginalis. The histologic examination confirmed the diagnosis. Ducuing and Rigaud. 108 The patient was 33 years of age. She entered the service of Chamayou at the Hotel Dieu. Gave a previous history of tuberculosis. Bipara. Pulmonary tuberculosis. Irregular menstruation and profuse purulent leukorrhea were the chief symptoms referable to the genital tract. Examination shows the external os to be the seat of an irregular ulceration, affecting chiefly the right side of the TUBERCULOSIS OF THE CERVIX 179 cervix. The cervix was increased in size. Biopsy revealed the true character of the lesion. Panhysterectomy was performed. Animal inoculation from the cervix produced tuberculosis. The patient recovered. Williams. 109 Case 1. Aged 63 years, multipara; death from pul- monary tuberculosis. Autopsy showed advanced phthisis and tuberculous pelvic peritonitis. Uterus slightly enlarged. Anterior cervical lip hyper- trophied and adherent to the adjacent vaginal wall. A number of ulcers were present. They were irregular, sharply cut, possessed slightly raised edges and a base studded with grayish semitransparent granulations. Extension to the adjacent vagina had occurred. The diagnosis was verified by histologic examination. Case 2. Bipara, aged 36 years. Chief symptoms backache and pro- fuse leukorrhea. These symptoms were of several months standing. Painful and scanty menstruation. The cervix was lacerated and felt indurated, and was reddened and the seat of an ulcer, the base of which was yellowish gray, with indurated and sharply cut edges. This bled easily to touch. A caseous cast filled the ulcer, which, when removed, left a nodular bleeding base. Histologic verification of diag- nosis. Palliative treatment, but patient still under treatment when report was made. Patient had some lung condition some time prior to the appearance of the genital symptoms, and was a delicate woman. Maly. 110 The patient was a single woman, 21 years of age. Pul- monary tuberculosis had been present for some time and she was weak and anemic. The chief symptom referable to the genital tract was leukor- rhea, which was moderately profuse, offensive, and occasionally blood tinged. A pelvic examination showed the hymen unruptured and the portio vaginalis the seat of a papillary, fungus-like outgrowth, which was covered with discharge and which, on touch, was soft, friable, and bled easily. Hysterectomy was performed and the diagnosis verified by histologic examination. Tate. 111 Aged 36 years. Nullipara, married. Had an operation for the removal of tuberculous glands of the neck 6 years ago. Had an attack of pelvic peritonitis three years ago. Dysmenorrhea for one year. Examination showed cervix enlarged and the cervical canal extended into a large cavity filled with a soft, friable growth. A portion of this was removed digitally, and histologic examination showed tuberculosis. The uterus was enlarged and the appendages involved. Vagino-abdominal hysterectomy and bilateral salpingo-oophorectomy. The ulcer did not extend above the internal os, and the portio vaginalis was fairly normal. Recovery. Tedenat. 112 The patient was a woman, 26 years of age, who had 180 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS been married three years, and had never been pregnant. Amenorrhea. The chief local symptom was discharge. The cervix was the seat of a polypoid vegetative outgrowth, which was moderately soft to the touch. The diagnosis of cancer Mas made clinically, and the correct etiology of the lesion was only determined by histologic examination. The lesions were cauterized and treated with formalin, with some im- provement. Addisell. 113 The author exhibited a specimen of a uterus from a tuberculous woman, in which tuberculosis was present and was his- tologically demonstrated as extending from the cervix to the fundus. Bender. 63 The patient, aged 34 years, entered the clinic September 25, 191 1. The family history was negative for tuberculosis. There was a previous history of measles, scarlet fever, and articular rheumatism, and an acute pneumonia fourteen years ago. She was a tripara, the puer- perium had been normal, and the children were healthy. The chief geni- tal symptom was a profuse, purulent malodorous discharge. This had been present for four years, but of late had been increasing in amount. There was pain in the lumbar region. The kidneys were normal, as were the lungs. Examination revealed some tenderness over the lower abdomen. Pelvic examination showed the cervix enlarged and engorged with blood, reddish in color, and on the posterior lip was an irregular, granular, pink- ish ulcer partially covered with exudate. The ulcer was moderately soft and friable. On account of the possibility of cancer, biopsy was per- formed, and when the character of the lesion was determined, a trache- lectomy and dilatation and curettage was performed. Recovery was un- eventful, and the patient was well 2 years later. The diagnosis was finally verified by histologic examination and animal inoculation. Popow. 114 The patient was a multipara, 39 years of age, who pre- sented herself, suffering from a necrotic ulcer of the cervix. Macro- scopically the lesion was suggestive of cancer, and biopsy was performed. On account of the advanced character of the lesion and involvement of the corporeal portion of the uterus and also of the adnexa, a vaginal hysterectomy was performed. In addition to the cervical lesion, there was a tuberculous focus in the anterior uterine wall near the left cornu. Tuberculous salpingitis and endometritis was also present. Histologic verification. Stone 115 merely mentions a case operated upon by Dr. Cole. Stone states that he examined the specimen and that there was no tuberculosis found in any other portion of the genital tract or any history or physical signs of tuberculosis in any other portion of the body. The patient's husband was a strong healthy man, and there was no tuberculous family TUBERCULOSIS OF THE CERVIX 181 history. The source of the infection could not be determined, but Stone states that it was without doubt a primary tuberculosis of the cervix. Nicolo. 116 Case I. The patient was 24 years of age. Menstruated first at 14 years. Was always regular, but scant. She was married at 18 years. Had suffered from cough and other symptoms of pulmonary tuberculosis for some time. The chief symptoms referable to the genitalia were discharge and bleeding. The bleeding was of the metrorrhagic type and often followed slight trauma, such as coitus, etc. The discharge was purulent and frequently blood stained. Examination showed a fungoid, ulcerating mass, occupying the position of the cervix. The uterus was antiflexed and movable. Operation — Recovery. Histologic verification of diagnosis. Case 2. The patient was a married woman, 40 years of age, who had had a number of children. She had pulmonary and laryngeal tuber- culosis, and gave a history of lupus. The genital symptoms were sug- gestive of carcinoma — purulent, frequently blood streaked discharge, and irregular bleeding, especially following trauma. Examination showed a fungoid, ulcerating, friable mass originating from the cervix. His- tologic verification of the diagnosis. Kynoch. 117 The patient was a married woman of 45 years of age, who had a family history of tuberculosis. For 3 months there had been irregular hemorrhages per vagina. Pelvic examination showed the external genitalia and vagina normal. The portio vaginalis was nor- mal in appearance and the os patulous. The cervical canal, especially the anterior surface, was the seat of an eroded lesion. Many papillary outgrowths were present. These were stated not to have been friable, but bleeding followed manipulation. Biopsy was performed, followed by a vaginal hysterectomy. The diagnosis rests upon histologic veri- fication. A tuberculous endometrium was also present. Rossle. 118 Autopsy case. Eighty-seven years, tuberculosis of the lungs, fundus of the uterus, and anterior commissure. An ulcerative tuberculous lesion was present in the cervix. Moore. 119 Age 27; married; negative family history; husband sound ; nullipara ; normal menstruation. For four weeks observed spot- ting after trauma. Diagnosis, cancer. Correct diagnosis made by biopsy. Vaginal hysterectomy. Recovery, but considerable foul leukorrhea and induration of vault of vagina fifteen months later. Histologic examina- tion showed tuberculosis of tubes, fundus of uterus, and cervix, the oldest lesion being in the tubes. In the uterus and cervix the tuberculosis is limited to the mucosa. This case was of the miliary type. 182 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS For additional cases of tuberculosis of the cervix the reader is re- ferred to the chapter dealing with lesions of the vagina. TUBERCULOSIS OF THE BODY OF THE UTERUS Tuberculosis of the uterus may occur as an endometritis, a myome- tritis, or a perimetritis, or combinations of these lesions may be pres- ent. By far the most frequent structure of the uterus to be attacked by this infection is the corporeal endometrium. Infection of the endo- metrium is the second most frequent site for genital tuberculosis. Endometritis. — Like other forms of genital tuberculosis, endo- metritis is almost always secondary to tuberculosis elsewhere in the body. Gordeler, 120 in a series of 4,620 postmortems, observed one case of apparently primary tuberculous endometritis, that occurred in a woman 68 years of age. Our experience points to the fallopian tubes as the source of infection in the large majority of lesions of this local- ity. A study of our cases has shown that in not a single instance in our series has the endometrium been involved without a concomitant in- fection of the tubes. Furthermore, our cases seem to show that the tubes harbor the primary genital lesion, and that from them the dis- ease spreads as a descending infection, generally by continuity, to the endometrium of the body of the uterus. Contamination of the endo- metrium by leakage of the infected tubal contents through the intra- mural portion of the fallopian tube doubtless also accounts for a cer- tain percentage of cases of tuberculous endometritis. The fact that, in a definite proportion of specimens, only the endometrium in the im- mediate neighborhood of the uterine end of the tube has been involved, the lower portion of mucosa of the body of the uterus being normal, is significant. This is especially likely to be the case in early cases, for, as the disease advances, the entire mucosa often becomes invaded. As in endometritis, the result of organisms other than tubercle bacilli, how- ever, the entire mucosa is not as a rule uniformly attacked, irregular areas of well defined inflammation being scattered with other areas either less inflamed, or even normal endometrium. When, however, an endo- metritis is present, the mucosa in the cornua of the uterus is nearly al- ways invaded, and usually the seat of the more advanced inflammation. Numerous authorities have observed tuberculous endometritis with- out tubal involvement. As has been stated, this has not occurred in any of the cases comprising our series, and it is a generally accepted fact that, if the endometrium is the seat of a tuberculosis, the tubes are TUBERCULOSIS OF THE CERVIX 183 involved in the great majority of cases. This is an important point to be considered in the treatment of genital tuberculosis. In other words, where a tuberculous endometritis is present, the tubes are also involved in the large proportion of cases, and this fact should be taken into con- sideration in the treatment of the disease, as the endometritis cannot be cured if constant reinfection is occurring from above. In at least two cases of our series the chief symptom was leukorrhea, the symptoms resulting from the tubal lesion being of minor subjective importance as compared with those arising from the uterus. Tuberculous endometritis may be of either the (1) miliary, or (2) caseous or ulcerative variety, the former being by far the most frequent, in the proportion of 4 to 1 in our series. In this variety macroscopic lesions are not always present, although thickening and reddening of the endometrium are often observed. Tubercles can be seen in some specimens with the naked eye, but in many they are inconspicuous or even undiscernible except with the microscope. Characteristic lesions, although not by any means always present in the tubes, are, however, much more frequent than in the interior of the uterus. In the caseous variety the etiology of the lesion can generally be determined by the macroscopic appearance of the specimen. The thickening and redden- ing of the mucosa, with perhaps here and there actual ulcer formation, and the characteristic cheesy particles adherent to the endometrium should always at least suggest this form of infection. As in other forms of endometritis, more or less involvement of the underlying myometrium usually occurs, and in the advanced cases, especially of the caseous variety, the uterus is often enlarged, a well marked myometritis being present. On the other hand, in early cases, especially of the mild type, the uterus is often normal in appearance, and only upon close histologic examination will any involvement of the myometrium be found, and only then in the muscle fibers immediately underlying the infected endometrium. Tubercles may, of course, be present upon the peritoneal surface in any of the varieties of uterine tuberculosis. One or two cases have been observed in which the myo- metrium was the only portion of the uterus involved. Such instances are, however, of extreme rarity and may be regarded as pathologic entities. As has been stated, a well defined perimetritis is a frequent lesion and a common accompaniment of tuberculous salpingitis. Kromer 20 has recorded an unusual case, in which a tuberculous process had perforated the posterior uterine wall, forming a communication be- tween it and Douglas' cul de sac. As in tuberculosis of the tubes, no age is immune. The disease, how- 184 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS ever, most frequently occurs during the active sexual life of the indi- vidual. Symptoms. — The symptoms resulting from a tuberculous endo- metritis are generally more or less masked by those produced by the accompanying salpingitis. Less frequently the converse is true. As in- fection of the endometrium is usually antedated by a similar infection of the tubes, a close analysis of the history of the case will usually show that symptoms resulting from disease of the latter structures have occurred before those directly attributed to the uterus. The symptoms resulting from the endometritis alone are in no way pathognomonic of tuberculosis, but are common to all forms of chronic endometrial infec- tions of the uterine mucosa. Leukorrhea, pain or tenderness in the lower abdomen are perhaps the most frequent. The leukorrhea is not usually characteristic; in advanced cases of the caseous variety cheesy particles may be present in the discharge and, where observed, are al- ways suggestive of this form of infection. The leukorrhea naturally varies markedly in different cases, but is usually moderately profuse, at first thin, in the latter stages becoming purulent, and in some in- stances, especially when ulcers are present in the uterine mucosa, blood stained. The presence of leukorrhea without evidence of infection of the lower genital tract, especially in the young and virginal, should suggest the likelihood of tuberculosis as an etiologic factor. In gonor- rhea the discharge is chiefly cervical in origin. Tuberculosis exhibits a strong tendency to limit its downward spread to the internal os, so that the discharge in these cases is corporeal in origin, although naturally more or less mixed with the normal cervical mucus. The leukorrhea consists of secretions from the uterus and to a lesser extent from the cervical glands, epithelium debris, and leukocytes. In some instances cheesy particles consisting of tuberculous debris may be present. Tuber- cle bacilli are usually present in the discharge, but are often few in numbers and difficult to demonstrate. It would seem, however, that the examination of the discharge for tubercle bacilli, either by animal inoculation or by smear methods, or by both, in suspicious cases is a means of diagnosis which has not been fully taken advantage of by many observers. Pain is by no means a constant or reliable symptom. Pain and ten- derness in the region of the uterus are, however, suggestive of the oc- currence of an endometritis. How much of the dysmenorrhea which these patients suffer from is due to an actual uterine involvement, how much to the usual accompanying adnexal lesions, is difficult to determine. Pulmonary tuberculosis itself often produces dysmenorrhea, and as TUBERCULOSIS OF THE CERVIX 185 many of these patients are the incumbents of pulmonary lesions, this may account for some of the cases of this condition. Certain it is that cases of pelvic peritonitis of tuberculous origin nearly always suffer from dysmenorrhea, usually of the congestive type, the pains appearing some hours or days before the appearance of the flow, continuing for the first few days, and being of a dull, heavy aching character in the lower abdomen and lower lumbar and sacral region. Menstrual irregu- larities, both as to periodicity and amount of flow, are of frequent oc- currence, but are probably more the result of the primary lesion or of the ovarian involvement than of the actual endometritis. Diagnosis. — A positive diagnosis is practically impossible, unless tubercle bacilli can be demonstrated in the discharge or the tissue is ex- amined histologically. The absence of evidence of other forms of in- fection, the presence of a tuberculosis in other parts of the body, ex- treme youth or virginity, are all suggestive of this form of tuberculosis. As has been stated, uterine involvement is usually secondary to adnexal lesions, so that much of what will be said regarding the latter condition applies to tuberculous endometritis. Treatment. — All forms of local application are valueless in this variety of infection. Curettage, followed by the application of some bactericide, such as the tincture of iodin or formalin solution, or in severe cases, hysterectomy, are the two forms of treatment which offer the best hope of cure. Curettage alone is not indicated. The tubes are nearly always involved, and, unless an operation is directed towards them at the same sitting, an acute exacerbation of the salpingitis is likely to occur. For this reason curettage should immediately precede all operations for tuberculous salpingitis, but is usually contra-indicated under other circumstances. Vaporization has been employed by some operators. The introduction of live steam into the uterine cavity is not without danger. The author believes that the risks attending this form of treatment are greater than in curettage or even hysterectomy, and that the results are not so satisfactory. One of the chief disadvantages of vaporizing is the difficulty of actually controlling the steam and as- certaining the exact depths to which the tissues are being destroyed. The endometrium in these cases varies quite widely in thickness and what would be sufficient steam to boil off the mucosa in one case might only destroy the superficial layer in another. Of prime importance in these cases is the treatment of the adnexitis, and the question of whether or not the uterus shall be removed depends largely upon the type of operation practiced upon the tubes and ovaries. When it is necessary to remove both ovaries, nothing is gained by the 186 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS conservation of the uterus. The actual condition of the uterus, whether enlarged, etc., must also be taken into consideration, as well as the age of the patient, the condition of the primary lesion, and many other points which will be considered under the treatment of tuberculous adnexitis. The author believes that in all cases, regardless of symptoms, a curettage and iodinization of the uterus should precede all operations upon the adnexa in tuberculous cases. LITERATURE i. Raynaud. Arch. gen. de med. 1831. 26:486. 2. Virchow, R. Virch. Arch. 1853. 5 404. 3. Kiwisch. Klinik. von Frage. 1854. v. 1. 4. Chaton. Rev. de gyn. et chir. abd. 1908. 12 1947. Also Rev. int. de la tuberc. 1908. 14:401. 5. Rigal. Bui. soc. med. des hop. 1876. 6. Spath. These de Strassburg. 1885. 7. Mosler. Inaug. Dis. Strassburg, 1883. 8. Broucha. Rev. de gyn. et de chir. abd. 1902. 6:295. 9. Klobb, I. M. Pathologische Anatomie der Weiblichen Sexual- organe. Vienna, 1864. p. 193. 10. Kaufmann. Ztschr. f. Gebh. u. Gyn. 1897. 37:118,123. 11. Mich ales. Beitr. z. Gebh. u. Gyn. No. 14. 12. Brooks, W. H. B. Tr. Obst. Soc. London, 1903. 45:185. 1904. 46 :265. 13. Ferrari, P. L. Ann. di ost. e gin. 1903. 25 :i69, 456. 14. Beyea, H. D. Ann. de gyn. et d'obst. 1900. 54:169. Also Cong, internat. de med., sect, de gyn. 1900. p. 316. 15. Zweigbaum. Berl. Klin. Woch. 1886. No. 22:443. 16. Frankel. Jhrb. d. Hamburg, stskans. 1893, 1894. 17. Montanelli, G. La gin. 1907. 4:647. 18. Lepitit. Bui. soc. anat. de Paris. 1892. 19. Veyrat, H. These de Paris. 19 10. 20. Kromer. Monschr. f. Gebh. u. Gyn. 1908. 26:45. 21. Fabricius. Gynecologic. Paris, 1908. p. 180. 22. Glocner. Beitr. z. Gebh. u. Gyn. 1901. 5 : part 2. 23. Frank. Monschr. f. Gebh. u. Gyn. 1899. v. 10. 24. Broye. These de Paris. 1903. 25. Rigal. Bui. soc. med. des hop. May, 1879. 26. Thiercelin. Bui. soc. anat. de Paris. 1889. 2y. Martin, J. Arch. prov. de chir. 1905. 14:471. TUBERCULOSIS OF THE CERVIX 187 28. Menetrier. Bui. soc. anat. de Paris. 1886, 1889. 29. Lannes-Dehore, L. These de Lyon. 1905. 30. Alterthum. Monschr. f. Gebh. u. Gyn. 1902. No. 26. 31. Morlitte, C. Arch, di ost. e gin. 1901. pp. 6b, 649, 714. 32. Schutt. These de Kiel. 1889. 33. Meyer, A. Arch. f. Gyn. 1893. 45 : 5°4. 34. Giglio, G. Ann. di ost. e gin. 1892. 14:105. 35. Sippel. Monschr. f. Gebh. u. Gyn. 1905. 36. Cotte, G. Gaz. des hop. 1907. 80:1227. 37. Cova, E. Gyn. Rundsch. 1910. 4:318. 38. Patel, M. Ann. de gyn. et d'obst. June, 19 12. 39. Murphy, J. B. Tuberculosis of the Female Genitalia and Peri- toneum. Chicago, 1903. 40. Pollosson, A., et Violet, H. Rev. de gyn. et chir. abd. 1906. 10:205. 41. Bouilly. These de Leuret. 1903. 42. von Franque. Ztschr. f. Gebh. u. Gyn. 1897. 37: No. 2. 43. Nauard. These de Paris. 1892. 44. Lewers, A. H. N. Jr. Obst. Gyn. Brit. Emp. 1902. 1 ^76, 632. 45. Cornil, V. Jr. des conn. med. prat. 1879. 1 :i3i. 46. Emanuel. Ztschr. f. Gebh. u. Gyn. 1893. v. 29. 47. Ressegna. Arch, di ost. e gin. 1903. 12:554. 48. Vitrac. Arch, de med. exp. 1898. 10:295, 314. Also, Ann. de gyn. et d'obst. 1898. p. 32. 49. Hofbauer. Arch. f. Gyn. 1898. 50. Cornil. Jr. des conn. med. prat. July, 1883, 1888. 51. Deuville. These de Paris. 1887. 52. Bouffe. These de Paris. Also, Rev. de gyn. et de chir. abd. 1908. 12 ^47. 53. Petit-Dutaillis. La gyn. Feb., 1913. 17:65. 54. Patel, M. Rev. de gyn. Aug., 1912. 55. Haultin, F. N. W. Edinb. Med. Jr. 1913. 11:231. 56. Smith, A. J. Tr. Roy. Acad. Med. Irel. 1904. 22 :250. 57. Vineberg, H. N. Am. Jr. Obst. 1903. 42:98. 1908. 57:652. 58. Lorrain et Chaton. Bui. soc. anat. de Paris. 1907. p. 649. 59. Horrocks, P. Proc. Roy. Soc. Med., Sec. Gyn. 1907. p. 66. 60. Garkisch, A. Deutsch. Med. Woch. 1907. 32:208. 33:991. 61. Zweifel. Centrbl. f. Gyn. 1890. 62. Hamolle, H. Prog. Med. Obst. in Centrbl. f. Gyn. 1877. No. 15. 63. Bender, X. Rev. de gyn. et de chir. abd. 191 1. 17:193. 1914. 12 :29. 188 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS 64. Peham, J. La Gyn. 1908. 12:180. 65. Santi, E. La gin. 1909. 6:257. 66. Deletrez. Bui. acad. roy. de med. de Belg. 1907. 21 :6<48. Also, Ann. de gyn. et d'obst. 1908. 5:26. Also, Rev. mens. de gyn., obst., ped. 1908. p. 16. Also, Ann. de l'inst. chir. de Brux. 1908. 15:33. Also, La gyn. 1908. 12:178. 6y. Everling, K. Berl. Klin. Woch. 1909. 46:1446. 68. Croft, E. O. Jr. Obst. Gyn. Brit. Emp. 1902. 1 ^39. 69. Cullen, T. S. Carcinoma of the Uterus. New York, 1900. P- 193- 70. Driessen, L. E. Ned. tijdscr. v. verl. en. gyn. 1898. 9:66. 71. Baudet. Toulouse med. 1908. 10:193. y2. Young, E. E. Tr. Obst. Soc. Lond. 1906. 48:286. 73. Nebesky. Monschr. f. Gebh. u. Gyn. 1905. 22 : No. 5. 74. Matthews, F. S. N. Y. Med. Rec. 1898. p. 872. 75. Buscarlet. Bui. soc. anat. de Paris. 1890. 76. Galabrin. Tr. Obst. Soc. Lond. 1906. 48:300. yy. Reverdin. Rev. med. de la Suisse Rom. 1895. 78. Uhland. Inaug. Dis, Tubingen, 1886. 79. Laboullene. Elements d'anatomie pathologique. 1879. Ab- stracted by Chaton, No. 4. 80. Parrot. Quoted by Chaton, No. 4. 81. Reclus. These de Daurios. 1889. 82. Haby. Monschr. f. Gebh. u. Gyn. 1907. 83. Godard. Bui. soc. anat. de Paris. 1867. 84. Mayor. Bui. soc. anat. de Paris. 1881. Also, Prog. med. 1882. 85. Adenot. Gaz. hebd. de med. 1902. 86. Boldt, H. J. Tr. N. Y. acad. med. 1902. 87. Cheron, J. Rev. med-chir. des mal. des fern. 1886. 8:82. 88. Chiarabba. Gior. di gin. e di ped. 1904. 22:341. 89. Cornil, V. Bui. soc. med. des hop. 1879. 90. Fernel. These de Paris. 1887. 91. Frerichs. These de Nauard, No. 43. Quoted by Chaton, No. 4. 92. Gummert. Monschr. f. Gebh. u. Gyn. 1903. 93. Gottschalk, S. Arch. f. Gyn. 1903. 70:1. 94. Haidenthaler. Wien. Klin. Woch. 1890. 3 1655. Also, Centrbl. f. Gyn. 1891. 15:76. 95. Holmes, C. London Med. Gaz. 1830. Quoted by Chaton, No. 4. 96. Knauer, K. K. Monschr. f. Gebh. u. Gyn. 17:554. 97. Lionville. Bui. soc. anat. de Paris. 1873. 98. Rivilliod. Bui. soc. anat. de Paris. 1884. TUBERCULOSIS OF THE CERVIX 189 99. Richelot, L. G. Chirurgie de l'uterus. Also, La gyn. 1905. 10:481. Also, Compt. rend. soc. d'obst., gyn., paed. 100. von Hauschka. Wien. Klin. Woch. 1901. 101. Schultze. Gyn. Helv. 1905. 5:135. 102. Thompson. Lancet. 1872. 103. Lisfranc, Clin. chir. de la Pitie. 1842. 2:661. 104. Thiry. Presse med. Beige. 1852. 4:1. 105. Walther. Monschr. f. Gebh. u. Gyn. 1897. 106. Weigert. Virch Arch. 1876. 107. Winter. Centrbl. f. Gyn. 1887. 108. Dunning, J., et Rigaud. Provence med. 191 1. 22:284. 109. Williams, J. D. Brit. Med. Jr. 1895. 1 :968. no. Maly, G. W. Monschr. f. Gebh. u. Gyn. 1907. 26:219. in. Tate. Tr. Obst. Soc. Lond. 1904. 46:138. 112. Tedenat. Cong. Franc, de chir. 1905. 113. Addisell. Jr. Obst. Gyn. Brit. Emp. 1905. 8:348. 114. Popow. Russky. vratch. 1906. No. 12, 13. 115. Stone. Am. Jr. Obst. 1910. 61:98. 116. Nicolo, R. di. Arch. Ital. di gin. 1914. 17:61. 117. Kynoch, J. A. C. Brit. Med. Jr. 1903. 2:962. 118. Rossle. Verhl. d. Deutsch. Gesel. f. Gyn. 1911. 14:441. 119. Moore, G. A. Surg., Gyn., Obst. 1919. 29:1. 120. Gordeler, G. Beitr. z. Klin. d. Tuberk. 1913. 28 : No. 3. The following bibliography should be consulted, for many of these papers contain reports of cases, but were not included in the above list, as the author has been unable to obtain references to some of the original reports : Ajello, A. Rif. med. 1900. 3:615 (primary case). Amann, J. Monschr. f. Gebh. u. Gyn. 1902. 16:586-630. Archambault. Gaz. de gyn. 1902. Attilio. Med. Blat. 1906. Basso. Ann. di st. e gin. 1905. Baumgarten. Berl. Klin. Woch. 1904. Beaulin. Ann. de derm, et syph. 1903. p. 54. Beyea, H. D. Am. Jr. Med. Sc. 1901. v. 122 : No. 6. Brouardel. These de Paris. 1865. Cayla. Contribution a l'etude de la tuberculose du col de l'uterus Bordeaux, 191 2, Gounouilhou. Chaton. Bui. Soc. anat. de Paris. 1904. Cousyn. Sem. gyn. 1901. 190 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Cruveilhier. Quoted by Chaton, No. 4. Daurios. These de Paris. 1889. Delaunay et Darre. Gaz. des Hop. 1905. Dervaux. These de Lille. 1902. Deichmann, W. G. Uber einen Fall von Primarer, Papillarer Tuberkulose an der Portio Vaginalis Uteri. Leipzig. 1910. R. Noske. Fintecus, D. Rev. int. de la tuberc. 1913. 23 :33c Gastany. These de Montpellier. 1905. Giel. Inaug. Dis. Erlangen. 1881. Gorowitz, M. La tuberculose genitale chez la femme. These de Paris. 1900. Hartz. Monschr. f. Gebh. u. Gyn. 1902. v. 16. Heiberg. Centrbl. f. Gyn. 1892. Jerie, J. Sborn. lek. 1908. 9:1. Jerie, J. Rev. de med. Tcheque. 1908. 1 :20. Kaposi. Jr. de med. de Bordeaux. 1888. Kribich. Soc. Viennoise de derm. May 8, 1901. Kuttner. Beitr. z. Klin. Chir. 1913. 13:583. Labadie-Lagarre et Leguen. Traite med-chir. de gyn. 1904. Lannes-Dehore, L. Contribution a l'etude de la tuberculose du col de l'uterus. Lyon. 1905. Lassar. Soc. Viennoise de derm. 1891. Le Denu. Sem. gyn. 1901. Leuret. These de Paris. 1903. Limville. Bui. Soc. Anat. de Paris. 1873. Martin. Monschr. f. Gebh. u. Gyn. 1902. v. 16. Muret. Rev. med. de la Suisse Rom. Dec, 19 10. p. 1050. Naudin, L. Contribution a l'etude de l'ulceration du col de l'uterus. Paris. 1885. 7:616-623. Also, 1886, 8 :i34-i46. Popoff. Inaug. Dis. St. Petersburg. 1898. Pozzi. Traite de Gynecologic 1907. Schenk. Beitr. z. Klin. Chir. 1896. 17:526. Schulze-Smiarkovska, H. Uber einen Fall Tuberkuloser Er- krankung der Portio Vaginalis. Zurich. 1904. A. Mark- walder. Sinety, de. Gaz. med. de Paris. 1883. 5 :489. Taylor. Lupus of the Cervix Uteri and Female Genitalia, New York. 1888. J. H. Vail & Co. Thebierge. Ann. de derm, et syph. 1896. p. 1374. Vassmer. Arch. f. Gyn. v. 57. TUBERCULOSIS OF THE CERVIX 191 Voight. Arch, f . Gyn. 69 : No. 3. Weyl. tiber Localisierte Tuberkulose des Collum Uteri. Giessen. 1904. R. Lange. Williams, J. D. Med. Press and Circ. 1894. 58:228. CHAPTER IX TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES Fallopian tubes and ovaries anatomically and symptomatically considered together — Predisposition — Routes of transmission — Histologic examination — Factors — An- alysis of cases — Study of acute and chronic stages — Duration of acute stage — Characteristics of chronic stage — Other forms of infection — Tuberculin an aid to diagnosis — Differential diagnosis between tuberculous, gonococcal, and strepto- coccal pelvic inflammatory disease — Family history — Prognosis — Cases — Methods of treatment — Bibliography. General Considerations. — Tuberculosis of the fallopian tubes is a comparatively frequent form of infection, whereas true tuberculosis of the ovaries is relatively infrequent. However, when tuberculosis of the tubes is present, a peri-oophoritis is a common accompaniment. For this reason, and because the two organs are so closely associated, both anatomically and symptomatically, tuberculosis of these structures will be considered together. Tuberculosis of the tube, like tuberculosis of the other parts of the genital tract, is usually secondary to a tuberculous focus elsewhere in the body, pulmonary tuberculosis being by far the most frequent seat of the primary disease. Next to the lungs, the peritoneum, osseous system, lymph glands, and intestines are perhaps the most frequent sites of the primary infection. In a series of thirty cases from the gynecological de- partment of the University of Pennsylvania Hospital which have been studied, thirteen showed well marked pulmonary lesions. Of the thirteen, involvement of one lung was present in nine, and in the remainder both lungs were affected. In all thejpulmonary lesions were quiescent, and in none was the disease advanced. The material from which these statistics were formulated was based upon operative cases only. It has not been our custom to operate upon patients in whom the pulmonary lesions are either acute or advanced, and for this reason the foregoing statistics are some- what misleading. Albrecht and Schlimpert, 1 in a series of autopsies on women, found that the primary source of the genital infection was as follows : lungs, 73 per cent; intestines, 20 per cent; bones, 4 per cent; peritoneum, 2 per cent. 192 TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES 193 The frequency with which the tubes are affected in tuberculous fe- males has been analyzed in previous pages. In the series of postmortem records from the Henry Phipps Institute studied by the author these organs were found macroscopically diseased in about 7 per cent of cases. This closely corresponds with figures given by other observers. It should, however, be remembered in considering postmortem records that, as a rule, no histologic examination was made of the fallopian tubes, unless these structures presented macroscopic lesions. Williams 2 has very properly pointed out that occasionally histologic examina- tions reveal tuberculosis in macroscopically normal tubes, and this fact should be taken into consideration when considering the above figures. On the other hand, postmortem records, unless confirmed by histologic examinations, may be misleading, in that gonococcus or other pyogenic organism may produce pathological processes in the fallopian tubes of tuberculous women, and, unless the diagnosis of tuberculous salpingitis is confirmed by a microscopic examination, may cause a misconception regarding the type of the infection present. As has been stated, tuberculosis of the tubes is, in the great majority of cases, secondary to tuberculosis in some other parts of the body. A few undoubted primary cases of tubal tuberculosis, however, have been recorded. Thus, Macnaughton-Jones 3 records the history of three cases all of which he regards as primary in the tubes; two of these cases were unilateral, and in one both tubes were involved. Muller, 4 Spanton, 5 and Calzolari 6 have also recorded the histories of cases of primary tubercu- losis of the fallopian tubes. In Calzolari's case the disease was appar- ently transmitted by coitus from an infected husband. A negative ophthalmo-reaction was present subsequently to operation. Murphy 7 relates a similar case. Purefoy 8 states that 18 per cent of cases are primary. Our own studies have led us to believe that primary tubercu- losis of the tubes is a rare condition, and that a careful study of the case will nearly always reveal a primary lesion, or the history will point to a previous infection by the tubercle bacillus. In our series there were two cases in which the fallopian tubes were the only demonstrable seat of tuberculosis in the body ; both these patients are well and show no evidence of infection since operation, which in one case took place three, and in the other five, years ago. Great care, however, should be exercised before a case is pronounced as primary, for, as is well known, latent foci may be present in other parts of the body which are undemonstrable by any known means, or the primary lesion may even have undergone a complete resolution. Not a few gynecologists and surgeons even deny the existence of primary genital tuberculosis. A 194 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS sufficient number of cases have, however, been studied at autopsy, and reported, to prove the existence of a primary infection, although it is certainly extremely rare. Furthermore, the existence of a primary in- fection of the genital tract has been demonstrated by animal experi- mentation. (See Chapter V.) The tubes are the portion of the genital tract by far the most fre- quently attacked by tuberculosis. It is generally stated that the tubes are involved in 90 per cent of the cases of genital tuberculosis. In our laboratory, where all operative material is subjected to a routine his- tologic examination, the tubes have been found to be involved in all. It would seem, therefore, that 90 per cent is an under rather than an over estimate. Tuberculosis is nearly always for the genital tract pri- mary in the tubes, and almost invariably secondary to tuberculosis else- where in the body. . Mayer 9 states that of 40 cases of tuberculosis in the abdomen, in 21 the disease was situated in the adnexa, and in 19 was definitely peritoneal in origin. Tuberculous salpingitis constitutes from 4 to 12 per cent of all tubal infections. This proportion varies somewhat in different clinics. In the laboratory of gynecology of the University of Pennsylvania this form of infection was demonstrated in 7.3 per cent of all pelvic infection. Andrews 10 places the proportion at 1 to 3 per cent, Menge 1X at 9 to 10 per cent, Kronig 12 at 7 to 8 per cent, Pankow 13 at 22 per cent, Heynemann 14 at 11.7 per cent. Hurden 15 reports that, of 1,001 cases of salpingitis collected from the Johns Hopkins Hospital Reports, 109 were tuberculous. Williams, 2 from the obstetrical department of the same institute, reports 4 per cent of all cases of salpingitis due to the tubercle bacillus. The ovaries are comparatively rarely the seat of a true tuberculous oophoritis, although peri-oophoritis in the presence of tuberculous sal- pingitis is the rule rather than the exception. In our series of 31 cases, true oophoritis was present in 4 cases, peri-oophoritis in 7, while of the remaining 20, 5 showed well marked retention cysts. It is difficult to account for the normal structure of such a large proportion of ovaries from cases of tuberculous salpingitis, especially when it is considered that the mucosa of the tube is practically always involved. The result is the formation of considerable irritating secretion, which is poured out in the peritoneal cavity, as instanced by the adhesions found about these tubes. Another factor which would seem to favor the infection of the ovaries is that in tuberculous salpingitis, the abdominal ostium exhibits a marked tendency to remain patulous, thereby offering an opening for the escape of the tubal contents, which drip down over the ovaries inter- TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES 195 mittently, often for prolonged periods. Even if the intact surface of the ovary were able to withstand the infection thus brought in contact with its surface, the normal rupturing of graafian follicles would, it might be thought, offer an avenue for infection. Furthermore, the fact that the infecting microorganisms in tuberculous salpingitis are often con- veyed to the tubes by the blood or lymph stream and the close anatomical relationship of the blood supply to the tubes and the ovaries would appear to favor infection of the latter. Practically, however, the ovaries are comparatively rarely infected, much less frequently than in pelvic infec- tion from the gonococcus or other pyogenic organisms. It would seem, therefore, that the ovaries must possess some inherent immunity to this form of infection. The fact that the ovaries are involved in a relatively small proportion of cases is of extreme importance when considering the surgical treatment of tuberculous pelvic inflammatory disease. When actual involvement of the ovarian stroma occurs, it is usually the result of infection gaining access to the ovaries through a ruptured follicle, the lesion often being an abscess of a corpus luteum. Horizontow 16 believes that the stroma and cortical layer of the ovary are most frequently at- tacked. Todorsky 17 has especially emphasized the gravity of ovarian complications, and believes that abscesses and even fistulas not infre- quently follow. As a result of the peri-oophoritis which is so frequently present, re- tention cysts are often an accompaniment, and malposition of the ovaries due to adhesions and secondary edema is not uncommon. Martin has especially called attention to the hypertrophy of the ovaries occurring in these cases. Hypertrophy has not been frequent in our series. Pri- mary tuberculosis of the ovaries is extremely rare, even more so than a similar infection of the tubes. Senni 19 has, however, reported the history of such a case. The question of predisposition towards tuberculosis of the genital tract and especially of the fallopian tubes is a subject which has promoted considerable study of recent years. As the disease is secondary in the large proportion of cases, a primary focus in some other part of the body is perhaps of chief importance. Preexisting inflammation is undoubtedly also a predisposing factor. In tubal infection in general mixed infections are by no means uncommon. It is impossible in some cases to determine whether a gonococcal infection is superimposed upon a tuberculosis, or whether the reverse is the case; most authorities believe that the latter is the more frequent condition, and that once the tubal mucosa is altered by a gonococcal inflammation, an excellent soil for the development of 196 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS tuberculosis is prepared. In the Pathological Laboratory of Gynecology at the University of Pennsylvania 30 cases of tuberculosis of the tubes have been examined by the author. In 20 of these the history or his- tologic appearance of the specimen was sufficiently pronounced to cover this point, 6 of the 20 cases had apparently been preceded by a Neisserian infection. Owing to the fact that the bacteriologic tests have not been carried out in a routine manner in many of these cases, it is, however, im- possible to definitely determine this point. Simmonds 20 was one of the first to point out the relationship between preexisting inflammation and tuberculosis of the tubes. Saulman 21 and Schuchardt 22 have also empha- sized this point. Bandelier and Roepke 23 state that marked redundancy and folding of the plica of the tubal mucosa, stagnation of the tubal secretion, and poor blood supply are also predisposing factors. These latter causes, however, appear to the author to be somewhat theoretic and unproven. Trauma in rare instances may play a predisposing part in tuberculous infection, as it undoubtedly does in other parts of the body. The nor- mal fallopian tubes, however, owing to their protected position, are rarely the subject of wounds or injuries from without. Whether or not heredity plays a predisposing part is difficult to determine. It is, however, doubt- ful. Sellheim 24 and Schiffmann 25 believe that hypoplasia of the genital tract favors the development of tuberculosis. The age of the patient is undoubtedly an important factor, patients of certain ages apparently exhibiting a greater tendency to immunity to this type of infection than do others. Thus, women past the menopause are comparatively rarely attacked by this form of tuberculosis. Chil- dren are by no means immune. Bruning 26 has collected from the litera- ture 44 cases of genital tuberculosis occurring in the young, to which he adds 2 of his own. In the majority of these the tubes were affected. Allaria 27 has analyzed 19 cases, all of which are reported as primary genital infections. Chaffey, 28 Silcock, 29 and Collingworth 30 have also recorded cases of tuberculous adnexitis occurring in children. In our own series of 30 cases, the ages varied from 18 to 41, 2 being under 20 years, 1 5 between 20 and 30, 1 1 between 30 and 40, and 2 be- tween 40 and 50. In Cummins' 31 series of 21 cases the ages were as follows: 1 case between 10 and 15, 2 between 15 and 20, 7 between 20 and 25, 3 between 25 and 30, 4 between 30 and 35, 2 between 35 and 40, 1 between 40 and 45, 1 between 45 and 50. No age is immune. This disease is, however, most frequent during the active sexual life. Symptoms. — The symptoms produced by tuberculous adnexitis are by no means characteristic, and differ to no marked degree from those TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES 197 produced by other microorganisms. Pain, tenderness, sometimes slight enlargement of the lower abdomen, dysmenorrhea or other menstrual disturbances, sterility, dyspareunia, leukorrhea, constipation, nausea, vomiting, and evidences of a local peritonitis with fever, leukocytosis are among the most important. Hegar 32 has divided the disease into two stages, one in which the pelvic organs can be identified by palpa- tion, and one when they are matted together, forming an indistinguish- able mass. For the purpose of study, however, the division into the acute and chronic stages seems more satisfactory. Patel 33 divides tubal lesions into four classes: where tubal lesions are the most prominent; where general peritonitis is the most prominent ; where the ovarian lesion is the most prominent ; where local complications are the most prominent. Under the last heading, Patel mentions peritonitis causing intestinal ob- struction, spontaneous evacuation of an abscess into the intestine, ureter, bladder, vagina, uterus, or through the skin. Murphy 34 states that, unless there- is a mixed infection, there is a strong tendency for the tube to stay open and that, while this condition exists, the course of the dis- ease is similar to that of recurrent appendicitis — a period of relief or even good health followed by a sudden attack of pain, nausea, vomiting, local tenderness, fever and often a discernible effusion in the peritoneal cavity — but that, when the tube is closed, the recurrent type of symptoms is not present. This is undoubtedly correct in theory. Practically, however, it seems probable that the closure of the distal end of the tube is often temporary and that, as a result of a lighting up of the infection, which produces an increase in the intratubal pressure, or of trauma, etc., formerly en- capsulated pus or other secretion within the tube oozes out through the external abdominal ostium and sets up a fresh attack of pelvic peritonitis. The tube may subsequently become walled off or the abdominal ostium again close and result in an amelioration of the subjective symptoms. In other cases the tubal opening may become permanently closed and this, as Murphy states, results in more or less permanent lessening of the symptoms. Evens 35 states that in a definite proportion of cases the previous history shows that there have been obscure attacks of peritonitis during girlhood, and that these are not infrequently followed by amenorrhea. In some cases the pelvic symptoms are preceded by those of a general peritonitis, often of the ascitic variety, which clears up and leaves behind a more or less well marked pelvic inflammatory disease ; or the reverse may be the case. See chapter on General Tuberculous Peritonitis. Acute Stage. — During this stage, the patient exhibits the usual 198 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS symptoms of an acute pelvic peritonitis, the severity of the attacks vary- ing with the individual cases and with the extent of the lesion and the previous duration of the disease. Thus, if the inflammatory processes are entirely walled off from the general peritoneal cavity, the symptoms are less marked than where an inflammatory tube is pouring forth an irritating secretion into the general peritoneal cavity. The subjective symptoms are similar to those of metritis, except that, where the tubes are involved, the pain and tenderness are more marked and are chiefly observed in the ovarian region. As tuberculous salpingitis is usually bilateral, pain is generally complained of on both sides of the uterus, the entire lower abdomen being tender. Bumm 36 and Menge 37 have very properly pointed out that tubal infections in general are more painful than are similar infections of the uterus. In the latter case the pain is often a marked symptom only at the menstrual periods. As a general rule, the onset of the symptoms, resulting from tuberculous adnexitis, is less marked and more insidious than in the ordinary forms of pelvic inflammatory disease. The symptomatology of tuberculous adnexitis is difficult to define, because of the numerous structures which may be involved and which may in themselves produce special symptoms. Thus, if a tuberculous tube becomes adherent to the bladder, vesical irritability and other symp- toms suggestive of a cystitis are likely to occur, whereas, if the tube be plastered against the rectum, painful defecation occurs and as a result constipation frequently follows. Cuturi, 38 as a result of experiments, states that when the bladder is in contact with a diseased tube, the former not infrequently shows a tuberculous cystitis at the point of contact. This may be a localized or a general cystitis. The author has observed two such cases. Unless this complication is borne in mind, the danger of a mistaken diagnosis, and considering the case one of renal tuberculo- sis, is not unlikely. The onset of the disease is frequently marked by a chill, followed by nausea, vomiting, malaise, headache, elevation of temperature, and in- creased pulse rate. The temperature varies during the height of the dis- ease from ioo° to 105 F., 101 or 102. 5 F. being perhaps the average evening rise. A blood count may show a moderate leukocytosis, which, however, is usually lower than in other forms of pelvic inflammatory disease or the white count may be normal. The appetite is lost and the usual symptoms of fever are present. The severity and duration of the attack vary markedly in different cases, and, as in gonococcal infections, the local symptoms are only a moderately reliable indicator of the extent of the disease. When pulmonary tuberculosis is present, the coughing TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES 199 often markedly increases the pelvic pain. In some cases observed by the author this has been a distressing feature of the case. An attempt has been made in the study of our cases to determine if, in tuberculous cases, the tubes were especially prone to be attacked at any particular stage of the menstrual cycle. The data which were obtained showed that cases might be attacked at any time, but that invasion of the tube appeared to be most frequent during the end of the second and the beginning of the third week following the beginning of the menstrual period, in this way differing from the gonococcal cases, in which exten- sion to the tubes is prone to occur at or immediately following menstrua- tion. One striking point brought out by our study was that, in the large proportion of our cases, 55 per cent plus, there was, or had shortly before been, an accentuation of the primary lesion in the lungs or elsewhere, just prior to' the onset of the pelvic symptoms. In many cases the ex- acerbation of the primary lesion was slight, but careful questioning and examination showed that it had been present frequently. The duration of the acute stage is uncertain, but as a rule this period lasts longer and is more resistant to palliative treatment than are the in- fections produced by the ordinary pyogenic organisms. A general peritonitis may either precede or follow the tubal infec- tion. The former class of cases will be considered in a subsequent chap- ter. The possibility of a general involvement of the peritoneal cavity following the tubal infection is a very real one. When the susceptibility of the peritoneum to the action of the tubercle bacillus is taken into con- sideration, and the vast number of tubercle bacilli which are present in the tubal secretion, much of which is being passed out into the peritoneal cavity, it is only remarkable that more cases of general tuberculous peri- tonitis do not result. Some cases run an acute or subacute course from the onset, rapidly developing a general peritonitis, and terminate fatally. No hard and fast rule can be laid down in this respect. As a general rule, however, it would seem that those cases which are depleted as a result of a primary focus of the disease at the time of onset of the pelvic symptoms offer less resistance and are more subject to a general peritonitis or a fatal termination than are those patients in whom the tubal involvement occurs early and who are in good general condition at the time of the beginning of the pelvic infection. Examinations during the acute stage will show more or less disten- tion of the abdomen, but, unless there is a general peritonitis, the enlarge- ment tends to be limited to the lower portion. Tenderness and rigidity are especially marked over the affected areas. Smith 30 has called atten- tion to the behavior of the abdominal cutaneous reflexes in acute condi- 200 GYNECOLOGICAL AXD OBSTETRICAL TUBERCULOSIS tions within the abdomen and pelvis. The reflex is tested by striking the skin over the suspected area with some blunt instrument, often the blunt end of a pencil. Further reference to the subject of abdominal cutaneous reflexes may be found in the works of Pflasterer, 49 Miiller and Seidel- mann, 41 Rosenbach, 42 Van Gehuchten, 43 Striimpell, 44 Bodon, 45 Jamin, 46 Sicard, 47 and Rolleston. 4S A pelvic examination will reveal the uterus either normal in size or slightly enlarged, and induration can be felt in one or usually both va- ginal fornices. The cervix is more or less fixed, and attempts to move it cause pain. This is a valuable diagnostic sign of all varieties of pelvic inflammatory disease. An inflammatory mass, varying according to the extent and character of the lesion, from slight thickening, induration or indistinct sense of resistance, to a tumor the size of a grape fruit or even larger, will be found occupying the region of the appendages. As a rule the lesions are not especially massive, except in advanced cases, when enormous masses composed of the inflammatory adnexa, omentum, intestines, and collections of serum or pus, may be present. During the acute stage, owing to tenderness and tympanites, it is generally impossible accurately to outline the adnexal lesions. Chronic Stage. — The chronic stage can usually be traced to an acute attack, but occasionally the disease is subacute from the onset and fol- lows an almost chronic course from the beginning. Indeed an insidious onset is more frequent in this than in any other variety of pelvic inflam- matory disease. As in the acute stage, the symptoms vary markedly with the individual case. As a rule, to which many exceptions occur, the dis- ease tends to run a prolonged chronic course, interspersed with acute or subacute attacks. Marked exacerbations are thought by some observers to occur only in the presence of mixed infection. The general health is as a rule poor, usually more so, perhaps, as a result of the primary lesion than actually caused by the pelvic trouble, although there is no certainty in this respect. These patients therefore are apt to be thin, losing weight, and often run a slight evening temperature, especially at the menstrual periods. As a result of adhesions, purulent material may be walled off and result in long periods of latency. Secondary anemia of varying severity occurs in over 80 per cent of cases. As has been stated, the symptoms resulting from the pelvic lesions are by no means characteristic of tuberculosis, but are more or less com- mon to pelvic inflammatory disease in general. Menstrual disturbances are usually present and may be either due to an accompanying endometri- tis, or metritis, or to ovarian involvement. Pulmonary tuberculosis in itself, without pelvic involvement, is prone to produce menstrual disturb- TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES 201 ances, a subject which will be considered in detail in a subsequent chap- ter. An analysis of our 30 cases showed that all suffered more or less from menstrual disturbances; in 1 case amenorrhea had been present for three months and menstruation had been scanty and irregular for nine months; 15 showed some tendency towards irregularity and scantiness of flow ; in 14 the flow was normal, or increased in amount ; in 5 the periods were too frequent; in 27 more or less dysmenorrhea was present, and in 21 this was quite a marked feature. Although the character of the dysmenorrhea may vary, it is usually of the congestive type. It gen- erally begins 12 to 48 hours or even more before the appearance of the menstrual flow and becomes less severe after the second or third day. The pain is of a dull, heavy, aching character, is worse over the lower abdomen, and is generally accompanied by backache and malaise. During the dysmenorrhea the general tenderness over the lower abdomen is in- creased. Slight tenderness and enlargement of the inguinal lymphatic glands is sometimes present at this time. The dysmenorrhea in these cases may result from the primary lesion, from congestion of the diseased pelvic organs, especially the endome- trium, may be ovarian in origin, or from a combination of these causes. Barbour and Watson 49 believe the dysmenorrhea is usually ovarian in origin, and is caused by a subalbugineal castration. It is noticeable that in all our cases in which the flow was increased in amount there was either an ovarian involvement or a well defined tuberculous endometritis, or both, showing that salpingitis alone has little or no effect upon the regularity or amount of the menstrual flow. This is in accordance with the findings of Boldt, 50 who states that in tubal disease, when not asso- ciated with ovarian lesions, the menstrual flow is not likely to be changed. In a series of 45 cases of tuberculous salpingitis, Baisch 51 observed men- strual disturbances in 50 per cent. It is probable that in these cases there was some ovarian involvement in the majority of cases. As tuberculous salpingitis is usually bilateral, sterility is usually the result, despite the fact that in more than half the cases at least one tube is patulous. In this connection, however, it is important to remember that the tubercu- lous tube exhibits a remarkable tendency to remain patent, much more so than do tubes affected with any other variety of infection. Pain is usually a more or less pronounced symptom, although Kelly 52 remarks upon the frequent absence of this symptom in children. The pain is usually general over the lower abdomen and is usually intensified at the menstrual periods. Defecation is often painful, especially in those cases in which the appendages are adherent to the rectum. As a result of this pain a constipated habit is often acquired. The accumulation of 202 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS hard feces within the pelvis tends in time to augment the pelvic inflam- mation, and in this manner a vicious circle is established. In some pa- tients the symptoms resulting from the sluggish action of the bowels con- stitute in themselves a marked feature in the case. If the inflamed adnexa lie anterior and are adherent to the bladder, vesical symptoms, such as frequent micturition and dysuria, are more or less pronounced, and unless a pelvic examination is made, the condition may be mistaken for an uncomplicated case of cystitis. Distention of the bladder and emptying of it may also cause pain. Backache, chiefly in the lower lum- bar and sacral regions, is not infrequent, and frontal or occipital head- aches may occur. During the chronic stage fever as a result of the pelvic lesions is often absent, although an evening rise, especially in cases in which there is pulmonary involvement, is very characteristic. Not infre- quently in fairly quiescent cases the rise will be but slight, often not more than a fraction of a degree. A slight evening rise is an extremely suggestive symptom. A slight rise in temperature following a pelvic examination is common to all types of pelvic inflammatory disease and is a valuable diagnostic symptom in those cases in which the pelvic lesions are small or palpation difficult. During the chronic stage there is usually no leukocytosis or only a slight increase above the normal. All the symp- toms are likely to be worse in the afternoon after the patient has been upon her feet, and are ameliorated by rest in bed. The discomfort is increased by exercise and pressure, such as may be produced by tight clothing about the waist or lower abdomen. The vermiform appendix is secondarily involved in a definite propor- tion of cases, and as a result tenderness over McBurney's point is not infrequent, but is perhaps less often present than in gonococcal infections. In our 30 cases of tuberculous salpingitis more or less appendiceal in- volvement was present in 10; in 3 of these tuberculous appendicitis was present, and in 7 peri-appendicitis. As has been mentioned, the severity of constitutional symptoms varies widely in different cases. They are more frequent and pronounced in these than in the gonococcal variety of chronic pelvic inflammatory disease. This is due to the fact that in a large majority of cases the constitutional symptoms are due not alone to the pelvic lesions, but are also often caused by the primary infection in the lungs or elsewhere. The patient is usually more or less incapacitated and tires easily. Usu- ally loss of weight and general ill health are present, although cases vary markedly in this respect. Tenderness over the lower abdomen is often marked, and in severe cases the gait may be almost characteristic, the patient walking slowly, stooping forward, often inclining to one side or TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES 203 the other, a hand being placed over the site of the pain. These patients may be observed to lower themselves carefully into a chair, and are apt to sit stooping forward, often bending towards the side of greatest pain. The same cautious action is observed when the woman arises out of the chair and at all times care is exercised to guard the abdomen from trauma or jolts of any kind, such as getting out of a street car, or going down steps, etc. During the latter maneuver the patient is likely to step down somewhat sideways, one step at a time, in the meantime holding on to the hand rail, somewhat after the manner sometimes adopted by young children. Dyspareunia is usually present, and as a result of prolonged suffering and general ill health, neurasthenia not infrequently results. Abdominal palpation reveals the presence of resistance and tender- ness over the affected areas, and in thin subjects or where the lesions are massive a tumor may be sometimes felt in one or both ovarian regions. Vaginal examinations show induration and tenderness in one or both vaginal fornices. The cervix is more or less fixed and attempts to move it in any direction cause pain in the ovarian regions and along the broad ligaments. The uterus is often in retrodisplacement and adherent, and in those cases where there is a metritis it is enlarged. The tube and ovary are often bound together, forming an indistin- guishable, adherent, tender, inflammatory mass, over which, in cases of large accumulations of fluid, fluctuation may be elicited. Fluctuation is more likely to be noticeable in thin patients, and in those cases in which massive lesions are present. More often fluctuation is absent and the tumor has a hard elastic feel. There may be bulging into one or both vaginal fornices. In some cases the ovary can be palpated as a separate structure, either normal or increased in size. In many cases, however, it cannot be differentiated until the abdomen is opened. As a rule the condition is bilateral, although frequently the pathologic process is more massive on one than on the other side. In our series 28 cases were bilateral, 2 were unilateral, and even the latter, owing to the difficulty in macroscopic diagnosis, are doubtful, as in these 2 cases the tubes appeared entirely normal and were not re- moved, the character of the infection being unsuspected by the surgeon at the time of operation. The longer the duration of the case, and the more acute the symptoms, the more massive are the lesions likely to be. Occasionally small lesions will produce marked symptoms and the con- verse may also occur, especially when the collections are serous in char- acter and the general peritoneal cavity is uninvolved. Occasionally nodules can be felt in the Douglas pouch, and, when present, are very suggestive of this type of infection. The differentiation between puru- 204 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS lent, serous, and hematogenous collections is extremely difficult by pal- pation alone. In purulent cases a slight rise of temperature of a half or one degree, following examination, is significant. On palpation a pyosal- pinx frequently imparts a hard resistant sensation to the examiner's finders, whereas serous collections are more elastic and often less ad- herent. Hydrohematosalpinges give the same general sensation on pal- pation as do simple serous accumulations. The typical retort shape, often assumed by non-purulent tubal accumulations, sometimes acts as a guide in determining the variety of the lesion present. This shape is relatively frequent in cases of tubal tuberculosis, in which form of in- fection the lesions are nearly always most marked in the ampulla of the tube, and not infrequently the inner two thirds of the organ, being com- paratively normal or somewhat drawn out, forms a sort of pedicle. Such tubes may undergo torsion and upon palpation may be mistaken for cys- tic ovarian neoplasms. The rare cases of hematosalpingitis not due to tubal pregnancy impart a soft doughy feel to the examiner's finger. Occasionally, especially in the early stages, the tubes are small and soft, and in these cases the demonstration of salpingitis by means of pal- pation is extremely difficult. Even after the administration of a general anesthetic, this may be almost impossible. Tenderness over the tube and fixation of the ovary are always significant. It is especially in these cases that an accurate history is of great importance in arriving at a correct diagnosis. Diagnosis. — As in most other affections, whether of the genital tract or elsewhere, the correct diagnosis may be either easy or extremely difficult to arrive at. The fact that a pelvic peritonitis is present is usually easily ascertained. The determination of the variety of infec- tion is, however, in many cases more difficult. Not infrequently the pelvic symptoms are more or less masked by those produced by the pri- mary lesion. In some cases an absolute diagnosis is impossible, and a tentative diagnosis, arrived at by exclusion of the ordinary forms of infection, is the best that can be done. As a rule, the onset is more in- sidious than in the other forms of pelvic inflammatory disease, and, as has been stated, these symptoms are often overshadowed by those pro- duced by the primary lesion. Von Franque relates instances in which the first symptom has been sterility, and warns against treating women in general for this symptom, without first excluding this form of infection. In the case of the gonococcal type of infection, the fact that the patient is a married woman or one of loose morals, and the evidence of gonorrhea in the lower genital tract are points which put the examiner on his guard for this variety of infection. In the case of streptococcus or staphylococ- TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES 205 cus infection the fact that these usually follow the emptying of a preg- nant uterus, whether at or before term, or succeed some intra-uterine manipulation, the sudden onset, the high temperature, the severity of attack in general, are suggestive of these organisms. In a definite pro- portion of patients the incumbents of tuberculous salpingitis, none of these symptoms are present, and this fact in itself is very suggestive of a tuberculous infection. A history of pleurisy and susceptibility to bron- chitis is always suggestive of this form of infection. A definite pro- portion of cases is secondary to tuberculous peritonitis or to osseous lesions. Enlarged lymphatic glands in the neck are present in some patients. Indeed the presence of tuberculosis in any other portion of the body is suggestive. In a certain percentage of cases, however, no evidence of the primary focus is present. It should also be borne in mind that tuberculosis is one of the most frequent forms of infection, and be- cause a woman has a tuberculous pulmonary lesion this does not prevent a gonococcal or other variety of pelvic infection. When a salpingitis occurs in a virgin, the chances are largely in favor of its being tubercu- lous in origin, and if, in addition, the disease is bilateral and associated with a demonstrable primary lesion, such as a pulmonary tuberculosis, the diagnosis is almost certain. The existence of a chronic cough should in all cases put the examiner on his guard for this form of salpingitis. Tuberculous salpingitis may occur in young girls and children before menstruation, and although gonococcal vulvovaginitis in rare instances results in ascending infection involving the tubes, it is comparatively rare as compared to tubal lesions in children caused by the tubercle bacillus. The time of onset of the initial symptom of the pelvic trouble is also some aid in determining the type of infection. In the gonococcal cases the spread to the body of the uterus and to the tubes nearly always fol- lows a menstrual period and less frequently the emptying of a pregnant uterus or intra-uterine manipulation, whereas in tuberculosis this is not commonly the case. Furthermore, tuberculosis of the tubes tends to be somewhat less acute and painful as a general rule than does the Neisse- rian infection. The fact that tuberculous salpingitis is distinctly less amenable to local and general treatment, such as copious hot douches, rest in bed, and the regulation of the bowels, is worthy of note and is also a suggestive point. From appendicitis the disease can usually be readily differentiated by its bilateral involvement, the presence of a primary focus, its pain low in the abdomen, the induration of the broad ligament, the presence of adnexal lesions as determined by palpation, the absence of marked tenderness over McBurney's point, and the absence of a history of in- 206 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS discretion in diet, etc. So also the history and finding on examination differ quite markedly from those usually observed in tubal pregnancy. In 4 of ii cases examined, the author has been able to demonstrate tubercle bacilli in the leukorrheal discharge. Cummh.s 53 has reported good results by this method. He is careful to obtain the secretion for examination from the depths of the cervical canal, as an additional safe- guard in eliminating the smegma bacillus. The endometrium is at least partly involved in many cases of advanced tubal tuberculosis. In these cases, therefore, it is only a matter of persistence to find the tubercle bacillus in the discharge. Care must necessarily be observed to exclude the smegma bacillus. Meyer-Rugg 54 is of the opinion that only in ex- ceptional cases are bacilli found in the secretion. Orthmann 55 has been able to demonstrate tubercle bacilli in 42 per cent of cases. Doubtless animal inoculation, if carefully carried out, would prove of value in this connection, but the time required for such diagnostic methods is a dis- tinct drawback, and, as the treatment is likely to be operative, no matter what form of infection is present, this nullifies the value of the method. The examination of the discharge by staining methods is naturally only of value in positive cases, the failure to demonstrate this organism by no means excluding the presence of tuberculosis. Hohne 56 has advocated evacuations of pelvic fluid by puncture and animal inoculation of the material thus obtained. This procedure may be of value in certain cases, but certainly is not advisable as a routine diagnostic method. Undoubt- edly valuable information may be obtained by the examination of such material, when the operation of vaginal incision is indicated from a clinical standpoint. Sellheim 24 recommends the histologic examination of portions of the uterine mucosa for evidence of tuberculosis. Tuberculin has been employed as an aid to the diagnosis. Pankow 57 states that he observed a focal reaction in three cases of non-tuberculous pelvic inflammatory disease, but that in one of these the reaction may have been caused by menstruation. Sahli 58 has emphasized the point that the sensitiveness to tuberculin is increased for a few days prior to menstruation. Beer 59 states that focal reaction in the absence of tuber- culosis is exceptional. Mohr 60 is of the opinion that a negative response excludes tuberculosis, but Beer thinks a general plus and a focal response is practically invariably due to a focal tuberculosis, and such a response locates the diseased area. A general minus focal response is of no prac- tical value, as the most careful examination cannot positively exclude a tuberculous focus in other parts of the body, which may give the general reaction. Tuberculin should not be employed at or near the menstrual period. TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES 207 Differential Diagnosis Between Tuberculous, Gonococcal and Streptococcal Pelvic Inflammatory Disease Tuberculous Gonococcal Streptococcal 1. Often a family history of tuberculosis ; 25 per cent (Lock 61 ). 1. Family history of tuber- culosis incidental. 1. Family history of tuber- culosis incidental. 2. Any age, although most frequent between 20 and 35. The most frequent cause for pelvic inflammatory disease in childhood. 2. Most frequent during active sexual life. Rare at other times. 2. Most frequent during active sexual life. Rare at other times. 3. Rarely primary. Nearly always a primary le- sion elsewhere in the body. The latter may be quiescent or reso- lution may have oc- curred. Close ques- tioning will nearly always elicit history pointing towards tu- berculosis in other parts of the body; a history of lung, in- testinal, peritoneal, bone or joint disease very suggestive. 3. No history of tubercu- losis elsewhere in the body. (In this con- nection it must be re- membered that tuber- culosis is an extreme- ly frequent disease and that persons suf- fering from it are by no means immune to other forms of pelvic infection.) If pres- ent, it is incidental. 3. No history of tubercu- losis elsewhere in the body. If present, it is incidental. 4. General health often impaired as a result of primary lesions. 4. General health good ex- cept as impaired by pelvic lesions. 4. General health good prior to onset of pel- vic infection. 5. Relatively as frequent in the virgin as in those in whom deflo- ration has occurred. 5. Extremely rare in vir- gins. 5. Extremely rare in vir- gins. 6. Onset of pelvic attack often between men- strual periods. 6. Pelvic attack usually follows a menstrual period and less fre- quently the emptying of a pregnant uterus or intra-uterine ma- nipulation. 6. Pelvic attack nearly al- ways follows the emptying of a preg- nant uterus or intra- uterine manipulation of the pregnant or parturient uterus. 7. Generally a gradual in- sidious onset. Previous history pointing to primary lesion in other part of the body. 7. Onset more severe than in the tuberculous. Previous history of leukorrhea, urethritis, and bartholinitis. 7. Onset severe and evolu- tion of symptoms rapid. Often intro- duced by a chill fol- lowed by hyperpyrexia. Usually a history of pregnancy and intra- uterine manipulation. 208 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Tuberculous 8. Temperature not as a rule high, and in ex- ceptional cases may be normal. Slight evening rise of 0.5 or 1 degree. Fever of- ten . continues over long periods. 9. Pulse in proportion to temperature. 10. Respiration often af- fected as a result of the primary lung le- sion. 11. Menstrual disturbances frequently antedate pelvic symptoms. Scanty menstruation or even amenorrhea not infrequent. 12. Sterility frequent. (Dis- ease is usually bilat- eral.) 13. Pelvic pain less. Onset often masked by symptoms of the primary infection. 14. Pain and tenderness usually bilateral. 15. Primary lesion else- where in the body. No bartholinitis, ure- thritis, or cervicitis. 10. The first portion of the genital tract attacked is the tubes, the en- dometrium being sub- sequently invaded. In other words, genital tuberculosis is a de- scending infection, so that bilateral symp- toms antedate leukor- rhea and other symp- toms of endometritis. Gonococcal 8. Fever of ioo°-io2° usually during attack, generally continues 5 to 10 days, and is fol- lowed by a period of normal temperature. 9. Pulse in proportion with or lower than would be expected with the temperature. 10. Respiration in propor- tion with temperature. 11. Menstrual disturbances follow pelvic infec- tion. Flow usually increased. 12. Sterility frequent, but often follows preg- nancy. (The so-called one child sterility.) 13. Pain more marked fea- ture. 14. Often more or less lo- calized to one or other ovarian region and becomes bilateral in later stages. 15. Evidence of gonorrhea in lower genital tract. Leukorrhea always present, usually yel- low and purulent or mucopurulent. 16. Is an ascending infec- tion. First the lower genital tract, then the mucosa of the body of the uterus, and from thence the ad- nexa, so that leukor- rhea and urethritis, etc., antedate the pel- vic symptoms. Streptococcal Hyperpyrexia, I0i°- 105° F. 9. Pulse rapid and often out of proportion with fever. Often of bad quality. 10. Respiration in propor- tion with temperature. 11. Menstrual disturbances, if present, follow pel- vic infection. 12. Sterility relatively in- frequent. Disease chiefly attacks cellular tissue of broad liga- ments. 13. Pain more marked. 14. May be either unilateral or bilateral. 15. No gonorrhea, but evi- dences of recent preg- nancy. Leukorrhea usually present and often thin and watery. 16. Sudden onset and rap- id involvement of ovarian structures. In- fection gains access through cervix or uterus. TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES 209 Tuberculous 17. Tubercle bacilli may be demonstrated in the leukorrhea. (Nega- tive findings do not exclude tuberculosis.) 18. Cervix normal. 19. Uterus normal in size and consistency. Slight symmetrical en- largement is however not rare. 20. Cellulitis not marked. Primary pelvic infec- tion in tube. 21. Nearly always bilateral. 22. Palpable lesions of the tubes are in the. nor- mal location of these organs, unless the tubes have prolapsed into Douglas' pouch, etc. 23. Tubes sometimes nodu- lar, and this charac- teristic may sometimes be demonstrated by bimanual examina- tion. 24. Both ovaries likely to be adherent, but marked enlargement not frequent. 25. A small but demon- strable amount of free fluid in peritoneal cavity often present during height of at- tack. Gonococcal 17. Gonococci may be dem- onstrated. (Nega- tive findings, unless frequent and careful- ly performed, do not exclude gonorrhea, especially in the chronic stage.) Gon- ococci usually readily demonstrated in the acute stage. 18. Cervix, reddened area surrounding the ex- ternal OS. 19. Normal or somewhat enlarged. 20. Cellulitis not marked. Infection chiefly in tube. 21. Often unilateral, es- pecially in the early stages of the disease. 22. Palpable lesions of the tubes are in the nor- mal location of these organs, unless the tubes have prolapsed into Douglas' pouch, etc. 23. Nodular character of tubes less frequent. Often sausage shaped. A small adherent ov- ary may however sim- ulate nodule. 24. One or both ovaries may be adherent. En- largements more fre- quent than in tuber- culosis. 2^. No free demonstrable fluid. Disease chiefly limited to pelvis. Streptococcal 17. Streptococci may b: demonstrated. No tu- bercle bacilli or gon- ococci present. Cervix, softened, patu- lous, and often ex- hibits evidence of re- cent pregnancy. 19. Usually enlarged (sub- involution). 20. Broad ligament chiefly involved. Tubes, if diseased, are second- arily so. May be either unilateral or bilateral. Chief lesions are lower in pelvis than either of the other forms. Base of broad liga- ment nearly alwa;. - thickened and tender and firmer than nor- mal. Cervix fixed and attempt to move it causes marked pain. Tubes not nodular, of- ten soft and edema- tous. 24. One or both ovaries of- ten enlarged and seat of adhesions or ab- scesses. 25. A well marked general peritonitis may result. In other cases no demonstrable free fluid is present m peritoneal cavity. 2io GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Tuberculous 26. May be positive for va- rious tuberculin tests. 27. Leukocytes normal in number or leukocyto- sis not marked (ex- cepting where large collections of pus are present, where mixed infection has oc- curred). Gonococcal 26. May be positive for gon- ococcal complement fixation test. 27. Leukocytosis during acute attack. the 28. A well marked anemia, often as a result of the primary lesion, often present. 29. No organism in the blood of the general circulation. 30. Runs a slow prolonged course. If death oc- curs, it is usually due to tuberculous lesions other than of the genital tract. 31. Resistant to palliative treatment as usually applied to cases of pelvic inflammatory disease. 2,2. In a definite proportion 32. General peritonitis ex of cases results in, or is followed after weeks or months by, a general tuberculous peritonitis. 28. Hemoglobin varying with stage of disease and individual case. Less marked anemia than in tuberculosis. Streptococcal 26. Unless tuberculosis or gonorrhea is present in conjunction with the streptococcic in- fection, the foregoing tests are negative. 27. Well marked leukocyto- sis usually present. 29. No organism in the blood of the general circulation. 30. Acute attack usually lasts from 5-10 days, continuing at varying intervals over period of years. More or less invalidism. Rarely terminating fatally. 31. Palliative treatment nearly always results in at least temporary improvement. tremely rare. 28. Anemia often marked, especially stages. late 29. Blood cultures quently positive. fre- 30. Sharp acute course, usually ending in complete recovery or death. 31. Palliative treatment of- ten eurative. 32. If general peritonitis oc- curs, is of an acute severe type and orcurs during the course of the pelvic infection. Prognosis. — In considering the prognosis it must be remembered that in the great majority of cases the genital infection is secondary. The primary lesion must, therefore, be as thoroughly studied as the pelvic, and is usually of grave importance. Statistics, moreover, unless from a large series of cases, and compiled with extreme care regarding the extent and location of the primary lesion, are apt to be misleading. Indeed so many factors enter into the question of a prognosis that ordi- nary statistics are practically valueless. All the points of the case must be carefully studied and the prognosis based upon the findings in the individual patient. The age of the patient, the duration, course, character, TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES 211 extent, individual disposition, the social and financial standing of the patient, are all points which must be considered, both as regards the course of the primary as well as of the genital condition. Mayer 9 is of the opinion that the presence of high fever before operation is an extremely unfavorable sign. Fever, either the result of a primary or of the pelvic lesion, is undoubtedly an unfavorable sign, and is generally an indication for delay in operative intervention. Of 22 cases of tubal tuberculosis — > in none of which was there a general peritoneal involvement, all of which were operated upon at least 3 years prior to our study, while some of them had been operated upon 12 years ago, and all were traceable — 16 are alive. About 2/3 of this series of cases were ward patients and are therefore presumably unable to follow out an ideal course of postoperative hygienic treatment. Baisch 51 states that, of no cases of tuberculosis of the peritoneum or genital organs occurring at the University of Tubingen during the ten years prior to his report, 40 died within four years after treatment ; there were no recurrences after this period. Five-sixths of the fatal cases died during the first year. Of 55 cases of tuberculous salpingitis, 13 cases were not operated upon; of these 8 died, in 4 the general condi- tion was too grave to warrant operation, and in 4 others pulmonary lesions were advanced. Five improved under expectant treatment, but only 1 was cured. Of 32 patients treated surgically, 9 died, 3 from peritonitis following injury to the rectum, 1 from bronchopneumonia; 5 died after leaving the hospital, one from pulmonary tuberculosis; 13 cases were cured. In 6 new inflammatory tumors appeared. The per- centage of recurrence was highest in those patients in whom only one tube was removed, the other appearing normal at the time of the operation. Evens 35 has reported a series of 23 cases of adnexal tuberculosis, in which conservative operations were performed when possible, and the uterus removed only for special indications. There were 2 postoperative deaths, 1 from postoperative hematemesis and 1 from septic nephritis ; 16 of the remainder were traced and did well; in 3 there was good operative recovery, but they were subsequently lost sight of. One patient died one year after operation from the primary lesion, and one case required a second operation for the removal of a previously conserved tube. Ollivier 62 has recorded the histories of a series of 1 16 cases of genital tuberculosis. Of these there were 9 operative deaths, 8 died later on, 19 were lost sight of and 80 were alive at the time of the report, some as long as 10 years after the operation. Bovis and Olow 63 report the histories of 55 cases. One died shortly 212 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS after operation and 3 others within 18 months; 43 of the series were able to work 1 to 15 years after operation. Mannheimer 64 reported a series of 22 cases operated upon, with 1 death. Twenty of these patients were followed subsequently and it was found 1 had died of pulmonary tuberculosis, and 1 was ill from the same cause. Eight of these 22 pa- tients died within 5 years of pulmonary tuberculosis. Lindquist, at the same meeting, records the results of operation in 20 cases. Of these there were 14 normal recoveries, and 6 left the hospital with fistulas — no operative mortality. Frolich at the same time reports 50 cases, 2 operative deaths and 2 who died subsequently ; 13 were improved, 29 were well, and 4 untraced. Kronig 65 believes the prognosis in genital tuberculosis should be extremely guarded. Geist 66 reports 28 cases with 3 operative deaths. There were 2 deaths subsequently, due to pulmonary tuberculosis. A number of fistulae developed and the average stay of these patients in hospital was 6 weeks; 13 of the patients were discharged from the hospital well, and 12 improved. Schlimpert 67 states that, in 2,173 postmortem examinations upon tuberculous individuals, 73, or 3.5 per cent, had some form of genital involvement. Simmonds, 20 in 6,000 postmortems upon women, found the genital organ involved in 1.33 per cent. In none of Schlimpert' s or Simmonds' subjects was the genital tuberculosis the cause of death, and in only 3 cases did the subjects come to the mortuary from gynecological wards. Although undoubtedly the genital lesions are of secondary im- portance in comparison with the tuberculous foci in other parts of the body, the author's experience does not by any means bear out the result of Schlimpert and Simmonds, as severe and even fatal lesions have been observed by him in a considerably higher proportion than found in the statistics above quoted. This is undoubtedly due to the character of the material from which the observations have been made. Desgouttes and Ollivier 68 believe that intestinal lesions, particularly those of the small intestine, have an especially unfavorable bearing upon the prognosis in cases of tuberculous adnexitis. These authors state that the prognosis depends very largely on the extent to which the intes- tines have become involved. When there are no intestinal adhesions, the operation is comparatively simple and safe. When only the large intestines are involved, all adhesions, both of the pelvic organs and peritoneum, should be freed with the greatest care. When the small intestines are involved, the prognosis becomes less favorable. The operative mortality from the operation per se in properly selected cases of tuberculous salpingitis is not greater than in other chronic tubal infections. The fact that pulmonary lesions are often present does, TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES- 213 however, markedly increase the operative risks. The subject of anesthesia and surgical intervention in general in phthisical individuals will be considered in a subsequent chapter. The most favorable results are ob- tained in those cases in which it is possible to remove the entire intra- peritoneal focus of the infection. All these patients exhibit a tendency to continue subacute or chronic symptoms despite palliative treatment, and the dangers of a subsequent general peritonitis are always present. The tendency to resist palliative treatment is a sign of considerable diagnostic value and is a point which has not been sufficiently emphasized. When an ordinary case of pelvic inflammatory disease is observed which does not show improvement under palliative treatment, a tuber- culous origin should be suspected. It is true that many tuberculous patients do show improvement, but the proportion is smaller than in the commoner varieties of adnexitis. The simple evacuation of a pelvic abscess is much less favorable than when the entire intraperitoneal focus of infection can be removed, and nearly always a more prolonged con- valescence may be expected in cases of tuberculous origin than in those of other forms of pelvic infection. In such cases chronic fistulas are prone to result. Indeed, Hannes 69 is of the opinion that vaginal incision is of little or no value in the case of tuberculous pelvic inflammatory disease. The final outcome of operative cases after leaving the hospital is less favorable in tuberculous cases than in those due to other varieties of microorganisms. First, these patients must face the dangers of the primary lesion, the possibilities of the development of a general or local tuberculous peritonitis, the former being a not infrequent result, as well as the development of new secondary lesions. The dangers of local recur- rence are especially great in those cases in which tubercles are observed in the peritoneal cavity which are not removable at operation, and in those cases in which one tube only is excised. In our laboratory tuberculosis has been demonstrated in every specimen of macroscopically normal tube removed in conjunction with tuberculous salpingitis of the opposite side. It should not be inferred from this that every case is bilateral, but there is evidence to show that this infection is usually bilateral, even when one of the tubes is macroscopically normal, and the leaving of such a tube certainly increases the risks of a local recurrence and also for the develop- ment of a general tuberculous peritonitis. Mayer-Rugg 54 is of the opinion that genital tuberculosis is rarely the developing point for a tuberculous peritonitis. As has been stated, however, in a certain pro- portion of cases studied by the author, a general peritonitis has developed, 214 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS so that we feel that this complication must be considered. Albrecht and Schlimpert 1 found that, in a series of cases of general tuberculous peri- tonitis occurring in women, 12 per cent originated in the genital tract. The possibility of a general miliary tuberculosis developing subsequent to operation must also be considered. In our series of cases this has never developed. A study of the literature of this subject shows that such a complication may take place and is most prone to occur after operating upon acute cases. Excluding the operative mortality, the great majority of fatal cases occur in the first year. Treatment. — The question of the form of treatment to be employed for cases of tuberculosis of the adnexa is still somewhat in doubt. No rule of thumb can be formulated, and each case must be judged indi- vidually. Whether operative intervention or palliative treatment will give the best results can only be decided after studying the particular case. At the risk of repetition, it must be emphasized that these cases are usually secondary, and therefore the condition of the primary lesion is of the utmost importance. All patients, the incumbents of pelvic inflammatory disease in which a tuberculous origin is suspected, should be subjected to an extremely rigid physical examination, in which the entire body should be carefully studied. According to von Franque, renal tuberculosis is comparatively infrequent as an accompaniment of tuberculous salpingitis; however, this complication was present in one of our cases, and we believe the kidneys should be carefully investigated in all cases, and that a cystoscopic exam- ination is indicated in all ; and, should any doubt exist, the ureters should be catheterized. Fortunately, renal tuberculosis in combination with genital lesions is less frequent in women than in men. Although primary cases of tuberculous pelvic inflammatory disease do occur, they are so rare and the difficulty in making such a diagnosis is so great that for practical purposes it is safe to regard all cases as secondary, and so treat them. As actually observed, cases of tuberculous adnexitis may be divided into three (3) classes — 1st, those in which there is an active primary lesion ; 2nd, those in which there is a demonstrable but non-active primary lesion ; and 3rd, those in which no primary lesion can be diagnosed with certainty. Class 1 comprises those cases in which there is an active primary lesion, and should not, as a rule, be submitted to operation, the exception being patients in whom some palliative operation is performed to relieve pain or other symptoms, such, for example, as the vaginal incision for the evacuation of pus in a large pelvic abscess. Extensive operations should certainly never be performed. Under proper hygienic and other treatment, the primary lesion may improve and the TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES 215 case may eventually come under the heading of class 2. Operations during the acute stage of the primary lesions are doubly hazardous, because of the dangers of dissemination of the infection by the actual operative procedure and the lessened resistance exhibited by those patients. General anesthetics are contra-indicated in the presence of active pulmon- ary lesions. Spinal anesthesia may be necessary in some of these cases. Minor and even certain major operations may in some cases be performed under local anesthesia, but the general rule to be adopted in cases in which there is an active primary lesion is non-operative interference, and this is particularly true when the primary lesion is in the lung. Proper hygienic and medicinal treatment is the course to be recommended for this class of patients, together with appropriate measures indicated, directed towards the pelvic condition. Class 2 comprises those cases in which the greatest difficulty will be encountered in deciding the best form of treatment. Here each case must be carefully studied individually. It must be remembered that the operative risks in these patients is much greater than in the ordinary patient. The chief points to be considered are the extent and character of the primary lesion and the actual danger to the patient from the genital lesion, the amount of discomfort produced by the latter, and the type of operation required to alleviate or cure the disease of the genitalia. As has been elsewhere stated, the subject of anesthesia in tuberculous patients will be considered in detail in a subsequent chapter. It is ob- viously a very important one in these cases. Our experience has been that ether anesthesia for patients with moderately small non-active pulmonary lesions has not proved exceptionally hazardous. However, this danger is a real one and must be considered. Certainly all cases belonging to this class should be carefully studied for a considerable period of time before operative intervention is decided upon, the exception to this being when the operation required is of a life saving character or can be performed under local anesthesia. Class 3 will generally be treated as ordinary pelvic inflammatory disease and the diagnosis of the tuberculous origin of the condition will often only be made after the abdomen is opened, or in the laboratory when the histologic specimens are examined. All patients belonging to this class, in whom tuberculosis is suspected, as for example, if pelvic inflammatory disease be diagnosed in a virgin, should be treated as if they belonged to class 2, and every precaution to prevent an exacerbation of a possibly existing primary lesion should be adopted. Little has been said regarding spontaneous cure of tuberculous sal- 216 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS pingitis. This undoubtedly occasionally occurs, but is rare, although partial resolution is not infrequent. Beyond question, many tuberculous women suffer from mild adnexal lesions, which do not require or receive any local treatment, and which subsequently undergo partial resolution. This is amply proven by a study of postmortem material from tuberculous women in whom some 7 per cent show evidence of salpingitis. It is true some of these may not have been of tuberculous origin, but even in series of autopsies verified by histologic or bacteriologic examination, a definite proportion of tuberculous tubal lesions is found in women in whom they were unsuspected during life, this showing that in these cases the symptoms of the pelvic disease were either masked by those of the primary lesion, or were of such a mild character that attention was not directed to the pelvic condition. At present doubtless far more can be accomplished by hygienic and general measures than in the past, and it is of the utmost importance that all tuberculous patients should receive a long course of postoperative care and observe the usual rules for tuber- culous patients. This applies to all classes and is usually best carried out in a sanatorium. Out door life, forced feeding, etc., are, generally speak- ing, of utmost importance, and are quite if not more beneficial to the patient who has been suffering from a tuberculous pelvic inflammatory disease than is the operation. A preliminary treatment of this character prior to the operation should also be advised in the majority of cases, certainly in all cases in which the pelvic lesions are not materially depleting the strength of the patient. Rollier 70 has treated 700 patients suffering from various forms of sur- gical tuberculosis by exposure to the sun's ray at Leysin, Sweden, during the past 9 years, and is convinced of the benefits to be derived from this form of treatment. He keeps his patient in the open air practically all the year. As in pelvic lesions, the result of microorganism other than tubercle bacillus, operations should be avoided during the active stage of the disease, and in all cases the patient should be subjected to a course of preliminary local treatment similar to that now generally adopted for non-tuberculous pelvic inflammatory disease. In cases of pelvic inflam- matory disease of tuberculous origin it is sometimes difficult to determine if fever is being continued by the primary or by the pelvic lesion. Prac- tically this is of no great importance, as patients should not as a rule be subjected to operation in whom hyperpyrexia is present. The local treatment consists of rest in bed, preferably in the Fowler position, the regulation of the bowels by means of mild cathartics or enemata, the application of heat to the lower abdomen, and the employment of frequent, TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES 217 copious, hot vaginal irrigation. Hofmeier, 71 Bumm, 36 and Freund 72 advise cold during the acute stage. The application of heat to the lower abdomen, together with copious, hot douches is of great benefit and tends to promote absorption. Heat may be applied by means of hot sand bags, the weight being regulated to suit the comfort of the individual patient, or large hot poultices, rubber coils containing hot water, or a hot water bag may be employed. An electrically heated pad, such as can be procured in instrument supply stores, is the best means of applying heat to the abdomen. In any case the heat should be applied as constantly as possible, a temperature of no° to 120 F. being maintained. A good working rule in this respect is to have the application as hot as can be comfortably borne by the patient. Under this form of treatment, combined with proper feeding and hygiene, many cases will improve, and in a small proportion no operative treatment will be required. In the presence of extensive pelvic lesions of doubtful origin it has been, however, the author's experience that cases O'f tuberculous adnexitis are less susceptible to palliative treatment than any other form of pelvic inflammatory disease. Indeed, in some cases in which the primary lesions have been quiescent and difficult to detect, the continuance of the acute symptoms, as indicated by fever, increased pulse rate, pelvic pain, etc., after a moderate trial of the palliative treatment, has been the first symptom which has suggested the correct diagnosis of the cause of the condition. Findley is a strong advocate of non-operative measures in the majority of cases. He states that in many cases the symptom complex complained of is often due not to the pelvic lesion, but to the primary lesion, and that tuberculosis of the genital organs in itself rarely causes death. Dysmenor- rhea is frequently the result of pulmonary tuberculosis. The general peritonitis which sometimes follows in these cases is not necessarily sec- ondary to the salpingitis, but may result from a hematogenous infection from the primary focus in the lungs or elsewhere. Findley emphasizes the facts that operation may awaken a latent primary focus and result in a general dissemination of the disease, that the mortality is relatively high, that there is at least some tendency towards self limiting of the genital lesions, so that when operation is necessary it should be as con- servative as possible, especially in young patients. Patel and Ollivier 74 have reported the results of operations on 121 patients, all of whom were operated upon since 1900. In their series the abdominal route was more satisfactory than was the vaginal. As the result of their obser- vations, they believe it unwise to save a uterus, if the ablation of the adnexa is necessary, and, on the whole, favor hysterectomy with bilateral 218 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS salpingo-oophorectomy in the majority of cases. They think that the artificial menopause induced by the removal of both ovaries is less severe in tuberculous than in ordinary patients. Berkley's 75 studies offer further evidence showing the advantages to be derived from hysterectomy in these cases. In his series of tuberculous adnexitis cases the uterus was involved in 29 per cent, and von Franque 76 found a similar proportion. At the German Congress held in Munich in 191 1 the majority, including Zweifel, Stockel, Gottschalk, Menge, Kiistner, Fehling, Opitz, Wertheim, Startz, and Sarvey, supported operative treatment for the majority of cases. Sellheim and Herff were less favorable to operation. Sippel recommended operation when conservative treatment failed. Nearly all warned against operation during the acute stage of the pelvic lesion, and urged that a thorough search for, and study of, the primary lesion be made prior to deciding upon the course of treatment. Von Franque believes that the majority of cases of genital tubercu- losis should be operated upon, not because there is any imminent danger of life, except in exceptional cases, but following the principle which applies to all forms of surgical tuberculosis, in which it is good surgery to remove the infected focus as far as is possible. Medical treatment, he believes, is slow, uncertain, and costly, and since the chief function of the genital organs is lost at any rate, it seems rational to remove them in order to prevent a further spread of the process. In 66 per cent of the cases permanent cures can be obtained. If the pelvic lesions are extensive, von Franque recommends a radical operation, removing the uterus and both adnexa; but if only the tubes are macroscopically involved, a bilateral salpingectomy is performed, leaving the uterus and ovaries. Murphy 7 also recommends operation in all cases of tuberculous salpingitis when the general condition does not contra-indicate it. He advises sparing the ovaries when possible, and stigmatizes the routine removal of the uterus as a pernicious practice. Patel 33 states that patients with tuberculous salpingitis as a rule do badly if not operated upon. Operative Treatment. — Presuming that operative interference has been decided upon, the type of operation to be performed is the next point to be considered. Shall the operation be conservative or shall the entire uterus and adnexa be removed. Much depends upon this point and many factors enter into the problem. Like similar treatment in other forms of pelvic inflammatory disease, no hard and fast rules can be formulated regarding this point. Our first object is to cure the patient. It is im- portant to consider what structures within the pelvis are diseased. Statis- tics have shown that in over 90 per cent of cases of pelvic inflammatory disease, the tubes are involved. In the series of 30 cases from which our TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES 219 studies have been made, this proportion reached 100 per cent. In the great majority of cases the involvement was bilateral, and even when one tube appeared macroscopically normal or only showed a few adhesions, histologic examinations usually revealed a more or less well marked invasion. Observers have shown that the endometrium is involved in about 20 to 30 per cent of cases. Extensive involvement of the myometrium is comparatively rare. On the other hand, true ovarian involvement is rather infrequent, although peri-oophoritis is more com- mon. Our custom is to conserve the uterus and one or both ovaries when possible. In common with most American gynecologists, we favor the abdom- inal route when operating upon cases of pelvic inflammatory disease, the single exception to this being in those cases in which pus can be evacuated without traversing the peritoneal cavity, as in the case of an abscess pointing into the vagina. In the case of tuberculosis the abdom- inal route is especially to be desired, as a close inspection of the pelvic organs is of the utmost importance. Minto 77 has performed a series of animal experiments, which he believes shows that oophorectomy is advisable. In these tests control animals in which the ovaries were not removed succumbed in all cases earlier than did those in whom oophorectomy was performed. Interesting as this series of experiments was, we do not believe that the results are analogous, or should be applied to the treatment of women. The author has elsewhere 78 stated at length the advantages of ovarian conservation when these structures are not hopelessly diseased. The fact that, in tuberculous pelvic inflammatory diseases, the ovaries are rarely actually invaded by the tubercle bacillus is added reason for this conservatism. Much will naturally depend upon the age of the patient and other circum- stances surrounding the individual case. The case is, however, exceptional where at least one ovary cannot be safely saved. Whether one or both tubes should be removed is often difficult to determine. Many factors, however, point to the advisability of bilateral salpingectomy as the routine procedure. With both tubes macroscopically diseased, even if one shows nothing more than adhesions, we believe that both should be removed. When one tube is diseased and the other is macroscopically normal, histologic examination of the latter often shows it to be the seat of a salpingitis. For this reason, a general radical attitude regarding the routing removal of both tubes in cases of tuberculous pelvic inflammatory disease is, we believe, to be encouraged. The only advan- tage in tubal conservation is to prevent sterilization. The advantages of fertility are less urgent in the tuberculous than in other forms of infec- 220 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS tion. Tuberculous endometritis is an accompaniment of tuberculous salpingitis in about 20 or 30 per cent of cases. The disease exhibits a tendency to especially involve the endometrium in the cornu of the uterus and to limit itself to areas about the internal os. The fact that in 1 in every 4 or 6 cases there is endometrium involvement is a strong argument in favor of a routine supravaginal hysterectomy. On the other hand, the operative mortality due to the prolongation of the operation, and the actual severity of the procedure is somewhat greater in hysterectomy than in bilateral salpingectomy. After the abdomen has been opened the same operative indications should govern the surgeon as in the ordinary inflammatory case, with the exception perhaps that greater radicalism as regards surgery of the tubes is indicated in this form of infection. The patient is sterilized by the removal of the tubes, but better conservative surgery can be performed by leaving the uterus, as the ovarian blood supply is less likely to be impaired. This is an extremely important factor in ovarian conservation; so, unless the uterus is macro- scopically diseased, its conservation is advisable for this as well as for the preservation of the menstrual function. The sterilization of these patients in those cases in which the disease has not already accomplished this result is, as a general rule, less of a calamity than in the ordinary case of pelvic inflammatory disease, as pregnancy in the tuberculous patient is unadvisable in most cases and often extremely detrimental to the general health of the individual. When it is necessary to remove the uterus, a supravaginal hysterectomy is preferable to a panhysterec- tomy provided that the cervix is uninvolved. In a previous chapter the rarity of cervical tuberculosis has been shown. Result of Operative Treatment. — The tendency towards the forma- tion of fistulae of divergent varieties is greater in tuberculous than in non-tuberculous patients, and for this reason drainage should rarely be employed. The immediate operative mortality is not particularly great in properly selected cases, but the end results are less favorable, often owing to the development of other secondary lesions or to the lighting up of the primary focus. LITERATURE 1. Albrecht und Schlimpert. Quoted by von Franque. No. 76. 2. Williams, J. W. J. Hopk. Hosp. Rep. 1893. 3 '&7- 3. Jones, H. M. Proc. Roy. Soc. Med., Sec. Gyn. 1907. 1 1177. 4. Muller. Deutsch. Med. Woch. Aug. 18, 1909. TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES 221 5. Spanton. Brit. Med. Jr. 1884. 1:881. 6. Calzolari, M. Arch, di ost. e gin. Jan. 1908. 7. Murphy, J. B. Tuberculosis of the Female Genitalia and Peritoneum. Chicago, 1903. 8. Purefoy, R. D. Med. Press and Circ. 1908. 136:399. 9. Mayer. Gyn. Rundsch. 5 : No. 19. 10. Andrews. Quoted by Gilmore in Am. Jr. Obst. 1910. p. 592. 11. Menge. Centrbl. f. Gyn. 1894. p. 24. 1895. 19799. 12. Kronig. Bakterien der Weiblichen Canals. Leipzig, 1897. 1 :264. 13. Pankow. Quoted by de Bovis. No. 63. 14. Heynemann. Ztschr. f. Gebh. und Gyn. 1912. 70: No. 3. 15. Hurden, E. Gynecology and Abdominal Surgery. Philadelphia, 1907. 1 :i68. 16. Horizontow, N. I. Ztschr. f. Gyn. u. Urol. 191 1. 30: No. 52. 17. Todorsky, O. These de Montpellier. 19 13. 18. Martin. Soc. de sc. Med. de Montpellier. Jan., 1909. 19. Senni. Gazz. degli osp. March, 1904. 20. Simmonds. Arch. f. Klin. Med. 1886. v. 38. 21. Saulman. Centrbl. f. Gyn. 1892. 16:335. 22. Schuchardt. Arch. f. Klin. Chir. 1892. 44:448. 23. Bandelier, B. and Roepke, O. A Clinical System of Tubercu- losis. London. 1913. 24. Sellheim. Centrbl. f. Gyn. 1902. No. 43. 25. Schiffmann, J. Arch. f. Gyn. 1914. 103: No. 1. 26. Bruning, H. Monschr. f. Gebh. u. Gyn. 1902. 16:144. 27. Allaria. Arch. ped. 1904. 21:710. 28. Chaffey, W. C. Tr. London Path. Soc. 1885. ^6:303. 29. Silcock. Tr. London Path. Soc. 1885. 36. 30. Collingworth. Tr. London Path. Soc. 1885. 36. 31. Cummins, H. H. Am. Jr. Obst. 19 14. 69:44. 2,2. Hegar. Deutsch. Med. Woch. 1897. No. 45. 33. Patel, M. Rev. de Gyn. Aug., 191 2. 34. Murphy, J. B. Jr. Am. Med. A. 1912. 58:13. 35. Evens, W. Jr. Obst. Gyn. Brit. Emp. 1913. 24:241. 36. Bumm. Ther. d. Geg. 1909. p. 51. 37. Menge, K. Handb. d. Geschlkr. 191 o. 38. Cuturi. Ann. des mal. des org. gen-ur. Sept. 1, 19 10. 39. Smith, R. R. Surg. Gyn. Obst. 1914. 19:504- Also, Jr. Mich. Med. Soc. Dec, 191 3. 40. Pflasterer. Quoted by Geigel in Deutsch. Med. Woch. 1892. p. 166. }) » 222 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS 41. Muller und Seidelmann. Munch. Med. Woch. 1905. No. 28. 42. Rosenbach. Rev. neur. 1904. p. 58. 43. Van Gehuchten, A. La nevraxe. 1900. p. 249. ^4. Strumpell. Text Book of Medicine. 1901. 45. Bodon. Zentrbl. f. d. Ges. Chir. 1898. No. 5. 46. Jamin. Deutsch. Med. Woch. 1904. p. 1088. 47. Sicard. Presse Med. 1905. No. 19. 48. Rolleston. The Brain. 1906. 1 199. 49. Barbour, A. H. T. and Watson, B. P. Jr. Obst. Gyn. Brit. Emp. 1911. 20:105. 50. Boldt, H. J. Jr. Am. Med. A. 1912. p. 101. 51. Baisch, K. Arch. f. Gyn. 1908. 84:345. 52. Kelly, H. A. Gynecology and Abdominal Surgery. New York, 1907. 1 :824. 53. Cummins, H. H. Phys. and Surg. 1912. 34:202. 54. Meyer-Rugg, H. Schweiz. Rundsch. Med. 1914. 14:525. 55. Orthmann. Deutsch. Kong. d. Gyn. u. Obst. Miinchen, 191 1. 56. Hohne. Monschr. f. Gebh. u. Gyn. 1912. 35:48. 57. Pankow. Zentlbl. f. Gyn. 1907. 58. Sahli. Tuberkulin Behandlung. 1910. 59. Beer, E. N. Y. Med. Rec. 84:652. 60. Mohr. Munch. Med. Woch. 1906. 61. Lock, N. F. Jr. Obst. Gyn. Brit. Emp. 1912. 22:1. 62. Ollivier. These de Lyon. 191 1. 63. Bovis, E., et Olow, J. Tr. North. Cong. Goteborg. July, 1916. 64. Mannheimer. Tr. North. Cong. Goteborg. July, 19 16. 65. Kronig. 14th Conv. Deutsche Gesel. f. Gyn. 66. Geist, S. H. Interst. Med. Jr. 1916. 23:1043. 67. Schlimpert. Quoted by Lewis in Springfield Med. Jr. 1912. 21 1291. 68. Desgouttes, C. L., et Ollivier, R. Lyon Med. 1913. 120: No. 11. 69. Hammes, W. Ergebn. d. Chir. u. Orth. 1913. 6:609. 70. Rollier. Monschr. f. Kindhk. 1913. 11: No. 8. 71. Hofmeier. Deutsch. Med. Woch. 1909. p. 2249. 72. Freund, H. Ther. Montschr. 1911. 25:157. 72>- Findley, P. Med. Her. 1913. 32:181. 74. Patel, M., et Ollivier. Rev. de Gyn. 1913. 20: No. 1. 75. Berkley, C. Jr. Obst. Gyn. Brit. Emp. 1913. 3:34. TUBERCULOSIS OF THE FALLOPIAN TUBES AND OVARIES 223 76. Von Franque. Pathologie unci Therapie der Genital-Tuberc. des Weibs. In Wurtzb. Abh. a. d. Gesgeb. d. Prakt. Med. 19 13. No. 45. yy. Minto. Gin. Mod. Dec, 1910. 78. Norris, C. C. Gonorrhea in Women. Philadelphia, 191 3. Saunders. CHAPTER X UNUSUAL MANIFESTATIONS AND REMOTE COMPLICATIONS Tuberculosis and neoplasms — Ways of occurrence — Etiologic relation to cancer — His- tologic similarity of certain forms of tuberculous salpingitis to carcinoma of fallopian tube — Types — Cases recorded — Tuberculosis and non-malignant tumors of the genital tract — Accidental or coincidental combinations — Pseudoneoplasms — Etiology — Infection of adenomyomata of uterus — Cases — Ovarian cysts — His- tology — Summary — Tuberculosis of uterus causing pyometra — Illustration — Tu- berculous tubal lesions — Torsion of tuberculous pyosalpinges — Factors — Action of diaphragm in cases — Rupture of tuberculous pyosalpinges — Collected statistics — Rupture of pyosalpinx in adjacent hollow viscera — Necessity for thorough pelvic examination — Extension of tuberculosis from pelvic lesion to other distinct areas — Tuberculous lesions in hernial sacs — Histologic study — Cases cited — Bibliography. TUBERCULOSIS AND NEOPLASMS A combination of tuberculosis and neoplasms of the genital tract may occur in one of two ways. A tuberculosis may be implanted upon a genital tract already the seat of a neoplasm and involve the tumor either on the surface, or, less frequently, in the substance of the new growth. The reverse may occur, that is, a neoplasm may develop from the genital tract already the seat of a tuberculosis. The pathological process resulting from either of these combinations may be identical, as far as the macro- scopic and microscopic examination is concerned. In other words, a combination of a tuberculosis and a neoplasm may be purely an accident, the one having no relation to the etiology of the other. On the other hand, if, for example, it is found that the coexistence of cancer of the fallopian tubes and tuberculosis is more frequent than would occur from a mere accident from these two conditions, another explanation must be sought for. From a study of a large series of cases and of the literature bearing upon this subject, it would appear that, as far as cancer of the fallopian tubes is concerned a preexisting chronic inflammation, such as is produced by tuberculosis, bears at least some etiologic relation to the occurrence of cancer; and this is what would be expected, if the Ribbert theory of preexisting irritation, lessened resistance from preexisting inflammation, etc., is taken into consideration. When, however, a tuber- 224 UNUSUAL MANIFESTATIONS AND REMOTE COMPLICATIONS 225 culosis of, let us say, the endometrium and tubes exists in a patient in combination with a uterine myoma, the accidental occurrence of the two conditions is probably the explanation. For, although little is known regarding the etiology of uterine myoma, the concurrences of these two conditions are of no greater frequency than would be expected from the incidence of these lesions. Given, however, a tuberculosis of the body of the uterus and a cervical carcinoma, the occurrence of these two condi- tions is less clear, and while still the theory of accidental occurrence of a carcinoma upon a tuberculous uterus is the most probable, the constant irritation to the cervical mucosa resulting from the discharge incident to the preexisting endometritis may in some degree be an etiologic factor. Tuberculosis and Carcinoma. — Harris, 1 Oertel, 2 Wolf, 3 Schwalbe, 4 Cone, 5 Pepper and Edsall, 6 and many others have recorded the existence of tuberculosis and cancer in organs other than the genital tract. The researches of Levin 7 to some extent bear out the Ribbert theory regard- ing the etiology of carcinoma. This author showed experimentally that healthy testes of the rat withstood implantation of the Flexner-Jobling tumor, but that when certain irritants were primarily applied a "take'' was almost constant. Kellert 8 states that certain observers have con- cluded that tuberculosis and its toxins are more or less directly to be considered in the etiology of cancers. Dixon, Smith and Fox 9 have apparently proved in animals that under certain conditions the tubercle bacillus and its products may stimulate epithelium to abnormal growth. As regards the combination of tuberculosis and carcinoma in the genital tract, it seems safe to assume that they may occur accidentally in the same patient, or the inflammation may be, at least to some extent, a causative factor in the production of the cancer ; the latter is especially likely to be the case when the neoplasm develops directly upon the tuber- culous process, as in the case of tuberculosis and carcinoma of the fal- lopian tubes, and much less probable when the two occur at distant parts of the genital tract. An example of both these types occurring in one patient has been reported by Lipschutz. 10 The patient was a multipara, forty-four years of age, in fairly good general condition. She suffered from pain in the back and lower abdomen. The uterus was irregularly enlarged, in retro- position and adherent. A diagnosis of myoma with adhesions was made. Supravaginal hysteromyomectomy and bilateral salpingo-oophorectomy were performed. The uterus was as large as a man's fist and contained a number of intramural myomata. The right fallopian tube at the ampulla passed into a tumor the size of a hazel nut. Histologic examina- tion of this showed it to contain typical tubercles, and sections from the 226 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS center of the tumor revealed the presence of a papillary carcinoma. Lip- schutz believes that the carcinoma in this case developed on an old tuberculous salpingitis. There was no recurrence five years afterward. Von Franque 1X has also recorded the history of a case in which a cancer developed on an old tuberculous lesion. In studying 16 cases of the coincident occurrence of cancer and tuberculosis in the genital tract which he has collected from the literature, this author states that in none was there positive evidence that the carcinoma preceded the tuberculosis, whereas in 9 the reverse was known to be the case. In only 1 of these specimens did the cancer actually develop in a tuberculous process, in 7 the two were closely adjacent, and in 5 they were some distance apart. The fact that carcinoma is an acute condition and that tuberculosis is essentially a chronic one, would, even apart from any acceptation of Rib- bert's theory, to a large extent explain the preexistence of tuberculosis in many of these cases. Similar cases are recorded by L'Esperance, 12 Devic, 13 Kaufmann and Wallart, 14 Lady Barret, 15 Maikoff, 16 Glockner, 17 and others. Von Franque n has recorded the history of another case, in which a carcinoma developed in a fallopian tube, the seat of an old salpingitis. In this specimen, however, the tumor originated from a point in the mucosa apparently free from the tuberculous process. D'Halluin and Delral 18 have reported the history of an interesting case in which the uterus and adnexa were fused into an inflammatory mass the size of a man's fist. The fundus of the uterus was the seat of an adenocarcinoma and was surrounded by a tuberculous endometritis. The authors believe that the cancer developed from tuberculous granu- lations. Nassauer 19 has reported the history of two cases in which tuberculosis of the endometrium coexisted with carcinoma of the cervix, and Wallart 20 has described a case of carcinoma of the cervix coexistent with a similar infection. In the chapter on Pathology attention has been called to the histologic similarity of certain forms of tuberculous salpingitis to carcinoma of the fallopian tube. Especial care should be exercised in histologically differentiating these conditions. Tuberculosis and Non-Malignant Tumors of the Genital Tract. — Combinations of tuberculosis and non-malignant neoplasms of the genital tract are by no means infrequent, and for the most part should be viewed as accidental or coincident combinations. In our series of cases of tuberculosis of the genital tract this infection has been present twice in conjunction with ovarian neoplasms, and once with a uterine myoma. Kelly, 21 in the examination of 1,800 uterine myomata, observed UNUSUAL MANIFESTATIONS AND REMOTE COMPLICATIONS 227 one case in which there was tuberculous endometritis, and another in which the adnexa were tuberculous. The frequency of these combinations is of importance from a clinical viewpoint, as in most cases the predomi- nance of symptoms and the findings on palpation point to a diagnosis of the tumor, and as a result the tuberculosis may be overlooked, unless care is exercised. A careful histologic examination is therefore indi- cated, for, as pointed out in a previous chapter, special postoperative treatment is indicated to all tuberculous patients. In a smaller proportion of cases the symptoms resulting from the tuberculosis will be found to mask those produced by the tumor. Pseudoneoplasms are frequent in certain forms of tuberculosis, as in salpingitis ischmaia nodosa or in some of the hypertrophic forms of this infection, such as are sometimes observed in the cervix, vagina, or external genitalia. In tuberculous peritonitis pseudo tumors are of frequent occurrence. Tuberculosis of the lower genital tract from the internal os downwards frequently pro- duces lesions which, upon clinical examination alone, closely simulate true neoplasms. The ulcerative lesions of the cervix, vagina, and external genitalia are especially likely to be mistaken for carcinomata, and the hypertrophic forms may easily be mistaken for other tumors. As has been stated, tumors of the genital tract may be accidental, and may develop either primarily or secondarily to the tuberculosis. They may spring from the area attacked by the tuberculosis or may arise from a distant and uninfected area. The tumor actually attacked by the tuberculous process may be invaded in one of two ways, either the surface of the tumor may be involved, or the actual substance of the neoplasms may be infected. The former is much the most frequent and is apt to occur when a tuberculous peritonitis or even only a salpingitis is present, in combination with any intraperitoneal pelvic tumor. The ordinary glandular ovarian cyst seems especially subject to this form of tuberculosis. On the other hand, and of less frequent occurrence, the substance of the tumor may be actually invaded by the tuberculous process. The etiology of this latter form of infection probably occurs in two ways, from without or by a blood or lymphatic infection from within. In the case of ovarian cystadenomata the infection from without is the most frequent, either as a direct extension from the capsule of the tumor to the underlying stroma, or the infection finds an avenue of ingress through a ruptured follicle, as more or less normally rupturing graafian follicles may occur in these tumors. In the case of infection of an adenomyoma of the uterus, especially of the diffuse variety, a direct extension by continuity from a tuberculous endometritis is probably the avenue of contamination in the great majority of cases. 228 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Ivins 22 reports the history of a case of an adenomyoma of the fallopian tube attacked by tuberculosis. In this case the uterine end of the tube was the seat of a firm, smooth nodule the size of a hazel nut, which upon histologic examination revealed the above condition. In the examination of tubal adenomyoma combined with tuberculosis care must be exercised, lest the nodules sometimes produced by a simple tuberculous salpingitis be confused with a true new growth. Miss Ivins believes her case to have been one of a true tumor combined with a tuberculosis. Von Franque, 11 and Parsons and Glendining 23 have reported the his- tories of cases in which the specimens closely resembled true adenomata, but in which the tumor-like formation was probably the result of tuber- culosis, and not of a new growth. Schutze 24 has described a rare speci- men, in which the cervix was the seat of an adenocarcinoma; a diffuse adenomyoma of the uterus was present, and distributed more or less dif- fusely throughout the latter tumor and especially involving its connective tissue were many typical tubercles. Many psammoma-like bodies were present in the wall of the uterus. Dickson 25 has observed a specimen of uterine adenomyoma invaded by tuberculosis. Multiple myomata were present and all but one of the tumors presented a number of cheesy necrotic areas, the largest of these having a diameter of 5 cm. Two small subperitoneal tumors were converted into white, necrotic material with a consistency of putty. Tuberculosis of the tubes and endometrium was also present. Kelly 26 remarks upon the extreme rarity with which tuber- culosis is found complicating large myomatous uteri. Heinrich, 27 and Violet and Perrin 28 have reported the histories of such cases. Grun- baum 29 has described a case of a large uterine adenomyoma, in which the tumor tissue was permeated with small tubercles. The myometrium con- tained many tuberculous foci undergoing cheesy degeneration. Tubercu- losis of the endometrium and lungs was also present. Grunbaum believes the infection was a hemogenic one to the endometrium and from thence by direct extensions to the tumor. Archambault and Pearce 30 report the history of a case in which an adenomyoma of the uterus showed typical tubercles. One tube was the seat of a tuberculous salpingitis, the other tube and the endometrium were normal. A pulmonary tuberculosis was also present. The authors believe a direct hemogenic infection from the lungs occurred, and that this case was not therefore the result of a spread from the endometrium, as is usually the case in these specimens. Kelly and Cullen 31 have also recorded a case of an adenomyoma of the uterus invaded by a tubercu- losis. The tubes showed advanced tuberculous salpingitis. These authors report another case of tuberculosis of the uterus associated with a myoma. UNUSUAL MANIFESTATIONS AND REMOTE COMPLICATIONS 229 Grad 32 has recorded the history of a case in which there was a bilateral tuberculous salpingitis and an ovarian cyst. The latter sprang from the right ovary and the tuberculous tube was drawn out and adherent over the surface cyst. Pewsner, 33 Logothetopoulos, 34 Prussmann, 35 Polloson and Violet, 36 Meriel, 37 and Poncet and Leriche 38 have described cases in which ovarian cystadenomata were invaded by tuberculosis. Poncet and Leriche go so far as to say that they believe tuberculosis to be a definite etiologic factor in the production of certain cystic tumors of benign type. They call attention to the frequency with which latent tuberculosis is found in patients who have goiter, and cite the works of several authors who have shown by histologic examination that simple goiters are often tuberculous. They believe that adenomatous prolifera- tion is one of the ways in which the thyroid reacts to tuberculosis. They state that cysts of the ovary showing no specific tuberculous lesions are frequently found in connection with tuberculosis of the fallopian tubes. A number of cases are cited. They conclude that these are due to inflam- matory tuberculosis of the ovaries, which react to the tuberculous process by the formation of cysts. They do not imply that all ovarian cysts are the result of tuberculosis, but think tuberculosis is one of the causes. That tuberculosis may produce cystic lesions is well recognized, but perioophoritis is a frequent condition and actual tuberculous oophoritis is more frequent than formerly supposed. Both these conditions may lead to the formation of retention cysts, but that cystadenomata or other forms of true new growths are the result of tuberculosis, or even that the presence of tuberculosis predisposes to the formation of ovarian neoplasms is certainly far from proven. As has been stated, a study of our material in the laboratory of gynecological pathology at the Uni- versity of Pennsylvania and of the literature pertaining to this subject does not seem to the author to bear out the assertion that ovarian neoplasms showing tuberculous invasion are more frequent than can be explained on the grounds of the purely accidental combination of these conditions. Furthermore, this opinion is strengthened by a study of the histology of these tumors. To summarize, it may be stated that the etiologic relationship between cancer and tuberculosis is not definitely proven, but preexisting inflam- mation is apparently at least to some extent an etiologic factor. Car- cinoma of the fallopian tube is in itself an infrequent tumor, but its relative frequency occurring with tuberculosis is at least suggestive. The assumption of Von Franque 1X that the tumor rarely springs from the area actively affected by tuberculosis must be taken with some reserva- tions. Tuberculosis affecting the fallopian tubes usually begins in the 230 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS ampulla, and this portion of the tube usually presents the most char- acteristic histologic picture of the infection. "We believe, however, that in most cases where the disease is moderately advanced, at least all or nearly all the mucosa of the tube is involved, and the fact that, in a given specimen, tubercles were not found near the uterine end is no proof that this part of the tube had not been invaded to some extent. Filiomyomata are rarely attacked by tuberculosis, whereas adenomyomata are frequently invaded by a direct extension from the mucosa and less rarely by a hemogenic or lymphatic infection. Ovarian cystadenomata, when occurring in conjunction with peritoneal or pelvic tuberculosis, are fre- quently attacked ; usually only the capsule of the tumor being involved, less frequently and as a result of long standing or virulent infections or occurring perhaps as a result of a hemogenic or lymphogenic infection, the substance of the tumor is invaded. In all cases care must be observed to differentiate between the pseudoneoplasm, which may be produced by tuberculosis alone, and true tumor. Tuberculosis of the Uterus Causing Pyometra. — It is generally conceded that pyometra rarely occurs except in cases of malignancy. In a few instances, however, this condition has been observed in conjunction with tuberculosis, most frequently with tuberculosis of the cervix, although occasionally an extensive corporeal endometritis may result in shutting off of the cervical canal and the consequent formation of a pyometra. Targett 39 presents an illustration of a pyometra, the endometrial cavity being much dilated and the myometrium markedly thinned. The endometrium was reddened and presented many small punctate ulcers, most of which were superficial. The abstracted reports of a number of cases of pyometra, the result of cervical tuberculosis, are presented elsewhere. Schiffmann 40 reports the history of an inter- esting case in which a woman, who had never menstruated or been pregnant, presented herself suffering from tuberculous adnexitis. An operation was performed and the patient died of a purulent peritonitis. A postmortem showed that she had suffered for years, probably since childhood, from a tuberculous metritis, which had resulted in occlusion of the canal and which accounted for the amenorrhea. Torsion of Tuberculous Tubal Lesions. — Tuberculous pyosal- pinges are subject to the same accidents, as are similar pathologic proc- esses the result of microorganisms other than the tubercle bacillus. Indeed, torsion is perhaps more frequent in tuberculosis than in other forms of pyosalpinges. Anspach 41 has especially emphasized this point.' The tendency of tuberculosis of the fallopian tubes to produce large retort shaped lesions, often comparatively free from adhesions, and the UNUSUAL MANIFESTATIONS AND REMOTE COMPLICATIONS 231 extreme chronicity of the disease, are all factors which make torsions more likely to occur. The fact that the ovaries are less likely to be severely attacked in tuberculosis than in other forms of pelvic infection, thus theoretically lessening the extent of the adnexal adhesions, may also be a factor in the somewhat more frequent occurrence of torsion in tuber- culosis than in other varieties of infection of the fallopian tubes. For- tunately, torsion, even in tuberculous cases, is a rare complication. Hydrosalpinges are, by reason of their frequent retort shape and often relative freedom from adhesions, more prone to torsion than are actual pus producing lesions. The exact etiology of torsion of inflammatory uterine adnexa is difficult to determine, but is probably largely influenced by the same factors as are known to produce so frequently similar acci- dents in cases of ovarian neoplasms. Among the causative agents, there- fore, are length of pedicle, irregularity in shape of the tumor, flaccidity of the abdominal walls, alternative filling and emptying of the bladder and rectum, peristaltic movements of the intestines, and rapid alterna- tions in the intra-abdominal pressure, such as are produced by pregnancy, labor, paracentesis abdominis, alternate distention and evacuation of the intestines, sudden, unusual or constrained movements of the body as a whole, such as stooping, turning to get out of bed, vomiting, trauma, falls, jolts, administrating of enemata, gynecologic examinations, and pressure of the abdomen against a hard object, such as a wash tub, etc. Bell 42 lays particular stress upon the action of the diaphragm in these cases. Payr 43 has directed attention to another, and which he believes to be an important factor in the production of torsion. This author believes that venous stasis in the pedicle, especially of small freely movable tumors, may cause them to twist. The veins in many such pedicles are extremely tortuous, much more so than the arteries, and, as a result of intense congestion, impart a spiral motion to the tumor; as twists occur the stasis becomes increased and a sort of vicious circle is formed. Payr's article contains a number of illustrations. The ovarian veins are normally unusually tortuous, so that the foregoing theory is particularly applicable to torsion of inflammatory tumors of the adnexa. Symptoms. — Torsion of the inflammatory tube, like torsion of ovarian neoplasms, may be acute, the twist more or less completely shut- ting off the blood supply and resulting in gangrene or rupture ; or it may be chronic, causing a disturbance of the blood supply and a mild exacerba- tion of the symptoms, followed by a remission, and later followed by other twists, any of which may be acute. Any degree of variation 2^2 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS between these extremes may occur, the symptoms usually depending upon the degree of torsions and impairment of circulation. A previous history of pelvic inflammatory disease is usually present. Not infrequently a history of some causative factor may be obtainable, although sometimes this is absent. In 63 per cent of the recorded cases of torsions due to all forms of infections, the patients have been kept under observation for a time before operation, showing that in a definite proportion the symptoms at the onset were not very alarming. Many of these cases were at first mistaken for exacerbations of a pelvic inflam- matory disease. The seizure is almost invariably ushered in by an attack of severe sharp pain over the seat of the lesion. This is accompanied by more or less marked symptoms of shock or collapse, followed shortly by the evidence of acute pelvic peritonitis, which not infrequently becomes general ; nausea, vomiting, hyperpyrexia and elevation of the pulse rate, with the accompanying evidence of peritonitis, develop. Vesical dis- turbances, such as retention of urine or irritability and frequency of urination, are frequently observed. Examination reveals the presence of a more or less tender, fluctuant tumor, which is generally pelvic in loca- tion. In cases in which a pelvic examination has been made prior to the attack, the change in shape, size and consistency of the tumor will be of aid in arriving at the correct diagnosis, as, subsequently to the torsion, the tube becomes larger, more tender and more tense, and possesses a somewhat more limited range of mobility. The enlargement is sometimes quite marked. The opposite adnexa are usually found to be the seat of an inflammatory lesion. A satisfactory pelvic examination can rarely be performed without an anesthetic, owing to the tenderness and rigidity which is generally present. Diagnosis. — The correct diagnosis of torsion of an inflammatory uterine appendage is extremely difficult and rarely made (Bell, 42 Anspach 41 ). For practical purposes, however, the character of the symptoms and the local findings are nearly always sufficient to call for immediate operative intervention in the severe forms. The condition is frequently mistaken for a torsion of a small ovarian cyst or, when upon the right side, for an acute appendicitis. An important point is to determine between the lighting up of a previously chronic inflammatory condition and a torsion or twisted hydrosalpinx. If a hydrosalpinx undergoing torsion is mistaken for an ovarian cyst or an acute ap- pendicitis, no great harm is done, as both require immediate surgical intervention. If, however, the condition is mistaken for an exacerbation of a previously existing pelvic inflammatory disease, much valuable time may be lost and a general peritonitis develop. The history of Anspach's UNUSUAL MANIFESTATIONS AND REMOTE COMPLICATIONS 233 case is as follows : Age, 26 ; symptoms, simulating acute appendicitis. Operation revealed a long, retort shaped right tube containing blood and pus, twisted two and one half times in the direction of the hands of a watch. Salpingo-oophorectomy was performed. Recovery. Subsequent to the operation the patient complained of pain in the left ovarian region, and a few months later a second operation showed a similar shaped fal- lopian tube on the left side. Microscopic examination proved the latter to be tuberculous in origin. The origin of the infection in the right side was probably similar, but this point could not be positively deter- mined because of the dense infiltration with blood and numerous hemor- rhagic infarcts, which were present, incident to the torsion. Von Meerdervoort 44 has reported a case occurring in a patient 24 years of age. Symptoms of pelvic disease and pain in the lower abdomen had been present for 5 years. At operation bilateral suppurative tubal lesions were found; the right tube was the seat of a torsion. Histologically, both pyosalpinges proved to be tuberculous in origin. Ross 45 has reported the history of a case in which the symptoms appeared suddenly after cranking a motor car. The diagnosis before operation was acute appendicitis, and an emergency operation was performed. Both tubes were found to have been converted into tuberculous pyosalpinges and the right was twisted. Sampson 46 has also recorded the history of a case occurring in a patient 21 years of age. The attack was sudden in onset and simulated the symptoms produced by the torsion of an ovarian cyst. Operation revealed bilateral pyosalpinges, with torsion of the right tube. Supravaginal hysterectomy, bilateral salpingectomy, and right oophorectomy was performed. Histologic examination proved the tuberculous origin of the inflammation. Treatment. — Immediate operative intervention is required in all cases of torsion. Rupture of Tuberculous Pyosalpinges. — This is a comparatively rare accident, but may occur to any pyosalpinx. As a result of adhesions to surrounding structures the tubal contents may be discharged into the intestinal tract, bladder, uterus, peritoneal cavity, or even through the abdominal wall. Rupture is most likely to occur into the rectum or sigmoid flexure or into the peritoneal cavity. The latter is the form of rupture usually meant by most writers when the term "rupture of a pyosalpinx" is referred to. Ruptures may occur spontaneously or may be the result of direct trauma, such as blows, kicks, falls, rough pelvic examinations, coitus, or labor ; violent peristalsis, straining at stool, may also in certain cases produce rupture. It is probable that, if pelvic inflam- matory disease did not usually produce sterility, rupture during preg- 234 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS nancy or labor would be much more frequent. As the growing uterus rises out of the pelvis to which inflammatory tubes are densely adherent, considerable traction is sure to be caused. A drawing out and thinning of the tube follows, which, if in itself it does not cause rupture, produces a lesion by the aid of which a small amount of trauma is sufficient to produce the accident. Indeed, under such conditions Gonsolin 47 states that tubes, both ends of which are patulous, may rupture as a result of traction. Labor itself may cause rupture. Bovee 48 in 19 10 collected statistics from fifty-five cases of rupture without reference to the type of infection, and submitted a history of an additional case from his own practice. In the majority of the ruptured cases there was no assignable cause for the rupture. Undoubtedly the acute exacerbations of chronic lesions tend to produce a condition favorable for rupture, as at these periods more secretion is excreted into the closed off tube, resulting in an increase of intratubal pressure, and the acute inflammation tends to weaken the abscess walls. Rupture usually takes place in the ampulla of the tube. No rule can be formulated as to the size of a pyosalpinx in which rupture is most likely to occur. In many of the reported cases the tubes have been small. Naturally, those specimens in which the walls are thin and friable are most prone to this accident. Adhesions in some cases probably play an important part. Symptoms. — These vary widely in different cases, depending upon the virulence of the organisms and the locality into which the pus escapes. A previous history of pelvic inflammatory disease is generally obtainable. In twenty-nine of thirty-one cases of rupture into the peritoneal cavity, without regard to the type of infection, collected by Bonney, 49 in which an accurate history was obtainable, the onset was abrupt and violent and the evolution of the symptoms rapid. At the time of accident a sharp pain at the site of the rupture generally occurs, usually followed by nausea and vomiting. The temperature may be normal or subnormal for a few hours, and the pulse rapid and weak; pallor, sweating, and other symptoms suggestive of an internal hemorrhage are frequent. The tem- perature soon rises and other evidences of a general peritonitis become manifest. The disproportion between the pulse rate and temperature in the early stage, together with the history of sharp pain perhaps occurring during straining at stool, trauma, etc., followed rapidly by the evidences of peritonitis, are very suggestive of this accident. Diagnosis. — If a pelvic examination has been made prior to the rupture, examination subsequently will reveal the altered shape of the tube, which is found collapsed and flaccid, whereas previously it may have been easily defined as a tense inflammatory mass. From a practical UNUSUAL MANIFESTATIONS AND REMOTE COMPLICATIONS 235 standpoint, however, this test is of little value in the average case, because, even if the surgeon has made a pelvic examination prior to the rupture, the rupture itself is likely to produce so much pain and tenderness that, on examination shortly after the accident, accurate outlining of inflam- matory masses is extremely difficult or impossible without the aid of an anesthetic. One of the chief dangers from the rupture, as well as from torsions, is that the accident may be mistaken for a simple exacerbation of a previous inflammatory disease and therefore treated palliatively. In both these conditions it is of the utmost importance that operative meas- ures be employed without delay. The fact that the degree of mortality bears a direct ratio to the time elapsing after the accident and before the operation is amply proven by Bovee, 48 Bonney, 49 Boldt, 50 and all others. In these cases, the aim of the surgeon should be to make the diagnosis and operate before the onset of the general peritonitis, which is almost sure to follow a rupture into the general peritoneal cavity. The previous history of the case, the acute onset, are usually sufficient to exclude the ordinary exacerbation of a pelvic inflammatory disease. When the lesion is on the right side, not infrequently these cases have been mistaken for an acute appendicitis. Torsion or rupture of an ovarian cyst may also be readily confused with this condition. Fortunately these conditions require immediate operative intervention, so that a mistake in diagnosis under such circumstances is not of vital importance. Treatment. — As previously stated, the treatment should be imme- diate operation. The type of operation employed will naturally vary with the individual case. Rupture of a Pyosalpinx into Adjacent Hollow Viscera. — Accord- ing to statistics this is probably a more frequent accident than is generally thought, and doubtless its diagnosis is often overlooked by the keenest observers. The opening may be direct into the bowel or may be indirect, the tubal opening leading into the bowel through a walled off fistular tract. The former is the more frequent. In a series of tubal cases operated upon in the University Hospital in the last ten years, a number of cases of this kind have been observed, and, in all, ruptures had taken place between the tube and the lower large bowel. In all the point of rupture occurred in a portion of the tube which was adherent to the intestine and no general peritoneal involvement had occurred. Tubal abscesses of tuberculous organs are perhaps more prone to this accident than are like conditions, the result of other types of infection. The rupture of a pyosalpinx into the bladder is less frequent than into the bowel, probably because, owing to the anatomic situation of 236 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS the tube, vesical adhesions are less frequent than are similar lesions to the bowel. Violet and Chalier 51 believe that this is not an uncommon condition. They state that the opening into the bladder may be direct (tubovesical) or indirect (peritoneovesical), a soft, caseous mass inter- vening between the tube and bladder. The former is the more frequent. Under such circumstances the resulting cystitis and the recovery of tu- bercle bacilli from the urine may lead to a diagnosis of renal infection, unless a careful cystoscopy examination is performed. Kutschner 52 has described a case of this character, in which a tuberculous pyosalpinx perforated into the bladder and simulated bilateral renal tuberculosis. Israel 53 reports the history of a similar case. The patient was a young woman suffering pain, failing health, loss of weight, night sweats, fever, dysuria and pyuria. On the first urine analysis tubercle bacilli were demonstrated; in the next two examinations no tubercle bacilli were found. Cystoscopic examination revealed the presence of a cystitis and a tuberculous ulcer on the right side of the bladder. Both ureters were catheterized and tubercle bacilli were demonstrated by animal inoculation from each. A second catheterization was performed with the same results. The diagnosis of bilateral renal tuberculosis was made, and an unfavorable prognosis given. The patient was sent to a sanatorium ; 1 1 months later she presented herself apparently in perfect health, having gained 38 pounds. . Cystoscopic examination at this time showed a cystitis, but the ulcer had disappeared. Renal palpation was negative. An exam- ination under ether revealed a mass situated to one side of the uterus. The examination was followed by hyperpyrexia and pelvic pain, which persisted for a day or two. A few days later another urethral catheteriza- tion was performed, special care being taken to avoid contamination of the catheters in the bladder. Normal urine was obtained from both sides. The diagnosis was now clear. Abdominal section showed a tuberculous pyosalpinx, the lumen of which communicated with the interior of the bladder by a hollow band four or five centimeters long. Following the operation the pyuria disappeared, but tubercle bacilli continued to be present in the vesical urine for six months. The patient was reported well two years later. This case illustrates the necessity for a thorough pelvic examination in all cases, and illustrates also how easily the most experienced may be misled by such findings as reported above. Aurray 54 has described three somewhat similar cases. He states that tubovesical fistulas rarely heal spontaneously. Violet and Chalier 51 report the histories of three cases of this kind and urge the necessity for UNUSUAL MANIFESTATIONS AND REMOTE COMPLICATIONS 237 operation in such cases, recommending that generally an abdominal hysterectomy and bilateral salpingo-oophorectomy be performed. The portion of the bladder surrounding the fistulous opening should be ex- cised and the bladder closed. Vaginal drainage is indicated in most cases of rupture. 1 Extension of Tuberculosis from Pelvic Lesion to Other or Dis- tant Areas. — In a previous chapter the extreme rarity, but occasional occurrence, of primary genital tuberculosis has been pointed out. If it is accepted that primary genital tuberculosis exists, it therefore follows that extensions from such a focus may occur. This, however, is more theoretic than practical, because of the rarity of primary genital lesions. Under certain circumstances, such as loss of continuity of the vaginal mucosa, or chemical irritation, either local or distant tuberculous lesions may be produced by the introduction into the 'vagina of virulent tubercle bacilli. It is of importance to recognize that distant lesions, such as pulmonary tuberculosis, etc., may in rare instances be thus produced. It should, however, be emphasized that such results only occur under special conditions which favor infection, and are by no means the rule. As has been stated in the resume of primary and secondary genital tuberculosis, what frequently does occur is that there is a well marked secondary genital lesion, and the primary lesion in the lungs or elsewhere has undergone partial resolution, or is of such small size that its clinical demonstration is almost impossible with any degree of certainty. In these cases it seems probable that an active pelvic lesion may be the focus for an extension of the infection, even to a distant part of the body ; and especially is this true if operation is performed, as the trauma may open up avenues for infection and break up what formerly were walled off collections of infectious material. Brett 55 has described a case of miliary tuberculosis which he believes had its origin from a tuberculous metritis. It seems likely that occasionally a spread of infection to distant portions of the body may occur as a direct result of an operation for genital tuberculosis. Tuberculous Lesions in Hernial Sacs. — The fact that tuberculosis of the adnexa usually produces adhesions which in themselves tend to prevent the inflammatory structures from entering hernial sacs by limiting the range of mobility is one of the chief reasons for the rarity with which diseased tubes and ovaries are found in hernial sacs. Cullen, 56 *A more extensive resume of the subject of torsion and rupture of inflammatory uterine adnexa, without regard to the type of infection, may be found in the author's previous work, "Gonorrhea in Women," Phila. and 'London, 1913, pp. 319-355. This includes literature and an abstract of cases, 238 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Gopel, 57 Le Nouene 58 have recorded instances where inflammatory adnexa were discovered in hernial sacs. In none of these cases is the type of infection definitely stated. Constantini 59 has described a case of tuberculous infection actually invading an inguinal hernia, in a woman 40 years of age. Morrison 60 has also recorded the histories of a series of cases in which tuberculosis occurred within hernial sacs. He states that in the Royal Hospital for Sick Children, Edinburgh, tuberculosis was present in 2 per cent of all herniae. The condition is much more frequent in children than in adults. In 1906 Cotte, 61 in a study of a series of such cases, found 25 per cent occurred in children under 5 years of age. Either the hernia or the tuberculosis may be the primary lesion, the former being the most frequent. Jennesco 62 believes that the tuber- culosis within the hernial sac generally precedes the abdominal tuber- culosis, and not vice versa, as might be supposed. Morrison, 60 however, believes the latter condition the most frequent, and states that there is no postmortem record which shows abdominal tuberculosis absent when present within the hernial sac. He further points out that abdominal tuberculosis is sometimes difficult to recognize and may, therefore, be overlooked. The occurrence of tuberculosis in hernial sacs may be viewed as purely accidental, and is of interest chiefly on account of its rarity. The interior of the sac may present any of the changes common to tuberculous peritonitis, the variety in most cases corresponding with that present within the abdomen. The most frequent variety is that in which the peritoneum is thickened, congested, and studded with grayish tubercles. Perhaps, as a result of gravity, the fundus of the sac is prone to be the area chiefly attacked, although in some recorded specimens the chief changes have been present in the neck of the sac, evidently as a result of a direct extension from within. The sac frequently contains more or less fluid, the characteristics of which vary with the type of the peritonitis present. In some of the recorded cases the peritoneum of the hernial sac has been literally covered with tuberculous granulations, and in others the caseous or the fibrinous variety has been observed. In some specimens the interior of the sac has been filled with an almost indistin- guishable mass, macroscopically resembling cicatricial tissue. Adhesions between the peritoneum and the other coverings of the hernia are fre- quent, and as a result these herniae are often irreducible. Morrison 60 states that in the Children's Hospital at Edinburgh 75 per cent of the children coming for treatment suffer from some form of hernia. This author states that in his series it was impossible to demonstrate abdom- inal tuberculosis by clinical methods in more than 36 per cent of cases UNUSUAL MANIFESTATIONS AND REMOTE COMPLICATIONS 239 in which the infection was present in hernia. The end results in Mor- rison's series showed that 3(11 per cent) of the 27 cases died of tuber- culosis, and 3 more were seriously ill at the time of writing - . The prog- nosis is, therefore, grave. Maylard 63 directs attention to the fact that the symptoms of the tuberculosis are often subservient to those of the hernia. He cites a case occurring in a child 2 years of age, in which the tuberculosis was dis- covered accidentally when operating on the hernia. Similar cases have been recorded by Wallace, 64 Kennedy, 65 and Owen. 66 In these cases the symptoms of peritonitis were mild and the existence of infection might not have been discovered except for the operations which were performed for the hernia. Maylard 63 states that the infection in his case might have subsided and its presence never have been known, but for the hernia which required operation. Tuberculosis and Syphilis. — Whether the occurrence of these two types of infection is purely accidental, or whether the one in any way predisposes towards the other, is still somewhat undetermined. Pick and Handler 67 state, in presenting a series of cases studied by them, that 31 per cent of the deaths of these syphilitic patients were due to tuberculosis. Tuberculous Wound Infection. — This is by no means of rare occurrence, and is particularly likely to occur when drainage is employed and tuberculous material left behind. An instance of this type of infec- tion is often observed in the fistulas following nephrectomy for tuber- culosis of the kidney. Occasionally the same thing occurs after operation for tuberculous peritonitis or adnexitis. A rare complication is that observed by Edebohls. 68 This author operated upon a patient for bilateral tuberculous pyosalpinges. A miliary tuberculosis of the peritoneum was present. The wound healed satisfactorily, but subsequently a tuberculous infection of the cicatrix developed, which required a second operation. Tuberculous Salpingitis as an Etiologic Factor in Tubal Preg- nancy. — The fact that salpingitis is a frequent etiologic factor in the production of tubal gestation is well known. Fehling 69 reports the results obtained in 170 cases of early extra-uterine pregnancy, in nearly half of which, when a careful examination was possible, the opposite adnexa were found diseased. Cones, 70 in an analysis of 202 cases of ectopic pregnancy, found that 83 per cent were accompanied by inflam- matory lesions. The author 71 found 59 per cent of a series of 64 cases to have been preceded by inflammation. Numerous other statistics could be quoted bearing out the etiologic relationship which exists between preexisting salpingitis and tubal pregnancy. As tuberculosis constitutes a definite proportion of all tubal infections (about 7 per cent), it is but 240 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS natural to expect to find it a not infrequent etiologic factor in the causa- tion of tubal pregnancy. The facts that tuberculosis of the tubes is usually bilateral, chronic in character, and that in this type of infection the tubes tend to remain patent longer than in the ordinary forms of infection, all are points which make the likelihood of tubal implantation of the gravid ovum likely in this variety of infection. Croom 72 has reported the history of an advanced extra-uterine pregnancy •complicated with not only a tuberculosis of the tubes, but also of the peritoneal cavity. Appendicitis and Tuberculosis. — Silvestri 73 found manifestations of tuberculosis in 45.63 per cent of 103 persons with appendicitis. Peri- appendicitis, as an accompaniment of tuberculous salpingitis or peritonitis, is frequent and has occurred in a large percentage of our cases. LITERATURE 1. Harris, W. H. Jr. Med. Res. 1913. 29:471. 2. Ortel. Jr. Med. Res. 1912. 25:503. 3. Wolf. Forts, d. Med. 1895. No. 18. 4. Schwalbe. Virch. Arch. 1897. 149. 5. Cone. Arb. a. d. Path-anat. Inst. z. Tub. 1894. u. 2. 6. Pepper and Edsall. Am. Jr. Med. Sc. 1897. 114. 7. Levin, I. Jr. Exper. Med. 1912. 15:163. 8. Kellert, E. Jr. Am. Med. A. 1914. 63:1819. 9. Dixon, Smith, and Fox. Penn. Health Bui. 191 1. No. 24. 10. Lipschutz, K. Monschr. f. Gebh. u. Gyn. 1914. 39: No. 33. 42:41. 11. Franque, O. Ztschr. f. Gebh. u. Gyn. 191 1. No. 27. 1912. 69: No. 2. 12. L'Esperance, E. S. Proc. N. Y. Path. Soc. 17: No. 6, 8. 13. Devic. These de Lyon. 1894. 14. Kauffmann und Wallart. Ztschr. f. Gebh. u. Gyn. 1904. 15. Barret, Lady. Quoted by L'Esperance. No. 12. 16. Maikoff, S. Medits. Oboz. 1914. 80: No. 19. 17. Glockner. Zentrbl. f. Gyn. 1904. p. 702. 18. d'Halluin et Delval. Bui. et mem. soc. anat. de Paris. July, 1910. 19. Nassauer. Centrbl. f. Gyn. 1895. No. 29. 20. Wallart. Ztsch. f. Gebh. u. Gyn. u. 1. 21. Kelly, J. K. Brit. Med. Jr. 1905. 2:712. UNUSUAL MANIFESTATIONS AND REMOTE COMPLICATIONS 241 22. Ivins. Jr. Obst. Gyn. Brit. Emp. 191 1. 19:266. 23. Parsons, J., and Glendining, B. Proc. Roy. Soc. Med. 3 1238. 24. Schutze. Ztschr. f. Gebh. u. Gyn. 60: part 3. 25. Dickson. Am. Jr. Obst. 1906. 53 799. 26. Kelly, H. A. Operative Gynecology. 1899. 2:381. 27. Heinrich. Monschr. f. Gebh. u. Gyn. 1908. 27 : No. 4. 28. Violet et Perrin. Soc. des. sc. med. de Lyon. June 8, 1910. 29. Grunbaum, E. Arch. f. Gyn. 81 :383. 30. Archambault, J. L., et Pearce, R. M. Rev. de gyn. et de chir. abd. Jan. and Feb., 1907. 31. Kelly, H. A., and Cullen, T. S. Myomata of the Uterus. Phila- delphia and London, 1909. p. 335. 32. Grad, H. Am. Jr. Obst. 1910. 60:95. 33. Pewsner, C. These de Lyon. 1913. 34. LoGOTHETOPOULOS. Zentrbl. f. Gyn. 1908. p. 377. 35. Prussmann. Arch. f. Gyn. 1904. 36. Polloson et Violet. La. gyn. 1914. 18:66. 37. Meriel, M. E. Bui. soc. d'obst. et de gyn. de Paris. 1913. 2:732. 38. Poncet, A., et Leviche, R. Lyon chir. 1913. 11: No. 1. 39. Targett, J. H. Brit. Med. Jr. 1903. 2:959. 40. Schiffmann. Arch. f. Gyn. 1914. 103 : No. 1. 41. Anspach, B. M. Am. Jr. Obst. 1912. p. 553. 42. Bell, R. H. Jr. Obst. Gyn. Brit. Emp. 1904. p. 514. 43. Payr. Arch. f. Klin. Chir. 1902. 68:501. Also Ztschr. f. Chir. 1906. 85 :392. 44. Von Meerdervoort. Med. tijdscr. v. verl. en gyn. p. 175. Ab- stracted in Frommel's Jhrber. 1905. p. 209. 45. Ross. Am. Jr. Obst. 1906. 54:653. 46. Sampson, J. A. Am. Jr. Obst. 1912. p. 271. 47. Gonsolin. These de Lyon. Quoted by Lamoreaux. Arch. gen. de chir. Jan., 19 10. 48. Bovee, J. W. Surg., Gyn., Obst. 19 10. 10:405. 49. Bonney, C. W. Surg., Gyn., Obst. 1909. 9:542. 50. Boldt, H. J. Am. Jr. Obst. 1889. 22:262. 51. Violet et Chalier. Rev. de gyn. et de chir. abd. Feb., 1909. 52. Kutschner, H. Inaug. Dis. Berlin, 1913. 53. Israel. Deutsch. Med. Woch. 1913. 39:2295. 54. Aurray, M. Arch. mens, d'obst. et de gyn. 1914. 3:195. 55. Brette, M. Lyon Med. 1914. 46: No. 18, 19. 56. Cullen, T. S. J. Hopk. Hosp. Bui. 1906. p. 152. 57. Gopel. Zentibl. f. Chir. 1896. 23. 242 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS i> 59 8. Le Nouene. Gaz. de gyn. 1903. 15 : 337- Constantini. Bui. et mem. soc. anat. de Paris. 1914. 89:48. 60. Morrison, J. T. Clin. Jr. 1914. 43 '-609. 61. Cotte, G. Rev. de gyn. et de chir. abd. 1906. 10:981. 62. Jennesco. Rev. de chir. 1891. 11:185,455. 63. Maylard, A. E. Brit. Jr. Tuberc. 1909. 3:45. 64. Wallace, C. Tr. Med. Soc. London. 1906. 29:401. 65. Kennedy, A. E. Lancet. 1900. 2:581. 66. Owen, E. Lancet. 1902. 2:1106. 67. Pick und Bandler. Tr. 7th int. cong. derm. syph. 68. Edebohls, G. M. Am. Jr. Obst. 1892. 25 :96. 69. Fehling, H. Arch. f. Gyn. 92. 70. Cones, W. P. Bost. Med. Surg. Jr. 191 1. 164:677. 71. Norris, C. C. Gonorrhea in Women. Philadelphia and London, I9L3- Croom, J. H. Jr. Obst. Gyn. Brit. Emp. 1914. 25: No. 4. /*= 73. Silvestri, T. Rif. med. 1920. 36: No. 2. CHAPTER XI PREGNANCY AND TUBERCULOSIS History — Fertility of the tuberculous — Frequency — Physiology of pregnancy bearing on course of tuberculosis — Organs affected — Puerperium and its bearing upon course of tuberculosis — Susceptibility of pregnant women — Strain of lactation. — Condition of children of tuberculous mothers — Infant mortality — Influence of pulmonary tuberculosis on course of pregnancy — Influence of pregnancy on course of pulmonary tuberculosis — Tubercle bacilli in mother's milk — Tuberculin as diagnostic and curative agent — Law regarding marriage of tuberculous persons — Indication for induction of abortion prior to fifth month — Results — Consultation and precaution prior to induction of abortion — Choice of operation — Sterilization — Anesthetic — Technic of operation (during first two months) — Convalescence — Technic and choice of operation for emptying uterus from second to fifth month — Pregnancy after fifth month — Delivery of tuberculous patients — Cesarean section — Puerperium, treatment during, nursing — Influence of pregnancy upon tuberculous lesions other than the lungs — Bibliography. HISTORIC From the early ages the subject of pregnancy in tuberculous patients has attracted marked attention. Among the early papers devoted to this subject are especially noteworthy the contributions of Horn, 1 Succow, 2 Herrieux, 3 Robert, 4 Grisolle, 5 Dechambre, 6 Tott, 7 Dubreuille, 8 Lassegue, 9 Warren, 10 Thomas, 11 Caresme, 12 and of Ortega. 13 The latter is a report of 132 pregnancies, of which 95 went to term, 28 were premature, and 9 aborted. Third pregnancies were rare. Ortega believed pregnancy ex- erted a deleterious influence on the course of the tuberculosis. Other interesting contributions to this subject exist, reference to many of which may be found in the article of Malsbary* 14 from which much of the foregoing information has been obtained. In reviewing the early literature of pregnancy in the tuberculous, it is interesting to find that pregnancy was for many years believed to exert a favorable influence on the course of pulmonary tuberculosis. This is probably due to the fact that gestation tends somewhat to increase the weight of the woman. This is, however, generally only temporary, and after the fifth or sixth month rapid advancement of the disease is likely to occur. As early as 1862 Gassner 15 commented upon this finding. The 243 244 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS increase in weight occurs chiefly in the latter months of pregnancy, the normal gain being from 1600 to 2500 grams a month (from 3 to 5 pounds). According to De Lee, 16 the increase in weight is due to in- creased assimilation of the fetus and the secundines, the storing up of fat and albumin, the accumulation of water, especially in the lower ex- tremities, and increase in the amount of blood. When pregnancy occurs in the tuberculous woman, very frequently, even in those cases in which pregnancy ultimately exerts an unfavorable influence, no deleterious re- sults occur, or at least become manifest during the early months of gestation. Fertility in the Tuberculous. — No practicing physician can have failed to observe the frequency with which pregnancy occurs in the tuber- culous. This fact has led many observers to conclude that persons af- fected with tuberculosis are unusually fruitful, and that, as a result of the disease, the sexual appetite is increased. Numerous observations have been recorded to bear out this assertion. Sexual intercourse is often practiced even by those in whom the disease is advanced. Simmonds 17 has reported a case in which a man had intercourse with his wife on the day on which he died from an advanced pulmonary tuberculosis, and numerous other somewhat similar instances have been recorded. The fact that, as the result of treatment, many tuberculous patients are idle may have some bearing on the increased sexual desire. Posthumous children are frequent among the tuberculous. Be the reasons what they may, it appears to be certain that tuberculosis, even when moderately advanced, does not materially decrease the sexual appetite nor interfere with fertility. Cornet 18 quotes a number of cases in which the sexual appetite was apparently increased in the later stages of tuberculosis, but does not accept these as proof of an increased sexual appetite; he believes that, because of bizarre nature, observers are unduly impressed by them. He holds that, in the majority of cases, as the disease progresses the sex- ual desire is diminished. However this may be, the fact remains that pregnancy in tuberculous women is of extremely frequent occurrence, and, as stated, it seems to be an assured fact that the disease itself exerts little or no influence on conception. Tuberculosis itself is essentially a disease due to faulty hygiene; the latter is the most common among the ignorant and poor, a class in whom fertility is notorious. Although the fertility among the poor is probably largely the result of ignorance regarding the methods of preventing con- ception, the fact remains that pregnancy and tuberculosis frequently co- exist. Frequency. — In 1913 Bacon 10 stated that 32,000 tuberculous women PREGNANCY AND TUBERCULOSIS 245 become pregnant annually in the United States, and that between 44,000 and 48,000 women of the child bearing age die of tuberculosis every year. Probably 25 per cent of the latter have reached the puerperal state, or, in other words, 11,000 or 12,000 tuberculous pregnant women die an- nually. This writer believes that 33 per cent of pregnant tuberculous women die in less than one year following labor. He points out that these data show only a part of the important bearing which pregnancy has upon tuberculosis. Besides an increased mortality among tuberculous pregnant women, the latter are a source of infection to the family and an important factor in the spread of the disease. The Physiology of Pregnancy as It Bears Upon the Course of Tuberculosis. — The deleterious influence of pregnancy on tuberculous women is well known, and many theories have been advanced to explain this fact. During pregnancy the woman carries a double load, and, as the gestation advances, the drain upon her strength becomes more and more marked. Although pregnancy is a physiologic process, and one that the healthy woman is well able to bear, when it occurs in a patient whose resisting powers are weakened by disease, the extra stress may be sufficient to overbalance her resistance, and, as a result, the disease may progress rapidly in a woman who had heretofore held her own, or who had even been successfully combatting her infection. This is true of all diseases, but especially is it so of tuberculosis. Many of the physiologic changes that occur as the result of pregnancy, and that are commonly pointed out as the cause for the injurious action of pregnancy hardly appear of sufficient importance during the early stages to account for the rapid progress of the disease frequently observed at this period. The author believes that further study of this subject is required to explain why so many cases of early pregnancy show an exacerbation of the tuber- culous condition. Some of the physiologic reasons commonly referred to as exerting a deleterious influence on pregnancy, and which are doubtless important factors in the latter months of gestation, are the following: Lungs, — During the latter months of pregnancy a change occurs in the shape of the lungs, although their capacity is but little altered; the organs become shorter and broader as the result of upward pressure of the gravid uterus ; the diaphragm is pushed up, and the lungs are some- what retracted to the sides, thereby exposing a larger part of the heart. These changes are more marked in primiparae than in multiparae, the abdominal walls in the latter being lax. Respiration becomes more of the costal type, owing to restriction of the movements of the diaphragm. The respiratory rate is increased — from 24 to 26 a minute — and more 246 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS carbon dioxid is excreted- (Vejas 20 ). During the last two weeks of pregnancy, when "lightening" has occurred, the foregoing phenomena are less marked. In the late months of pregnancy the condition just de- scribed may in all justice be pointed to as an etiologic factor in causing aggravation of the pulmonary disease. But in the early months of gestation this is not the case. During the strain of labor more or less congestion of the lungs occurs. It can readily be understood how such straining efforts may exert a deleterious effect upon pulmonary lesions. Larynx. — The frequency with which laryngeal involvement occurs in the pregnant tuberculous woman has been commented upon by most observers. Malsbary 14 has suggested that some relationship may exist between this and the so-called "genital spot." Bretteuer has called atten- tion to the relationship between the "genital spot" and dysmenorrhea. Hofbauer 21 has demonstrated that, as a result of pregnancy, there is an increased congestion of the larynx, affecting especially the false vocal cords, and that there is also a slight cellular infiltration of the tissue in this location. Hofbauer also states that, in the normal pregnant woman, the mucosa of the larynx becomes reddened and swollen, so that a step from the physiologic to the pathologic is not unlikely. Circulatory System. — Heart. — It was formerly believed that, as the result of pregnancy, the heart became hypertrophied. Stengel and Stanton 22 showed that this was not the case, and that the increase in dullness to the left was not the result of hypertrophy of the left ventricle, or of any special increase in work, but that it was caused by the upward and outward displacement of the organ. These observers state, how- ever, that in labor there is probably some dilatation of the right ven- tricle, but they believe that there is no material change in the blood pres- sure prior to or following labor. De Lee 16 asserts that in 25 per cent of cases a systolic murmur is present over the base of the heart. Norris 23 is of the opinion that the displacement of the heart tends to cause a kink- ing of the large vessels, thus adding to the work demanded of that organ. Wiessner 24 believes that this explains the occurrence of accidental pul- monary murmurs. Norris states that in normal pregnancy the blood pressure rarely exceeds 120 mm. of mercury and, if taken between pains in the second stage of labor, it varies between 130 and 150 mm. After delivery the normal values are established. During the uterine contractions of active labor the pains, as well as the intra-abdominal compression, cause a much higher blood pressure than is present in the interim between the pains. Heynemann 25 observed a fall of from 60 to 90 mm., following the birth of the child. During the early months of pregnancy many ill nourished women, PREGNANCY AND TUBERCULOSIS 247 and especially those in poor circumstances, suffer from a form of chloran- emia (De Lee). The condition is very frequent among the poorer classes of phthisical patients — "Virchow's physiologic leukocytosis." Dietrich 26 has in the main confirmed these findings. The blood changes during pregnancy are not marked, and probably exert little influence on tuberculosis, except in those patients who are anemic and whose natural resisting powers are diminished as a result. As is well known, the ductless glands exhibit special activity during preg- nancy. Digestive Tract. — More or less vomiting or nausea occurs in about 50 per cent of pregnant women. This is especially likely to occur in neurotic subjects, and during the early months of pregnancy. When vio- lent straining occurs, the blood pressure is raised and unusual pressure is exerted upon the lung tissue. This condition must, therefore, be con- sidered when the cause for the exacerbation of pulmonary lesions is sought. Brooks and Leuckhardt, 27 in their recent investigations, have shown that although vomiting does not always produce a marked increase in the blood pressure, sharp rises often occur. These investigators state that during the vomiting sudden and severe oscillations of the blood pressure are of frequent occurrence, and that these may cause rupture of a blood vessel that would not occur with the same degree of pressure but with slower changes. As the result of these studies, they also show that the danger to the vascular system during vomiting is not minimized, but that the responsibility is shifted from hypertension to the sudden variations in the condition of the circulatory apparatus. If the vomiting becomes so serious as to interfere with nutrition, its deleterious action on the course of the tuberculosis is most marked. All who have studied pulmonary tuberculosis agree that the phthisical patient requires an abun- dance of nutritious food. If sufficient food cannot be taken, or if assimi- lation is interfered with, a great handicap is placed upon the tuberculous woman. Kidneys. — Throughout pregnancy there is a tendency toward renal disturbances, and lesions of these organs are subject to exacerbations. This is especially injurious to tuberculous patients. Other Changes Incident to Pregnancy. — Many other changes occur as the result of pregnancy, but a large part of these cannot be held responsible for the aggravation of the pulmonary lesions. Exactly why pulmonary tuberculosis is so prone to exacerbation during pregnancy is difficult to explain, except upon the broad ground that pregnancy in itself throws an added burden upon the general system, and that this may in some cases be enough to overthrow the balance of resistance on the part 248 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS of the patient. Hofbauer (quoted by Bandelier and Roepke 28 ) believes that the increased predisposition to tuberculosis in the pregnant woman is due to a reduction of the lipolytic quality of the serum with advancing gestation, hyperglycemia and certain physiologic causes, such as hyper- emia, increased lymphatic flow, and peribronchial infiltration. Sergent 29 suggests that the chlorosis, anemia, decalcification, excessive excretion of phosphorus, and adrenal insufficiency incident to the gravid condition, are determining factors. Davis 30 states that with the growth of the fetus a large part of the iron is appropriated from the mother's blood, and that this, together with the drain on the maternal calcium, are factors that tend to deplete the woman's strength. Davis also directs attention to the changes that take place in the ductless glands, and states that even in normal pregnancies 6 per cent of women suffer from hyperthyroidism. "It is a significant fact, at present not explained, that the Abderhalden test for early pregnancy gives a positive reaction in non-pregnant patients who have tuberculosis. Evidently the disturbances in the blood caused by pregnancy are closely allied to those of tuberculosis. It seems reason- able to suppose that the combination of these two conditions increases the pathologic condition." Friedrich's 31 experiments upon rabbits did not show that lipoidemia favored the dissemination of tuberculosis; in fact, they indicated the contrary. Puerperium. — Fraught with more danger than pregnancy itself is the puerperium, and here a definite basis for the exacerbation of the tuberculous condition which so frequently occurs at this time can be determined. The patient has already suffered the strain of pregnancy, and has undergone whatever deleterious effects this exerted. The strain- ing and increased blood pressure incident to labor are probably frequently sufficient to break down minute, partially healed pathologic processes, and thus convert closed lesions into open ones. As a result, hitherto partially or entirely encapsulated tubercle bacilli are liberated in more or less large numbers. Many free organisms are thrown into the blood stream, thus accounting for many of the cases of miliary tuberculosis that have been reported as occurring at this period. The actual physiologic ex- haustion following a difficult labor is also a contributing factor in many cases. The congestion of the lungs incident to labor must likewise be taken into consideration. The prolonged muscular exertion, the physical exhaustion of labor, the possible loss of blood, or the effects of a general anesthetic, if one has been used, are also factors that must be taken into account. Tuberculosis of the placenta has been described. It suffices here to state that tubercle bacilli have been found in the placentae of tuberculous PREGNANCY AND TUBERCULOSIS 249 parturients by some observers in 40 per cent of cases. In the author's series, virulent tubercle bacilli were positively demonstrated in about 5 per cent of a series of cases comprising patients in various stages of the disease. Tubercle bacilli are prone to be present in the placenta of women suffering from an active lesion, and especially if hyperpyrexia or pyrexia is present. The organisms are much more likely to be present at term than in the immature placenta. Placentae containing tubercle bacilli may be, and frequently are, microscopically normal. In cases in which tubercle bacilli are present in the placenta it is but reasonable to suppose that organisms are also present in the decidua. If this theory is accepted, it follows that the contraction? of the uterus incident to labor must force out a definite number of virulent organisms into the circulatory blood strea«m, and that, thus liberated, these tubercle bacilli may in turn set up new lesions and cause an exacerbation or the development of a miliary form of the disease. Von Bardeleben 32 considers this so serious a cause for trouble that he recommends performing cesarean section and the excision of the placental sites prior to the onset of labor, for the double purpose of preventing the labor pains, which may squeeze out the organisms, and the removal of the possibly infected placental site. For the latter reason some operators recommend excision of the placental site and sterilization of the patient by ligation of the fallopian tubes, by partial or total sal- pingectomy, or by supravaginal or panhysterectomy, this being done for the purpose of preventing subsequent conception. Another reason why exacerbations are so frequent during the puer- perium is that lighting up of the pulmonary process has really started during the pregnancy, but has had time only to advance to such a stage as to attract definite attention by the time the puerperium has been reached. Lactation. — Lactation, particularly when the woman is below par, as most of the tuberculous are, is also a very definite added strain, and may in itself be sufficient to lower the woman's resisting powers to such a point as to exert an unfavorable influence on the course of the disease. As early as 1887 Hanau 33 pointed out the dangers of auto-infection. He asserted that the excessive straining, etc., induced expectoration, which was frequently drawn into hitherto uninfected pulmonary areas, only to set up fresh lesions there. The dangers of aspiration in such cases are undoubtedly real. Susceptibility of Pregnant Women to Tuberculosis. — The author believes that, as a general rule, pregnancy, and especially the puerperium exerts an unfavorable influence upon the course of tuberculosis. Whether the normal pregnant woman is more susceptible to infection by the tubercle 250 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS bacilli is still an open question. It is certain that a definite proportion of women apparently contract the disease during either pregnancy or the puerperium. This is particularly true of the wives of tuberculous men living amid unhygienic surroundings. Whether this is due to an in- creased susceptibility at this period, or to the added strain on the general system is not known, but both are probably contributing factors. The change in the general routine of life incident to pregnancy, and the lessened amount of fresh air and lack of exercise indulged in by preg- nant women may to some extent also be causative factors in some cases. Doubtless many cases in which the disease is apparently contracted dur- ing pregnancy are in reality exacerbations of hitherto mild and unsus- pected lesions, and as the disease progresses, clinical symptoms become manifest, with the result that the condition is attributed to an infection occurring during pregnancy. Fisbberg 34 found that, of 286 married tuberculous women, 107 or 37.4 per cent first noticed their pulmonary symptoms after one or more pregnancies had occurred. Jacob and Pann- witz, 35 in 337 tuberculous women, found that 25 per cent traced the origin or the exacerbation of their condition to pregnancy. Trembley 36 states that of 240 cases of tuberculous married women, 151, or 63 per cent, gave a positive history of the disease originating or becoming defi- nitely recognizable either during pregnancy or the puerperium. Turban (quoted by Schauta 37 ) found that 29 per cent of tuberculous women who had borne children attributed the onset of their condition to preg- nancy or the puerperium. Funk, 38 in a series of 200 married women suffering from pulmonary tuberculosis, found that 30 per cent first noticed symptoms either during or shortly following pregnancy. Grisolle 5 observed that, in a series of 2^ cases of tuberculosis in pregnant women, there were apparently many instances in which the disease developed during gestation. The average duration of the pul- monary symptoms in this series was g l / 2 months. Maragliano 39 found that 59 per cent of tuberculous women who had been pregnant first no- ticed severe symptoms during gestation or in the puerperium. Of 100 cases, Funk 38 found that in 43 the first symptoms of the pulmonary lesion became manifest during pregnancy or shortly afterward. The average age of these patients was 35.7 years. Combining these results, we find, in a series of 963 cases, 42 per cent first noticed the pulmonary symptoms during pregnancy or lactation. The important points in the study of this condition are the prognosis and the treatment. Notwithstanding the frequency of these cases and the amount of study that has been devoted to them, comparatively few PREGNANCY AND TUBERCULOSIS 251 valuable statistics have been formulated — too few, in fact, to permit the drawing of any hard or fast rules. The reasons for this are obvious, as so many factors enter into each case — the virulence of the infection, the stage of the disease, the type of infection, the resistance of the patient, her social standing, mode of life, ability and intelligence to submit to treatment, and the advancement of the pregnancy are all vital factors, to be considered in each case. Additional difficulties encountered in the compiling of statistics are that, with reference to the pulmonary condi- tions, special diagnostic skill is required, and even when this is had the most experienced may vary widely, since the personal equation enters largely into these cases. For present purposes, only broad statements will be made. In studying this subject, we cannot escape the fact that no fixed rule can be formulated that will apply to all cases, but that each case must be considered individually. All points bearing upon the indi- vidual case must be carefully weighed before a prognosis can be made or a line of treatment can be instituted. In considering the prognosis and treatment, a question that immediately arises in the investigator's mind, and in the minds of the prospective parents, is the probable condition of the child. Although this is only of secondary importance to the health of the mother, it is a point that must and should be definitely considered. Condition of the Child of Tuberculous Mothers. — The subject of placental and congenital tuberculosis has been dealt with somewhat in de- tail in a previous chapter and only a brief review will be given here. It may be accepted that pregnancy or labor tends to produce a tuberculous bacillemia, and that, although this may be infrequent, as a result virulent tubercle bacilli may reach the placenta. The further the pregnancy is advanced, and the more active are the pulmonary lesions, the more likely is this to be the case. Tubercle bacilli are present in the placenta far more frequently than was formerly believed. The fact that virulent tubercle bacilli are present in the placenta is, however, no conclusive proof that a congenital infection exists. Although tubercle bacilli are not infre- quently present in the placenta of tuberculous women, congenital tubercu- losis is, nevertheless, an extremely infrequent disease, only 4 undoubted cases of this condition being recorded in the literature. Investigators have demonstrated that congenital tuberculosis may be produced in a small proportion of cases by animal experimentation, but even here the conditions can hardly be compared with those that occur in the pregnant woman, as the amount of culture of tubercle bacilli introduced into the pregnant animals is far in excess of what could possibly occur in the woman. It must be admitted, however, that congenital tuberculosis does occur occasionally in man, the condition being so rare, however, that for 252 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS practical purposes, it need not receive serious consideration. In nearly all children affected with tuberculosis the infection is a postnatal one. The question as to whether or not the children of a tuberculous mother exhibit a greater or lesser susceptibility to this form of infection is of much greater importance, and is, unfortunately, still an unsettled point. Many convincing arguments may be arranged on both sides. The author believes that such children may show a hypersusceptibility to infection. This, however, is probably not marked. The number of tuberculous in- fants under one year of age who are the offspring of tuberculous parents would seem to be a strong argument in favor of this belief. Stutz 40 states that the children of tuberculous mothers are constitutional weak- lings. The high mortality among these children is probably dependent more upon the unhygienic environment and often motherless condition to which these children are exposed, than to any hereditary predisposition. As has been stated, it is still a mooted point whether children of a tubercu- lous mother exhibit a hypersusceptibility to tuberculosis. The fact that the mortality among infants of tuberculous mothers is far greater than that among children of healthy progenitors has been substantiated, and should be taken into consideration when the question arises of perform- ing the so-called "therapeutic abortion." Thus Sergent 29 states that 68 per cent of children of tuberculous mothers die. Parry 41 is of the opinion that 50 per cent of these infants die during early months of life. Pankow and Kupferle 42 states that 54.5 per cent of these infants die under one year of age. Zirkel (quoted by Pankow and Kupferle) places the mortality at 58 per cent; Deibel 43 at 78 per cent; Weinberg (quoted by Pankow and Kupferle) at 78 per cent. Fellner, 44 in a series of 289 children, found that 24 per cent died at birth or shortly afterward. Sil- berman's 45 infant mortality was 28 per cent; Dirner's 46 37.5 per cent, and Sergent's 29 68 per cent. Parry 41 states that 50 per cent of children of tuberculous mothers die during early months of life. In a series of cases of laryngeal tuberculosis Glas and Kraus 47 found that 60 per cent of infants died within a short time after birth. Trembley 48 asserts that the offspring of tuberculous parents are weak and display a tendency toward tuberculosis. Jacobi (quoted by Polak and Matthews 49 ) states that 70 per cent of infants succumb during the first year. Weinberg 50 places the proportion at 67.9 per cent ; Zirkel 51 at 58 per cent. Thus it is seen that the combined results of 14 observers show that there was an average infant mortality of 58.83 per cent among children born of tuber- culous mothers. Miller and Woodruff 52 examined 150 children of tuberculous parents and found 51 per cent positively tuberculous. Floyd and Bowditch 53 found 66 per cent. Kunreuther r ' 4 also emphasizes the PREGNANCY AND TUBERCULOSIS 253 unfavorable prognosis for children of tuberculous mothers, and records one family in which there were 6 children, of whom 3 had died of tubercu- losis and all the others were infected. Bacon 19 estimates that, of the 10,000 children under 5 years of age who die annually in the United States of tuberculosis, 7,500, or 75 per cent, are born of tuberculous mothers. Armand-Delille 55 studied a series of 787 children born or living in 175 families, one or more members of which were tuberculous. Of these children, 323 were placed in the country and did well : 396 were not removed from their infectious surroundings, and of these 238 devel- oped tuberculosis. From this can be seen the postnatal danger to which the child of a tuberculous mother is exposed. Doubtless a large proportion of the mortality of the children is the result of death or invalidism of the mother, which often leaves the child without adequate care. Many of the infants of tuberculous mothers are bottle fed even during the mother's life, and the mortality among such children is naturally high. Kings ford 56 reports the result of his study of 339 post-mortem records of children who had died of tuberculosis. Of these, 162 had died during the first two years of life, and 270 during the first five years. These records show how fatal tuberculosis is in the young. Many authorities believe that the children of a tuberculous mother are constitutional weaklings. In this the author concurs only to a limited extent. The author has seen large healthy children born from mothers in the last stages of the disease. Some possible causes for the high infant mortality other than constitutional weakness have already been suggested. It is, however, probable that, if a large series of such infants was com- pared with a series from normal women, the former would be found smaller and weaker in the average, and this would probably also be the case, if a series of infants of anemic or otherwise weakened but non- tuberculous women were studied. In other words, it does not seem prob- able that tuberculosis exerts any specific action on the infant other than would be produced by any other weakening condition. Influence of Pulmonary Tuberculosis on the Course of Preg- nancy. — The question of sexual desire in the tuberculous has already been discussed. The fertility of the tuberculous is a subject of great im- portance. The tuberculous woman is quite as likely to conceive as is the normal woman. Indeed, Shauta 37 believes that tuberculous women are especially fertile, and states that he has found it necessary in some of his cases to induce abortion two or three times in a year. On the other hand, Pinard 57 is of the opinion that pregnancy is relatively uncommon among women affected with an active tuberculosis. Even in advanced 254 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS cases, in this author's experience, conception is relatively infrequent. The employment of methods to prevent conception, so often adopted, is prob- ably the chief reason why pregnancy does not occur even more frequently. After conception has taken place, the tuberculosis has little influence on the course of the pregnancy per se. As the result of excessive coughing, progressive anemia, or hyperpyrexia, abortion or premature labor may occasionally occur, but this is comparatively infrequent. The tuberculous gravida is probably especially prone to develop such complications as renal disturbances and gastric disorders. Excessive vomiting and renal insufficiency may in themselves bring about abortion or miscarriage. As has been stated, in advanced cases abortion or miscarriage occasionally occurs, and this is especially likely to take place just before a lethal ter- mination of the disease and in the event of a laryngeal involvement. Both De Lee 16 and Williams 58 state that the disease does not predispose to premature interruption of pregnancy, unless the pulmonary lesion be of the florid or fulminating type. In such cases the cough and hemoptysis, fever, vomiting, tuberculous infection of the placenta or decidua, placental hemorrhages, etc., may precipitate a premature labor. In Glas and Kraus's 47 series of cases, 28 per cent of patients with laryngeal tubercu- losis suffered premature labor. Funk, 38 in a series of 100 cases of preg- nancy and tuberculosis, compiling from the total number of pregnancies, observed 7.4 per cent of miscarriages or abortions; and in a later series the same authority found that, among 200 cases, miscarriage or abortion occurred in 18 per cent of cases. Landouzy's (quoted by Pinard 57 ) ex- periments tended to show that animals inoculated with tuberculosis before pregnancy takes place go to term, but that when inoculated during preg- nancy, they may abort, the effect depending upon the virulence of the microorganisms. Influence of Pregnancy Upon the Course of Pulmonary Tubercu- losis. — Of even more importance than the life of the unborn child is the question of the influence pregnancy will have upon the course of the tuberculosis in the woman. Before undertaking the systematic study of this condition, and influenced only by the literature and a few personal observations, the author was of the opinion that too much stress had been laid upon the deleterious influence of pregnancy upon tuberculosis, and he believed further that most of the German investigators were far too pessimistic in their prognosis regarding these cases, and that their mode of treatment was far too radical. Within the last 9 years the author has examined all pregnant tuberculous women coming to the Henry Phipps Institute for treatment. The physical and bacteriologic examinations have been performed by skilled internists, and careful histories of the PREGNANCY AND TUBERCULOSIS 255 pulmonary condition, weight, and general health have been kept. After delivery each case was kept under observation by a social worker, who visited the patient in her home. Following the puerperium an endeavor was made to have each patient return to the Henry Phipps Institute for further treatment. New charts showing the pulmonary condition were then made. The infants received the necessary treatment, and the mothers were instructed in special clinics as to the proper hygiene, etc. A pelvic examination was also made in each case. In this way, in the majority of cases, the condition of the patient has been under observation for at least one year. Ninety per cent of the patients at the Phipps Institute are foreigners, and are, as a general rule, an extremely ignorant class, and therefore unsatisfactory patients. Granting that these patients are, as a rule, unfavorable subjects for treatment, notwithstanding the excellent work done by the social service department of the Phipps Institute, no observer can fail to be impressed with the unfavorable influence often exerted by pregnancy on the course of pulmonary tuberculosis. Again, however, we can only generalize, and must once more emphasize the fact that each case must be studied individually. In the author's series not a few cases of advanced tubercu- losis, which were first seen in the middle or later months of pregnancy, withstood well the test of the later months of pregnancy, labor, and the puerperium, and were at least as well six or nine months after delivery as they were at the sixth or seventh month. Such cases are, however, the exception. The author recalls one case of advanced bilateral pulmonary tuberculosis that had been bedridden at the Phipps Institute for three months, and that had suffered frequent and profuse hemorrhages. These were especially frequent and copious during the ninth month. The patient was removed to the author's service at the Maternity Hospital, where everything was held in readiness for the performance of cesarean section, as it seemed almost certain that the strain of labor would induce an ex- cessive hemoptysis. This patient was nursed carefully and when the labor started spontaneously, a modified form of twilight sleep was in- duced. No hemoptysis occurred ; the first stage was normal, and as soon as complete dilatation had occurred, delivery was effected with forceps. This patient improved and was alive one year later, although still suffer- ing from the pulmonary disease. This case is cited merely to show how difficult it is to foretell just what is likely to occur in a given case. The frequent hemoptysis that was present during the last months of preg- nancy, often brought on by a slight attack of coughing, made it seem likely that, with the onset of labor pains, a copious hemorrhage would take place, and while the labor was made as easy as possible for the pa- 256 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS tient, no hemorrhage of any kind occurred. Furthermore, the outlook for this patient, even if she did survive the hemorrhages, was extremely unfavorable. As a matter of fact, the dreaded puerperium was passed without any exacerbation, and improvement set in almost immediately after delivery. In the series of cases here under discussion, unfortunately, only too frequently the reverse occurred. Patients who were apparently favorable subjects suddenly developed symptoms of a marked exacerba- tion of the disease; in some cases this occurred during the pregnancy, but in more it took place during the puerperium. A number of authorities argue that, since no one can tell which ap- parently favorable cases will do well and which will do badly, the correct treatment of all cases of early pregnancy is, therefore, to empty the uterus, and thus be on the safe side. However, the reverse is also true, although, unfortunately, in a much smaller percentage of cases. Even the test of pregnancy is no certain criterion, as even the cases that do well during this period may suffer severe exacerbations during the puer- perium. Furthermore, apart from an exacerbation of the pulmonary condition, it seems but logical to assume that obstetric complications will develop more frequently in these patients than in normal individuals. In nearly all tuberculous patients, forceps, version, or some other form of operative delivery is indicated, and this in itself tends to increase the likelihood of sepsis, lacerations, and other complications, and thus to increase the maternal and infant mortality. The anemia and general weakened con- dition of many of the mothers also constitute a factor in increasing the proportion of dystocia and other obstetric complications. Lebirt 59 found that pregnancy had a bad influence on the course of tuberculosis in 75 per cent of cases. Deibel 43 found this to occur in 64 per cent of cases; von Rosthorn 60 in 70 per cent. Von Bardeleben 32 found this to be true in 71 per cent, and states that 47 per cent of these patients died during pregnancy, labor, or the puerperium. In all von Bardeleben's mild cases there was more or less, sometimes only slight, aggravation of symptoms during pregnancy or the puerperium ; in most of these cases the acute symptoms subsided, at least to some extent, in from 8 to 12 months. In this series 16 per cent were presumably closed lesions when the pregnancy occurred, 12 per cent were severe or acute cases, and all exhibited an aggravation of the disease, especially toward the close of pregnancy. Heiman's collected statistics (quoted by Schauta 37 ) showed that pul- monary lesions grew worse during pregnancy in 73.4 per cent. Pankow and Kiipferle 42 found that 94 per cent of their cases of active pulmonary lesions grew worse. Reiche 61 observed ill effects in JJ per cent, and PREGNANCY AND TUBERCULOSIS 257 Freund (quoted by Pankow and Kiipferle 42 ) in 38 per cent of cases. Of Lobenstine's 62 10 cases, all grew worse and only 4 survived labor for 3 months. Fellner 44 and Schauta 37 found that quiescent or mild chronic cases that had been well for a considerable period prior to pregnancy, suffered a relapse in 68 per cent of cases, Pradell's (quoted by Schauta 37 ) findings were even less favorable. In a series of 1035 cases he found that 95 per cent grew worse. Kunreuther 54 also emphasizes the dangers incident to this condition. Merletti 63 found that 50 per cent grew worse during pregnancy ; von Rosthorn, 60 70 per cent ; Kamina, fj4 50 per cent. Schauta 37 states that in tuberculous guinea pigs pregnancy dis- tinctly shortens the life of the animal. Schauta quotes the authorities from German sanatoria to the effect that only 25 per cent of tuberculous women were able to work 4 years after childbirth, and that all these are by no means cured cases. Albeck, of Norway (quoted by Schauta 37 ), found that of 16 cases, all of which were treated in private sanatoria and were, therefore, presumably receiving excellent treatment, 6 died within 15 months. Essen-Moller (quoted by Schauta 37 ) reports that death or aggravation occurred in 50 per cent of his series of sanatorium patients. Schauta states that in at least 75 per cent of all cases the disease was aggravated as the result of the pregnancy. Ebeler, 65 from a study of 32 cases, recommends the immediate emptying of the uterus uncondition- ally in every stage and in any month of pregnancy. Parry 41 reports that in her series of 38 cases, all of which were of the severe type, 50 per cent died within 2 months after labor. Fellner 44 observed a general maternal mortality of 9 per cent. Osier quotes Dubois to the effect that "If a woman threatened with tuberculosis marries, she may bear the first accouchement well, the second with difficulty, and the third never." Alals- bary 14 found the highest mortality among primiparae. Bacon 19 esti- mates that 33 per cent of tuberculous women who become pregnant die in less than one year following labor. Hoffman &6 found that the greatest mortality among tuberculous women was between the ages of 15 and 45 years (195.5 P er 100,000 population), whereas in men the highest mor- tality was between 45 and 64 years (254 per 100,000), indicating that many tuberculous women die as the result of pregnancy and childbirth. Schlimpert 67 asserts that the greatest number of deaths from tuberculosis during pregnancy occur in childbed. In reviewing the foregoing statistics, a number of facts must be taken into consideration. A certain number of cases of pulmonary tuberculosis will exhibit exacerbations, even when not pregnant, and this proportion must be deducted from the figures here given when considering the in- fluence of pregnancy upon the course of the disease. On the other hand, 258 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS statistics compiled from maternity hospitals, from which patients are discharged in one or two weeks after labor, no further trace being kept of them, are misleading in that no note can thus be made of the exacer- bations occurring in the late puerperium or during lactation. In this class belong the majority of statistics compiled from English and American hospitals. Only when the cases are carefully followed for at least six months (some authorities assert for two years or more) can accurate figures be obtained. Owing largely to their registration laws, the oppor- tunities for the German to gather figures were exceptional, and their sta- tistics are therefore valuable. In reviewing the literature on this subject, it must also be remembered that in Catholic countries the general feeling against the induction of abortion must be taken into consideration, and doubtless influences the view of many operators. Laryngeal Tuberculosis. — This variety of tuberculosis has its onset with very great frequency during pregnancy, and always influences the prognosis unfavorably. In the author's experience this complication oc- curred most commonly in cases in which the pulmonary lesions were active. However, Bandelier and Ropke 28 state that laryngeal tubercu- losis frequently appears when the pulmonary condition is showing few symptoms. During pregnancy laryngeal tuberculosis exhibits a marked tendency to extend, and this despite any form of treatment. Von Soko- lowski 68 states that he has observed cases of laryngeal tuberculosis that have endured pregnancy without developing any serious complication. In the author's experience this, however, is exceptional. Milligan 69 states that laryngeal involvement occurs in from 33 to 40 per cent of all cases of pulmonary tuberculosis. The clinical manifestations are difficulty in talking, due to weakness of the vocal cords, the voice becoming low and hoarse; the patient com- plains of a feeling of fullness or tickling in the larynx, and there is a frequent desire to clear the throat ; usually there is more or less difficulty in swallowing. Any symptom suggestive of this complication demands immediate investigation and a laryngoscopic examination to determine with certainty the condition present. Milligan states that hyperemia of one vocal cord often precedes for some time a more definite involvement. As has been stated, when laryn- geal involvement occurs, the prognosis becomes extremely grave. Whether or not this complication shall be regarded as an absolute indi- cation for the interruption of pregnancy, will be discussed under the head of Treatment. Local treatment of laryngeal tuberculosis is often of little avail. Some authorities recommend an application of 25 per cent argyrol or of some bland lotion; gargles, the swallowing or holding in the mouth PREGNANCY AND TUBERCULOSIS 259 of bits of ice, and the application of an ice bag externally may give tem- porary relief. Vagni 70 recommends electrocauterization. In extreme cases trachelotomy may be demanded as a life saving measure. Glas and Kraus 47 state that where there is ulceration, with relative stenosis, trachelotomy may materially improve the laryngeal condition. Healed laryngeal lesions are prone to undergo exacerbations, if pregnancy takes place. Practically all authorities recognize the gravity of laryngeal involve- ment in tuberculosis. Fellner, 44 in his series of 289 cases, had a maternal mortality of 44 per cent. Of 231 cases of laryngeal tuberculosis collected from the literature by Lobenstine, 62 200 died during pregnancy, in labor, or soon after — a mortality of 86 per cent. In this series of cases spon- taneous abortion and premature labor were not infrequent. Raspini 71 emphasizes the ill effects of laryngeal involvement. In the combined mortality statistics from all series of deaths from tuberculosis and preg- nancy, cases of laryngeal involvement constitute a very definite percentage. The death rate among the infants of these patients is about 60 per cent. Imhofer 72 reports a mortality of from 86 to 90 per cent in those cases in which laryngeal involvement ocurs; Kuttner, 90 per cent; Stockel, Lasogna, 74 Pankow and Kiipferle, 42 Lubliner, 75 von Sokolowski, 68 and others give practically similar figures. Influence of Lactation on the Course of Pulmonary Tuberculosis. — It is generally conceded that lactation exerts an unfavorable influence on the course of pulmonary tuberculosis. In practically all our cases the child has been taken from the mother and fed from the bottle, or, in a few instances, with a wet nurse. Among the extremely ignorant, bottle feeding is undoubtedly attended by a high infant mortality. In a few instances in our series breast feeding has seemed the lesser of two evils. In those cases the mothers were of a class that would make bottle feeding extremely dangerous, and in all these women the pulmonary lesions were mild and there was little or no expectoration. In two additional cases of which we have record the mothers began nursing their children after discharge from the Maternity Hospital, despite instructions and warnings. In both these cases the infants succumbed in less than one year, and both apparently from tuberculosis, although this is not certain, since post- mortems were not obtainable. Clinical signs of the disease were, how- ever, present in both instances. Tubercle Bacilli in the Maternal Milk. — The question as to whether the mother's milk is likely to contain tubercle bacilli is of at least theoretic interest. In 9 examinations performed by the author by means of animal inoculation, no tubercle bacilli were demonstrated. The 260 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS milk used for these experiments was obtained in each case less than one week after labor, except in one instance, in which a mother with a closed lesion was nursing a three months old infant. Two of the remaining cases w r ere in the last stages of the disease, whereas the remainder were moderately advanced. Tubercle bacilli have, however, been demonstrated in the mother's milk by Escherich, Rabinovitsch, and Kempner (all quoted by Malshary 14 ). Auche 76 has published an interesting article on this subject. Bandelier and Ropke 28 state that the danger of transference of tubercle bacilli to the infant through the mother's milk is a very real one. These authors believe that the milk of a tuberculous woman contains a toxin that lowers the resisting powers of the child. Cornet 18 was able in rare instances to demonstrate the presence of tubercle bacilli in the milk of tuberculous women. Recent investigations show that tubercle bacilli are found in the milk more frequently than was formerly believed, especially in the miliary variety of the disease, or in those patients suf- fering from an acute exacerbation. It is, therefore, undesirable to feed the infants with mother's milk, even if this be obtained by means of a breast pump. The chief danger of breast feeding to the child is, how- ever, due to accidental contamination, such as occurs from infected fingers carrying tubercle bacilli to the child's mouth, either directly or from in- fection of the nipples. Kissing and handling of the infant by the mother are a fertile source of infection, and these accidental contaminations are probably much more likely to occur than is a direct transference of the disease by tubercl-bacilli-bearing milk, and probably constitute the chief danger of nursing. Tuberculin in Pregnancy. — Kalabin (quoted by Schauta 37 ) recom- mended tuberculin in the treatment of these cases. More recent investiga- tors have not, however, confirmed the value of this remedy. Martin (quoted by Schauta) considers that a positive ophthalmic reaction in the tuberculous pregnant woman is a favorable sign, as indicating the pres- ence of a sufficient number of antibodies to protect the patient against extensive invasion. Veit, Kraus (quoted by Bandelier and Ropke 28 ), and Kaminer (quoted by Schauta 37 ) believe, as does also the author, that the test is of no value as a diagnostic sign. The cutaneous reaction is also valueless as a prognostic aid. Even for diagnostic purposes the cutaneous test during pregnancy becomes less certain in its results. Tuberculin During Lactation. — Palmer 77 states that he has for some time employed tuberculin as a diagnostic agent in certain cases. He has used it guardedly in this way from time to time, and has wit- nessed many pronounced reactions, without the slightest disturbances in the infant, although he is convinced that in at least three instances the PREGNANCY AND TUBERCULOSIS 261 breast fed infants were clinically tuberculous at the time the mothers were given the test. This observer reports one case in which the administra- tion of tuberculin to the mother was followed by a definite exacerbation in the infant, from which it died ten days later. He concludes that in this case it hardly seems possible that sufficient tuberculin could have reached the child to cause the slightest disturbance, and he is inclined to attribute the exacerbation in the infant to coincidence. Nevertheless, he directs that extreme caution should be employed in giving tuberculin to nursing mothers. Schlosmann 78 has employed the test in 49 nursing mothers ; in 18, or 36.8 per cent of these there was more or less reaction, but in none was the child affected in any way. At best, the use of tuberculin is not without danger. Prophylactic Measures. — Many authorities believe that tuberculous individuals should not marry. As a general rule, marriage is more harm- ful for tuberculous women than for tuberculous men. Indeed, many men appear to improve after marriage. The danger to their wives and possible progeny must, however, be taken into consideration. We believe that, as a general principle, it is correct to advise the tuberculous woman against marriage, but a hard and fast rule to this effect cannot be laid down. Certainly marriage should be advised against in the presence of any active lesion, no matter how limited in extent. On the other hand, it seems too radical an attitude to forbid the woman with a small, non-active closed lesion, which has been in abeyance for two or three years, to marry. Recent investigations seem to show that, at least among the intelligent, marital infection of tuberculosis is less frequent than was formerly be- lieved. Should a husband or a wife become infected, sanatorium treatment is advisable, at least for a time, not only for the good of the patient, but as a means of protecting the family and in order that the patient may learn prophylaxis, to guard others. Some authorities recommend sterili- zation of the tuberculous wife, if it becomes necessary to empty the uterus on account of the disease. This the author believes to be unjustifiable, except in exceptional circumstances. Knopf 79 is of the opinion that every man who has an active pulmonary tuberculosis should undergo a vasec- tomy, and that a bilateral salpingectomy should be performed upon every affected woman. Law Regarding the Marriage of Tuberculous Persons. — That the healthy individual who marries a tuberculous person runs some risk of contracting the disease is well known, the risk varying in degree with the type of the lesion and the intelligence of the contracting parties. The Supreme Court of New York (Special Term, New York County, Sobol 262 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS vs. Sobol, N. Y. Supp. 248; Reference from /. Am. Med. A. 1915, 64 1024) holds that fraudulent concealment of tuberculosis by a person en- tering into the marriage relation is ground for annulment of the mar- riage. In this case it was established that the defendant had been treated for tuberculosis prior to the time of his marriage, and that he knew that he was suffering from the disease. Before marriage he explained to the plaintiff that his symptoms were manifestations of a cold. Furthermore, it appeared that, a few days subsequent to the marriage, the defendant's condition was such as to require the services of a physician, who diagnosed the case as tuberculosis. The court based its opinion on the concealment of the disease, the possible effect on one of the contracting parties, and upon their posterity. The court deemed it proper, in view of the wide spread prevalence of tuberculosis and the disastrous consequences to those who suffer from it, to take judicial notice of its characteristics, for the purpose of this discussion. There can be no doubt that the disease is of an infectious nature, and that close association with a person thus af- flicted, unless attended with great care, exposes those coming in contact with such persons to the danger of infection. Tissier 80 states that the induction of abortion is against the law in France, and that it can be performed only by "the intelligent tolerance of the civic authorities." The chief danger to the tuberculous married woman is pregnancy, and her safest plan, regardless of the nature of her lesion, is to avoid con- ception. Although this may be a great hardship to her and her husband, there can be no doubt regarding the truth of this assertion. We have previously endeavored to emphasize the necessity for individualizing in the case of these patients. Occasionally a case may occur in which the lesion is limited in extent and has been inactive for not less than two years. Under such circumstances, if the patient is intelligent and able to avail herself of proper treatment and supervision, and if she is especially desirous of having a child, conception is justifiable. These cases are, how- ever, exceptional, and even under the most favorable circumstances such a patient materially increases the risk of bringing on an exacerbation of her disease. If one or two children are living at the time that the woman becomes infected or seeks advice, conception is best advised against in all cases. It is impossible to escape the fact that any form of pulmonary tuberculosis, no matter how limited in extent, is especially prone to become aggravated during the pregnancy and the puerperium. Some cases may do well, and, as the result of a limited experience, the physician may easily be led to underestimate the dangers of pregnancy. Unfortu- nately, despite the most painstaking study, we are as yet unable to deter- mine with certainty which case will bear pregnancy and the puerperium PREGNANCY AND TUBERCULOSIS 263 well, and which will fare badly. No positive prognosis can, therefore, be given in the case of an individual patient. At times, even those cases that appear most favorable will result disastrously, and occasionally, though unfortunately only in a small proportion of instances, the reverse will be the case. The safest plan for the woman, therefore, is to avoid concep- tion. In those exceptional cases in which conception has been counte- nanced, strict hygienic measures must be enforced, and the woman kept under close observation and examined at frequent intervals by an experi- enced internist. Dice 81 divides the non-active cases into two classes — first, the early cases, in which the patients are apparently cured, where the tuberculous process is arrested, and secondly, those in which the disease is fairly well advanced, but has been inactive for two or more years. Even in the most favorable cases, Dice advises against pregnancy, unless there has been a period of quiescence of not less than two years, and even in such cases he believes the dangers of pregnancy are by no means small. In deter- mining the extent of the pulmonary lesions in non-active cases, the X-ray has been found a valuable aid. Treatment of Pregnancy and Tuberculosis. — As a matter of fact, the physician is frequently not consulted regarding the advisability of either marriage or conception, and often sees the case for the first time after pregnancy has taken place. This is especially true of the ignorant classes, and even the intelligent are as yet not sufficiently educated upon this point. If pregnancy has taken place, the most important point to be decided is, shall the uterus be emptied, and if so, what are the indications for performing abortion. A General Hygienic and Dietary Treatment. — All cases of preg- nancy occurring in tuberculous women should be subject to a rigid hygienic and dietary treatment. This should be instituted as soon as tuberculosis is diagnosed, but it is especially important if pregnancy occurs. The pregnant tuberculous woman needs every possible aid in combating her infection. She should, therefore, be placed under the care of a physician who understands this special form of treatment. If it be- comes necessary to interrupt the pregnancy, there should be as little break in the hygienic regime as possible. If the weather is at all suitable, the patient's convalescence will be more satisfactorily accomplished by placing her out of doors. This is particularly true of those cases that have been accustomed to an out of door life prior to the operation. Even in the most favorable postoperative cases, the hygienic regime should be con- tinued for at least three, and preferably for six or more months, following the termination of pregnancy. 264 GYNECOLOGICAL AXD OBSTETRICAL TUBERCULOSIS These cases are, as a rule, best treated in a well conducted sanatorium, and if the operation cannot he performed there, the patient should be removed to such an institution as soon as possible after the opera- tion. EUagxosis. — Presuming that the diagnosis of tuberculosis has been established beyond doubt, the question of the attitude the physician shall assume is of the greatest importance. In cases of early pregnancy the diagnosis of the latter condition is somewhat difficult. Too much atten- tion must not be paid to amenorrhea as a diagnostic sign, as this is not an infrequent symptom in tuberculosis. In our series of 214 cases of tuberculosis in which the menstrual changes were especially studied, total amenorrhea was present in 5 per cent of cases, and scanty or irregular flow was observed in an additional 53 per cent of patients. Schauta 37 states that the opinion of the medical world regarding the treatment of pregnancy in the tuberculous may be divided into three groups : the first, the French school, which admits the unfavorable effect of pregnancy on the course of pulmonary tuberculosis, but declines to induce abortion, placing its hopes for success upon diet, hygiene, etc. ; the second group, which consists of those who individualize, and who induce abortion if the tuberculosis is advancing, but if it is not, employ general treatment and supervision ; and the third, which considers tuberculosis an uncondi- tional indication for abortion. The author is not in accord with any of these groups, but believes that the attitude toward any given case must depend upon the conditions surrounding it. In considering the subject, many factors must be taken into consideration, among the most important of which are the advance- ment of the pregnancy and the character of the pulmonary lesion, the social status of the patient, her intelligence, whether she is able and willing to observe proper hygienic and dietary precautions, her financial condition, her mental attitude, the question of whether she already has one or more children. These considerations are all factors of the utmost importance, and should be weighed carefully before determining upon the treatment to be instituted. No hard and fast rules that will be applicable to all cases can, therefore, be laid down. In the early months of pregnancy, with a rapidly advancing pulmonary lesion, there can be no question that the induction of abortion should be performed without loss of time, and this is also true if laryngeal involve- ment occurs. On the other hand, given a similar case in the late months of pregnancy, little can be gained by the induction of premature labor. The most dangerous period — the puerperium — will occur in any event, and under such circumstances it is usually better to direct all one's efforts PREGNANCY AND TUBERCULOSIS 265 toward establishing the well being of the child, as in any event the mother is probably doomed. Speaking on the broadest general lines, the cases of pregnancy in the tuberculous may be divided into two groups, according to the advance- ment of the gestation, the first group consisting of those cases seen prior to the fifth month, and the second, those encountered from the fi'fth month on. Indications for the Induction of An Abortion in the Tuberculous Prior to the Fifth Month. — The writer believes that in the presence of an extensive lesion, even in the quiescent stage, or even of a small active lesion, the uterus should be emptied at once. This also applies to those cases in which laryngeal involvement of any degree is present. The de- velopment of secondary tuberculous lesions in parts of the body other than the lungs is also an indication for this procedure in most cases. Excessive vomiting, renal insufficiency, and other complications of preg- nancy may, as in the normal woman, constitute indications for emptying the uterus. It must be remembered that the tuberculous woman has less resisting power than the uninfected one. Our object is to maintain her powers of resistance at their highest point — in other words, to improve her general health. This is of the utmost importance. Gastric dis- turbances or other complications that might be borne by the normal woman may be sufficient to lower the tuberculous patient's resisting pow- ers to such an extent that an exacerbation may occur. For this reason, intervention should be employed considerably earlier in the tuberculous woman and for a milder degree of complications than in the normal woman. Loss of weight is not in itself an indication for the induction of abortion. It is, however, a danger signal of great practical value. Veit 82 rightly lays special stress upon the prognostic value of a loss or a gain in weight. Women who lose weight in the latter months of pregnancy often succumb during the puerperium. As a general rule, the earlier the intervention, the better is the prognosis. A much more difficult point to determine is the attitude of the physi- cian toward the patient with a quiescent lesion of moderate or small extent. Here the patient must be studied individually, and the points previously referred to considered. It must be remembered that in every such case the woman runs an added risk by allowing the pregnancy to continue. It is conceded that intervention in the early months of preg- nancy is productive of at least moderately good results, but that interven- tion in the latter months of gestation is of little value. One of the chief dangers, therefore, in these cases is that the patient may do well until about the sixth or the eighth month, when it is too late to do any good 266 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS by emptying the uterus. As has previously been stated, it is impossible to determine with certainty which case will, and which will not, do well. On the other hand, it is by no means justifiable to advise the induction of abortion in every case. As a general rule, the longer the lesion has been inactive, the better is the prognosis. Lesions of limited extent and those that have never shown very marked activity are also more favor- able. A factor of the utmost importance is whether or not the patient is in a position to obtain proper hygienic and dietary treatment. If, during the course of observation, and prior to the fifth month, evidence of laryngeal involvement or an exacerbation of .any sort arises, the safest plan is to advise immediate intervention. Results of Abortion in the Tuberculous Prior to the Fifth Month. — Owing to the many factors that enter into their compilation, statistics are sometimes misleading. The types of cases from which they are drawn, and the closeness with which they are afterward followed, are points that tend to increase or to nullify their value regarding this con- dition. What may be considered justifiable grounds for emptying the uterus by one authority may not be so regarded by another. The physi- cian who routinely recommends intervention in all cases will naturally be able to show better figures regarding maternity mortality than will the one who waits for the onset of an exacerbation before emptying the uterus. Pradella (quoted by Schauta 37 ) attempted to classify his cases of pregnancy, regardless of their degree of advancement, according to the extent of their pulmonary lesion. He found that, in the first stage of tuberculosis, emptying of the uterus had a beneficial effect in 89 per cent of cases, in the second stage in 83 per cent, and in the third stage in 25 per cent. In cases of tuberculosis in the first stage and less than one month pregnant, Pradella found that the induction of abortion was successful in 91 per cent of all cases. Kaminer (quoted by Schauta 37 ) takes a more pessimistic view. He states that he has never seen a cure, but believes abortion to be of value in the early months of pregnancy. In moderately advanced or far advanced cases, however, he expects but little success, but believes that early abortion tends to limit the extent of the disease. The interruption of pregnancy, even in early cases, is not always followed by improvement. Veit, 82 Kronig, 83 and von Rosthorn 60 believe in individualizing each patient, and hold that the pregnancy should be interrupted in the early months in the event of untoward symptoms aris- ing. Veit 82 very properly declares that successive abortions are quite as injurious as one or two pregnancies, especially if the latter have been properly treated. This author has collected 347 cases, in which abortion PREGNANCY AND TUBERCULOSIS 267 was performed in patients with active lesions. Of these, 56.7 per cent were benefited, and the remainder were unimproved. Veit quotes von Bardeleben to the effect that only 50 per cent of the latter's active cases were improved. He points to the fact that in active cases abortion may be followed by the development of miliary tuberculosis. Trembley 84 reported 29 cases in which abortion was performed, with one recru- descence. Edgar 85 believes that it is best to assume the attitude of the alarmist in these cases. Knopf 79 states that the more of these cases he sees, the more inclined he is to favor radical treatment. Werner 80 re- ported 60 cases operated upon in Wertheim's clinic. Of these, 1 died from hemorrhage, 1 died of tuberculosis 4 months after leaving the hos- pital, and 4 were but little benefited. In none of these cases had the pregnancy advanced more than five months. All had active pulmonary lesions, laryngeal involvement, were in poor physical condition, or suf- fered from some other complication. Of 25 patients of this series, all of whom had been operated upon not less than one year before, 1 died of tuberculosis, 20 were feeling well, and in 4 the symptoms of the disease were either unimproved or had grown worse. Bossi 87 urges rapid mechanical dilatation, followed by curettage, and states that 40 cases terminated by this method before the sixth month gave good results. Schrerschewer 88 reports favorable results obtained in 10 cases operated upon by Bumm, and in 1 from the Marburg Klinik. Hysterectomy was performed in this series. Hoist, 89 as the result of his experience, recommends abortion in all cases, if the lesions are active. Schauta 37 takes a radical stand, and states: "We are of the opinion that in every case where tuberculosis is definitely diagnosed, the indication is to bring on abortion. Inasmuch as in at least 75 per cent of all cases the disease advances during pregnancy or in the puerperium, and as the time for advancement is uncertain, one may proceed to treat it too late. It is preferable to sacrifice the life of the child, which is in any case of doubtful value in the conditions present." Zirkel, 90 on the basis of Hofmeier's work, recommends the induction of abortion if there is a noticeable loss of weight or an aggravation of the symptoms, and Sergent 91 recommends that it be induced only in excep- tional cases. Von Franque 92 believes that abortion should be induced only when it can reasonably be expected to improve the patient's condi- tion. Stutz 40 reports 32 cases, and recommends immediate emptying of the uterus in every stage of tuberculosis and in any month of pregnancy. In 75 per cent of the cases in the first and second stages of tuberculosis there was a marked improvement in the objective symptoms after the termination of pregnancy. In patients in the third stage the prognosis is 268 GYNECOLOGICAL AXD OBSTETRICAL TUBERCULOSIS always bad. Von Bardeleben 32 states that there is a mortality of 2.54 per cent among incipient cases, in whom the uterus is emptied prior to the fourth month of pregnancy. Among ordinary cases from the fourth to seventh month similarly treated the mortality is from 20 to 25 per cent, and in advanced cases the death rate reaches 50 to 80 per cent. Pankow and Kupferle 42 state that the results in the Freiberg Frauen Klinik are relatively good, if abortion is performed early, but that a mortality of 40 per cent follows the operation when it is performed in the second half of pregnancy, and upon patients suffering from active lesions. Permin 93 urges the early interruption of pregnancy when progressive lesions are present. Crede and Holder 94 believe that the progress of the pulmonary condition can be divided into two stages, the first stage being marked by an infiltration of the apices and the parenchyma of the lungs, and by catarrh of the apices; the second, by cavity formation, the formation of infarcts, hemoptysis, and infiltration, sometimes of the entire lobes. In the first stage, if the woman is well nourished, there is no indication for the induction of abortion. The patient should be kept under observa- tion and examined at intervals by an internist ; she should be given sana- torium treatment and especial care during labor and in the puerperium, and she should not be allowed to nurse her child. If the patient in the first stage is poorly nourished, she may become worse during pregnancy ; labor is likely to be difficult, and in the puerperium exacerbations are of frequent occurrence. In these latter cases, if the pregnancy is of only a few weeks' duration, abortion should be induced; if it is of more than a few months' duration, it should be allowed to continue. The patient should be under constant supervision, breast feeding, and future conception prevented. In the second stage of the disease pregnancy is particularly dangerous. In spite of this, Crede and Horder believe that the treatment should be the same as that for a poorly nourished woman in the first stage of the disease. These cases should be indi- vidualized. In borderline cases abortion must often be performed. Sellheim 95 reports the result of operation upon 10 patients, none of whom were pregnant more than 5 months. The results in all cases were good. Stutz 40 reports 15 cases, 14 of which were operated upon through the vagina with good results. Peterson 9G is of the opinion that preg- nancy exerts a harmful influence on the course of tuberculosis, less when the disease is of the fibroid type, and is especially dangerous if a pleurisy or pneumonia should develop. He recommends individualizing all cases and advises inducing a premature delivery in some cases. McPherson 97 believes that, in every case of incipient tuberculosis accompanying preg- PREGNANCY AND TUBERCULOSIS 269 nancy, the pregnancy should be terminated. Jellett 98 believes that pul- monary tuberculosis is not influenced unfavorably by pregnancy. Rabnow and Reicher," Kohne, 100 and Cohn 101 are of the same opinion. Rabnow and Reicher report a series of 10 cases occurring in working women, all of whom had active lesions. Of these pregnancies 7 are reported to have had no injurious effects. Cohn reports that of 58 cases, 53 were apparently no worse for their pregnancies, and Kohne found the same to be true in 10 out of 22 cases. It is worthy of note that these writers do not report having followed their cases for very prolonged periods subsequent to childbirth. Van Tussenbroek, 102 from a study of the mortality statistics from Amster- dam and other Dutch cities, arrives at the conclusion that the mortality from tuberculosis during the first 6 months of pregnancy was increased; during the later months it was diminished in such degree that the two were about even. She found that the mortality from tuberculosis dur- ing the year following pregnancy was about equal to the general mor- tality among tuberculous women who had not been pregnant. The gen- eral opinion that the death rate is increased by pregnancy is, therefore, not borne out by this investigator's studies. These results are not in accord with the author's experience, or in fact with those of most observers. Permin 93 urges the necessity for terminating pregnancy when the disease is advancing. Williams ° 8 as- serts that the harmful influence exerted by pregnancy upon the course of tuberculosis is generally conceded. He believes that abortion should be performed on primiparae when the disease becomes manifest in the early months of pregnancy, but admits that premature labor is of little value. Davis 30 treats his patients individually, and in the early months of pregnancy he empties the uterus on the first symptoms of an exacerbation in previously mild or quiescent cases. When a patient with an active lesion, who is just holding her own, becomes preg- nant, Davis believes that, in the majority of cases, abortion should be, induced. The combined statistics of twenty-one observers, comprising nearly 1,000 cases, show that JJ per cent of women were benefited by the induction of abortion. The percentages vary from 20 to 97. The diversity of opinion regarding the treatment of this condi- tion is evidence in itself that no ideal plan has as yet been evolved. It will be noticed, however, that the general trend of opinion is toward interruption of pregnancy in the early months of gestation, and toward non-operative treatment in the second half. The author believes that the wise obstetrician will familiarize himself with the results obtained by others, and carefully consider the source and the methods employed in 270 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS compiling the statistics; that he will then individualize his cases, and empty the uterus only when it is necessary, and will not allow his natural repugnance to the performance of this operation to influence him to the detriment of his patient. The average obstetric surgeon's dislike to performing abortions is not only natural, but laudable, and may in some instances be guided by his religious belief. He may, however, comfort himself and his patient's family by following the advice of Pinard, 57 who emphasizes the fact that abortion is induced only to save the mother's life, for, as the mother will probably die without operation, abortion must, therefore, not be viewed as murder, since the child will perish if the mother dies. Unfortunately for this argument, in a certain propor- tion of cases the mothers live at least until the puerperium is reached. Furthermore, even apparently unfavorable cases may do well without intervention. Consultation and Precautions to be Adopted Prior to Empty- ing the Uterus. — Before deciding to empty the uterus, a consultation should always be held. If any doubt exists as to the certainty of the diagnosis of tuberculosis, an experienced internist should be called in. The services of a competent bacteriologist will prove an additional safe- guard. Freund 103 very properly believes that, in general, the internist's duties should consist in giving information to the gynecologist in regard to the pulmonary condition, but that he should not be the one to decide whether or not abortion should be performed, since the obstetrician or the gynecologist often has greater experience in this particular. The entire procedure should be conducted as openly as possible ; the family of the patient, and in most cases, the patient herself, should be informed of what is about to be done. No loophole for subsequent criticism should be left. The prognosis should in all cases be guarded, for benefit may not accrue from emptying the uterus, and the family should be so informed, and the true state of affairs explained to them as nearly as possible. ^Yith the patient herself a more optimistic view is justifiable. Choice of Operation. — If having decided upon intervention and having obtained consultation, and secured the consent of the family and of the patient, the next question to be determined is the method of op- eration. No matter what method is selected, the operation should be per- formed by an experienced obstetric surgeon in a well equipped hospital. Sterilization of the Tuberculous. — Many methods of operation have been advised, some operators advocating the vaginal and others the abdominal route. Some advocates of the former method and many of the latter recommend sterilization of the patient by one method or an- other, for the purpose of preventing future conception. Among these PREGNANCY AND TUBERCULOSIS 271 are Sellheim, 95 Stutz, 40 Werner, 86 von Franque, 92 Schottelius, 104 Hohne, 105 Ebeler, 65 von Bardeleben, 32 Martin, 106 Pankow and Kup- ferle, 42 Schauta, 37 Kunreuther, 54 and others. Many operators also ad- vise excising the placental site. The author believes that routine steriliza- tion is entirely unjustifiable, regardless of the method employed. Cer- tainly, in the average case in which the uterus is emptied there is some hope of effecting a cure of the tuberculosis. Some operators advise the performance of a type of operation that permits, if it should be desired at a subsequent time, of the reconstruction of the genital tract, so that conception may take place. Gauss 107 recommends effecting steriliza- tion by means of the X-rays. It is claimed for the X-rays that they may be applied so as to produce either temporary or permanent sterilization. Pincus 108 employs atmocausis for the purpose of producing sterilization. Without being influenced by a desire to evade the responsibility of the operation of sterilization, it is better to place the responsibility of preg- nancy upon the woman or upon her husband by advising them to avoid subsequent pregnancies, unless the pulmonary condition improves suffi- ciently to permit pregnancy to be carried out with a reasonable degree of safety. The operation of sterilization necessitates opening the peri- toneal cavity and prolongs the operative procedure, two factors that can- not fail materially to increase the mortality incident to the operation. Dice 81 advises that the surgeon safeguard himself from subsequent crit- icism or legal responsibility by obtaining a written agreement from the parties concerned before sterilizing the patient. Anesthesia. — The question of inducing anesthesia is of vital im- portance in these cases, and will be dealt with in detail in a subsequent chapter. It will suffice here to state that in every case of very early preg- nancy — under 4 or 6 weeks — curettage can often be performed without the aid of an anesthetic, a hypodermic injection of a 1/4 grain of morphin and 1/150 grain of scopolamin often being sufficient for the purpose. If the patient is a highly nervous woman, a few whiffs of nitrous oxide gas may be necessary. When performing this operation upon a con- scious patient, the author has sometimes employed a weak solution of cocain or eucain applied to the cervical canal by means of cotton soaked in the solution. In inactive cases Anderes 109 reports that at the Zurich clinic chloroform and oxygen in combination is employed, and in the presence of active lesions, spinal anesthesia is used with excellent results. As a general anesthetic nitrous oxide gas should be the choice. The anesthetic must, however, be carefully administered. Technic for Emptying of the Uterus Prior to the End of the Second Month of Pregnancy. — Prior to the eighth or ninth 2J2 GYXECOLOGICAL AND OBSTETRICAL TUBERCULOSIS week, curettage is the operation of choice, the uterus being emptied as nearly as possible at one sitting. If the cervix is unusually rigid, a preliminary cervical pack of gauze may be employed for from twelve to twenty-four hours prior to the operation. Since the operation should be as complete as possible, thorough dilatation of the cervix is to be obtained. Thorough dilatation increases the speed with which the op- eration may be performed, and also conserves the blood. The latter is of especial importance, and as no bleeding occurs until the dilatation is completed, this procedure should be thoroughly performed. After the uterus is empty, or as nearly empty as seems advisable, a gauze pack should be inserted. Shortly before the completion of the operation ergotin or pituitrin may be administered. The pack may be removed in about five minutes, and, if little bleeding occurs, it need not be reintro- duced. In the majority of cases, however, more or less oozing will con- tinue, and as it is highly important that the blood be conserved, the pack should, under such circumstances, be reintroduced, special care being taken to pack the fundus of the uterus firmly. Carelessness in this re- spect may result in a firm stopper of gauze in the lower uterine segment, above which more or less blood may accumulate. The pack should be removed within twenty-four hours. Any remnants of membrane, etc., that may have escaped the curet usually come away with the gauze at this time. It is generally advisable to administer ergot or ergotin at six-hour intervals for the first twenty-four or thirty-six hours sub- sequent to the operation. The Fowler position is beneficial, materially aiding by gravity in the drainage of the uterus. If it is not employed, or if there is an extreme retrodisplacement of the uterus, the patient should be encouraged to turn frequently upon her side or her face, thereby aiding uterine drainage. Thorough emptying of the uterus at the time of operation is far preferable to simple breaking up of the gestation sac, and means a more rapid convalescence and, in the long run, the conservation of blood. Convalescence. — At the end of twenty-four hours, or immediately, if the weather conditions permit, the patient should be removed out of doors and the hygienic and dietary treatment suitable for pulmonary tuberculosis continued. If the operation has been complete, the patient can usually be out of bed by the fifth day; if the gestation has been an early one, she may leave her bed the third day, provided no compli- cations have arisen and the pulmonary condition permits. Technic of Operation After Pregnancy Has Advanced Be- yond the Second Month. — Vaginal hysterotomy is as a rule the most PREGNANCY AND TUBERCULOSIS 273 satisfactory operation for emptying the uterus. It is performed as follows : The field is prepared, as for any vaginal operation. The cervix is pulled down, the bladder stripped off by blunt dissection, using the fingers wound with dry gauze until the peritoneum comes into view. The peritoneal cavity is not opened. At this stage it is advisable to administer a hypodermic of ergotin. The cervix is then split in the median line. The membranes then bulge into the wound and are incised. The amniotic liquor is evacuated and the fetus is delivered. A large gallstone forceps applied to the fetal head is an excellent instrument with which to accomplish this step of the opera- tion. A little care at this point of the operation will save much time. If the fetus is decapitated or an attempt is made to remove it piece- meal, much time is lost. After the delivery of the fetus, the placenta and membranes can be easily removed manually. The cervix is sewn up with interrupted No. 1 catgut sutures, and the vaginal mucosa su- tured in its original position. The uterus is then replaced manually in its normal position and the patient catheterized. To guard against unnecessary loss of blood, traction should be made upon the cervix continuously, the operator depending as much as possible upon blunt dissection. It is important that the operation be confined as much as possible to the median line, as thus a less vascular field is encountered. Few ligatures will be required, although hemostasis should be thor- oughly carried out. In one instance, in the case of a firm unyielding vaginal outlet with a high cervix, which could not be brought down, in an elderly primipara, the writer employed episiotomy, choosing this operation in preference to working through a small, unyielding opening that would have prolonged materially the vaginal hysterotomy. The after care of these patients is the same as that advised for those upon whom curettage has been performed. The patients are out of bed on the fifth day. The author has had uniform success with this operation. The operation requires from fifteen to twenty minutes, and it is surprising how little bleeding occurs. A preliminary dose of morphin, sometimes supplemented with scopolamin, is advisable. The patient should be placed in the lithotomy position, and the vagina and ex- ternal genitalia prepared for operation prior to the administration of the anesthetic. (See Chapter on Anesthesia in Pulmonary Tubercu- losis.) This operation has these advantages, that there is little bleeding, that the uterus can be completely emptied at the time of operation, that there is no subsequent oozing from the uterus, and convalescence is 274 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS rapid. The further the pregnancy is advanced, the more decided is the indication for vaginal hysterotomy. It must, however, be remembered that emptying the uterus in this class of cases is only indicated when the pulmonary lesions are active or advanced. Both of these are conditions which make the adminis- tration of a general anesthetic especially hazardous. In some cases, therefore, dilatation of the cervix, and simple breaking up of membranes and removal with placental forceps or a large dull curette of as much as possible of the products of conception, is the safer plan. This oper- ation has the further advantage that it can be performed under a local anesthetic, or in many cases without anesthesia of any kind, although the preliminary administration of morphin and scopolamin facilitates the procedure. The disadvantages of this operation are that it is almost impossible to entirely empty the uterus, and in many cases nothing more than breaking up of the membranes can be accomplished. The opera- tion is, therefore, followed by considerable free oozing which continues for two or three days, and in the end results in the loss of considerably more blood than does the vaginal hysterotomy. The uterus usually has to be packed with gauze and often it is necessary to renew this pack- ing. This causes discomfort to the patient and adds to the danger of infection. Furthermore, the convalescence is prolonged. Conservation of the patient's strength is of special importance; the prolonged convalescence, with the loss of blood incident to this operation, and the danger of a general anesthetic necessary with the vaginal hys- terotomy, are the factors which must be weighed against each other in the choice of a method of emptying the uterus. In the hands of a skilled surgeon vaginal hysterotomy is in the author's opinion usually the lesser of the two evils, although cases will be encountered in which the mere breaking up of the membranes is clearly indicated. As in all cases of abortion, retrodisplacement is to be guarded against. The same methods applicable to the non-tuberculous patient are satisfactory in these cases. The Treatment of Pregnancy Advanced Beyond the Fifth Month in the Tuberculous. — As a general principle, it may be stated that when the pregnancy has advanced beyond the fifth month, little benefit will be derived by the patient from the induction of abortion, since in any event the most dangerous period for the pregnant tuberculous woman will not be avoided. The puerperium will occur in any event, and the interruption of pregnancy in many cases only means shortening the life of the patient, for in advanced cases death is likely to occur at this time. Furthermore, although the life of the child is still of secondary PREGNANCY AND TUBERCULOSIS 275 importance to that of the mother, nevertheless from the fifth month on, the fetus must receive more consideration than in the early months of gestation. After the fifth month of pregnancy, little can be accom- plished beyond enforcing a strict general hygienic and dietary regime, and adopting the general treatment usually employed for the tubercu- lous. The indications for treatment are along general lines and every effort must be made to maintain the patient's strength and to improve her general health. Such patients are best treated in a sanatorium until the time for labor approaches. In mild or moderately advanced cases miscarriage or premature labor rarely occurs, but in advanced cases pre- mature labor is not uncommon. As the time for delivery approaches, it is usually preferable to place these patients in a maternity hospital. Some patients may stand the last few months of pregnancy well, but the risks are, however, great. In desperate cases, as a rule, the mother's condition is of secondary importance, as she is doomed in any event, and every effort should be directed toward establishing the well being of the child. The author has previously advanced the opin- ion that the high mortality among the infants of tuberculous mothers is due not to any congenital infection or special predisposition, but is chiefly the result of improper surroundings, faulty hygiene, bottle feed- ing, and the motherless condition of these children. Delivery of Tuberculous Patients. — Attention has been previously directed to the dangers of labor for the patient suffering from pulmonary tuberculosis. To recapitulate, chief among these are muscular exertion, exhaustion, increased and sudden changes in the blood pressure, the possibility of breaking down healed or partially healed pulmonary lesions, resulting in the liberation of virulent tubercle bacilli into the blood stream, edema of the lungs, and the squeezing out of organisms into the general circulation from the placental site and hemoptysis. Polak and Matthews 49 state that mild cases going to term may be com- pleted without causing alarm, while in advanced cases labor may be tedious, prolonged and fraught with many dangers to the mother, as, e.g., dyspnea, cough, hemoptysis, impending cardiac failure, pulmonary edema, pneumothorax, and, rarely, general dissemination of the infection through the lungs. These authors state that mild inactive pulmonary tuberculosis seems to have no effect per se upon the puerperium, hemor- rhage is no greater and involution is not retarded. In the more ac- tive and progressive cases, there is apt to be excessive hemorrhage, and involution may be tardy. These ill effects are no doubt due to the general asthenic condition of the woman at this time. With these dan- gers in mind, the general principles of the conditions of labor and de- 276 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS livery can readily be formulated. The delivery should be conducted with two ends in view, the birth of a living child, and the expenditure of as little physical exertion and strain on the part of the mother as pos- sible. Mild cases may stand labor well, but advanced cases are likely to be complicated by dyspnea, cough, hemoptysis, pulmonary edema, cardiac collapse, and pneumothorax. If the general condition is weak- ened, labor is slower, and the second stage is likely to be prolonged. Occasionally, in patients desperately ill, it may be necessary to in- duce premature labor in the interests of the child. In the mild quiescent case induction of premature labor is rarely indi- cated unless some obstetric complication exists, such as a contracted pelvis, etc. It should, however, be remembered that a decided effort should be made to shorten labor and to make it as easy as possible for the woman. With these points in view, premature labor, like the in- duction of abortion, should be performed after the seventh month for a lesser degree of obstetric complication than it should be in the non-tuberculous woman. Furthermore, these patients should not be per- mitted to go beyond the time. The induction of labor can be per- formed without anesthesia, and does not in itself aggravate the pul- monary lesion. It is, therefore, indicated, if the patient does not fall into labor at term, as the patient is thus spared the obstetric difficulties attendant upon the delivery of an oversized child. In cases exhibiting evidence of activity of the pulmonary lesion the induction of labor two weeks before term is generally advisable. Cesarean Section. — This operation may be indicated as a life saving measure for the mother or for the child. In one case the author per- formed cesarean section on a mother nearly at term, who had advanced bilateral ulcerative pulmonary lesions and who, during the last month or two of pregnancy, suffered from frequent profuse hemoptysis. Slight coughing or exertion often produced marked hemorrhages. Four weeks before the date of her expected accouchement this patient was removed from The Henry Phipps Institute, where she had been bedfast for two months. Almost the first labor pains produced a hemorrhage. Cesarean section under spinal anesthesia was performed without loss of time, and a living infant was the result. The mother survived the operation nine months. In this case it is almost certain that, without operation, the mother would have died during the first stage of labor, and the infant would also have been sacrificed. The author has on two occasions performed cesarean section strictly in the interest of the child. In the first of these cases the mother was almost moribund ; spinal anesthesia was employed, and the patient died while the abdominal PREGNANCY AND TUBERCULOSIS 277 wall was being incised. The second case was one of advanced bilateral ulcerative lesions with frequent hemorrhages. This patient was in the worst possible physical condition, and it seemed certain that the physical exertion of labor, no matter how guarded, would result in death of the woman. The heart sounds were fairly good, and cesarean section under spinal anesthesia was performed. The woman survived the oper- ation three months. In both instances, the infants were fortunately saved, both were well and apparently normal when six months old. Cesarean section should be performed only in desperate cases. In the average case, the general indications are especial care and symptomatic treatment in the first stage of labor, and the use of the for- ceps during the second stage. Preparations for a rapid delivery should be made and the instruments sterilized and the operating room in condition, so that an immediate operative delivery can be performed at any stage, if it should become necessary. Assistants and nurses should be at hand. The author believes that even the shock of a cesa- rean section is preferable to permitting a pulmonary edema to become advanced. Naturally, such an operation should be performed only in the first stage of labor, and only under very exceptional circumstances. As a rule, no such radical measures are required. If edema of the lungs sets in, labor must be terminated without delay, if either the mother or the child is to be saved. In the event of such an occurrence in the second stage of labor, extraction, either with the aid of forceps or by version, should be performed. No time should be lost, but the child should not be sacrificed by too great haste. It is likely that, in any event, the mother is doomed. The author makes it a routine measure in these cases, as soon as the labor pains become pronounced, to inhibit them with doses of morphin sulphate, usually gr. 1/16, given at frequent intervals to overcome at least the extreme severity of the pains. In no event should the drug be pushed to its physiologic limit. In a few cases a modified twilight sleep has been employed. After the completion of the first stage, operative delivery is generally indicated ; at such times it may be necessary' to administer the hypodermic more frequently, and occasionally even a few whiffs of nitrous oxide may be given. The latter should, however, be avoided if possible. Nevertheless, in most cases brief, light anesthesia is pref- erable to a prolongation of hard labor. In order, as much as possible, to minimize the dangers of a congenital infection in the child through the squeezing out of tubercle bacilli that may be present in the placenta, the cord should be ligated as soon as possible, and the interval of waiting for cessation of pulsation omitted. 278 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Puerperium. — If the weather conditions are at all favorable, the puerperium should be passed out of doors. In any event, the break- in the regular hygienic regime incident to labor should be as short a one as possible. Every effort should be directed toward improving the patient's general condition and maintaining her resisting powers. As soon as possible the patient should be removed to a well conducted sana- torium. Unfortunately, this is attainable only in a small proportion of cases. For the intelligent, and for those willing and able to adopt the proper hygiene and obtain efficient medical supervision, this is less necessary. It is, however, the poor and ignorant who especially require sanatorium treatment, and these are the patients who are, as a rule, unable to, and in some cases unwilling to, profit by it. Nursing. — Breast feeding should be forbidden, and the infant should be kept in a separate and preferably distant room from the mother. Although this may be a great hardship for the Woman, the less she sees of the child the better it is for it. The infant is put at once upon the bottle with a suitable milk mixture, or given a foster mother. The only treatment the mother's breast requires is a support- ing bandage. This need not, and in fact should not, be too tightly applied. For the first day or two after the milk comes the breast will feel heavy and full. In a small proportion of cases this will be more than a discomfort, and will amount to actual pain; in the latter case a few small doses of morphin may be administered. An ice bag over the breasts usually does much to diminish the discomfort. After the second or third day following the appearance of the milk, no further discom- fort is experienced, the flow of milk ceases and involution of the breasts sets in. Under no circumstances should massage or the breast pump be employed. These merely tend to prolong the period of discomfort. Tight bandaging increases the discomfort and has no effect upon the milk secretion. Belladonna and other local applications of similar char- acter are of doubtful value. A dose or two of Epsom salts and mod- erate restriction of liquids are probably of benefit and may be employed for a day or two. The Influence of Pregnancy upon Tuberculous Lesions Other Than Those of the Lungs. — Tuberculous lesions other than those of the lungs are frequently secondary infections, the primary focus being often situated in the lungs. For this reason a careful examination of the lungs should be made in every case. A negative chest examination and history, although it does not exclude a pulmonary lesion, at least shows that, if such a lesion is present, it is of limited extent and inactive and for practical purposes is not of sufficient gravity to influence the PREGNANCY AND TUBERCULOSIS 279 prognosis very materially. Its possible existence should, however, be borne in mind, and an examination made at regular intervals to de- termine the pulmonary condition. In general, tuberculous lesions other than those of the lungs are less affected by pregnancy than are pul- monary lesions. Tuberculosis of the osseous system is but little influenced by preg- nancy. Pinard 57 states that he has never seen a case of bone tubercu- losis aggravated by pregnancy. The author has seen a number of such cases, including lesions of the hip and spine, none of which has been unfavorably influenced by pregnancy. Tuberculous peritonitis, by involving the pelvic peritoneum, fre- quently results in sterility. Indeed, these patients are so ill that inter- course does not generally take place, and it follows that pregnancy is rare. Peritonitis may, however, develop during the course of preg- nancy, but even here it does not appear to be of a more severe type than in the non-pregnant. Abortion and premature labor are not, however, infrequent. Schmidt 110 has recorded the results obtained from the oper- ative treatment of 37 cases of peritoneal or genital tuberculosis, and in each case these results were excellent. Delassus 1X1 reports the history of a remarkable case. The patient, who' was 29 years of age, was op- erated upon for a tuberculous peritonitis. Later, the abdomen was opened during the course of a herniotomy, at which time the peritoneal condition was found to be improved. Eighteen months later she was delivered spontaneously of an 8-pound infant. Benestad 112 has re- corded the history of a case of acute tuberculous peritonitis occurring during the puerperium. Oppenheimer 113 states that women who have had a nephrectomy performed for renal tuberculosis and subsequently become pregnant run a decided risk not only of kidney insufficiency, but, if a tuberculous cystitis persists, the infection may spread rapidly to adjacent organs. Esch (quoted by Davis 30 ) has seen such patients stand the strain of pregnancy well. Davis reports a case in which eclampsia occurred. The patient survived, however, and was in good condition one month later. Treatment. — As a general rule, cases of tuberculosis affecting other portions of the body than the lungs should be treated along general sur- gical lines. In all cases hygienic and dietary measures are of prime importance. From the surgical aspect, intervention may be required. In the event of an acute exacerbation of the lesion, the induction of abortion may be indicated, just as it may in any other disease. The benefits to be derived from it are, however, much more questionable than when the lungs are attacked. Should pulmonary symptoms be- 280 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS come manifest, the treatment is the same as that previously suggested under the head of pulmonary tuberculosis and pregnancy. During labor the cases should be carefully guarded and every effort made to conserve the patient's strength. To this end, forceps should usually be employed during the second stage of labor and measures be taken to prevent undue exertion and exhaustion. As a general rule, these moth- ers should not nurse their infants ; the safer plan, for both mother and child, is to institute artificial feeding at once. LITERATURE i. Horn, E. Erfahrungen iiber die Natur und Behandlung der Phthisis Puerperalis. Arch. f. med. erfahr. 1804. 6:86. 2. Succow, G. C. F." Historia phthisicos pulmonalis purulentae in femme gravida ortoe et post partum sponte sanatae. Jena, 1822. 3. Herrieux, E. Union med. 1847. 1 1138. 4. Robert, A. Union med. 1847. J : T 4°- 5. Grisolle, A. Arch, gen de med. 1850. 22:41. 6. Dechambre, A. Gaz. med. de Paris. 1851. 6:639. 7. Tott, C. A. Ztschr. f. Gebh. 1851. 30:223. 8. Dubreuille, C. Rev. med. franc, et etrang. 1851. 2:649. Also, Ann. de med. beige. 1852. 1:21, 366. 2:98. Also, Bui. acad. med. 1851. 22:14. 9. Lassegue, P. De l'influence de la grossesse et de l'etat puerperal sur la marche de phthisic Paris, 1856. 10. Warren, E. The Influence of Pregnancy on the Development of Tuberculosis. Philadelphia, 185 1. Also same paper under title Does Pregnancy Accelerate or Retard Development of Tuberculosis of the Lungs in Persons Predisposed to This Disease? Jr. Med. Sc. 1857. 34:87. 11. Thomas, T. G. N. Y. Jr. Med. 1859. 12:238. 12. Caresme, A. A. Recherches cliniques relatives a Tinfluence dc la grossesse sur la phthisie pulmonaire. Paris, 1866. 13. Ortega, S. These de Paris. 1876. 14. Malsbary, C. E. Am. Jr. Obst. 1905. 57:28. 15. Gassner. Monschr. f. Gebh. 1862. 19:1. 16. De Lee, J. B. The Principles and Practice of Obstetrics. Philadelphia and London, 1913. P. 107. 17. Simmonds, M. Arch. f. Klin. Med. 1886. 38:571. PREGNANCY AND TUBERCULOSIS 281 18. Cornet, G. Tuberculosis. New York and London. 1904. p. 416. 19. Bacon, C. S. Jr. Am. Med. A. 1913. 61 750. 20. Vejas. Arch. f. Gyn. 1912. v. 95. 21. Hofbauer. Zentrbl. f. Gyn. 1908. p. 1196. 22. Stengel and Stanton. Univ. Penn. Med. Bui. Sept., 1904. 23. Norris, G. W. Blood Pressure. Philadelphia and New York. 1914. p. 347. 24. Wiessner, M. Uber das Verhalten des Blutdruckes wahrend der Menstruation. Leipzig, 1904. 25. Heynemann. Ztschr. f. Gebh. u. Gyn. 1913. 74:854. 26. Dietrich. Arch. f. Gyn. 191 1. 94:394. 27. Brooks, C., and Leuckhardt, A. B. Am. Jr. Phys. 1915. 36:104. 28. Bandelier and Ropke. A Clinical System of Tuberculosis. London, 19 13. 29. Sergent, E. Rev. prat, d'obst. et de paed. 1914. 27:47. 30. Davis, E. P. Ther. Gaz. 1915. 39:153- 31. Friedrich, M. Arch. f. Gyn. 1913. 101 :376. 32. von Bardeleben, H. Deutsch. Med. Woch. 191 1. p. 764. Also, Berl. Klin. Woch. 1912. No. 37. 33. Hanau. Ztschr. f. Klin. Med. 1887. 12:1. 34. Fishberg. N. Y. Med. Jr. 1909. 2:1166. 35. Jacob und Paunwitz. Entstehung und Bekampfung der Lungen-tuberkulose. 1901. 36. Trembley, C. C. Tuberculosis and Pregnancy. Saranac Lake, 1912. 37. Schauta, F. Monschr. f. Gebh. u. Gyn. 191 1. 33:265. 38. Funk, E. H. Med. Clin. No. Am. 1918. 2:803. Also, Ther. Gaz. 1915. 39 :I 58- 39. Maragliano. Gac. d. osp. 1899. 14:1193, 1225. 40. Stutz, G. Ztschr. f. Gebh. u. Gyn. 1913. 73 -397- J 9 l 4- 6 :87. 41. Parry, A. Am. Jr. Obst. 1914. 70:94. 42. Pankow, O. R., und Kupferle, L. Die Schwangerschaftsunter- brechung bei Lungen und Kehlkopftuberkulose. Leipzig, 191 1, G. Thieme. 43. Deibel. Inaug. Dis. Heidelberg, 1899. 44. Fellner, O. O. Wien. Med. Woch. 1904. p. 11 58. 45. Silberman. Quoted by Malsbary, No. 14. 46. Dirner. Quoted by Malsbary, No. 14. 282 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS 47. Glas, E., and Krause, E. Klin-ther. Woch. 1908. No. 50. 48. Trembley, C. C. Jr. Am. Med. A. 1909. 53:989. 49. Polak and Matthews. Surg., Gyn., Obst. 1915. 21:235. 50. Weinberg, W. Beitr. z. Klin. d. Tuberk. 1908. 11:299. 51. Zirkel, K. Wurzburg, 1908, F. Standeraus. 52. Miller and Woodruff. Jr. Am. Med. A. Mar. 27, 1909. 53. Floyd and Bowditch. Bost. Med. Surg. Jr. Feb., 19 10. 54. Kunreuther, M. Berl. Klin. Woch. 1914. 51 :i628. 55. Armand-Delille. Am. Jr. Obst. 1912. 2:664. 56. Kingsford, L. Lancet. Sept. 24. 1904. 57. Pinard. Ann. de gyn. et d'obst. June, 191 2. 58. Williams, J. W. Obstetrics. New York and London, 1903. 59. Lebirt. These de Paris. 1909. 60. von Rosthorn. Wien. Med. Woch. 1908. No. 50. 1909. No. 1. 61. Reiche. Munch. Med. Woch. Sept. 19, 191 1. 62. Lobenstine. Am. Jr. Obst. 1913. 67:363. 63. Merletti. Arch. Ital. di gin. 1904. 2 14. 64. Kamina. Deutsch. Med. Woch. 1901. 35 :587. 65. Ebeler, F. Prakt. Ergeb. d. Gebh. u. Gyn. 1914. 6:87, 443. 66. Hoffman. Pub. Dep. Med. Jefferson Med. Coll. 1914. 67. Schlimpert. Arch. f. Gyn. 90:121; 1911. 94:863. 68. von Sokolowski, A. Ztschr. f. Lar. u. Rhin. 1909. 2 1575. 69. Milligan, W. Brit. Jr. Tuberc. 19 12. 70. Vagni, D. A. Sem. med. 191 5. 22:24. 71. Raspini. La gin. 1913. 10:249. 72. Imhofer, R. Prag. Med. Woch. 1914. 39:3. 73. Kuttner, A. Ann. des mal. de Tor., du lar. 1907. 33:445. 74. Lasogna, F. Arch. ital. di otol. 1914. 25 :io. 75. Lubliner, L. Med. i Kron. lek. 1910. 45 489. 76. Auche, M. B. Jr. de med. de Bordeaux. 1914. p. 93. • yy. Palmer, G. T. Jr. Am. Med. A. 1915. 64:1312. 78. Schlossmann. Monschr. f. Gebh. u. Gyn. 19 13. 17:1311. 79. Knopf, A. N. Y. Med. Rec. June, 1906. 80. Tissier. Arch. mens, d'obst. et de gyn. 1913. 2 :52. 81. Dice, W. G. Am. Jr. Obst. 1915. 71 ^97. 82. Veit. Versl. Deutsch. Naturf . u. Arz. in Cassel : Abt. f . Gebh. u. Gyn. Sept. 21, 1913. 83. Kronig. Versl. Deutsch. Naturf. u. Arz. in Cassel : Abt. f . Gebh. u. Gyn. Sept. 21, 1913. 84. Trembley, C. C. Fr. Trans. N. Y. Obst. Soc. 1910. PREGNANCY AND .TUBERCULOSIS 283 85. Edgar, J. C. Am. Jr. Obst. 1913. 67:363, discussion. 86. Werner, P. Zentrbl. f. Gyn. 1913. 37:1581. 87. Bossi. Med. nuova. 1914. 5:19. 88. Schverschewer, D. Munch. Med. Woch. 1909. p. 2656. 89. Holst, M. Munch. Med. Woch. 1905. p. 417. 90. Zirkel, K. Munch. Med. Woch. 1908, p. 1802. 91. Sergent, E. Presse med. July 5, 1913. 92. von Franque, O. Wiirzb. Abhl. a. d. Gesgeb. d. Prakt. Med. 1913. 14:1. 93. Permin, G. E. Hosp-tid. 1914. 57: No. 28. 94. Crede und Holder. Tuberkulose und Schwangerschaft. Elfte Int. Tub. Kong. Ber. 1913. p. 372. 95. Sellheim. Monschr. f. Gebh. u. Gyn. 1913. 38: No. 2. 96. Peterson, R. The Practice of Obstetrics. Philadelphia and New York. 1909. McPherson, R. Am. Jr. Obst. 191 5. 71 :3C>3. Jellett, H. A Manual of Midwifery. London, 1910. p. 569. Rabnow und Reicher. Deutsch. Med. Woch. 191 1. 37:^019. Kohne. Beitr. z. Klin. d. Tuberk. 191-1. 21:17. Cohn. Beitr. z. Klin. d. Tuberk. 1913. 26:71. Van Tussenbroek, C. Arch. f. Gyn. 101 : No. 1. Freund, H. Gyn. Rundsch. 1914. 7:313. Schottelius. Beitr. z. Klin. d. Tuberk. 20: No. 2. Hohne. Med.-Klin. Feb. 23, 191 3. Martin. Sam. Klin. Vortr. 1912. No. 665. Gauss. Zentrbl. f. Gyn. 191 1. 35:1004. Pincus. Centrbl. f. Gyn. 1902. No. 8. Anderes, E. Monschr. f. Gebh. u. Gyn. 1914. 6:87. Schmidt. Ztschr. f. Gebh. u. Gyn. 1913. 73: No. 2. Delassus, M. Rev. prat, d'obst. et de gyn. 1913. 21 : No. 2. Benestad, G. Norsk mag. f. laeg. 1914. 75 : No. 9. Oppenheimer. Monschr. f. Gebh. u. Gyn. 1914. 4°: No. 1. CHAPTER XII MENSTRUAL DISTURBANCES IN CONJUNCTION WITH PULMONARY TUBERCULOSIS Classification according to types — General considerations — Etiology — Theories advanced — Later observations — Chief indication for treatment — Dysmenorrhea — Clinical reports — Use of tuberculin — Scanty menstruation — Statistics — Irregular scanty flow — Amenorrhea — Cases studied — Menorrhagia — Vicarious menstruation — Periodic hemoptysis — Cases cited — Leukorrhea — Influence of menstruation on temperature in pulmonary tuberculosis — Cause — Consideration — Precautions in- stituted — Bibliography. GENERAL CONSIDERATIONS Menstrual disturbances frequently occur in conjunction with pul- monary tuberculosis. As a general thing, they tend towards a lessening of the loss of blood, increase in pain, or both. The disposition towards a scanty flow may occur at any time during the course of the pulmonary disease, but is most frequent in advanced or acute cases. Dysmenor- rhea, on the other hand, is common, even in the early stages of the dis- ease. In 234 ambulatory cases of pulmonary tuberculosis observed by the author, all of whom were free from pelvic disease, and whose ages vary from 17 to 39 years (the average being 28 years), the following menstrual disturbances were observed : Per cent Normal 23 Abnormal JJ Dysmenorrhea 72 severe 30 Scanty flow, fairly regular 53 Irregular, scanty 10 Amenorrhea 5 Menorrhagia 0.8 Vicarious menstruation 0.43 Macht's ] findings are in accord with our own. Classified according to the ordinary types of menstruation, Macht found : 284 MENSTRUAL DISTURBANCES 285 Per cent Regular, no change in 51.6 Amenorrhea (scanty or complete) 27.3 Irregular (some menorrhagia or amenorrhea) 8.3 Menorrhagia 4.6 Pregnant (in which amenorrhea could be ac- counted for on grounds other than tuber- culosis) 4.4 Menopause (artificial or otherwise) 3.8 In considering the menstrual disturbances resulting from pulmonary tuberculosis, it is important to remember that even in the normal woman the standard is a variable one; what is normal for one individual may readily be abnormal for another. Dysmenorrhea is also a relative symp- tom, the amount of pain which will keep one patient in bed may be but little complained of by another. The patients comprising our series have all been personally interviewed by the writer and particular care has been exercised to obtain an accurate menstrual history. It should be stated that some of the cases of diminished flow gave a history of a previous period of increase in flow, as a rule preceding for a short period the lessening of the flow. Etiology. — Before considering in detail the various menstrual dis- turbances, a study of their etiology is advisable. At the outset it must be remembered that not only are many of these symptoms, such as dysmenorrhea, scanty flow and menorrhagia, relative symptoms, for which the normal standard can be obtained only by studying the indi- vidual patient, but also that, even when pathologic in their degrees, they are present more or less frequently in otherwise healthy non-tuber- culous women. For this reason special care must be observed in classi- fying the various symptoms, and judgment must be exercised before declaring that in any given case the menstrual disturbance is the result of tuberculosis. Needless to state, all cases in which there are distinct pelvic lesions should, with possibly one exception, be excluded. The possible exception is of those cases of pulmonary tuberculosis suffering from menstrual disturbances and complicated by hypoplasia of the genital organs. Whether or not there is a relationship between hypo- plasia of the genital organs and tuberculosis will be considered sub- sequently. All patients were excluded from this series who were suffering from a combination of tuberculosis and some other disease when the char- acter of the latter might in itself influence menstruation. 285 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS The following theories have been advanced to account for the menstrual disturbances which occur in these patients : (i) Thorn, 2 from a study of uteri from tuberculous patients, be- lieves that he has found an almost uniform atrophy of the uterus with degenerative changes in the blood vessels, in general, similar to those present in the senile organ. (2) That the menstrual changes are the result of a toxemia or analogous condition, which reacts upon the ductless glands, resulting in ovarian changes, either histologically or physically. Butner 3 believes that the menarche of the tuberculous girl is delayed or absent, not be- cause of an economic anabolism or conservatism of nature, but rather from a catabolic toxin being elaborated by the growth of the tubercu- losis, which has a selective action in some unknown way over menstrua- tion, probably by the influence of the toxin on the internal secretion of the ovary. (3) That pulmonary tuberculosis, when occurring im early life, tends to prevent the complete development of the genital organs, so that many of these patients have hypoplasia of the uterus, of the ovaries, or both. (4) That the menstrual disturbances are the result of a poor gen- eral condition incident to the tuberculosis, such as anemia, hydremia, general loss of strength, etc. Regarding the first theory, the author's studies have not confirmed the findings of Thorn. The more modern view, that in general the uterus is of secondary importance to the ovary in the function of men- struation, is apparently based upon a firmer scientific basis than is the older view that the uterus is chiefly to blame for abnormal changes in the menses. In regard to the second theory, De Jong 4 has studied the ovaries from a large series of tuberculous women and finds the external appear- ance of the ovary in these patients to be variable; it may be smooth or furrowed. Comparison of the ovaries of tuberculous and non-tubercu- lous women shows that there is no marked difference in size, but the weight of the former is less. Tuberculosis does not affect the number of primordial follicles, nor does it destroy them. De Jong does, however, believe that to some extent it prevents their proper development. This results in a lessened number of corpora lutea. This may, in part, account for the scanty menstruations and dysmenorrhea so common in these patients, as it is accepted that the luteum cells exert a definite influence upon menstruation. Poncet and Leriche 5 believe that many of the sclerotic lesions in MENSTRUAL DISTURBANCES 287 the pelvis are due to tuberculosis occurring in early life, such as micro- polycystic degeneration of the ovaries, fibrosis of the uterus, hydro- salpinx, and hypoplasia. The amenorrhea and other menstrual dis- turbances, they believe, are a direct manifestation of the disease and not merely the result of a general dyscrasia. Sessa, 6 in the study of the changes in ovaries of children result- ing from infectious diseases in children under 5 years of age, dead of acute or chronic infectious diseases, and who had exhibited no symptoms suggestive of ovarian disease, found no macroscopic change in the ovary. Microscopically, more or less pronounced changes were observed. In tuberculous patients there were generally interstitial changes with more or less infiltration by chronic inflammatory products. Grafenberg, 7 Schiffman 8 and others have pointed out the frequency with which underdeveloped uteri are present in these cases, Hegar, Merlitti, 9 de Rouville, 10 and others have emphasized the frequency with which genital tuberculosis occurs in hypoplastic organs. As genital tuberculosis is, in 90 per cent of cases, a secondary infection, these latter observations have definite bearing upon the subject under dis- cussion. In the entire series of 234 cases constituting the author's study, II patients had what might be termed "infantile uteri." Of these 9 exhibited more or less scanty flow, and 8 definite dysmenorrhea. Hypo- plasia of organs other than the genital tract is not especially frequent in the tuberculous. Furthermore, hypoplasia in the non-tuberculous is relatively a frequent condition. To prove this theory it would, there- fore, be necessary to show that hypoplasia of the genital organs was more frequent in the tuberculous than in the non-tuberculous. Nat- urally, hypoplasia of the genital organs can only be attributed to tuber- culosis when the infection has originated at a period prior to that in which the development of the genital organs occurs. The generally accepted theory that many cases of pulmonary tuberculosis are the re- sult of infection in early life, and only become manifest later, is, how- ever, to be considered. It appears, moreover, improbable that hypo- plasia of the genital tract should be attributable to these early infections, which are inactive. It is possible that a moderately active tuberculosis occurring at a period during or prior to the development of the genital organs may have some inhibiting action on the development of the uterus or ovaries. This, however, is not yet proven. The fourth theory, that the menstrual disturbances are the result of a general malnutrition, appears to afford the most probable explana- tion in the majority of cases. It is true that some patients, especially 288 GYNECOLOGICAL AXD OBSTETRICAL TUBERCULOSIS those suffering from dysmenorrhea, are often in comparatively good condition and exhibit little anemia. The majority of tuberculous pa- tients are distinctively below par, and often show more or less blood changes. Practically all the anemias produce menstrual disturbances. The menstrual disturbances usually accompanying chlorosis are in gen- eral strikingly similar to those occurring in tuberculosis. All the in- fectious fevers are prone to produce menstrual changes; menstrual dis- turbances are, therefore, only what would be expected in tuberculosis. It is probable that the menstrual disturbances resulting from pulmonary tuberculosis may be the result of a number of conditions, and that either the toxemia theory, or the general malnutrition theory may be applicable to certain cases. From our own findings we attribute little weight to the theory based upon a hypoplasia of the genital tract. In our series hypoplasia has been present, but not more frequently than might be expected in a series of non-tuberculous patients. Especially is it important to emphasize the fact that menstrual dis- turbances are more likely to occur in women under 35 years of age. If the pulmonary lesions become manifest after 35 years of age, severe menstrual disturbances are less frequent. In those in whom the tuber- culosis has appeared earlier, menstrual disturbances are likely to be somewhat lessened after this age. An early menopause is frequent in the tuberculous. In tuberculous girls the onset of menstruation is often delayed. Treatment. — As menstrual disturbance is so frequent in the tubercu- lous, these patients should be especially guarded at the time of the flow. The chief indication for treatment in all these cases should be directed towards the pulmonary condition, as it follows that, if the cause of the disturbance can be improved, the menstrual abnormality will tend to improve. As a general rule, they are better in bed for a few days prior to the flow, and for the first day or two of the menstrual period. The bowels should be regulated with great care, and if there is a tendency towards dysmenorrhea, especially if it is of the congestive type, a brisk purge is advisable. In cases of excessive flow, care should be observed to conserve the strength by checking an abnormal loss of blood. With- out exception, all such cases should be confined to bed during the period of greatest bleeding. Unfortunately many women suffering from tu- berculosis are, on account of their social surroundings, unable to stay in bed for two, three, or more days each month. Nevertheless, these patients should be advised against physical exertion at these periods, and if they cannot stay in bed or spend considerable portion of the time MENSTRUAL DISTURBANCES 289 upon a couch, should at least endeavor to guard against undue exertion. Dysmenorrhea. — In the series of cases studied from which these conclusions have been drawn, 72 per cent of patients complained of more or less dysmenorrhea. This in itself is not an unusual propor- tion. The researches of Tobler, 11 Schaffer, 1 - and others have shown that at least 70 to 75 per cent of otherwise normal women suffer more or less at the time of the flow. Schaffer found that dysmenorrhea severe enough to be classified as pathologic was present in 14 per cent of his cases. That 30 per cent of our series suffered from severe dysmenorrhea is, however, excessive. This latter group, consisting of 70 cases, was studied as to the type of dysmenorrhea present, with the following results : 5 cases were of a purely obstructive type of dysmenorrhea, i.e., the pain appeared simul- taneously with, or a few hours before, the onset of the flow, was cramp- like or expulsive in character, often simulating miniature labor pains, frequently temporarily relieved by the expulsion of a clot, and was most severe for the first third of the menstrual period. Forty-eight were plainly congestive in type, i.e., the pain began some time before the onset of the flow, in some instances two or three or more days, was of a dull, heavy aching character, experienced over the lower abdomen, sides, and back, and sometimes extending into the thighs, usually some- what relieved after the first day or two of the flow. The remaining 17 cases were of a mixed type and could not be classed as either pure con- gestive or expulsive dysmenorrhea, neither one nor the other type pre- dominating sufficiently to warrant classifying them with any degree of certainty. In the majority, however, the congestive symptoms were the most marked, the congestive type in the characteristic variety of dysmenorrhea resulting from tuberculosis. Simple dilatation or split- ting of the cervix is a failure in this type of case, as there is no stenosis of the canal, and, therefore, no indication for such an operation. Of our 70 cases of dysmenorrhea, 58 were under 35 years of age. As already stated, the characteristic type of tuberculous dysmenorrhea is the congestive type, and occurred with sufficient severity to constitute a definite symptom in one-fourth of all our cases. In not a few of our cases, dysmenorrhea has been the symptom of which the patient com- plained more than any other. In one patient it was so severe that on one occasion she attempted self-destruction. In 48 per cent of our cases, the dysmenorrhea appeared early in the course of the tubercu- losis, in this confirming the findings of Macht, 1 who observed 45.8 per cent of his cases of tuberculous dysmenorrhea develop during the first stage of the disease. Hollos and Eisenstein 13 found dysmenorrhea an 290 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS early symptom in tuberculosis. They point out the frequency with which it is present and urge a physical examination of the chest in all patients suffering from this symptom. In nearly half (48 per cent) of our cases the dysmenorrhea ap- peared early in the course of the disease, and is, therefore, a sign of some diagnostic importance, and should suggest the possibility of tuber- culosis being present in all patients suffering from dysmenorrhea. This is a point which should be emphasized. In not a few cases dysmenor- rhea is the chief symptom and will be the one from which the patient will seek relief. If an examination of the chest is not made and only the ordinary treatment for dysmenorrhea instituted, not only will the latter be unaffected, but the lung lesion may be given time to advance, and, if a dilatation under ether is performed, the latter may be the means of causing an exacerbation of the pulmonary condition. Expulsive Dysmenorrhea. — All tuberculous patients suffering from dysmenorrhea must be studied individually. When found suffering se- verely from a definite obstructive type of dysmenorrhea, in some instances a dilatation operation may be performed. Operative relief should, how- ever, be withheld, unless the dysmenorrhea is of an unusually severe type and the pulmonary lesion mild. In other words, these cases are to be treated as any other surgical case, complicated by a pulmonary tubercu- losis. Dysmenorrhea is never a fatal symptom, whereas operative inter- vention may be the means of lighting up the pulmonary condition. This type of dysmenorrhea is not of tuberculous origin and occurs merely incidentally in tuberculous subjects. Congestive Dysmenorrhea. — A more or less pure congestive type is common, and is frequently of sufficient severity to require treatment. Our experience at the Henry Phipps Institute in Philadelphia has shown that in general the severity of the dysmenorrhea waxes and wanes with the exacerbations or improvement of the pulmonary condition. If it is possible to build up the patient's general condition, so that she will show steady gain in weight, increased hemoglobin, etc., the dysmenorrhea be- comes less marked, whereas in the presence of an active pulmonary lesion and the patient generally going down hill, the dysmenorrhea is likely to become worse. Treatment. — The treatment of these patients should be along the lines of the treatment instituted for the tuberculous. Because the dys- menorrhea is prone to develop early in the course of the tuberculosis, and may in some cases be temporarily the dominant symptom, and because of the general tendency of surgeons and others to perform dilatative operations on all cases of dysmenorrhea, regardless of their origin or MENSTRUAL DISTURBANCES 291 type, it is of the utmost importance that tuberculosis be excluded before any operative measures are attempted. Eisenstein and Hollas 13 report that among 70 cases of dysmenorrhea, in 23 tuberculosis was demon- strated. Grafenberg 7 reports that at the Kiel Clinic all cases of dys- menorrhea are examined for tuberculosis, and not only is a physical ex- amination performed, but the tuberculin test is also employed. Of 30 patients tested by the latter means 21 reacted with fever, and all gave a general reaction attended with local exacerbations of the trouble for which they applied to the clinic. Grafenberg states that, should the test be followed by a general and local reaction, no operative intervention should be attempted, and quotes Prochownik's warning against curettage in cases of genital tuberculosis. He states that where there is no local reaction to the tuberculin test, operation may be safely employed. Operation offers little hope of relief in the congestive type of dys- menorrhea, regardless of its primary origin, and in the tuberculous cannot by any means be regarded as free from danger. Grafenberg calls at- tention to the frequency with which tuberculous patients in poor general condition suffer from dysmenorrhea. Eisenstein and Hollas 13 found a positive tuberculous skin reaction present in a large series of women suffering from menstrual disturbances. The latter observers report that in 22 cases of dysmenorrhea treated with tuberculin by the Spengler method, 16 were cured. The results in amenorrhea were reported as even more satisfactory. The author's experience with tuberculin has been too limited to draw- conclusions from it. In the great majority of cases, if the general health can be improved, the dysmenorrhea will improve. During the carrying out of the general hygienic and dietary 'treatment, these patients should have special treatment during the menstrual and pre-menstrual periods. At these times the patient should be confined to bed, or at least to a reclining chair. One or two purgations, accomplished either by Epsom salts or castor oil, are often of benefit in relieving the dysmenorrhea. These should be given so that they will act during the height of the pain. A warm soapsuds enema administered at this time is also of benefit. Hot applications to the lower abdomen also relieve pain. In severe cases small doses of phenacetin may be tried. Opium or its derivatives should be avoided, except under very exceptional circumstances. Scanty Menstruation. — Scanty menstruation was found to be present in 53 per cent of our series of cases. Friedrich 14 observed scanty men- struation, or complete amenorrhea, in 65 per cent of a series of 200 tuberculous women. In tuberculous patients with hypo-plastic uteri the flow is scanty from the onset of menstruation and manifests itself by a 2 9 2 GYNECOLOGICAL AXD OBSTETRICAL TUBERCULOSIS short period and scanty flow. The first 6 or 8 months after the beginning of menstruation, the periods are frequently delayed, the individual often menstruating but three or four times in the 6 or 8 months after the first menstrual period. The age of onset of menstruation in these patients is often somewhat later than normal. Galop 15 found that menstruation was established late and that a premature menopause frequently occurred. The tendency for scanty flow is very marked in pulmonary tubercu- losis. Sometimes this manifests itself by scanty flow, by short periods, by delayed periods, and even, in exceptional cases, by complete amen- orrhea. Frequently the scanty flow is preceded for a few months by menorrhagia. Scanty flow is not only common in those cases in which there is hypoplasia of the uterus, but in those cases in which the uterus is normal the flow usually is scant, if the pulmonary disease is active, and especially so if the general condition is poor. Scanty menstruation is in itself a symptom which rarely causes the patient much concern. It may be considered an effort on the part of nature to conserve blood and thus maintain the strength and resistant powers of the patient. Unfortunately, scanty menstruation is usually accompanied by dysmenorrhea, sometimes of a severe type, and for this reason the patients require treatment. Dysmenorrhea was an accompani- ment of scanty menstruation in 88 per cent of our series. Macht, 1 who apparently classifies scanty menstruation under amenorrhea, found that a large proportion of those cases occurred in young women. Macht's table is as follows : Under 20 years of age 32.5 per cent 20 to 30 39.0 30 to 40 23.9 Above 40 . . 4.6 Macht gives the following table showing the stage of the pulmonary lesion : 1st stage 45.0 per cent of 42 patients 2d. stage 14.0 3d stage 23.7 Patients reported dead at time of computing statistics. . 16.5 Friedrich, 14 in the series of 200 patients studied, found scanty men- struation or complete amenorrhea in the following proportion of cases : MENSTRUAL DISTURBANCES 293 1st stage 45 per cent of 42 patients 2d stage 64 90 3d stage 85 " 68 This report, while emphasizing the fact that scanty menstruation is com- mon in the early stages of tuberculosis, also shows that, as the disease advances, the menstrual disturbances become more frequent. This is in accord with the author's observations. Treatment. — Scanty menstruation in itself requires no treatment. If the general health of the patient can be improved, the flow usually be- comes more normal. Corpus luteum extract is of value in some cases. The trial of extract should be begun about 15 or 20 days before an ex- pected period, administering 5 grains 3 times a day and increasing 1 pill daily until 20 or 30 grains are taken in 24 hours. If organotherapy is to be of value, it proves itself so in the one treatment. If no benefits are derived, it is generally useless to repeat it. If the period is increased or the dysmenorrhea relieved, it may be repeated each month. Of chief importance is treatment directed along the lines of improving the general health. Irregular Scanty Flow. — This was present in ten per cent of our cases. In the advanced stages of pulmonary tuberculosis it is a frequent symptom, but may occur early. Amenorrhea. — Complete absence of menstruation, either of the pri- mary or secondary type, was present in 5 per cent of our series. Fried- rich 14 believes it a common symptom of pulmonary tuberculosis. This symptom frequently causes mental distress to the patient. In many cases all the subjective phenomena of menstruation are present, except bleeding. The secondary type frequently gives a history of scanty or irregular bleeding for a time preceding the complete cessation of the flow. This is a not infrequent symptom in advanced cases of pulmonary tuberculosis. Treatment. — This should be directed towards the improvement of the general condition. Corpus luteum extract is occasionally of value in the treatment of these cases. In the married woman pregnancy must be excluded. It must also be remembered in this connection, as well as with scanty flow,, that the menopause occurs somewhat earlier in tuberculous than in non-tuberculous patients. In a series of 21 patients in various stages of pulmonary tuberculosis, studied by the writer, the average age of the onset of the menopause was found to be 41 years. The average in non-tuberculous patients is about 47 years (Norris 18 ). Menorrhagia. — Menorrhagia was present in 8 per cent of our series 294 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS of cases. It is prone to occur in conjunction with those cases in which there is an irregular periodicity of flow. It rarely persists for prolonged periods, and frequently is followed after a few months by scanty flow, which persists. In some of our cases excessive flow was accompanied by marked dysmenorrhea. Menorrhagia rarely occurs after 30 or 35 years of age, unless associ- ated with a local lesion. It is not infrequently an early symptom of tuberculosis. This symptom was commented upon by Hand forth 17 as early as 1887. Treatment. — Like other menstrual disturbances the results of tuber- culosis, the treatment should be directed towards the pulmonary condition, and if this can be improved, the menstrual disturbances usually become normal. If, as in some of our cases, the flow is excessive, the patient should be kept in bed during the bleeding, and ergot or pituitrin admin- istered. Rest in bed is usually sufficient to control the bleeding. Exces- sive flow is to be combated more vigorously in the tuberculous than in the non-tuberculous, on account of the necessity for conserving the strength and resistant powers of the patient. If it is evident that the flow should be checked, this can be done by radiumization, small doses being employed so as to avoid the permanent menopause. By carefully graduated doses amenorrhea can be produced for a few months, and when the flow is reestablished, it is often normal in amount. Guillermin 18 recommends permanent sterilization and the production of the menopause by the roentgen rays in some cases. Whereas sterilization may be advisable in some cases, the production of the artificial menopause has definite disadvantages, and should be em- ployed only in carefully selected cases. Vicarious Menstruation. — In our series of 214 cases vicarious men- struation was present in 1 patient. This case was moderately typical. Macht x believes it more common than generally thought and observed 1 5 cases in her series. In our case menstruation was normal until 22 years of age. Tuberculosis became manifest at 20 years of age. The patient was in the first stage of the disease, which was, when first seen, quiescent, although the history indicated periods of mild activity. The patient was in moderately good physical condition, and there was no other, demon- strable cause for the menstrual phenomena, the genital tract being appar- ently normal. At 22 years of age and in the second year of her tuberculosis, the menstrual periods became somewhat more profuse than formerly. This continued irregularly, one or two periods being profuse and another scant for six months. Then the periods became very scant and lasted only one day. At time for the flow, there was a hemorrhage MENSTRUAL DISTURBANCES 295 from the bowel sufficient to necessitate wearing a pad, the blood being bright red and the bleeding painless. There was still a show of blood per vaginam for the first day. The bleeding per rectum continued in- termittently for two or three days. The usual menstrual molimina, tingling in the breast, etc., continued. Proctoscopic examination showed the rectum normal. The periodic bleeding from the bowel continued for six months and then ceased. Periodic hemoptysis in tuberculous patients has been frequently ob- served, not only in women, but also in men. Huguenin, 19 Macht, 1 and others record the histories of cases in which hemorrhages have occurred more or less regularly at varying intervals. In considering vicarious menstruation, it is important, however, to exclude all accidental or coin- cident hemorrhages. It is probable that many cases of supposed vicarious menstruation are incorrectly diagnosed. Macht x states that periodic hemorrhages at the menstrual periods have been recorded by Tiedman, 20 Scherer, 21 Kober, 22 Davis, 23 Flesch, 24 Ford, 25 Schlippe, 26 Mosig and Stern, 27 and others. In our case the bleeding was slight ; it may, however, be profuse. Macht 1 records a case of Dr. Brown's, in which the patient bled to death, despite the fact that there was improvement in the pul- monary condition. Flesch's case also terminated fatally. Vicarious menstruation may occur from any mucous membrane. Hemorrhages from the nose, throat, lungs, alimentary tract, kidney, breast, lips, have been observed by Hauptman 2S and Ventura. 29 Macht 30 records the history of a case which bled regularly from an ulcer in the breast. Treatment. — This is similar to that indicated for amenorrhea and scanty menstruation. If the amount of flow per vaginam can be brought up to the normal, the vicarious bleeding usually ceases. Leukorrhea. — Gallard 31 has referred to the occurrence of periodic leukorrhea. Leukorrhea in general is usually more profuse just before and after menstruation. Our investigations have not shown that leuk- orrhea is either more frequent or profuse in the tuberculous than in the non-tuberculous. No periodic leukorrhea other than the type above mentioned has been observed. The Influence of Menstruation on the Temperature in Pul- monary Tuberculosis. — As early as 1878, Goodman, 32 and later von Ott 33 and others have demonstrated that definite changes occur in the woman at the menstrual period. These changes are not only local, but affect more or less the entire economy. Goodman, von Ott, and others believe that among other changes for a few days prior to the appearance of the menstrual flow there is a slight rise in temperature, in pulse rate, 296 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS respiration, and the entire nervous system is somewhat more sensitive at this time. When the flow becomes established, all these conditions suddenly drop to a point somewhat below the normal line, from which time there is a gradual rise until the next menstrual period. An increase in temperature is, however, seen in many cases apparently normal. In considering the rise of temperature which often occurs at the men- strual period in pulmonary tuberculosis, the so-called Goodman-von Ott wave must be taken into consideration. From the observations of others, and from our own studies, however, it seems established that in the tuber- culous the tendency towards a rise in temperature is considerably above that present in the normal woman. This rise in temperature is of distinct diagnostic importance. According to Kraus, 34 it occurs in 66 per cent of cases. Weisse 35 observed a premenstrual rise in 40 per cent of cases. Only 32 per cent had normal temperature during menstruation. Han- sen 36 observed a premenstrual or menstrual rise in temperature in the majority of cases. The increase in temperature is thought to be due to a certain degree of exacerbation of the pulmonary condition, which is explained by a hyperpyrexia of the lungs. Macht x states that at the menstrual period all symptoms exhibit a tendency to become worse. Cough, expectoration, anorexia, general malaise, etc., become more mani- fest, while laryngeal involvement is prone to spread and physical signs become more marked. The exacerbation is usually transient, but may continue. Taking a basis of 99°F. as the standard, Weisse 35 found that 13 per cent of patients had a menstrual rise of temperature usually on the first day of flow, at times continuing over the second; (10 per cent in the first stage of tuberculosis, 15 per cent in the second stage, and 17 per cent in the third stage). Weisse's statistics were formulated from a series of 500 cases of active pulmonary tuberculosis, Riebold 37 found a rise in 12 per cent of cases. Sabourin 38 found the rise in temperature a frequent symptom. Scherer 21 observed a rise most frequently in ad- vanced cases. Noncher, 39 whose paper contains a valuable bibliography, found a rise in temperature in either the premenstrual or menstrual pe- riods in 50 per cent of cases, Kraus 34 in 66 per cent, Macht x in 40 per cent. Van Voornveldt 40 has recorded a case in which there was a regular intermenstrual rise which may be somewhat analogous to the mid scJimcrchcn occasionally observed. As seen from the above figures, the premenstrual rise is the most frequent. Postmenstrual rise of temperature was observed in but 24 per cent of cases and is an unfavorable sign. This rise in temperature may occur in otherwise normal, or may manifest itself as a higher rise in temperature at the menstrual period in MENSTRUAL DISTURBANCES 297 those patients who are experiencing more or less fever. The fever is generally highest in the evening. The rise in temperature may be present in mild as well as in advanced cases. Geisler's 41 suspicion was in one case first aroused towards an incipient lung lesion by these symptoms. A marked rise in temperature is usually an indication of an active lesion, and is an unfavorable prog- nostic sign. So frequent is a slight rise in temperature, that this symp- tom should warn of possible presence of pulmonary tuberculosis and calls for an examination. A previously silent case may give positive findings at or just before the menstrual period. The rise in temperature at the menstrual period in tuberculous women has been the subject of consid- erable study, papers having been devoted to it by Mantoux, 42 Riebold, 37 Sabourin, 38 Kraus, 34 Scherer, 21 Pel, 43 and others. Pregnancy, even in the early stages, not infrequently exerts an un- favorable influence on the course of pulmonary tuberculosis. When it is considered how closely the early stage of pregnancy resembles the physiologic process incident to menstruation, the etiologic relationship of the exacerbation which sometimes occurs at the menstrual periods can be readily understood. Menstruation may be viewed as a preparation of the genital tract for the implantation in the uterus of the fertilized ovum. The same congestion of the genital tract, the thickening of the endome- trium, the nervous phenomena, are common to both conditions and are the same as occur in pregnancy, but to a lessened degree. Menstruation has been well termed the abortion of the unfertilized ovum. From a practical viewpoint, the fact that at the menstrual period pulmonary tuberculosis is especially prone to exhibit exacerbations calls for especial care of all patients at this time. Rest in bed, or at least the reducing of all physical exertion, is of prime importance at this time. If dysmenorrhea or other menstrual disturbances are present, rest will serve a double purpose. All factors which are prone to exert an unfavorable influence on the pulmonary lesions should be avoided as much as possible. Thus, especial care should be exercised against "taking cold." Overheating should be avoided and the diet and bowels should be carefully regulated. These precautions should be instituted for a few days prior to, and for the first few days of, the flow. Such precautions are indicated in all patients suffering from pulmonary tuberculosis, but are especially called for in those patients who exhibit a rise in temperature at, or prior to, the men- strual periods. Macht 1 and other authorities warn against the adminis- tration of tuberculin at this time. Owing to the tendency toward exacer- bations of the pulmonary lesions at this time, operative intervention of 298 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS all kinds should be avoided. This applies to cases of frank pulmonary, tuberculosis, and also to suspected or incipient ones, thus including all tuberculous lesions of the genital tract, which are, in the majority of in- stances, secondary to pulmonary tuberculosis. LITERATURE Macht, D. I. Am. Jr. Med. Sc. 1910. 140:835. Thorn. Centrbl, f. Gebh. u. Gyn. 16:67. Butner, A. J. 111. Med. Jr. 1915. 27:92. DeJong, L. These de Paris. 1914. Poucet et Leriche. La gyn. May, 1910. Sessa, P. La ped. 1914. 22 :255. Grafenberg, E. Munch. Med. Woch. 1910. p. 515. Schiffman. Arch. f. Gyn. 1914. 103. Merlitti. Arch, di ost. e gin. 1901. p. 612, 649, 714. de Rouville, M. Bui. soc. d'obst. et de gyn. de Paris. 1914 P- 559- Tobler, M. Monschr. f. Gebh. u. Gyn. 1905. 22 : No. 1. Schaffer. In Veit's Handbuch. Hollas, J., und Eisenstein, K. Gyn. rundsch. 1907. No. 23. Also, Ztschr. f. Gyn. 1908. No. 44. Friedrich, M. Arch. f. Gyn. 1913. 101 1376. Galop, M. J. La gyn. 1913. 17:659. Norris, C. C. Am. Jr. Obst. 1919. Handforth. Brit. Med. Jr. 1887. p. 153. Guillermin, R. Rev. med. de la Suisse rom. 1919. 38: No. 7. Huguenin. Lungebluthunger. Cor-bl. f. Schw. artzte. 1898. 38 -97 '■ Tiedman. Inaug. Dis. Wurtzberg, 1842. Scherer. Brauer's Beitr. 6 :287. Kober. Berl. Klin. Woch. 1895. No. 2. Davis. Lancet. 1884. 11:782. Flesch. Centrbl. f. Gyn. 1890. No. 37. Ford. Am. Jr. Obst. 1899. p. 154. Schlippe. Brauer's Beitr. 8 Y2jj. Mosig et Stern. Rev. de la tuberc. Oct., 1907. Hauptman. Munch. Med. Woch. Oct. 29, 1909. Ventura, C. Gac. d. osp. 1907. No. 129. Macht, D. I. N. Y. Med. Rec. Feb. 29, 1910. MENSTRUAL DISTURBANCES 299 31. Gallard. Ztscht. f. Gyn. 1886. p. 561. 32. Goodman. Am. Jr. Obst. 1878. 33. von Ott. Intnat. Kong. Berlin, 1890. 34. Kraus. Wiess. Med. Woch. 1905. No. 13. 35. WeIsse, F. W. Beitr. z. Klin. d. Tuberk. 1913. 4:335. 36. Hansen, B. Beitr. z. Klin. d. Tuberk. 1913. 27:291. 37. Riebold. Beitr. z. Klin. d. Tuberk. 1899. 19:8. 38. Sabourin. Rev. de med. 1905. p. 275. 39. Noncher. These de Paris. 1906. 40. Van Voornvedlt. Ztschr. f. Tuberk. 1905. p. 543. 41. Geisler. Russky oratch. 1909. No. 3. 42. Mantoux. Rev. de la tuberc. Oct., 1905. 43. Pel, P. K. Berl. Klin. Woch. 1909. No. 38. CHAPTER XIII PULMONARY TUBERCULOSIS AND OPERATION Three distinct dangers — Choice of anesthetic — Classification of pulmonary tubercu- losis based on physical findings and constitutional symptoms — Subdivision into groups — Study of different stages of the disease — Spinal anesthesia — Precautions before operation — Importance of expert anesthetist — Convalescence — Results — Condition of pulmonary lesion six or more months after operation performed under general anesthetic — Statistical report — Bibliography. CLASSIFICATIONS Pulmonary tuberculosis is one of the most frequent diseases to which mankind is heir. When it becomes necessary to subject a patient suffering from this form of infection to operation, the individual so treated is exposed to materially greater risk than is the non-tuberculous patient. Three distinct dangers occur, which are not present in the non-tuberculous patient : ( i ) the operation itself may be the means of disseminating the infection either to distant and hitherto uninfected parts of the body, or it may result in an exacerbation of the pulmonary process; (2) if a general anesthetic is employed, this may light up the pulmonary lesion; and (3) to these dangers are added the fact that the tuberculous patient is generally below par, and possesses lessened resistant powers, and is therefore less able to withstand the dangers common to operation. When a general anesthetic has been employed and ill results follow, it is some- times difficult to determine whether these are the results of the anesthetic or the operation. From a practical standpoint, however, it is safe to assume that all tuberculous patients are less favorable operative risks, and operations upon them are followed by greater morbidity and a higher mortality than in non-tuberculous patients. As the risks are greater, the indications for operation should be well denned. In considering the subject, it is necessary to individualize all patients. Naturally, the graver the pulmon- ary lesion, the greater are the dangers incident to operation, and the more urgent should be the necessity for operation, before such is advised. Thus, in mild quiescent pulmonary lesions, operations may be advised to 300 PULMONARY TUBERCULOSIS AND OPERATION 301 do away with some discomfort which would never threaten the life of the patient, such, for instance, as a laceration of the peritoneum which is producing definite symptoms ; on the other hand, operation would never be justifiable for a similar gynecological lesion in a patient the incumbent of an advanced or active pulmonary tuberculosis. In the case of an operable cancer, however, great risks are justifiable, as it is known that the patient is doomed if the tumor is not removed, whereas the pulmon- ary lesion, even if advanced, may possibly be checked and held in abeyance for years, or even cured. Various classifications of pulmonary tuberculosis have been suggested; one of the most satisfactory is that of the American Medical Association, which depends upon a combination of the physical finding and the con- stitutional symptoms. This classification divides pulmonary tuberculosis into three stages. In Stage I are placed all incipient cases and those which present slight or no constitutional symptoms. The temperature is not over 100.5 F., pulse under 90, expectoration not more than 30 c.c. in the twenty-four hours. Physical signs limited to infiltration above the clavicles, if bilateral, or to above the second rib, if unilateral. Stage II comprises the moderately advanced cases. In this stage there are no marked local or constitutional symptoms. Marked dyspnea, extreme weakness, anorexia, tachycardia, are constitutional symptoms excluding the patient from this class. Physical examination must show that, if unilateral, not more than half of one lobe is involved; if bilateral, involvement even less, and there must be only slight or no evidence of cavity formation. Stage III includes far advanced cases, all those in which there are marked constitutional symptoms, and all those in which the physical examination shows consolidation of more than one lobe of the lung; if unilateral, advanced cavity formation, or all those cases which are advanced beyond Class II. Miliary tuberculosis is classified separately. It has been our experience that physical signs generally rather under- estimate the extent of the pulmonary lesion, and this is in accord with the statement of Brown. 1 First Stage. — In our work we have subdivided this class of cases into two groups. The first (group A) comprises those cases which are prac- tically free from subjective symptoms, and the only indication of a pulmonary lesion is that there are present slight physical signs and a suggestive history. To this group is added all cases of tuberculosis of the genital tract which present no evidence of pulmonary lesions beyond the fact that we know nearly all genital lesions are secondary and that, 302 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS where there is no other demonstrable primary lesion, it is probable that the lungs were the primary seat. It is true that by treating this latter class of cases as if they were the incumbents of a pulmonary lesion it is probable that a few primary genital lesions are included. However, this is erring on the side of safety. Patients presenting no subjective symptoms and only mild, quiescent or even suspicious physical signs stand even a general anesthetic well, and while the indication for operation should be somewhat greater than in those patients with normal lungs, nevertheless the risk is so slight that it is unjustifiable to allow these patients to suffer when operation -offers a reasonable hope of cure. Group B consists of patients in Stage I who are exhibiting symptoms such as mild cough, expectoration, slight fever, or a little acceleration of pulse. In these a decidedly more cautious attitude should be assumed. In this class of cases Brown's warning that the . physical signs often underestimate the pulmonary lesions should be borne in mind. The administration of ether to such a patient, while in the majority of cases is harmless, will, however, in a certain percentage of cases be the agency which will light up the pulmonary lesion or produce an extension of the disease. With this group of cases, therefore, when the surgical condition permits, it is preferable to advise a course of preliminary treat- ment in an endeavor to improve their pulmonary condition so that it will come under Group A. If this is impossible and operation is demanded, the patient should receive a preliminary dose of morphin gr. ]/\ to 1/3 with atropin gr. 1/150. Local anesthesia is the anesthetic of choice, and much may be done under local anesthesia, if a careful technic is devel- oped. If the entire operation cannot be performed under local anesthesia, it may be supplemented with nitrous oxid, which is decidedly preferable to ether. If deep anesthesia and relaxation is necessary, a few whiffs of ether may be employed during that stage of the operation in which it is required. Its use should be avoided if possible, and only enough given to obtain the desired effect, and a switch back to nitrous oxid made as soon as the conditions permit. Patients in Group A, Stage I, are treated as if in Group B, except that the operative indications are less strictly drawn and there is less hesitancy in resorting to a general anesthetic. Second Stage. — The indications for operation should be well defined, and, with few exceptions, operative intervention requiring a general anesthetic should be refused to patients in this stage of tuberculosis, unless surgical intervention is demanded as a life saving measure. The admin- istration of a general anesthetic is extremely hazardous, and ether espe- cially "dangerous. In this stage of the pulmonary disease every .effort PULMONARY TUBERCULOSIS AND OPERATION 303 should be made to employ only a local anesthesia. The preliminary administration of scopolamin with morphin is of advantage. Third Stage. — All that has been said regarding the danger of a gen- eral anesthetic in the preceding stage is doubly true in this group of patients. Most of these patients are doomed as a result of the pulmonary conditions, and an attempt to alleviate surgical conditions generally means hurrying the end. Choice of the Anesthetic. — The choice of the anesthetic to be em- ployed when operating upon patients suffering from pulmonary tuber- culosis is of the utmost importance. The choice naturally will be deter- mined by the character of the operation necessary, the character of the pulmonary lesion, and, to an appreciable degree, upon the skill of the surgeon. Spinal Anesthesia. — This form of anesthesia is sometimes advis- able in patients in the second and third stages. In the writer's opinion this form of anesthesia is in itself distinctly dangerous ; nevertheless cases "in advanced pulmonary tuberculosis which have to be operated upon in which local anesthesia cannot be employed are safer with spinal than with a general anesthetic. Some years ago the surgical literature was rife with enthusiastic reports of this form of "anesthesia, but, while it is still employed successfully by many operators who have probably attained especial skill in its use, its dangers and ill effects are now recognized. As a matter of fact, the cases of advanced pulmonary tuberculosis that demand operation are few in number, and it is to those in which local anesthesia cannot be employed that spinal anesthesia is especially valuable. The writer's experience with spinal anesthesia is limited to twenty-two cases, in all of which the Gellhorn technic was employed. Miliary tuberculosis is in itself generally a rapidly fatal disease, and surgical treatment is rarely if ever necessary. Hewitt 2 states that patients with old lesions stand anesthesia well. This authority recommends the use of the C. E. mixture, or the C. E. chloroform sequence, or open ether, preceded by the administration of atropin. He believes that nitrous oxid may also be safely employed in chronic cases, but should not be pushed so far as in the normal. Gwath- mey and Baskerville 3 recommend nitrous oxid as the anesthetic of choice, and warmed chloroform and oxygen as their second choice. They believe ether is contra-indicated. Magaw 4 states that these patients stand ether well. Precautions Before Operation. — Presuming that the diagnosis of pulmonary tuberculosis has been made, and a physical examination has shown that the case is in the first stage of the disease, what precautions 304 GYXECOLOGICAL AXD OBSTETRICAL TUBERCULOSIS can be taken to minimize .as much as possible the dangers incident to surgical intervention? In the gynecological department of the University of Pennsylvania it has been a rule that no tuberculous patients are sub- jected to operation, who are running a temperature of more than 99 ° F., unless the operation is very urgently demanded. Thus, in the case of ordinary tuberculous salpingitis in a patient exhibiting a slight evening rise of temperature, we believe that it is "usually safer to delay operation until such time as the temperature is normal. In the interval this class of patients should receive appropriate hygienic and dietary treatment, and, if it is thought that the fever may be caused by the pelvic .lesion, the usual palliative treatment for such conditions is instigated. Usually, after a week or two of such treatment, the temperature returns to the normal and the operation may be performed. With "this method, and with patients in the first stage of pulmonary tuberculosis, good results have been obtained. Occasionally a case will be encountered in which the fever continues, and under such circumstances a further delay is usually advisable. A sharp line must be drawn even in patients in the first stage of the disease, between those patients in whom the pulmonary lesions exhibit a tendency to be active and those in whom they are non- active. In the former the risks incident to operation are definite, whereas in the latter it has been our experience that they are small. The above treatment should be employed in all cases in which a pulmonary lesion is suspected, as, for example, when a virginal patient is found to be suffering from a pelvic inflammatory disease, as the majority of such cases are of tuberculous origin, and, even when the history and physical signs are negative for tuberculosis, it is safe to treat such patients as if they were the incumbents of an incipient pulmonary lesion. The administration of atropin combined with a small dose of morphin, prior to the administration of a general anesthetic, is advisable in all cases. The morphin quiets the patient, and, as a result, if a general anesthetic is necessary, it is better taken and less is required, and there is a lessened danger of straining, vomiting, etc., while the atropin lessens the secretion of mucus. An expert anesthetist should be at hand, if a general anesthetic is employed. It is of the utmost importance that these patients take the anesthetic quietly, and that they do not "fill up" with mucus during the course of its administration. While atropin and morphin are of distinct value in attaining these ends, an expert anesthetist is of even greater importance. This point cannot be too greatly empha- sized. Especial care should be exerted to avoid chilling of the patient while on the way to and from the operating room. PULMONARY TUBERCULOSIS AXD OPERATION 305 If a general anesthetic is employed, the operation should be performed as quickly as possible, so that the patient will not be under anesthesia longer than is absolutely necessary. If a general anesthetic is necessary, nitrous oxid is far preferable to ether. The author has had but little experience with chloroform and other varieties of general anesthetics. Gwathmey and Baskerville 3 especially recommend the employment of nitrous oxid in tuberculous patients. Some authors recommend that the ether fumes be warmed. This may be of value, but has not been employed by us in our work. The anesthesia should be as light as possible, only sufficient being administered to keep the patient under its influence. This is especially true if ether is employed. On the other hand, much harm may be done by a timid or inexperienced anesthetist, who allows a patient to come partly out of ether during the performance of the operation. This often means that the patient vomits or becomes "filled up" with mucus, and always means that the total amount of ether employed will be greater than if an even anesthesia has been administered. At the completion of the operation and before the patient has come out of the ether, it is generally a good plan to wash out the stomach, as this tends to prevent postoperative vomiting. Vomiting in tuberculous patients is to be especially avoided, owing to the increased strain put upon the lungs. Excessive straining or vomiting may be the means of breaking down hitherto incapsulated pulmonary lesions. The inspiration of mucus should also be especially guarded against, both during the administration of the anesthetic and while recovering from it. The operating room should be warm and chilling of the patient avoided. Convalescence. — In the tuberculous patient this should be especially guarded. Particular care to avoid chilling, exposure, etc., should be exerted immediately following operation. Vomiting and straining should be eliminated as much as possible by appropriate measures. The admin- istration of a small dose of morphin or codein as soon as the patient begins to come out of the anesthesia is usually advisable and may be repeated somewhat more frequently than with the non-tuberculous patient. With the above exceptions the subsequent surgical convalescence differs in no respect from the ordinary case. The latter treatment is that usually indicated for tuberculous patients in general, and is especially to be recom- mended for all operative cases. Results. — Doderlein and Kronig, 5 Zweifel, 8 Wahlander 7 and Mayer 8 have remarked an exacerbation following surgical intervention in tuberculous patients. Furniss 9 suggests that this reaction is not due to an actual dissemination of tuberculous material, but that the condition is owing to the "reactivation" of the tuberculous process by the tuber- 306 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS culin liberated by the disturbance of the operation. This reaction occurs about twelve hours after the operation, and is characterized by a rise in temperature of 2 to 4 F., an increase in pulse rate, general malaise, body ache and often headache. A reaction such as described by Furniss, but usually somewhat milder, has been observed in many of our cases. Again, a slight reaction is a certainty in many operative cases even when nontuberculous. In our cases this reaction has not, as a rule, been greater than in the nontuberculous. Weil 10 and others have reported cases of miliary tuberculosis following operation upon tuber- culous patients. Prochownik " has carefully examined 7 cases in which tuberculosis apparently followed gynecologic operation and resulted in death. These were simple operations, such as dilatation and curettage, reposition retrodisplaced uteri, salpingectomy, etc. In 5 of these patients there was no evidence of latent tuberculosis. Prochownik urges the necessity of a thorough examination in all suspected cases of pelvic inflammation which do not yield in a reasonable time to palliative means. As previously stated, our experience has been that in the mild cases, and properly safe- guarded, operation is comparatively safe. Kinghorn 12 reports a similar experience. Certainly, however, all tuberculous patients subjected to operation are exposed to definitely greater risks than the non-tuberculous, and this must be taken into consideration when deciding for or against the advisability of operation. Nearly all cases of tuberculosis of the female genital tract, peritoneum, or intraperitoneal organs are secondary, and, in the majority of cases, are secondary to pulmonary tuberculosis. The pulmonary lesions are frequently latent, but are nevertheless a source of danger, and should be definitely considered and safeguarded as far as possible. On account of the predominance of secondary lesions in many cases and the difficulty often experienced in demonstrating small quiescent pulmonary lesions, all cases of genital tuberculosis should be treated as if pulmonary involvement were known to be present. The following are the results attained in a series of 126 cases of pulmonary tuberculosis operated upon under general anesthesia for various gynecological conditions. Most of these patients had small quiescent pulmonary lesions, and in a few they were of the unsuspected variety, the diagnosis of tuberculosis having been made by histologic examination of the specimen removed at operation, no demonstrable primary lesion in the lungs or elsewhere having been present. All cases in which the end results have been studied have been followed for at least six months and many for much longer periods. In this series there were no operative deaths. PULMONARY TUBERCULOSIS AND OPERATION 307 CONDITION OF PULMONARY LESION SIX MONTHS OR MORE AFTER OPERATION PERFORMED UNDER GENERAL ANESTHESIA Stage of Pulmonary Lesion at Time of Operation Number of Cases Improved No Change Worse Dead 1st stage, group A 104 24 76 3 1 1st stage, group B 18 3 13 1 1 II stage 4 3 1 III stage Total 126 27 92 5 *2 CONDITION OF PULMONARY LESION SIX MONTHS OR MORE AFTER OPERATION PERFORMED UNDER NITROUS OXID AND OXYGEN ANESTHESIA Stage of Pulmonary Lesion at Time of Operation Number of Cases Improved No Change Worse Dead 1st stage, group A 54 15 39 1st stage, group B 10 2 7 1 II stage 2 2 III stage Total 66 17 48 1 CONDITION OF PULMONARY LESION SIX MONTHS OR MORE AFTER OPERATION PERFORMED UNDER NITROUS OXID, OXYGEN, AND ETHER ANESTHESIA Stage of Pulmonary Lesion at Time of Operation Number of Cases Improved No Change Worse Dead 1st stage, group A SO 9 37 3 1 1st stage, group B 8 1 6 1 II stage 2 1 1 III stage Total 60 10 44 4 T2 *One of these occurred three months after operation, and was due to an exacer- bation of the pulmonary condition directly traceable to the anesthesia. The other death occurred in a case in which the tuberculous origin of the pelvic lesion was only discovered during the course of the routine histologic examination of the specimen re- moved at operation. Six weeks after operation a tuberculous peritonitis of the ascitic variety developed, a second operation was performed, but death occurred fourteen weeks after the original operation. I These were two of our earlier cases, and with our present knowledge would not be given ether. 308 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS LITERATURE i. Brown, L. Jr. Am. Med. A. 1915. 64:1977. 2. Hewitt, F. W. Anesthetics and Their Administration. London. 1912. p. 163. 3. Gwathmey, J. T., and Baskerville, C. Anesthesia, D. Appleton and Co., New York and London. 19 14. p. 329. 4. Magaw, A. Mayo Clin. 191 1. p. 573. 5. Doderlein und Kronig. Operative Gynecology, 191 3. 6. Zweifel. Arch. f. Gyn. No. 93. 7. Wahlander. Inaug. Dis. 1893. 8. Mayer, A. Gyn. Rundsch. 191 1. No. 5. 9. Furniss, H. D. Am. Jr. Obst. 1913. 67:910. 10. Weil, F. Munch. Med. Woch. 19 10. No. 7. 11. Prochownik, L. Zentrbl. f. Gyn. 1913. 37: No. 7. 12. Kinghorn, H. M. Jr. Am. Med. A. 1916. 67:1842. CHAPTER XIV TUBERCULOSIS OF THE BREAST Histologic study of tuberculosis of the breast — Frequency — Primary and secondary infection — Routes of infection — Additional foci of disease — Predisposing causes — Age incidence — Statistics — Varieties — Confluent — Disseminated — Physical manifes- tations — Initial symptoms — History of cases noted — Tuberculosis of breast in com- bination with true neoplasms — Differential diagnosis between tuberculosis and cer- tain cases of chronic pyogenic mastitis — Results of postoperative treatment — Bibli- ography. HISTORICAL It is difficult to determine who was the first to describe this form of tuberculosis. In 1829 Sir Astley Cooper 1 wrote of a "scrofulous swell- ing of the bosom," which doubtless referred to this condition. In i860 Lancereaux 2 reported a case, the diagnosis being based upon macro- scopic findings. Johannet, 3 in 1853, and Valpeau, 4 in 1854, described this condition. Heyfelder 5 reported a case occurring in a man of 26 years. Horteloup, 6 in 1872; Poirier, 7 in 1883; Demme, 8 in 1889; Hebb, 9 in 1893; Khesin, 10 Schede, 11 in 1893; Ferguson, 12 in 1898; Parsons, 13 Delbet, 14 in 1892, and Ressigue, 15 also reported cases. In 1881 Dubar 16 reported a case verified by bacteriologic and histologic exam- inations. This is perhaps the first authentic case recorded. In 1883 Ohnacker 17 reported two cases, one of which was proven by animal inoculation. Frequency. — Tuberculosis of the breast is a rare form of infection. Among 196 specimens of various breast lesions in the gynecological laboratory at the University of Pennsylvania, there was one example of tuberculosis. A further analysis shows 91 malignant breast tumors, 75 benign breast tumors, 29 inflammatory lesions (other than tuberculous), and 1 tuberculosis. Deaver and Herman 18 observed five cases of tuberculosis of the breast in a series of 600 operative cases of mammary disease. This was less than 1 per cent of all cases and constituted 2.5 per cent of the benign lesions. Bloodgood 19 found tuberculous mastitis in 6 per cent of all benign lesions of the breast admitted to the Johns Hopkins Hospital. Scott 20 gives the following table: 309 310 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Acute mastitis (abscess) 380 cases Chronic 79 cases Benign tumors (including cysts) . 296 cases Malignant tumors 105 1 cases Tuberculosis 24 cases 1830 Scott thus found tuberculosis present in 1.3 1 per cent of a series of 1830 cases of mammary lesions, and in 3.17 per cent of the benign cases. Bull, 21 in 185 cases of mammary disease requiring amputation, observed one specimen of tuberculosis. . Thus, among 281 1 breast lesions, tuber- culosis was present in thirty-six, or about 1.31 per cent of all cases. Many cases are recorded as tuberculosis, in which the diagnosis is open to doubt. Durante 22 gives notes regarding 1 50 cases and adds 2 of his own. In some of these cases the diagnosis is not positively proven. No case should be considered to be of tuberculous origin unless on positive histologic or bacteriologic findings. In 1891 Roux 23 accepted 31 cases, at the same time recording 3 of his own. In 1904 Anspach, 24 in a careful review of the literature pertaining to this subject, was willing to accept 42 cases as authentic, to which number he added 1 of his own. Ten years later, Deaver and Herman, 18 taking Anspach's series as a basis, were able to collect 87 cases, to which they added 5 new ones. Powers, 25 Scudder, 26 Bartsch, 27 Scott, 20 Schley, 28 Geissler, 29 Brandle, 30 Tuller, 31 Bender, 32 and Miles 33 have contributed valuable articles to the literature of this subject. Primary and Secondary Forms. — Like a similar infection in other parts of the genital tract, tuberculosis of the breast may be either primary or secondary, the latter being by far the most frequent. Indeed, so rare is the former, that its existence has been denied by such authorities as Klebs (quoted by Deaver and Herman 18 ), Ribbert (quoted by Deaver and Herman 18 ), and later by Spediacci 34 and others. A number of cases of indisputable primary origin have, however, been recorded in recent years. Demme (quoted by Schmidt 35 ), Orthmann, 36 Kramer, 37 and others have recorded cases in which the organisms have gained entrance through abrasions about the nipple. Indeed it is claimed by Babes 38 that the tubercle bacillus is capable of passing through the normal skin. Certainly direct infection of the breast through abraded surfaces, such as cracks in the nipples, must be regarded as the most frequent avenue of infection of the primary variety. Deaver and Her- man 18 state that in rare instances tuberculous infection via the lactiferous TUBERCULOSIS OF THE BREAST 311 ducts incites a primary focus in the alveoli of the breast. Verneuil 39 and Verchere 40 have reported cases of ductile infections. In the case of secondary infection, the tubercle bacilli are as a rule carried to the breast by way of blood or lymph channels from more or less distant foci. In rare instances a secondary infection may perhaps result from a direct extension from a nearby focus. This is, however, relatively infrequent. In the lymphogenic form of infection any of the lymphatics of the axillary, cervical and retrosternal nodes and those in the neighborhood of ribs, sternum, pleura and larynx may play an important part. It must be remembered that probably the tubercle bacilli may in some instances pass through, or laterally to, lymph glands without the latter showing macroscopic or even microscopic in- volvements. It is well known that the cervical lymph nodes may be attacked by tubercle bacilli, which have gained entrance through the tonsils, and yet the latter may be apparently normal. Routes of Infection. — These naturally vary. In the primary variety a direct infection from without occurs, either through abrasions of the skin covering the breast or nipple, or perhaps, in rare instances, through the lactiferous ducts. Cracks in the nipple are the most frequent route of ingress of the primary form. The so-called primary secondary form of infection, which has been described in a previous chapter, is also possible, that is, a patient with a pulmonary lesion may cause an infection of this region by means of contaminated fingers, etc., the tubercle bacilli being on the hands or clothing, and these, brought in contact with a fissure of the nipple, may lead to a mammary tuberculosis. The route taken by the infection in the secondary case is less certain. In not a few cases of the latter variety, the axillary glands are attacked before the breast. In other cases, the routes have evidently been by way of the cervical lymph nodes. Cignozzi, 41 Bahaud, 42 Scott, 20 Brandle, 30 and Deaver and Herman 18 state that the cases in which the axillary nodes escape, merely support the well known pathologic fact that lymphatic nodes may transmit infectious organisms without becoming involved in the disease process. The most frequent route is probably by way of the communicating trunks between the retrosternal lymphatics and those of the breast. These branches follow the mammary branches of the internal mammary artery. In many cases it is impossible to determine the course by which the infecting organisms have reached the breast. Cases have been recorded as secondary to tuberculous arthritis by Khesin. 10 Abra- ham, 43 and Hardouin and Marquis, 44 but these are probably in many in- stances really secondary to small quiescent pulmonary lesions, the in- fection occurring through the lymph or blood channels. 312 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Among 29 cases collected from the literature by Deaver and Herman, and which were believed to be of the secondary variety, the following additional foci of disease were present : Cases Bilateral axillary lymphadenitis 4 Pulmonary tuberculosis 4 Cervical lymphadenitis 5 Tuberculous osteitis of the ribs '3 Tuberculous osteitis, bones of jaw and forearm 1 Axillary adenitis 5 Cold abscesses of forearm 1 Tuberculous infectious maxillary bone and cervical lymphadenitis 1 Pleurisy 1 Tuberculous osteitis of hip joint 1 Pulmonary tuberculosis and osteitis of phalanges ... 1 Entire axilla filled with tuberculous lymph nodes .... 1 Tuberculous osteitis of knee joint 1 29 Predisposing Causes. — As would be expected, tuberculosis of the breast is extremely rare in the male sex. Among the 150 cases collected by Durante, 22 6 occurred in men. Deaver and Herman found 10 cases and some of these are not positively proven. According to these authors, cases occurring in men have been recorded by Heyfelder, 5 Ferguson, 12 Ressigue, 15 Poirier 7 (quoted by Deaver and Herman), Hebb, 9 Schede, 11 Demme (quoted by Schmidt 35 ), Parsons, 13 and Khesin. 10 Age. — Tuberculosis of the breast is most frequent between 20 and 50 years of age, in other words during the period of active sexual life. A combination of the statistics of the primary and secondary cases previously recorded by Deaver and Herman shows the following results : Age Incidence 1 o to 20 years 5 20 to 30 years 19 30 to 40 years 23 40 to 50 years 16 50 to 60 years 7 60 to 70 years 3 Not mentioned I TUBERCULOSIS OF THE BREAST 313 Demme 8 has recorded a remarkable case, occurring in a male child four days old. Anspach 24 analyzed the reports of 40 cases and found that Per cent 28 were married 70 19 had borne children 47.5 12 were single 30 12 had hereditary taint 30 6 gave histories of trauma 30 8 suffered from mastitis during lactation 20 2 were directly inoculated 5 Deaver and Herman's statistics of both primary and secondary cases show : Cases Single 13 Married 45 Widowed 2 Males 2 Not mentioned 2 Multiparous 4° Parous 31 Many authors attempt to divide these cases into primary and second- ary. As already stated, we are of the opinion that the great majority are secondary, even those cases which are apparently primary; and because of this uncertainty no attempt has been made in our study to separate the two forms. In this connection Deaver and Herman's sta- tistics are of interest, in that, in their analysis of primary cases, 51.1 per cent were parous, whereas in the secondary cases only 27.5 per cent had borne children — a significant study, for in the primary cases direct infection occurs chiefly through a crack in the nipple, which lesion naturally would be expected to be much more frequent in the women who have borne children, as it is in the puerperium that abrasions at or about the nipples are most prone to occur. Trauma. — The actual part played by trauma is difficult to determine, but it is probably not as great as thought by some authorities. The general surgical principles, however, that trauma predisposes to tuber- culosis in those persons already infected, probably holds as true in tuberculosis of the breast as in other areas in the body, and undoubtedly, if a latent focus of tuberculosis is present in the breast, trauma is especially prone to light it up. Deaver and Herman found that 13.3 per cent of their primary cases gave a history of suppurative mastitis, 314 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Scudder, 20 18.8 per cent of his, and Von Eberts, 45 20 per cent, and of inflammation of some sort complicating lactation 42 per cent. In this" connection Scott's 20 case is a remarkable one ; the patient, aged 34, pierced her breast with a needle, suppuration followed, and a sinus per- sisted. An area of induration developed, and the axillary glands became enlarged. Tuberculosis was demonstrated by histologic examination of the tissue. Varieties. — Various classifications for the tuberculous lesions oc- curring in the breast have been suggested. Perhaps the most simple and satisfactory is that which divides them into (1) confluent and (2) disseminated. Confluent Variety. — This type results from either a pure or a mixed infection, the latter being by far the most frequent. In some cases only a small localized lesion is present, whereas in others the disease progresses until the entire breast is involved. In some instances the entire mammary gland is apparently spontaneously attacked. In ad- vanced cases fistulas, retraction of the nipples, profuse discharge with its accompanying pruritus, involvement of the axillary or other adjacent lymphatic glands occur. Discrete nodules varying in size are probably the most frequent lesions. The method by which the breast is attacked by the tubercle bacillus is similar to that usually observed in other organs, and differs only because of anatomic conditions present in this region. The invading organisms are generally enmeshed in the stroma of the gland. Here they develop and produce typical tubercles. These gradually increase in size, owing to the development of new tubercles forming in the periphery. In this way, in a variable length of time, usually some months, a palpable mass may be formed. In the meantime fresh areas of infection are probably developing in other portions of the breast, so that in the later stages multiple nodules varying in size are likely to be observed. At a still later stage, the center of the nodules may break down and the contents find its escape through the skin. This results in a sinus, which generally exhibits little tendency to close spontaneously. If examined at this stage, the sinus, of varying length, may be found leading down to an apparently small collapsed abscess cavity, the walls of which are usually hard and indurated. The pus is often yellowish, or brownish, and may contain cheesy particles. As a result of the irritating properties of the discharge, the skin is likely to be inflamed, especially in neglected cases. Owing to absorption of purulent material, the axillary lymphatics are nearly always enlarged. This is the most frequent variety. Sections of the breast may show tubercles in various stages of development. TUBERCULOSIS OF THE BREAST 315 The so-called "cold abscess," the result of an unmixed tuberculous infection, is rare. It presents the usual character of such a lesion, gen- erally as a smooth, fluctuant, elastic swelling covered by an intact and sometimes normal appearing skin and surrounded by little or no palpable induration. The veins under the skin are often dilated and visible. The axillary glands are rarely involved by this variety, unless sinuses are present. Sinuses from such abscesses are common and often persist over long periods. Disseminated Variety. — There is, as a rule, not much enlargement of the breast, and the nipple and skin covering the gland is normal. Scattered throughout the breast are isolated tuberculous lesions in various stages of advancement. These usually appear as small nodules, often whitish or yellowish in color, and may, on section, contain cheesy material. The course is generally extremely chronic. Scott has described a third variety, known as sclerosing tuberculous mastitis, which he com- pares to the fibroid form of pulmonary tuberculosis. To this list Ingier 46 has added a fourth form, to which he has given the name of mastitis tuberculosa obliterans. Various combinations of the several forms have been described. Scott states that in 10 of his 27 cases the most prominent histologic feature was a diffuse sclerosis. In 3 of these cases there were deep seated abscesses, whereas in 4 others superficial abscesses were present. In the remainder solid neoplastic lesions, which were at first mistaken for carcinoma, were removed. If it is to be employed at all, it is to the latter class of cases that the term sclerosing tuberculosis should be applied. More or less sclerosis is present in practically all cases. The writer has had the opportunity to examine only 6 cases of tuberculosis of the breast, but in all the sclerosis in varying degrees was more or less marked. Scott states that the true sclerosing mastitis is most likely to occur in elderly patients, and is, as in his case, commonly mistaken for cancer, a mistake which is generally not discovered until a histologic examination of the specimen has been made. This variety is analogous to the fibroid lesions which occur in the lungs. Tubercles are usually few in number or may be entirely absent. In the terminal stage the breast is small, hard and misshapen. The nipple is often retracted. Tubercle bacilli are present in small numbers. Fistulas rarely occur. In the obliterating tuberculous variety of mastitis of Ingier the chief lesions are present in the excretory ducts and the peri-acinous connective tissue, with but slight involvement of stroma. In a case, the history of which was recorded by Ingier, an ulcer was present which had destroyed the nipple and part of the adjacent tissue. The granulation had spread 316 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS inwards and involved the membrana propria of the smaller ducts and acini. Proliferation of duct epithelium was present in many fields and re- sulted not infrequently in obliteration of the lumen. The case is de- scribed as one of a primary infection of the ducts. Symptoms. — The majority of cases of tuberculosis of the breast are of the secondary variety. A careful examination will therefore usually reveal a primary lesion or a history suggesting a previous infection in some other part of the body. This most frequently occurs in the lungs, but may be found in other areas. Occasionally there is pain on deep inspiration, and in these cases an X-ray should be taken to determine the possible existence of a tuberculous osteitis of the ribs. A thorough search for a primary lesion should be made. The primary lesion may be well developed and obvious, or it may exist in a small quiescent focus, the demonstration of which is difficult or perhaps impossible. A history of tuberculosis in the patient's family, or exposure to infection, such as living with tuberculous individuals, is common. No age is immune, but the disease is most frequent during the active sexual life. It is more common in the married than in the single. Deaver and Herman's 18 combined statistics of the primary and secondary cases show the initial symptom was as follows : Lump 5° Tender lump 4 Lump in neck and breast I Hardening I Acute puerperal mastitis I Abscess rupturing spontaneously I Discharge from nipple I Pain 6 Swelling after trauma i Pain and hardening 3 Not mentioned 4 Lump in axilla I 74 This table shows that in the great majority of cases a swelling or tumor- like formation is the most frequent initial symptom. This was the first noticed by the patient in 55 or 74.32 per cent of 74 cases. The lump is usually painless, although trauma not infrequently lights up a more acute inflammation. Unfortunately the appearance of a lump is also the most frequent initial symptom of all breast tumors, and is therefore of little value in differential diagnosis. A moderate amount of pain may be TUBERCULOSIS OF THE BREAST 317 present and has been observed in about 36 per cent of cases. Retrac- tion of the nipples in the case of Dubreuil, 47 Verneuil, 30 and Warden 48 occurred respectively 1 1 months, 2 years, and 5 years prior to discovery of tumor. The disease usually runs a moderately rapid course, 10 or 1 1 months being the average duration prior to operation, it being in this respect more rapid in evolution than even cancer. The general condition of the patient is frequently good and even in the secondary cases is often fairly satisfactory. In the latter variety, the disease exhibits no especial tendency to occur in advanced cases of pul- monary tuberculosis, but frequently becomes manifest during the early stages of the primary lesion, in this way causing further difficulty in differentiating the primary from the secondary cases. Miles, 33 in his report of 6 cases of tuberculosis of the breast, records the history of 1 remarkable case occurring in a single woman 49 years of age and 2 years past the menopause, in which there was retraction of the nipple and an early discharge of a milk-like secretion through the nipple. There was no swelling or pain, but two years later the lower half of the breast was swollen, tender, and the seat of a tumor the size of a hen's egg. As already stated, the initial symptom is usually the discovery of a nodule or swelling in the breast. The most frequent location is in the upper outer quadrant, although any part of the breast may.be attacked. Among the 74 cases, 37 occurred on the right side and 28 on the left, 1 case was bilateral, and in 8 the side attacked was not mentioned. A few cases have been recorded in which both breasts were attacked simultaneously. Walther, 49 Chiavarelli, 50 Gilberti, 51 and Abraham 52 have observed cases in which first one breast and later the other became invaded. The condition of the skin in this series was as follows : Adherent 9 Reddened 3 Red and tender 1 Red and adherent 1 Adherent at areola 1 Darkened 1 Dimpled 1 Adherent and discolored at site of fistula 10 Ulcerated in axilla 2 Adherent and ulcerated 6 Hard and discolored 1 Abscesses of skin 1 318 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Fistulas were present in 37.4 per cent of the cases. In Scudder's 2C series fistulas were present in over 50 per cent, and nearly all had en- largement of the axillary lymphatic glands. Naturally the more advanced the case, the greater the likelihood of a fistula being present. The nipple was retracted in 38.6 per cent. Palpable enlargement of the axillary lymphatic gland was observed in 63.8 per cent of cases, occurring more frequently (72.4 per cent) in the secondary than in the primary cases. Scott, 20 in an analysis of 2.7 cases, found fistulas present in 35 per cent, a definite history of injury in 3 per cent, an acute onset in 6 per cent, skin adherent in 70 per cent, nipple retracted in 30 per cent, and enlarge- ment of the axillary glands in 60 per cent. The important symptoms, therefore, are the presence of a lump, fis- tulas, retraction of the nipple, and enlargement of the axillary lymphatic glands. These symptoms should put the surgeon on his guard for the possibility of tuberculosis in the breast, and when in addition they occur in a woman known to be suffering from tuberculosis elsewhere in the body, they must be looked upon as extremely suspicious. Tuberculosis of the Breast in Combination with True Neoplasms. — Tuberculosis and cancer in conjunction have been observed a number of times. Klose 53 has collected 17 cases, many of them not above sus- picion. He, however, reports 1 case, and Franco 54 has observed 2. Kil- lenberger, 55 Scheidigger, 56 Rodman, 57 Bauer, 58 Warthin, 59 Moak, 60 and Berger, 61 have also recorded cases. Tuberculosis of the breast has also been observed in combination with benign tumors. Revel 62 has recorded the history of a case of adenofibroma associated with tuberculosis. It will be observed that the clinical picture is in many cases by no means diagnostic. In considering the treatment, therefore, the variety of tuberculosis and the relative frequency of true tumors of the breast must be taken into consideration, and no valuable time should be lost in deter- mining the character of the lesion beyond possible doubt. Diagnosis. — Owing to the various forms in which tuberculosis of the breast may appear, and to its rarity, the diagnosis is often difficult, and it may readily be confused with a variety of conditions, among the most frequent of which are carcinoma, fibro-adenoma, retention cysts which have undergone suppuration, simple pyogenic mastitis either of the subacute or chronic form, and less frequently sarcomata and other malig- nant and benign tumors, syphilis and actinomycosis. Some forms of tuberculosis are indistinguishable from carcinoma prior to their removal. The age of the patient, other tuberculous foci, and in rare instances the demonstration of the tubercle bacilli in the dis- charge from the lesion are diagnostic points of value. The latter is of TUBERCULOSIS OF THE BREAST 319 course positive proof. In Duvergey's 63 case the diagnosis was confirmed by the demonstration of the tubercle bacilli by staining methods in the pus. In Delfino's 64 and Mantelli's C5 cases the diagnosis was made by guinea pig inoculation with pus aspirated from the abscesses of the breast ; and Davis 6G demonstrated the organism in the discharge from the nipples in a case of tuberculous mastitis. Biopsy may be of value in certain cases, but it should be remembered that malignant disease is far more frequent than tuberculosis, and when a suspicion of the latter exists, it is better to err on the side of radicalism than on that of prolonged palliative treatment. In the event of a suspicion of syphilis, the Wassermann re- action will naturally be of value, as well as the history and the exhibition of antisyphilitic remedies. Actinomycosis of the breast is an extremely rare condition, only a few cases being on record. If in the latter con- dition a sinus exists, the discharge may contain ray fungus. The differential diagnosis between tuberculosis and certain cases of chronic pyogenic mastitis, especially when sinuses have formed, is im- possible without laboratory methods. Under such circumstances smear preparations and animal inoculation should be made. As a final step, a small piece of tissue may be excised for histologic examination. More numerous and more typical organisms are likely to be present in the wall of the abscess than in the actual pus; for this reason when obtaining material for examination, it is usually advisable to lightly curette the walls of the abscess or of the sinus, rather than use the discharge only. In general, it should be remembered that tuberculosis is a rare dis- ease, whereas tumors of the breast are frequent. In doubtful cases and especially in patients presenting a lump in the breast who are at or near the cancer age, no valuable time should be lost in establishing the char- acter of the lesion under suspicion, beyond the question of a doubt, and it is under these circumstances safer to err on the side of radicalism rather than to run the risk of palliating a possible malignant neoplasm. Treatment. — This, as in other forms of tuberculosis, must depend largely upon whether the case be a primary or secondary one, and in the latter event upon the condition of the primary focus. In all primary cases operation offers an excellent hope of cure. The choice of operation must vary with the individual case ; in young patients, and when the lesion is localized, excision is probably the operation of choice. In older patients, or when the lesion is extensive, amputation of the breast is more satisfactory. In secondary cases the same lines of treatment hold good in the majority of cases, the danger of recurrence or of the development of other secondary foci must, however, be taken into consideration. These, however, are not especially great, certainly not more than encoun- 32 3 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS tered when operating upon other secondary lesions, such as tuberculous salpingitis or tuberculous bone disease, whereas the results of the opera- tion per se are usually good, and relief of symptoms permanent. With secondary cases which are associated with an advanced or active primary lesion, the surgeon must be guided by the conditions of the individual case and treat the patient accordingly. When enlarged axillary glands are present, these should be removed, and as such involvement is usually present, this is generally a necessary step to the operation. In those rare cases which present only a chronic abscess, incision and drainage are preferable to excision. Iodoform is often valuable. The postoperative treatment is important. It is safer to consider all cases as secondary ones and treat accordingly, for as has been stated, small primary lesions may be present which are almost impossible to demonstrate, and these may exhibit activity subsequent to operation. For this reason, all cases should receive an extensive course of hygienic and dietary treatment, preferably under the care of a skilled internist. The exhibition of tuberculin is recommended by many authorities, and probably has some value in increasing the resistant powers of the patient and aiding nature to overcome any small areas of tuberculosis which may have escaped the knife. Indeed von Eberts 45 advises tuberculin alone in early cases. End Results. — Owing to the paucity of material, accurate statistics regarding the end results are difficult to obtain. The immediate mor- tality and morbidity following operation is small. Recurrences have, however, been recorded by Stromberg and Kassagledov, 67 and by Rabin- sohn 68 . In 12 cases collected from the literature by Anspach, 24 4 were well 1 year after operation, 3 were not heard from, 1 died at the end of 3 years from an unknown cause, and the remainder were well, 8, 4, 3, and 2 years afterward respectively. In the primary cases, or in those secondary cases in which the primary focus of infection is small and quiescent, the results are as a rule excellent. In the secondary cases the results are naturally less satisfactory than in the primary, the prog- nosis in this class of cases depending largely upon the character, activity, location and extent of the primary lesion, and the patient's ability and willingness to adopt proper treatment. A number of instances have been recorded in which other secondary lesions, such as peritonitis, menin- gitis, or acute miliary tuberculosis have subsequently developed, but in these cases such complications probably occur independently of the mam- mary condition. TUBERCULOSIS OF THE BREAST 321 LITERATURE 1. Cooper, Sir A. Illustrations of Diseases of the Breast. London, 1829. 2. Lancereaux. Bui. soc. anat. de Paris, i860. 3. Johannet. Rev. Med. et Chir. 1853. 4. Velpeau. Traite des maladies du sein et de la region mammaire. i854. 5. Heyfelder. Deutsch. Klin. 185 1. 3:590. 6. Horteloup. Des tumeurs du sein chez l'homme. 1892. 7. Poirier. These de Paris, 1883. 8. Demme. Schmidt's Jhrb. 1891. p. 229. 9. Hebb. Tr. London Path. Soc. 1892-93. 44:123. 10. Khesin. Kir. 1909. 25 :552. 11. Schede. Deutsch. Med. Woch. 1893. l 9- P- I 3 l6 - 12. Ferguson. Jr. Am. Med. A. 1898. 30:1412. 13. Parsons. Brit. Med. Jr. 1907. 2:263. 14. Delbet. Quoted by Duplay and Reclus. Traite de chirurgie. 1892. 15. Ressigue. Alb. Med. Ann. 1909. 30:671. 16. Dubar. These de Paris. 1881. 17. Ohnacker. Arch. f. Klin. Chir. 1883. 28:366. 18. Deaver, J. B., and Herman, J. L. Am. Jr. Med. Sc. 1914. 147 :i57- 19. Bloodgood. In Kelly and Noble's Gynecological and Abdominal Surgery. Philadelphia and London. 1908. 20. Scott. St. Barth. Hosp. Rep. 1905. 40:97- 21. Bull. Q;uoted by Anspach. No. 24. 22. Durante, L. Policlin. 1914. 21: July. 23. Roux. These de Geneve. 189 1. 24. Anspach, B. M. Am. Jr. Med. Sc. July, 1904. 25. Powers. Ann. Surg. 1894. 19:159- 26. Scudder. Am. Jr. Med. Sc. 1898. 116:75. 27. Bartsch. Inaug. Dis. Jena, 1901. 28. Schley. Ann. Surg. 1903. 37:510. 29. Geissler. Deutsch. Med. Woch. 1906. 32:1780. 30. Brandle. Beitr. z. Klin. Chir. 1906. 50. p. 215. 31. Tuller. N. Y. Med. Jr. 1909. 32. Bender, X. Rev. de chir. et de chir. abd. 1915. 13:265. 33. Miles, A. Edinb. Med. Jr. 1915. 14:205. 34. Spediacci. Schmidt's Jhrb. 1895. 247:148. 322 GYXECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Schmidt. Ber. u. d. Thatkt. d. Jenner. Spit. Beru. 1889. Orthmann. Virch. Arch. 1885. p. 365. Kramer. Centrbl. f. Chir. 1888. 15. p. 867. Babes. Presse med. June 15, 1907. Verneuil. Prog. med. 1882. 10:580. Yekchere. These de Paris. 1884. Cigxozzi. Policlin. 1910. 17 :8i 1 ; Rif. med. 1910. 26:965. Bahaud. Gaz. med. de Nantes. 1906. 24:317. Abraham. These de Paris. 19 10. Hardouin et Marquis. Rev. de chir. 1908. 38:79. von Eberts. Am. Jr. Med. Sc. 1909. 138:70. Ingier. Virch. Arch. 1910. 202:217. Dubreuil. Gaz. hebd. des sc. med. 1890. 12:325. Warden. N. Y. Med. Rec. Oct., 1908. Walther. Bull, et mem. soc. d'anat. de Paris. 1906. 32:1076. Chiavarelli. Rev. ven. di sc. med. 1907. 47:424. Gilberti, P. Policlin. 1916. 23:321. Abraham. These de Paris. 1910. Klose. Beitr. z. Klin. Chir. 1910. 66:1. Franco. Virch. Arch. 1908. v. 193. Killenberger. Quoted by Klose, No. 53. Scheidigger. Ein Fall von Carcinom und Tuberkulose der Gleichen Mamma. 1904. Saurlander & Co. Rodman. Tr. 6th Intern. Cong, on Tuberc. 1908. Also N. Y. Med. Rec. 1908. Bauer. Uber Kombination von Carcinomen under Tuberkulose in der Mamma. Gottingen, 1912, L. Hoffer. Warthin, A. S. Am. Jr. Med. Sc. 1899. 118:25. Moak, H. Jr. Med. Res. 1902. 8:128. Berger. Rev. gen. de clin. et de therap. 1906. 20, p. 22. Revel. Trib. med. 1908. 4, p. 741. Duvergey. Jr. de med. de Bordeaux. 191 1. 51, p. 841. Delfino. Gac. d. osp. 1906. 2y.gyj. Mantelli. II. morg. 1910. 42:96. Davis. Med. news. June, 1897. Stromberg and Kasageldov. Russk. klin. arch. 1909. 25:512. Rabinsohn. Inaug. Dissert., Konigsbr. in Prague. 191 1. Loumeau. Gaz. hebd. des sc. med. 1917. 38:45. Victor, J. A. N. Y. Med. Rec. 1918. 94:829. Goxzalez-Marmol, D. Rev. med. cub. 1919. 30:209. Gulewood. Jr. Am. Med. A. 1916. 67:1660. CHAPTER XV TUBERCULOSIS OF THE PERITONEUM Early history — First authentic operation performed by Sir Spencer Wells — Primary intraperitoneal foci — Primary and secondary tuberculous peritonitis — Cases studied with view of determining primary lesion — Routes of infection — Pathology — Classi- fication of tuberculous peritonitis — Varieties — Acute miliary, ascitic, fibroplastic, and suppurative — Latent cases accidentally discovered — Frequency; special fre- quency among colored race — Variety of tubercle bacillus causing tuberculous peritonitis — Division into groups — Histologic study — Pseudotuberculosis of the peritoneum — Difficulties encountered in differentiating malignancy from tubercu- losis — Methods of treatment — Operative complications — Tuberculosis in hernia — Reformation of ascites following operation — Comparison of results of medical and surgical treatment — Bibliography. HISTORY The early history of tuberculosis contains comparatively few refer- ences to tuberculous peritonitis, despite the works of Bichat, Laennac, Bayle, and others, and it was not until the appearance in 1825 of Louis' dictum, to the effect that chronic peritonitis was usually of tuberculous origin, that the attention of the medical profession became seriously directed to the condition. The first authentic operation performed upon a patient suffering from tuberculous peritonitis is the now celebrated case of Sir Spencer Wells. On Christmas Eve, 1862, Wells operated upon a patient of the surgeon etcher, Mr. F. Seymour Haden. The anesthetic was administered by Clover, and Savage was an assistant. The patient was operated upon under the mistaken diagnosis of an ovarian cyst, and is referred to by Wells x in a subsequent publication. This patient recovered, but it remained for Kronig 2 to trace the sub- sequent history and thereby adduce the positive proof of an ultimate cure — the first authentic operative cure of this condition. Credit is also due Kronig for urging operative treatment for this condition, first in 1884 and later in 1890. The latter paper contained a report of 139 cases operated upon, 107 of which were improved or well 2 or more years subsequent to operation; and, comparing these results with those obtained by medical treatment, Kronig drew an analog}' between tuber- 323 324 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS culosis of the peritoneum and a similar infection of the joints, the opera- tive benefits of which were recognized. This paper was the starting point of the long continued discussion as to the relative merits of the medical and surgical treatment of tuberculous peritonitis. Stone 3 has called attention to the fact which has been overlooked by many historians that in 1884 Dr. Z. B. Adams of Farmington operated upon a patient for this condition, this probably being the first operation performed for tuberculous peritonitis in this country. It is worthy of note that the older writers considered this disease fatal, and it is only with the advent of more modern surgical treatment that a more opti- mistic attitude has been assumed. Intraperitoneal Foci. — Tuberculous peritonitis may be either local or general. General tuberculous peritonitis may, as a result of healing, result in a local peritonitis. The reverse is even more frequent. In a previous chapter the subject of tuberculous salpingitis and pelvic peri- tonitis has been discussed. It is difficult to separate these two conditions. In women, at least, the fallopian tubes are the primary intraperitoneal focus in the majority of cases. In some instances the disease apparently limits itself to the genital tract and to the peritoneum of the pelvis, never assuming the dimensions of a general peritonitis; in other cases it begins as a salpingitis and remains localized for a longer or shorter period, and finally becomes general. On the other hand, in not a few cases of macro- scopically localized tuberculous peritonitis which have undoubtedly originated in the tubes (primary intraperitoneal focus) there is a history which indicates that at some stage of the disease there has been a general tuberculous peritonitis, the salpingitis remaining after the general peri- tonitis has cleared up. Thus the intraperitoneal infection may begin as a salpingitis and subsequently develop into a general peritonitis; or it may begin as a general peritonitis which undergoes cure, but leaves behind a salpingitis. One or more attacks of general peritonitis may occur during the course of a tuberculous salpingitis, the former being the most frequent. Schlimpert, 4 in a long series of postmortems, found that among females 87.9 per cent of the cases of tuberculous peritonitis were secondary to genital lesions, the tubes being the infecting foci in the great majority of cases. It should be remembered that as long as a tuberculous focus remains in the peritoneal cavity, a potential factor in the production of a general peritonitis is present. Primary and Secondary Tuberculous Peritonitis. — Primary tuber- culosis of the peritoneum is an extremely rare condition — so infrequent, in fact, that before accepting such a case as authentic, a carefully per- formed autopsy is necessary. Cases reported without necropsy, although TUBERCULOSIS OF THE PERITONEUM 325 in some instances probably authentic, are open to doubt. Borschke 5 found 2 cases which he considered primary in 226 necropsies performed upon subjects dying of tuberculous peritonitis. The lungs were involved in 200. Hamman, 6 in 35 similar postmortems, observed 1 case in which the tuberculosis was limited to the peritoneum. In this case there was, however, an adhesion in the pericardium which may have been of tuber- culous origin. In this series pulmonary tuberculosis was present in 18 cases and in 29 either the pleura or pericardium was involved. Miinster- mann 7 found 1 case which he believed primary in 46 autopsies upon subjects dead of tuberculous peritonitis. So infrequent is primary tuber- culosis of the peritoneum, that its existence has been doubted by some authorities. The well proven fact that tuberculosis under certain cir- cumstances can pass through various tissues without producing definite lesions therein, is however proven by the carefully worked out postmortem results of many observers as to the existence of primary peritoneal tu- berculous lesions. Primary peritoneal tuberculosis is not to be con- fused with those not uncommon cases, in which the primary focus has undergone partial resolution or is in abeyance by the time the peritonitis has become manifest. It should not be forgotten that, in those' cases in which there are lesions of the peritoneum, as well as in other areas in the body, it is possible for the peritoneum to be the primary seat, and the other foci to be the secondary. Whereas this is theoretically possible, careful study has shown that this is rarely the case, the reverse being true in nearly all cases. The early history of peritoneal tuberculosis con- tains accounts of many cases of supposed primary tuberculous peritonitis. Thus Rokitansky 8 in 1855 was °f the opinion that many cases were primary. The primary form is probably less infrequent in young chil- dren than in adults. The lungs are the primary seat in the great majority of cases. Albrecht 9 studied 200 cases of peritoneal tuberculosis which came* to autopsy, with view of determining the primary lesion, with* the results shown in table on following page. Matteson 10 was able to demonstrate tuberculosis in other parts of the body in 50 per cent of his cases and in 75 per cent of those cases which were under 30 years of age. Matteson's statistics are drawn from operative material, and hence show a smaller percentage of secondary cases than would those formulated from postmortems. Among Ham- man's' 6 series of 150 cases, definite physical signs of pulmonary tuber- culosis were present in 34 patients, and 47 complained of cough. Routes of Infection. — The tubercle bacilli may gain access to- the peritoneum by means of the blood or lymph from either distant or ad- 326 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS Lungs Lymphatic glands Intestines Genitalia Bones Pleura Tuberculosis of serous membrane Unknown Women — per cent 46.1 20.8 7-7 12. 1 3-3 3-4 3-3 3-3 jacent foci, by contiguity or continuity from other foci, or by direct implantation upon the peritoneum, either from without, as an experi- mentally produced peritonitis in animals by way of the genital tract, or directly through normal structures, such as the intestine or lymph gland. The passage of tubercle bacilli- through normal tissue has been proven. Tubercle bacilli have also been demonstrated in the normal fallopian tube. Practically there are four chief routes by which infection occurs, ( 1 ) by the blood or lymph channels, (2) from an intestinal lesion through the walls of the gut, (3) from a tuberculous mesenteric gland, (4) from a tuberculous salpingitis. Tubercle bacilli' may pass through the in- testinal wall or through a lymph node without the latter being seriously affected. This is probably of rare occurrence. The acute miliary variety of tuberculous peritonitis is the result of a blood borne infection, a general tuberculosis usually being present as well as the peritoneal involvement. Tuberculous peritonitis in men is usually secondary to an intestinal lesion, often an ulcer, and generally located in the neighborhood of the cecum or vermiform appendix. In women, the fallopian tubes are the most frequent primary intraperitoneal site, although the appendix and cecum are not uncommon starting points. In children the infection most often occurs from an infected mesenteric gland. Allshut 21 believes that the path of infection is often from the peri- bronchial lymph tissue through the perforating lymphatics of the dia- phragm into the peritoneal cavity; chiefly by the retroperitoneal lymph glands, which he has found are generally involved. In rare instances, according to Goodrich, 12 infection may result from ulceration or infiltra- tion of the diaphragm. Apert (quoted by Goodrich 12 ) has recorded the history of such a case. The tonsils are often the entry way for the TUBERCULOSIS OF THE PERITONEUM 327 tubercle bacilli. Cummins (quoted by Goodrich 12 ) has recorded a series of cases in men, in which the infection appears to have been sec- ondary to a tuberculous epididymitis by way of the lymph, vessels of the spermatic plexus. As has been stated, among women the fallopian tubes are the* most frequent primary intraperitoneal focus. One of the characteristics of tuberculous salpingitis- is that the abdominal ostia of the tubes tend* to remain patulous, a point favoring the spread of the infection to the peritoneal cavity. The involvement is generally bilateral, and the' mucosa of the tubes is practically always involved, two conditions also favoring dissemination of the infection to the peritoneum. When a tuberculous salpingitis and peritonitis coexist, undoubtedly either may be the primary intraperitoneal focus. In 191 1 Kronig, 2 in the German Gynecological Kongress, upheld the view that, when these two conditions occurred to- gether, the peritoneum was most frequently the primary infection. Albrecht 9 stated that, as a result of over 10,000 autopsies and from clinical and experimental studies, he believed the two conditions frequently coexisted; in 33 per cent it would seem that the genital lesion was the primary one, but that the reverse was rarely the case, and when a hematogenous infection occurred, the tubes and the peritoneum were frequently simultaneously involved. Mayo 13 believes the fallopian tubes the most frequent intra-abdominal site in women, an opinion in which the author concurs. This belief is based upon the fact that in women tuberculous salpingitis is more frequent than is general tuber- culous peritonitis, tubercle bacilli being the exciting cause in from 4 to 8 per cent of all tubal inflammations, and that in the majority of cases of tuberculous peritonitis, a careful search in the history will reveal evidences of the existence of a salpingitis some time prior to the onset of the symptoms of the general peritonitis. Mayo 13 states that in a series of 18 cases, 11 were in women, in 9 of whom the origin was in the tubes, 1 in the vermiform appendix, and 1 unlocated; in the 7 men, 3 originated in the vermiform appendix, 2 in- the cecum; and 2 were unlocated. Mayo mentions 5 other cases- in which the lesions were most severe in the upper abdomen, the primary intraperitoneal site of which was unknown. These were in dlder patients. Kraus 14 has recorded the history of a case in which he believed that the adnexal in- fection was secondary to a tuberculous appendicitis. Pathology. — The lesions produced by the tubercle bacilli in the peri- toneum are in general similar to. those resulting from a similar infection in other parts of the body. They are, however, cfften modified owing to the peculiarities of the intraperitoneal viscera. At the point of implanta- 328 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS tion a typical tubercle is developed. In the miliary (hematogenous) variety of infection great numbers of tubercles- in various portions of the peritoneum, as well as elsewhere, develop simultaneously. When the inflammation results from the rupture into the peritoneum of a tuber- culous focus, the onset may be general. In the other fojrms of infection the lesion is probably, as far as the intraperitoneal condition is concerned, always local in its incipiency. The first area of infection" acts as a starting point from which other tubercles are developed, the mfection being sprread through the peritoneal cavity by peristaltic and respiratory movements, the peritoneal currents, gravity, etc. Tubercles are usually most numerous at or near the primary intraperitoneal foci. The tubercles may be seen as small, elevated, firm areas, varying in size from the micro- scopic to a few millimeters or more in diameter, and are found in various stages of development. Not infrequently nearby tubercles coalesce, and in this way massive lesions may be developed : in some instances these break down and abscesses of various size result. When actual suppura- tion takes- place, a mixed infection is nearly always present. As a result of the inflammation, peritoneal cysts may develop; these are often- thin walled and contain thin, clear, amber colored fluid. In some instances the contents are turbid and discolored : . Various lesions may develop*; adhesions are the most frequent. These may vary from a few light bands of adhesion situated at the point of the primary intraperitoneal focus, to great masses composed of plastered together intestines, omentum, or other intraperitoneal organs. In the cavities of such lesions, ab- scesses, cysts, fistulas leading to adjacent organs, or even artificial anastomoses may be present. With the exception of the fallopian tubes, the omentum is the most frequent intraperitoneal organ attacked. It may be adherent to some other focus, such as the cecum or tubes or, as is not infrequently the case, may be found rolled up, forming a more or less nodular sausage shaped mass, often lying diagonally or transversely in the upper peritoneal cavity. On palpation, the rolled up omentum often simulates a true neoplasm. As a result of adhesions, intestinal obstruction may result. Indeed, the fact that obstruction does not develop more often is remarkable-, when the frequency and character of the adherent masses often formed is taken into consideration. When much free fluid is present, adhesions are less likely to be a pronounced feature. In such cases the various organs are found floating free m the fluid and less chance for fusion is afforded. Occasionally a cure of the general tuberculous peritonitis occurs, but a local lesion, such as an ulcer or a salpingitis, persists. This may remain dormant, producing no, or only local, symptoms for a long period and may finally undergo resolution ; TUBERCULOSIS OF THE PERITONEUM 329 or may at some future time produce a focus for a fresh involvement of the general peritoneal cavity. Not infrequently in healed cases ad- hesions persist and may result in troublesome symptoms. Strangulation of the gut, stricture, or intestinal obstruction has been reported, and con- tractures resulting in painful traction upon various structures are of comparatively frequent occurrence. Many classifications of tuberculous peritonitis exist. Some authorities consider the various forms separate and distinct, whereas many believe them to be but different stages of the same thing. One of the most widely employed classifications is that which divides tuberculous peri- tonitis into the (1) acute miliary, (2) ascitic, (3) fibroplastic, and (4) the suppurative. To these Bryant 15 adds a fifth variety, the latent. Osier 16 classifies tuberculous peritonitis as follows : (1 ) serous, exudative or miliary, (2) nodular or ulcerative, (3) adhesive, fibroplastic or cystic, and (4) purulent. Acute Miliary Variety. — In this variety the peritoneal involve- ment is usually but an incident to a general infection, and for this reason, the symptoms of the peritonitis are often masked by those produced by the general infection. Death may supervene before the peritoneal in- volvement becomes pronounced. If the patient survives for a sufficiently long time, ascites, with its accompanying symptoms, develops. In some cases the clinical symptoms are suggestive of typhoid fever. This form of peritonitis is not as a rule amenable to surgical treatment. Ascitic Variety. — This is the most frequent form of tuberculous peritonitis. In an analysis of 500 cases by Wunderlich 17 the greatest number were found to be the ascitic variety. Stone, 3 in 122 cases, found fluid in the peritoneal cavity in 84, or nearly 69 per cent. Ham- man, 6 in a series of 122 cases, found fluid in the abdomen in 42 per cent. Among 103 cases which were operated upon or which came to post- mortem, 35 cases were of the ascitic variety. Baisch, ls in an analysis of no cases from the Tubingen Gynecological Clinic, found the ascitic variety by far the most frequent. In a series of 21 cases from the gyne- cological department of the Hospital of the University of Pennsylvania, about 60 per cent were of this variety. When the fallopian tube is the primary intraperitoneal focus, the resulting peritonitis is generally of the ascitic variety, so that this variety is especially common in women. Not only is the ascitic variety the most frequent, but it is also the variety which offers the best hope for surgical cure. As a general rule, the lower half of the peritoneal cavity is the area chiefly involved. This is especially the case in women because of the fact that the fallopian tubes are so frequently the primary intraperitoneal focus. The peri- 330 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS toneum is more or less thickly studded with tubercles in various stages of development. The peritoneum itself becomes thickened, hyperemic, and more or less destruction of the endothelial layer occurs. The omentum and the intestines, especially the small bowel, tend to become adherent, and may often be found glued together, forming tumor-like masses, which are generally pushed upwards by the exudate. The variety and character of the adhesions vary markedly. In addition to the omentum and intestines, masses may be composed of enlarged mesenteric glands, or pseudo tumors may be the result by fecal impaction. Probably, in those cases where the effusion occurs early the fluid is found more generally distributed and there are fewer adhesions, whereas, if exudate is formed late, adhesions are likely to be a pronounced feature. If the fluid is encapsulated, the walls of the cavity are, in part at least, com- posed of adherent coils of intestines, omentum, etc. As might be ex- pected, the character of the fluid varies considerably. It is usually clear, transparent, straw colored fluid, but may become cloudy or turbid from the admixture of various substances. Not infrequently considerable flocculent material is suspended in the fluid. From the admixture of blood the fluid may be reddish, dark brown or even black. If walled off cystic spaces are present, the fluid in some may be clear and in others discolored. When few adhesions are present and in the large cystic spaces the fluid is prone to remain clear, whereas in the small compart- ment degenerative changes are more likely to occur and result in a dark or turbid exudate. The fluid often contains a high percentage of lymphocytes. The amount of fluid varies considerably; as much as six or eight gallons have been observed. Fibroplastic Variety. — Of Wunderlich's 500 cases, 136 were of the fibroplastic variety. Mayo 13 believes that while the ascitic variety is the most common and is especially prone to occur in conjunction with lesions of the fallopian tube, the fibroplastic is more frequent as a result of appendiceal tuberculosis or in those cases in which the primary intra- peritoneal focus cannot be located. A mixed infection is often present and operative results are less successful than in the previously described form. Stone, 3 in 122 cases, observed 2>7 °f this variety. Hamman, 8 in 103 cases which came to operation or postmortem, observed 63 that were fibroplastic. Baisch, 18 in no cases, observed 22 which were of the fibroplastic form. The fibroplastic variety originates as a localized lesion, in adults often in the appendiceal region, and in children frequently from a tuberculous mesenteric gland. From the primary intraperitoneal focus the disease spreads, generally, however, exhibiting a tendency to remain localized ; TUBERCULOSIS OF THE PERITONEUM 331 adhesions of intestinal or omental origin often wall off collections of fluid. As the disease advances, caseation and ulcerations occur, and some authorities refer to this stage as the caseous or ulcerative variety. The ulcerations may perforate and result in a general peritonitis, or various forms of fistulas may occur. Massive inflammatory products may be present; the peritoneum is thickened and more or less profusely studded with tubercles in various stages of development. The diseased areas are covered with yellowish, whitish, or brownish gelatinous or fibrous material, often thickly plastered over the intestines and peritoneum. As a result of this process, more or less localized areas of a boggy or semiflocculent consistency, composed of adherent viscera and the fibrous exudate, are found. In some instances the fluid is more or less absorbed ; the endothelium however proliferates and the new tissue undergoes cicatrization, giving rise to firm adhesions which often result in fecal accumulations and may cause intestinal obstruction or stricture. In some localities the peritoneum may exhibit advanced evidence of the disease, and in others be comparatively or even entirely normal. Xothnagel 19 states that cicatricial contractions are specially marked in the mesentery and omentum. It is important to note that cicatricial masses develop more rapidly than the tubercles, so that the latter become encapsulated and may thus disappear and constitute a more or less complete cure. At operation or autopsy no tubercles may be visible, scar tissue being the only discernible evidence of the disease. Suppurative Variety. — This is a questionable variety and in nearly all cases is really an end stage of one of the other forms, the fibroplastic especially, tending to result in suppuration. In this variety any of the lesions previously described may be present. It is always the result of a mixed infection. Several varieties of pseudo-abscesses may be present. In some cases one area may be found to have undergone suppuration, while in others the mixed infection and the formation of pus have not yet taken place. The clinical symptoms are severe and the prognosis, either for surgical or medical treatment, unfavorable. Fortunately this variety is not frequent. Owing to the character of his material, , \Yunderlich's figures regarding suppurative tuberculous peritonitis are unusually high. Ten per cent of his cases were of the suppurative variety. Hamman 6 saw 5 cases in his series of 103 subjects, all of which were operated upon or came to postmortem. As the advantages of operative intervention have become more recognized, this form of peritonitis has become less frequent. Latent Variety. — Under this heading Bryant 15 describes those cases which are discovered accidentally. Stone 3 reports several instances 332 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS of deaths among patients apparently in perfect health and upon whom postmortem showed advanced tuberculous peritonitis. This variety is more frequent among men. Frequency of Tuberculous Peritonitis. — The tubercle bacillus is the most frequent etiological factor in the production of the chronic form of peritonitis. The frequency of the disease in men and women varies considerably. Operative statistics show that women are twice to four times as frequently attacked as men, but curiously enough post- mortem statistics show men more often affected than women. An ex- planation of this is said to lie in the fact that women are more frequently subjected to operation than are men. Tuberculous peritonitis is frequent among children. Cummins, in 3,405 postmortems, found 92 (2.7 per cent) cases of tuberculous peritonitis. From similar material Grawitz and Bruin (quoted by C. H. Mayo 13 ) observed 184 cases among 13,992 necropsies. Among 5,687 intraperitoneal operations performed in the Mayo clinic, 184 (3 per cent) were for some variety of tuberculosis. Hartel (quoted by Behle 20 ) found tuberculous peritonitis in 3-5 per cent of 27,000 postmortems, Friedrich (quoted by Behle 20 ) in 1.9 per cent. Schlimpert, 4 among 2,173 postmortems upon tuberculous subjects, found the peritoneum involved in 4.9 per cent. Albrecht (quoted by Behle, 20 ), in 2,155 necropsies upon tuberculous subjects, found peritonitis present in 10 per cent. In necropsies upon tuberculous subjects at the Henry Phipps Institute, peritoneal involvement was found present in 2 per cent of subjects, and in 5.9 per cent of all females. Nothnagel ia refers to statistics varying from 1.25 per cent to others as high as 16.16. The latter high estimate is given by Borschke. 5 Tuberculous peritonitis is frequent in the young. Thompson 21 found, over a period of 10 years, that some form of abdominal tuberculosis was present in from 1.67 to 4.51 per cent of all children in three large hospitals in the United King- dom. In the Mount Sinai Hospital however a much smaller percentage was encountered (0.044 P er cent), while in the same period of years in the Edinburgh Children's Hospital 3.70 per cent was observed. Caird 22 and Bovaird 23 also refer to the frequency among children. The disease is apparently more frequent in the United Kingdom than in this country. Faludi 24 has collected 306 cases which occurred in patients under 1 5 years of age. Of these nearly one half occurred between yy years of age. The incidence of sexes was nearly equal. Frequency Among the Colored Race. — Tuberculosis in general is well known to be very frequent among the colored race, and this variety of the infection is no exception. Goodrich, 12 Kelly, 03 and others have referred to the special frequency among these people, some authori- TUBERCULOSIS OF THE PERITONEUM 333 ties believing that the disease is twice as frequent among the colored as among the white. Variety of Bacillus Causing Tuberculous Peritonitis. — Barker 25 estimated that 50 per cent of tuberculous peritonitis was due to bovine tuberculosis. The English Commission on Tuberculosis in 191 1 placed it at 47 per cent and the German Commission at 63 per cent. It has been suggested that the bovine type gives a more favorable prognosis than the human. Prognosis. — From a practical viewpoint the majority of cases of tuberculous peritonitis may be divided into two groups, the one in which the prognosis is fairly good if the proper treatment is applied, and the other in which the prognosis is decidedly less favorable. Mayo - 6 has emphasized this division. The first group comprises those cases in which a definite anatomic starting point for the infection can be demonstrated and removed, such as is frequently observed in women when the fallopian tubes are plainly the primary intraperitoneal focus for the infection. The second group contains those cases in which the intraperitoneal focus of the infection is less well defined and in which, although a considerable quantity of fluid is present, it is contained in numerous compartments, and many adhesions have been formed. The character of the fluid is to some extent also a guide, the clear ascitic fluid being the most favorable. Numerous dense adhesions, the presence of pus, sinuses, extensive in- volvement of the entire peritoneal cavity, high fever, poor general con- dition of the patient, and grave primary lesions, such as extensive pul- monary involvements, are, on the other hand, unfavorable. Symptoms. — As has been stated, these depend upon the stage of the disease and the type of the lesion, and in some cases may be partially masked by the symptoms produced by the primary lesion. The disease occurs in women, chiefly in the child bearing period, although young girls are by no means immune. Alglave (quoted by Jacobson 27 ) refers to a remarkable case, which developed in an infant 3 days old. Death occurred on the sixth day and necropsy showed an advanced general tuberculous peritonitis. In a few cases seen by the author the disease has followed pregnancy in patients the incumbents of pulmonary lesions. Kelly 63 has noted similar occurrences. A history of trauma is present in a certain proportion of cases, but is a greater factor among men than among women. As a general rule, the symptoms are those of a chronic peritonitis, which, as the disease advances, are associated with well defined loss of strength, loss of weight, fever, especially in the evening, rapid pulse, increased respiration, and nausea or vomiting. Occasionally an acute 334 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS onset is observed, and when such is the case, the infection is prone to be of a more virulent type. The local symptoms vary widely ; in the common ascitic variety the presence of free fluid in the peritoneal cavity with its accompanying phenomena are the chief symptoms. The ascites generally shows a more or less marked tendency to become walled off in compartments. This is especially pronounced as the disease advances. Osier has drawn attention to the fact that some cases exhibit a subnormal temperature, in others the temperature may be normal. As the disease progresses the patients become pale and anemic. The amount of fluid which may be present varies greatly in different cases. Nothnagel 19 reports a case in which 11,500 c.cm. were removed. The amount of fluid occasionally varies in amount in the individual case, and it is not uncommon for patients themselves to remark on this fact. The shape of the abdomen is often somewhat pyriform in tuberculous peritonitis, rather than the flattened top and overlapping sides so commonly observed in other varieties of ascites. The fluid is generally somewhat yellowish, but is often dark from the admixture of blood. It may be clear or cloudy, or contain flakes of lymph or fibrin. In some instances it is milky and opaque. If sacculated, a different appearance of the fluid is often present in the different loculi, the contents of some being clear and straw colored, of others discolored. This difference in the character of the fluid is doubtless due to the stage of the disease in different compartments, to mixed infection, and to some extent to the parts involved. Ross 28 be- lieves a high percentage of lymphocytes in the ascitic fluid suggestive of tuberculosis, and an excess of endothelial cells, except in the very early stages, the reverse. Gibbert and Villaret 29 have expressed a similar opinion regarding the significance of numerous endothelial cells in the fluid. Old fluid is said to lose its bactericidal properties, and newly formed fluid to contain greater antituberculosis action. Edebohls attached considerable diagnostic significance to the occurrence of rounded plaque-like thickenings which are occasionally palpable on the anterior and lateral parenteral peritoneum. These vary from 1 to 8 cm. in di- ameter and feel not unlike urticarial wheals. They occur early in the course of the disease. Murphy 30 believed these to be hyperemic in origin. In the fibroplastic type the formation of one or more tumor- like masses is the prominent symptom. These masses are at first some- what movable, but later tend to become fixed. The masses are tender. When suppuration is present the general symptoms, such as fever, pain, and tenderness, are more marked. Pain is however a variable symptom, and is less pronounced in tuberculous than in other forms of peritonitis, and in some cases it may be entirely absent. More or less pain is, TUBERCULOSIS OF THE PERITONEUM 335 however, usually present. It must be remembered, also, that tuberculous peritonitis exhibits a tendency towards remissions and may become quiescent for prolonged periods, even without treatment of any kind, and, in a small percentage' of cases, may undergo spontaneous cure. Diarrhea, or alternate diarrhea and constipation, is present in many cases, especially if there are well defined intestinal lesions. For this reason, these cases are sometimes diagnosed as "intestinal indigestion." The spleen is frequently enlarged, but its demonstration is generally difficult. The liver may be enlarged, but is more often unchanged. The skin of the abdomen is tense, waxy, and enlarged veins are often present; pigmentation may occur, and is especially likely to be present on the face. This pigmentation may be so marked as to suggest Addi- son's disease (Osier, 16 p. 311). In women, in addition to a history pointing towards a bilateral salpingitis, scanty menstruation is often present. In Hamman's 6 series of 150 cases, 104 had pain, 42 vomiting, 48 constipation, 33 diarrhea, 4 alternating diarrhea and constipation, 6 blood in the stools, 1 1 pain in the chest, 47 coughs, 34 showed physical evidence of pulmonary tuberculosis, and in 30 cases dyspnea was present, loss of weight in 61, night sweats in 27. The leukocyte count showed 8 cases under 5,000, 38 cases between 5,000 and 10,000, 8 cases between 10,000 and 15,000, 8 cases between 15,000 and 20,000, and 3 cases above 20,000; 70 per cent were under 10,000 and 83 per cent under 15,000. Stone 3 has recorded very similar blood findings. When a higher leukocyte count than 15,000 is present, it usually indicates that com- plications are present. Jaundice, due to obstruction of the ducts, is not uncommon. Individual tubercles tend to become encapsulated by con- nective tissue, which contracts and then produces a cure. The connective tissue growth is said to be more rapid than is that of the tubercle. If in a given case this be true, a clinical cure results. This fibrous meta- morphosis may be often observed in histologic preparations, or even macroscopically at operation, or on the postmortem table. Connective tissue formation may result in partial or complete intestinal obstruction, or in dragging and distortion of the various intraperitoneal viscera, with resulting clinical symptoms. Adhesions may result from fibrous change in the tubercles but are probably more often due to an ordinary inflam- matory process. Diagnosis. — Many cases present a history or physical signs sug- gestive of the primary infection either in the lungs or elsewhere. Pleurisy as an accompaniment of tuberculous peritonitis has been observed in some cases, and a previous history of an old pleuritic infection is fre- quently present. A bilateral pleurisy is particularly suggestive. Peri- 336 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS cardial effusion may be present, but is infrequent among the cases ob- served by the author. In women a previous history indicating a chronic bilateral salpingitis is often obtainable. In a smaller percentage of cases a previous history pointing towards the appendix or cecum will be discovered, and in rare instances the upper abdomen will have been the seat of the primary intraperitoneal focus. This however is rare, and in the great majority of cases the fallopian tubes will have been the primary intraperitoneal seat. It should be remembered that in children the mesenteric glands are usually the primary intraperitoneal focus. As a general rule the diagnosis of tuberculous peritonitis presents no great difficulties. The chronic character of the peritonitis, the loss of weight and strength, the presence of either pus or encapsulated fluid Avithin the peritoneal cavity, the tender masses especially in all but the miliary and ascitic forms, the occasional diarrhea or diarrhea alternating with constipation in those cases in which the intestines are involved, the resistance of palliative treatment, the physical findings, and the previous history, usually render the diagnosis easy. An inflammatory mass in the pelvis followed by ascitis is very suggestive of tuberculous peri- tonitis, and is the condition present in many cases of tuberculous peri- tonitis in the female. Beale 31 has directed attention to the fact that the pain of tuberculous peritonitis is often relieved by pressure. This however is by no means true in all cases. The author has observed patients in whom there was no pain except on pressure. Monro 32 and others have referred to the frequency with which the omentum is found rolled up as a nodular transverse cord in the upper abdomen. This is especially common in the fibroplastic variety of the disease. This pres- ence of an omental tumor is very characteristic of tuberculous peritonitis, and with the exception of pelvic tumors this is the most frequent symp- tom. These so-called pseudo tumors are however generally multiple. Cancer and cirrhosis can usually be easily excluded, as can pelvic neo- plasms. Cirrhosis of the liver has been observed in conjunction with tuberculous peritonitis by Rolliston, Osier, and others, and when present, appears to markedly reduce the resistant power of the peritoneum to the tuberculous infection. From a general carcinomatosis of the peritoneum tuberculosis can be differentiated by the age of the patient, tuberculosis occurring early in life or during the childhood period, and carcinoma generally later. In carcinoma the disease is steadily progressive, and in tuberculosis the course is chronic. Elevation of temperature, pulse, respiration, and night sweats are more constant in tuberculosis than in cancer. Cirrhosis is frequently syphilitic in origin, whereas tuberculosis gives TUBERCULOSIS OF THE PERITONEUM 337 p. history or physical signs of tuberculosis elsewhere in the body. The physical changes in the liver in cirrhosis, the fact that cirrhosis is more frequent in men and is comparatively rare in early life, the blood picture, and the presence or absence of the Wassermann reaction are all diag- nostic aids. Cirrhosis can only be mistaken for the ascitic form of tuberculosis, and its differential diagnosis from it should be easy. In cirrhosis the abdomen usually presents the well known saddle bag appear- ance, the top being flat and the sides pouched out. In tuberculous peri- tonitis, on the other hand, a pyriform abdomen, of the shape often produced by a greatly overdistended bladder, or by an ovarian cyst, is not infrequently met with. The ascites of cirrhosis is generally free and few adhesions are present, whereas in tuberculosis the tendency for the fluid to become walled off into various sized compartments is pro- nounced, especially as the disease becomes advanced. As a result of this, movable dullness is less often present. From ovarian neoplasms, especially those producing ascites and ad- hesions, the diagnosis may be more difficult. Ovarian neoplasms gen- erally occur later in life, and there is an absence of previous history of tuberculosis elsewhere. Ovarian neoplasms, unless associated with definite inflammatory lesions, do not generally produce fever, hyper- pyrexia, night sweats, or intestinal disturbances. If they are asso- ciated with inflammatory lesions, there is generally a well marked leukocytosis, which is absent in tuberculosis. The ovarian cyst, even if bound down by adhesions or associated with ascites, will often give a history of a previously movable pelvic tumor without marked evidence of peritonitis; whereas the tuberculosis, even if it has been preceded by a salpingitis, will usually give a history of small bilateral inflammatory tumors and from the start has been accompanied by pain, tenderness, and elevation of temperature. In some cases the differential diagnosis between these two types of lesions is extremely difficult. A careful pelvic and abdominal examination will, however, generally clear up the case. Occasionally, when seen early in the disease, the onset may simulate typhoid fever. As a rule, the tuberculosis is more insidious in onset, and in any case the differential diagnosis should not be difficult. The author has seen two cases in which tuberculous peritonitis was asso- ciated with ovarian neoplasms, one of which was a cystic teratoma and the other a pseudomucinous cyst. In both, the outer surfaces of the tumors were studded with tubercles. In another of our cases the tuber- culous peritonitis was associated with a fibromyoma of the uterus, the tubes being tuberculous ; in another case a cervical carcinoma was present. Gallstones were found in still another case. Croom 33 has recorded the 338 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS history of a remarkable case associated with extra-uterine pregnancy, which had ruptured; there was also tuberculosis of the kidney. Tuber- culous peritonitis, occurring as it does in such varying forms, may be mistaken for many of the intraperitoneal diseases, among which may be mentioned sarcoma, omental and mesenteric cysts, renal tuberculosis, ectopic pregnancy, various diseases of the gallbladder, ulcer and cancer of the stomach, duodenal ulcer, appendicitis, ascites due to cardiac or renal disease, and the ascites which is sometimes present as an accom- paniment of the infectious diseases of childhood, or polyserositis. Ex- cept in extremely atypical cases, the differentiation from the above men- tioned conditions should not be difficult. Thomayer 64 believes that in the ascitic variety of tuberculous peritonitis there is a tendency for tympany to be more pronounced on the right side than in other diseases producing ascites, except cancer. It is stated that this is the case, because in tuberculous peritonitis the mesentery of the small in- testines draws them to the right, owing to their oblique insertion, the space thus formed on the left becoming filled with fluid. Whereas this sign is of diagnostic value, the reverse may be the case, and the greatest tympany present on the left side. In many early cases observed by the author, and in some advanced ones, the mesentery has not been markedly diseased and hence has not undergone contraction. The excretion of large quantities of indican, which is so characteristic of diffuse acute peritonitis, is absent in the tuberculous form of the disease. (Noth- nagel. 19 ) Pseudotuberculosis of the peritoneum is a rare disease, which, from the macroscopic appearance of the peritoneum, may be similar to a true tuberculosis, and in some recorded cases is said to have resembled it to some extent histologically. The etiology of this condition is obscure. It would appear in many cases to be due to a reaction of the peritoneum to foreign bodies. Ascites is rarely present, the disease usually simu- lating the fibro-adhesive form of tuberculous peritonitis. The foreign bodies may reach the peritoneum through rupture of cystic neoplasms, hydatids, or rupture of some portion of the gastro-intestinal tract. Cobb 34 has recorded the history of a case due to vegetable material. Alessandri 35 has had a similar case, in which the vegetable residue gained entrance to the peritoneal cavity through the perforation of a gastric ulcer. Meyer (quoted by Cobb 34 ) has seen a case due to cholesterin crystals from a ruptured ovarian dermoid. Hebbring (quoted by Cobb) has recorded the history of a case due to the tenia worm, which gained entrance to the peritoneum from the intestine. Egidi 36 has recorded the histories of cases of pseudotuberculous perito- TUBERCULOSIS OF THE PERITONEUM 339 nitis which have healed after war wounds of the chest and abdomen. The previous history, the absence of tuberculosis elsewhere in the lxidy should make the diagnosis easy in most cases. The tuberculin reaction is of little practical value in the diagnosis of tuberculous peritonitis and may even be misleading. In some cases animal inoculations may be of value, but this has the disadvantage of requiring considerable time. The tubercle bacilli are demonstrable by staining methods in the ascitic fluid in only a small percentage of cases, and even animal inoculation is not certain. Behle 20 states that animal inoculation is positive in only 50 per cent of cases. Paracentesis abdominis is more dangerous than a small incision; if the latter is performed the diagnosis can nearly always be made with certainty, and if any doubt exists, the excision and microscopic examination of a small piece of tissue will render it certain. Animal inoculation of the ground up diseased tissue will give positive results in practically all cases, and is much more reliable than the injection of the ascitic fluid. If a case exists in which the diagnosis is in doubt, a small incision can be made, under local anesthesia if necessary, and if the condition found proves to be one in which operation is indicated, this can then be performed. The author believes this a far preferable method to aspiration with a needle. With the present improved surgical technic, paracentesis abdominis for tuberculous peritonitis or for the diagnosis of vague intraperitoneal lesions, is no longer justifiable. It is less certain and more dangerous. Paracentesis abdominis is not only dangerous, in that grave injury may be done to the intraperitoneal viscera, but it is unreliable. Even if fluid is obtained and negative results obtained by both staining and by the time consuming inoculation methods, tuberculosis cannot be excluded with certainty. Indeed, open- ing the abdomen alone is not a certain method in all cases, but the per- centage of doubtful cases that may be so diagnosed is very much higher than by mere puncture, and if it be supplemented by histologic examina- tion of a small piece of excised tissue and by inoculation of the ground up particles into a guinea pig, it may be regarded as practically certain. Morris 37 relates the history of an instructive case exemplying the diffi- culties sometimes encountered in differentiating malignancy from tuber- culosis. He performed a laparotomy upon a patient, and on opening the abdomen found a condition closely simulating a general carcinomatosis. A small piece of tissue was excised, the abdomen closed, and an unfavor- able diagnosis rendered. A complete cure followed, which the patient attributed to Christian Science, which had been employed after leaving the hospital. Through a mistake, the piece of tissue excised at operation 340 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS had not been subjected to histologic examination. Subsequent examina- tion of this showed tuberculosis, and the cure was consequently due to opening the abdomen. But it is doubtful if the surgeon received credit for the cure. Treatment. — Whereas a definite proportion of cases of tuberculous peritonitis will not yield to either surgical or medical treatment, and whereas the hygienic and medical treatment is of the utmost importance" and should not be minimized, all cases of tuberculous peritonitis are essen- tially surgical, and the final decision as to whether or not to operate upon them should be left to the surgeon. In arriving at a decision as to whether or not to operate upon any given case, many points must be considered and the case carefully studied. Although physical examination of the patient to determine the location and condition of the primary lesion is of the first importance, two other factors besides the condition of the peritoneal lesion must be borne in mind, the first that a certain percentage of cases will recover without operation, and the second that, whereas surgery offers the best hope of a cure in many cases, the end results are, even at best, none too satisfactory. For this reason a conservative attitude should be adopted, and in the majority of cases a preliminary trial of palliative measures is the wisest course. Some surgeons attempt to specify definitely how long this palliative treatment should be tried, and recommend periods varying from 2 to 8 weeks and even longer. All cases should be in- dividualized and no hard and fast rule adopted. As long as the patient continues to improve, operation should be withheld. Unfortunately there are many cases which seem to arrive at a standstill or get definitely worse, and in these the proper decision is often difficult to arrive at. As a general rule, the ascitic variety yields definitely better results by operation and the removal of the primary intraperitoneal focus, than by any other form of treatment. Drainage is not indicated, and often leads to fistulas and mixed infection. The fibroplastic form is decidedly less favorable and must be judged individually ; in those cases in which there are great numbers of adhesions, much thickening of the peritoneum, and extensive involvement, especially if sinuses are present, the operative prognosis is poor, and if much purulent material is present, is decidedly unfavorable. Haggard 38 states he has never seen a recovery of such a case. The miliary variety is not operable under any circumstances, as death is practically certain from involvement of structures other than the peritoneum. Simple laparotomy will cure a certain percentage of cases, but if the primary intraperitoneal focus can be removed, this percentage will TUBERCULOSIS OF THE PERITONEUM 341 be definitely increased, as conclusively proved by Mayo 29 and others. It must be remembered that tuberculous peritonitis is frequent in the child bearing period and that, as tuberculous salpingitis is generally bilateral, the removal of the abdominal focus therefore means the steril- ization of the patient. Despite this fact, bilateral salpingectomy un- doubtedly offers the best hope of cure, and should be resorted to in most cases. The fact that there is a primary focus of tuberculosis present elsewhere, often in the lungs, and that pregnancy so often results disastrously to this class of patients, are added reasons for removing the fallopian tubes. Tuberculous peritonitis is generally associated with sterility. Of Baisch's 39 35 cases, all of whom were in the child bearing period, only 1 became pregnant subsequently. Tweedy 40 has also re- ferred to the frequency of sterility in these cases. The appendix and cecum should be inspected, and as a rule an appendectomy performed. The fibroplastic variety of tuberculous peritonitis is especially prone to originate from the iliocecal region, and in these cases the removal of this part of the bowel is indicated, when this can be performed without too great danger to the patient. In women, however, this is relatively infrequently the case. Except in the ascitic variety, a cure is rarely obtained, unless the primary focus is removed, and even in that variety the outlook is greatly improved, if such an operation is performed. Mere removal of the fluid may cure the ascitic form, as the old ascitic fluid loses its bactericidal properties. Even if fluid reaccumulates after operation, it is said to possess a higher opsonic index and thus a higher resistance to tuberculosis than the old fluid. The admittance of air to the peritoneal cavity has been suggested as the reason for cure in some cases, but more recent study tends to show that it is the removal of the old fluid and the formation of new which is the chief beneficial agent. In the older days, the late Joseph Price was in the habit of referring to this as the "sunshine operation." Other theories which have been from time to time advanced to explain the cures sometimes following simple laparotomy are evacuation of toxins in the exudate, hyperemia produced by the operation, light or oxygen intro- duced to affected area, proliferation caused by operation resulting in encapsulation of the tubercles. The so called floating theory has also been advanced. This theory is based upon the belief that the infection originates in or is kept up by the escape of infectious material from the abdominal ends of the patulous fallopian tubes, and that these re- main open because the fimbria are floating in the ascitic fluid. The re- moval of the fluid gives the tubes a chance to become sealed off and thus prevents the further escape of the infectious material. It is probable 342 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS that many factors enter into the cure of these cases. Murphy 30 believed that tuberculosis of the fallopian tubes rarely, if ever, caused closure of the tubes, unless a mixed infection was present. The important point which has been amply proven is that a definite percentage of cases will be cured by simple laparotomy, and a still greater proportion, if it is possible to remove the primary intraperitoneal focus of the infection. Operation offers but little to those cases in which there a-re numerous small pockets of fluid, much fibrin, and many adhesions, and will, in a certain percentage of cases, result in troublesome fistulae. Fortunately, according to Mayo, the adhesive variety is the most favorable for a spontaneous cure. Various applications to the peritoneum have been suggested. Judd 41 recommends hydrogen peroxid, followed by physiological normal salt solution. He further suggests that the hydrogen peroxid is of some diagnostic aid, in that it produces a frosted appearance of the peri- toneum, and, after flushing with salt solution, the tubercles stand out as pearly white bodies. Kocher 42 recommends swabbing tehe cavities with an iodoform and glycerin solution. Stocker, 43 as a result of animal experimentation, recommends the application of the tincture of iodin and concludes that the iodin exerts a definite curative action and that the danger of its resulting in adhesions has been overestimated. Other investigators have employed various antiseptics. Strong antiseptics to the peritoneum are generally contraindicated, and, except in small, walled off cavities, are, as a rule, to be avoided. Probably the advantages of the various agents, which have been from time to time advocated, have been somewhat overestimated and the bene- ficial results obtained are due more to the surgical measures instigated than to the particular form of application employed. The use of radium or X-ray has been tried in these cases. Our own experience has been that both these agents will very definitely cause an acute exacerbation of pelvic peritonitis in chronic cases, whether of tuberculous or other origin, and in some instances result in the production of an acute general peritonitis. Until the rationale and clinical results of this form of treat- ment have been more thoroughly established, we would hesitate to employ either of these agents upon the class of cases under discussion. Hygienic and medicinal treatment is of the utmost importance. In a considerable proportion of cases such a course, together with suitable palliative treatment directed towards the peritonitis itself will result in a cure or at least temporary improvement. A reasonable trial of the palliative treatment should be attempted, but should not be con- tinued too long. Not all cases are operable, but it is a poor principle TUBERCULOSIS OF THE PERITONEUM 343 to allow what was a comparatively mild case to be converted into a grave one. Prior to the more generally accepted view regarding the advisa- bility of operation in many of these cases, it was not uncommon for this to occur, and even today there is a tendency for the surgeon to receive all the advanced cases, many of which have been treated for prolonged periods by the internist. As a postoperative measure, hygiene and medicinal treatment are, if anything, of even greater importance than in the early stages of the disease. An attempt should be made to place the postoperative patient in the best possible surroundings. Hartel 44 recommends sanatorium treatment when possible. A suitable climate, good diet, and general hygienic treatment will greatly increase the num- ber of permanent cures. The fact must not be lost sight of that many of these patients have more or less definite pulmonary lesions, and, be- cause of the peritonitis, they are below par and therefore especially subject to an exacerbation of the lung condition. As a general rule the postoperative treatment should be continued for a prolonged period ; even after patients are apparently cured they should be urged to exercise proper hygienic measures, and should be kept under observation for at least two years. Even if they have not been sterilized, it is unwise for women to become pregnant for at least this period of time, and in many cases it is better for the patient to permanently avoid conception. This, however, is a point on which each case must be judged individually. Complications. — Operative. — What has been said in a previous chapter regarding anesthesia and operation upon tuberculous patients should be considered. Apart from the danger of the operation or anes- thetic lighting up preexisting, although perhaps quiescent, pulmonary lesions, these patients possess distinctly lessened resistance and bear operative trauma rather poorly. Owing to the character of the infection and the type of intraperitoneal lesions often encountered, fistulas of various kinds are especially prone to follow ill advised surgical pro- cedures. For this reason especial care should be adopted in dealing with adhesions, and drainage should rarely be employed. If fistulas result, they are extremely likely to be chronic and difficult to cure. These patients are particularly subject to wound infection, and every effort should be adopted to guard against this complication. Spencer 49 records the history of an unusual complication, which occurred in a girl 15 years of age. A sinus formed which communicated with an intermittent hematosalpinx and for one year blood was discharged through the sinus at each menstrual period. The condition was verified by operation. The same author reports the history of another case in a girl 18 years of age, in which there was a postoperative fistula through which blood 344 GYXECOLOGICAL AND OBSTETRICAL TUBERCULOSIS appeared at each menstrual period for some months following operation. Kaufmann 46 has recorded the history of cases in which there were several small utero-intestinal fistulas, and numerous instances are on record in which fistulas connecting with the intestinal tract or bladder have been present. Umbilical fistulas are of rare occurrence, but are less infrequent in tuberculous than in any other form of peritonitis. They are more fre- quent in children than in adults, and in the latter, when associated "with a peritonitis, are almost pathognomonic of tuberculosis. Achard and Leblanc 47 have recently reported the history of a case. Tuberculosis in Hernia. — This may occur, either as a localized infection, or as a part of a general peritonitis. Any of the various varie- ties may be present. The neck and bottom of the sac are especially subject to attack. It is more common in men and in children than in women. When starting as a localized infection it may become general. It usually produces periodic attacks of pain and is rarely diagnosed prior to opera- tion. Cornet 48 was probably the first to* report a case of tuberculosis in a hernial sac. Jonnesco, 49 Hagler, 50 and Bruns 51 were among the earlier observers of this condition. Reformation of Ascites Following Operation* — Reaccumula- tion of fluid, following its removal by operation, occurs in certain cases, but is less likely to result, if the primary intraperitoneal focus for in- fection is removed, than otherwise. Reaccumulation of fluid is not a contraindication to operation and many series of cases have been recorded in which ultimate cures have been attained only after repeated operation. Mayo 26 has recorded an instructive series of such cases, some of which have been operated upon seven times. Schley, 52 Murphy, 30 D'Urso (quoted by Jacobson 27 ), and others have reported similar experiences. Results. — -The older literature is replete with reports comparing the re- sults of medical and surgical treatment of this condition — the advantages of the one or other form seeming often to depend on whether or no the given series of cases was reported by an internist or a surgeon. Kronig was the first to call attention to the value of simple laparotomy, and in 1890 reported the results obtained in 139 cases, of which 84 recovered. Shattuck, 53 from material obtained from the Massachusetts General Hospital, showed a mortality of 68 per cent in cases treated medically, as compared with a mortality of 37.5 per cent among patients treated surgically. Gelpke 54 has recorded the results in a series of 71 operative cases, in which there were 4 deaths, as compared with a series of 51 cases treated by medical methods alone, in which there were 6 deaths. Some important statistics have been collected by Bircher, 55 who, in a TUBERCULOSIS OF THE PERITONEUM 345 series of 1,295 cases treated surgically, found 69 per cent of immediate cures, but that only 31 per cent were well a year or more after operation. Wunderlich, 17 among 176 cases treated surgically, found that only 26 per cent were well at the end of 3 years. It must, however, l>e taken into consideration that practically all these cases were the incumbents of a primary focus elsewhere than in the peritoneum, and that a definite proportion of those showing poor end results are doubtless due to this fact. Fenzer 56 states that 35 per cent are now cured by surgery, which were formally fatal, and Moynihan 57 presents even more favorable figures. Baisch, 39 in the study of no cases, found that 40 died within 4 years, about 5/6 of these succumbed in the first year, and that not one died after the fourth year. The cases studied were observed from 5 to 12 years. These, as well as other studies, show that the great ma- jority of fatal cases occurs in the first year following operation. In Baisch's series there were 22 cases of the fibroplastic variety; of these 1 1 were treated medically and 8 died ; among the 1 1 which were subjected to operation, there were 3 postoperative deaths, while 5 of the remaining 8 were well 5 or more years subsequently. Goodrich 12 states that 25 per cent of patients recover, if treated medically, and 80 per cent if treated surgically, but that of the latter, only 30 per. cent survive a 5 year period, 25 per cent perishing in the first year. Caird, 22 in 31 operative cases, observed 3 postoperative deaths, 10 were lost, sight of, 9 died, 9 were alive for periods varying from 2 to 9 years. Mat- teson, 10 in a series of 53 cases treated surgically, found that 23 per cent showed no improvement and. subsequently died. In none of these cases however did death follow immediately after operation, nor was any directly traceable to surgical intervention. Of 38 cases, the after his- tories of which it was possible to trace, 44 per cent were cured of the peritonitis. Russanoff reports 24 cases treated surgically, of which 9 remained well from 2 to 5 years after operation. Hartel, 44 after an ex- tensive review of the literature and an analysis of the end results obtained by medical and surgical treatment, states that the early results of surgery are better, but the longer the periods over which the cases are followed, the closer do the surgical results come to those obtained by medical treat- ment only. These conclusions are probably reached because of the fact that not a few of the cases treated surgically subsequently succumb, either to a recurrence -of the peritonitis, or to some other form of tuber- culosis. In this connection Ochsner 59 states that most cases are first treated medically, and when finally turned over to the surgeon, are in bad condition, and his 50 per cent of recoveries is in the worst cases, whereas, if he had had the case from the beginning, his percentage of 346 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS cures would be 75 per cent at least. Marckthurm (quoted by Senn and Friend 60 ) records 36 cases with 21 permanent cures. Rosch (quoted, by Senn and Friend 60 ) records 358 cases with 20 operative deaths; 70 per cent are reported as cured. Binnie 61 believes 30 per cent cured and 25 per cent improved by surgical intervention. Mayo 13 reports conclusions based upon 144 operative cases ; 59 were operated upon by the older meth- ods, 42 cured, 15 improved, and 2 deaths; in 58 cases the fallopian tubes were removed, 56 recoveries and 2 deaths ; in 27 the vermiform ap- pendix was tuberculous and removed, no deaths. Thus, among 144 operative cases, there was a surgical mortality of 2.JJ per cent. The prognosis in children, especially the very young, is less favorable than in adults. Dingwall-Fordyce, 62 in a series of 137 cases, found that in the majority the onset was prior to the fourth year, the earlier the onset the more severe the case ; the mortality in this series was 46 per cent among the bottle fed infants and 28 per cent among the breast fed. Free fluid was uncommon under 3 years of age. An analysis of 2356 cases treated surgically, some of which were not subjected to modern surgical methods, shows that there were 39 per cent of permanent cures. A more careful analysis, which includes only those cases in which the report states that they have been definitely followed for a period of 3 years or more, shows 31 per cent of permanent cures. Another 30 per cent are definitely improved, and about 36 per cent die of a recurrence of the peritonitis, of tuberculosis elsewhere in the body, or from intercurrent disease in the three year period following operation. The Mayo series of 144 cases treated surgically with 4 deaths is a fair presentation of the immediate surgical mortality in properly selected cases treated by modern surgical methods. LITERATURE 2 3 4 5 6 7 8 9 10 Wells, Sir S. Ovarian and Uterine Tumors. London, 1873. Kronig. Centrbl. f. Chir. 1890. p. 657. Stone, A. K. Bost. Med. Surg. Jr. 1908. 158:705. Schlimpert. Arch. f. Gyn. 191 1. 94:863. Borschke. Virch. Arch. v. 127. Hamman, L. J. Hopk. Hosp. Bui. 1908. 19:256. AI Onsterman-n. Inaug. Dis. Munich, 1890. Rokitansky. Handbook of Pathological Anatomy. 1855. 2:29. Albrecht. Deutsch. Kong, of Gyn. u. Obst. 191 1. Matteson, G. A. Prov. Med. Jr. 191 1. 12:6. TUBERCULOSIS OF THE PERITONEUM 347 11. Allshut, W. Tr. 6th Intern. Cong. Tuberc. 1908. 12. Goodrich, C. H. L. I. Med. Jr. 1910. 4:414. 13. Mayo, C. H. and Mayo, W J. A Collection of Papers. Phila- delphia and London. 19 12. p. 37, 45. 14. Kraus. Monschr. f. Gebh. u. Gyn. 1902. 15:2. 15. Bryant, J. D. Principles of Surgery. Philadelphia and London. I9I3- 16. Osler, Sir W. Principles and Practice of Medicine. New York and London. D. Appleton & Co. 1905. 17. Wunderlich. Arch. f. Klin. Gyn. 1899. 59:216. 18. Baisch. Munch. Med. Woch. Aug. 20, 1907. 19. Nothnagel, H. Diseases of the Intestines and Peritoneum in "Encyclopedia of the Practice of Medicine." Philadelphia and New York. 1907. 20. Behle, A. C. North West Med. 1914. Vol. vi. No. 1, p. 16. 21. Thompson. Brit. Jr. Tuberc. 1907. 1 :25c. 22. Caird, T. M. Edinb. Med. Jr. 1912. 1 :29s. 2T,. Bovaird, D. Arch. Ped. 1909. 26 432. 24. Faludi. Jhrb. f . Kindlik. 1905. 62 :304. 25. Barker, L. F. In Monographic Medicine. New York. 1916. D. Appleton & Co. 3 :684. 26. Mayo, W. J. Am. Jr. Med. A. April 15, 1905. 1918. 71 :6. 27. Jacobson, N. N. Y. St. Jr. Med. 191 1. 11:53. 28. Ross, A. E. Tr. London Path. Soc. 1906. 57:361. 29. Gibbert et Villard. Compt. rend. soc. de biol. 1906. 60:820. 30. Murphy, J. B. Tuberculosis of the Female Genitalia and Peri- toneum. Chicago, 1903. 31. Beale, P. Med. Press and Circ. 1909. 138:112. 32. Monro, J. C. In Keen's Surgery. Philadelphia, 1908, W. B. Saunders. 3 :748. 33. Croom, J. H. Jr. Obst. Gyn. Brit. Emp. 1914. 26:192. 34. Cobb, F. Bost. Med. Surg. Jr. 1907. 157:861. 35. Alessandri. Policlin. Aug., 1908. 36. Egidi, G. Policlin. 1920. 2y, No. 1. 37. Morris, R. J. Arch. Diag. 1914- 7:146. 38. Haggard, W. D. Jr. Tenn. St. Med. A. 1909. 2:126. 39. Baisch. Arch. f. Gyn. 1909. 84:345. 40. Tweedy, E. H. Jr. Obst. Gyn. Brit. Emp. 1912. 22:342. 41. Judd, A. N. Y. Med. Jr. 1911. 93:1222. 42. Kocher, T. A Text Book of Operative Surgery. London, 191 1. 43. Stocker, S. Schweiz. Rundsch. f. Med. 1913. 13:745- 348 GYNECOLOGICAL AND OBSTETRICAL TUBERCULOSIS 44. Hartel, F. Ergebn. d. Chir. u. Orth. 1913. 6:370. (Contains an extensive bibliography.) 45. Spencer, W. G. Brit Med. Jr. Jan.. 10, 1914. 46. Kaufmann. Arch. f. Gyn. 1887. 29:407. 47. Achard, C., et Lebdanc, A. Bui. et mem. soc. med. des hop. de Paris. 1918. 42:301. 48. Cornet, G. Tuberculosis. Philadelphia, New York, and London. 1904. p. 194. 49. Jonnesco. Rev. de chir. 1891. 11:185. 50. Hager. Arch. f. Chir. 1893. 15:316. 51. Bruns, P. Bietr. z. Klin. Chir. 1892. 9:209. 52. Schley. N. Y. Med. Rec. 19*12. 81 493. 53. Shattuck. Am. Jr. Med. Sc. 1902. 124:1. 54. Gelpke. Deutsch. Ztschr. f. Chir. 84:512. 55. Bircher, E. Die Chronische Bauchfell Tuberkulose, ihre Be- handlung mit Rontgenstrahlen. Aarau, 1907, Sauerlander. 56. Fenzer. Ann. Surg. Dec, 1901. 57. Moynihan, Sir Bi Surgical Operations: 1905. p. 89. 58. Russanoff, A. G. Dissertation. Moscow, 19 13. (Contains an extensive bibliography. ) 59. Ochsner, A. J\ Tr. Am. Surg. A. 1902. 20:191. 60. Senn, E. J., and' Friend, L. Principles of Surgery. Philadelphia, 1909. p. 546. 61. Binnie, J. F. Manual of Operative Surgery. Philadelphia, 1913. P- 438- 62. Dingwall-Fordyce, A. B;rit. Med. Jr. 1909. p. 761. 63. Kelly, H. A. Operative Gynecology. 1914. INDEX Abortion, Indication for, in the tuber- culous, 265 — result of, 266 — technic, 271 Abscess of the ovary, 195 Acid proof bacteria, differentiation from tubercle bacillus, 10 Adenitis, external genitalia, 113 — with vaginitis, 142 Adenofibroma, with tuberculosis of breast, 318 Adenomyoma and tuberculosis, 227, 228 Adnexitis — See Salpingitis Age of menopause in the tuberculous, 293 Amenorrhea, 291, 293 — in pulmonary tuberculosis, 284 Anesthesia and pulmonary tuberculo- sis, 300 — choice of, 302, 303 — classification of pulmonary lesions, 301 — importance of expert anesthetist, 304 — precautions, 303 — salpingitis, 215 — spinal, 303 Animals, congenital tuberculosis in, 58 Appendicitis and tuberculosis, 240 Ascites, 334 — character in peritonitis, 334 — general peritonitis, 329 — reformation after operation, 344 Bacillemia, tuberculous, 51 Bacillus leprae, 10 — differentiation from tubercle bacil- lus, 10 Bartholin's gland, tuberculosis of, 121. See External Genitalia Bartholinitis, 121 Biopsy — cervix, 6 — external genitalia, 6 — vagina, 6 Benign tumor and tuberculosis, 226 Bladder, rupture pyosalpinx into, 235 Blood picture in peritonitis, 335 Body of uterus, tuberculosis of, 182 Bone tuberculous and pregnancy, 279 Bowel, rupture pyosalpinx into, 235 Breast, tuberculosis of, 312 — age, 312 — bilateral, 317 — biopsy and other diagnostic meth- ods, 319 — cold abscess of, 314 — confluent variety, 314 — course of disease, 317 — diagnosis, 318 — disseminated variety, 315 — end results, 320 — frequency, 309 — frequency of fistula, 318 — general condition, 317 — historic, 309, 312 — in male, 309 — obliterative mastitis, 315 — predisposing causes, 312 — primary and secondary, 310 — routes of infection, 311 — sclerosing variety, 315 — ■ similarity to true neoplasm, 316 — symptoms, 316 — treatment, 319 — with true neoplasms, 318 — varieties, 314 Cancer and tuberculosis, 224, 225, 228, 318 Carcinoma and tuberculosis, 224, 225, 228, 318 Carcinoma differentiated from cervici- tis, 157 Caseous indometritis, 183 Cervicitis, 152 — age, 152 — biopsy, 6 — carcinoma, 228 — case histories, 160 349 350 INDEX Cervicitis, diagnosis, 156 — differential, 158 — differentiation from carcinoma, 157 — discharge, 153 ■ — frequency, 149 — hemorrhage, 153 — histologic simulating carcinoma, 22 — historic, 149 — interstitial, 156 — interstitial, pathology, 21 — location of primary focus, 151 — miliary, 156 — miliary, pathology, 22 — other portions of the genital tract involved with, 150 — pain, 153 — papillary, 155 — papillary, pathology, 21 — pathology, 20 — phthisis with, 151 — portion of cervix involved, 154 — predisposing causes, 151 — primary, 150, 151 — prognosis, 158 — pseudo neoplasms, 227 — salpingitis with, 150 secondary infliction, 150 symptoms, 152 treatment, 1 59 ulcerative, 155 ulcerative, pathology, 20 varieties, 154 — with endometritis, 150 Cervix, 149. See Cervicitis Cesarean section in the tuberculous, 276 Coitus, infection by, 97, 98, 99 Colored race, susceptibility, 331 Complications, peritonitis, 343 Confluent mastitis, 314 Congenital tuberculosis, 44, 251, 277 — animal experiments, 59 — case histories, 64 — definition, 45 — etiology, 46 — experimental criticism, 60 — fate of the congenitally infected, 63 — frequency, 55 — germinative, 49 — germinative spermatozoic, 46 — histology of placenta in relation to, 51 — historic, 45, 58 Congenital tuberculosis, in animals, 58 — literature, 87 — period in which transmission is most likely to occur, 61 — predisposing factors, 62 — summary, 85 — unfertilized ovum, 4 Convalescence in the tuberculous, 305 Corporeal endometritis, See Endome- tritis Corpus uteri, 182 — tuberculosis, 182 Curettage, 6 — diagnostic, 6 in endometritis, 7 Cystadenoma and tuberculosis, 227 Cystitis with salpingitis, 192. See Sal- pingitis — and pregnancy, 279 Decidua, See Deciduitis Deciduitis, 30 Diagnosis by staining method employ- ing exudate, 11 — cervicitis, 156 — differential, pelvic inflammatory dis- ease, 207 — endometritis, 186 — external genitalia, 115 — histologic summary, 12 — laboratory methods, summary, 12 — pelvic inflammatory disease, 204 — pelvic peritonitis, 204 — salpingitis, 204 — tuberculin in, 11 — tuberculin in salpingitis, 206 — vaginitis, 143 Diagnostic curettage, 6 Diagnostic excision, 6 lower genital tract, 6 Discharge, 6 — examination of, for diagnosis, 6 Dissemination from genital lesions, 237 Disseminated mastitis, 315 Drainage, in the treatment of salping- itis, 220 Dysmenorrhea, 185 — in pulmonary tuberculosis, 284, 289 — treatment, 290 Dyspareunia in vaginitis, 142 Eclampsia, 279 — tuberculosis, 279 INDEX 35i Endometritis, 182 — caseous, 183 — cervicitis with, 150 — diagnostis curettage, 7 — diagnostic examination of leucor- rhea, 7 — diagnosis, 185 — frequency, 182 — miliary, 183 — pathology, 23 — symptoms, 184 — treatment, 185 — ulcerative, 183 — varieties, 183 — varieties, pathology, 23 Ether, See Anesthesia Etiology, 109 — external genitalia, 109 — menstrual disturbances in pulmo- nary tuberculosis, 285 Examination of discharge for diagno- sis, 6 Excision, diagnostic, See Biopsy — diagnostic, lower genital tract, 6 Experimental congenital tuberculosis, 59 External genitalia, 108 — adenitis, 113 — age, 111 — biopsy, 6 — case histories, 119 — diagnosis, 17, 115 — etiology, 109 — frequency, 108 — frequency of primary, 108 — frequency of secondary, 108 — genitalia, 109 — historic, 108 — hypertrophic, 109, 114 — hypertrophic variety, pathology, 17 — modes of infection, 111 — parts most frequently involved, 114 — pathology, 15 — prognosis, 115 — pruritis, 113 — pseudoneoplasms, 227 — symptoms, 109 — trauma as predisposing cause, 109 — treatment, 116 — ulcerative, 109, 113 — varieties, 15, 109 Fertility in pulmonary tuberculosis, 244 Fetal tuberculosis, 50 Fetus, susceptibility, 50 — tubercle bacillus in, without histo- logic change, case reports, 81 Fever during menses, 295 Fibroplastic general peritonitis, 330 Fistula following operative treatment of salpingitis, 220 — following operative treatment, gen- eral peritonitis, 343 Frequency, of genital tuberculosis, 103 — external genitalia, 108 — of pregnancy and tuberculosis, 244 — of primary lesions producing geni- tal tuberculosis, 105 General peritonitis, See Peritonitis — ascites, 334 — ascites following operation, 344 — blood picture, 335 — character, 334 — character of fluid in, 334 — complications, operative, 343 — diagnosis, 335 — differential diagnosis, 336 — fistula following operation, 343 — frequency, 332 — hernia, 344 — in children, 332 — in the colored race, 331 — latent variety, 331 — mortality, 344 — paracentesis, 339 — peritonitis, 344 — prognosis, 333 — pseudotuberculous peritonitis, 337 — pseudo-tumors in, 336 — reformation ascites after operation, 344 — results, medical treatment, 344 — results, surgical treatment, 344 — symptoms, 333 — treatment, 340 — treatment, medical. 340 — treatment, surgical. 340 — variety tubercle bacillus causing, 333 Genital, historic, 3 Genital infection, 95 — primary, 95 — routes of, 95 Genital lesions and pregnancy, 279 Genital tuberculosis, pregnancy, 103 352 INDEX Genococcal salpingitis, 207 — differential diagnosis 207 Hemoptysis, 295 — periodic, 295 Hemorrhage due to cervicitis, 153 Hernia, 237 — tuberculosis in, 237 — tuberculous peritonitis, 344 Histologic methods, diagnosis, sum- mary, 12 — 'ulcerative form, tuberculosis of the external genitalia, 17 Histology of placenta in relation to congenital tuberculosis, 51 Historic, p. — — breast, tuberculosis of, 309 — cervicitis, 149 — congenital tuberculosis, 45, 58 — external genitalia, 108 — general, 1 — general peritonitis, 323 — genital, 3 Historic, 249 — lactation in the tuberculous, 249 — menorrhagia in the tuberculous, 294 — pregnancy and tuberculosis, 243 — tubercle bacilli in decidua, 58 — vaginitis, 140 Hydrosalpinx, 39, See Salpingitis — pathology 7 , 39 torsion, 231 Hyperexia during menses, 295 Hypertrophic forms, external genitalia, pathology, 17 Hypertrophic external genitalia, 109 — vaginal, 19, 142 — variety, external genitalia, 114 — variety, external genitalia, pathol- ogy, 17 Hypoplasia and tuberculosis, 287 Hysterotomy, pregnancy and tubercu- losis, 272 Infection, autogenital, 95 Infection, marital, 261 — predisposing causes, 103 — primary, experimental, 99 — primary genital, 95 — routes of genital, 95 — routes of, summary, 102 — secondary, 101 Infection, secondary frequency of pri- mary foci, 105 Interstitial cervicitis, 156 Intramural abscess, 27 — pathology, 27 — recorded cases, 27 Laboratory methods of diagnosis, sum- mary, 12 Latent general peritonitis, 331 Leukorrhea, 184 — demonstration of tubercle bacillus in, 11 — diagnostic examination of, 7 — due to cervicitis, 153 — in pulmonary tuberculosis, 295 — tubercle bacilli in, 184 Local anesthesia, See Anesthesia Mammary tuberculosis (also se;± Breast, tuberculosis of), 309 Marital infection, 261 Marriage, law regarding, 261 — of tuberculous women, 261 Mastitis (also see Breast, tuberculosis of), 309 Menopause, 293 — age of, in the tuberculous, 293 — in the tuberculous, 288 Menorrhagia in pulmonary tuberculo- sis, 284, 293 — treatment, 294 Menstrual disturbances and pulmo- nary tuberculosis, 284 Menstruation, influence on tempera- ture, 295 — etiology, 285 — frequency, 284 Miliary cervicitis, 156 — endometritis, 183 — peritonitis, acute, 329 — vaginitis, 142 — vaginitis, pathology, 143 Mortality, pelvic inflammatory disease, 211 — salpingitis, 211 Myoma and tuberculosis, 228 Myometritis, 182 — pathology, 26 Neoplasms, benign and tuberculous, 226 — differentiation from cervicitis, 157 INDEX 353 Neoplasms and tuberculosis, 224, 318 New growths, benign and malignant tuberculosis, 224, 318 Nitrous oxide, See Anesthesia Obliterative mastitis, 315 Oophoritis, 192 — pathology, 41 Operative treatment, salpingitis, 218 Operation and pulmonary tuberculosis, 305 — anesthesia, 300 anesthesia, chloroform, 305 anesthesia, choice of, 302, 303 anesthesia, ether, 303 anesthesia, local, 302 anesthesia, mixtures, 303 — ■ — anesthesia, nitrous oxide, 303 anesthesia, precaution, 303 anesthesia, spinal, 303 convalescence, 305 importance of expert anesthetists, 304 precautions, 303 results, 305 with complication of pulmonary lesions, 301 Organisms likely to be mistaken for the tubercle bacillus, 8 Osseous tuberculosis and pregnancy, 279 Ovarian abscess, 195 — pathology, 41 — tumors and tuberculosis, 227 Palliative treatment, salpingitis, 216 Papillary cervicitis, 155 Paracentesis abdominis in peritonitis, 339 Pathology, 15 — adenitis, inguinal in ulcerative tu- berculosis of the external genitalia, 17 — cervix, 20 histologic, 22 histologic picture simulating car- cinoma, 22 interstitial variety, 21 papillary variety, 21 miliary variety, 22 ulcerative variety, 20 — deciduitis, 30 — endometritis, 23 Pathology, endometritis, caseous, 24 miliary, 23 varieties, 23 — external genitalia, 15 hypertrophic variety, 17 ulcerative variety, 16 — general peritonitis, 327 — hydrosalpinx, 39 — hypertrophic variety, vaginal, 19 of the external genitalia, 17 — intramural abscess, 27 — miliary vaginitis, 19 — myometritis, 26 — oophoritis, 41 — ovarian abscess, 42 — perioophoritis, 41 — ■ perisalpingitis, 35 — placental tuberculosis, 31 — pyosalpinx, 38 — salpingitis, 34, 36 histologically suggesting carci- noma, 40 isthmica nodosa, 36 — ulcerative form of the external gen- italia, 16 — vaginitis, 18 hypertrophic form, 19 ulcerative variety, 18 Pelvic inflammatory disease, See Pel- vic Peritonitis Pelvic peritonitis, 192, See Salpingitis — diagnosis, 204 — mortality, 211 — operative treatment, 218 — prognosis, 209 — results of operative treatment, 211, 220 — treatment, 214 Periappendicitis and salpingitis, 240 Perimetritis, 182 Perioophoritis, 192, See Salpingitis, also Pelvic Peritonitis pathology, 41 Perisalpingitis, — pathology, 35 Peritonitis, 199 secondary form genital focus, 237 Peritonitis, general, 344 — acute miliary, 329 ascitic, 329 — classification, 329 fibro-plastic, 330 354 INDEX Peritonitis, general, classification, his- toric, 323 intraperitoneal foci, 324 mixed infection, 328 mode of development, 327 — pathology, 327 primary and secondary, 324 routes of infection, 325 and salpingitis, 326, 327 suppurative variety, 331 Peritonitis, pelvic, 192. See Salpingitis — diagnosis, 204 differential, 207 — mortality, 211 — operative treatment, 218 — prognosis, 209 — results of operative treatment, 211, 220 — treatment, 214 Peritonitis and hernia, 237, 344 — pregnancy, 279 Placenta, 249 — histology in relation to congenital tuberculosis, 51 — tubercle bacillus in, without histo- logic change, case reports, 81 Placental tuberculosis, 44, 50 — case histories, 73 — frequency, 44 — pathology, 31 — predisposing factor, 62 — summary, 86 Placentitis, See Placenta Precaution when operating, 303 Predisposing causes, cervicitis, 151 — external genitalia, 108 — to infection, 103 Pregnancy and genital tuberculosis, 279 — osseous tuberculosis, 279 — peritonitis, 279 Pregnancy and tuberculosis, 243 — abortion, 270 choice of operation, 270 consultation, 270 convalescence, 272 technic, 271 — care of child, 278 — cause for exacerbations, 245 — cesarean section, 249, 276 condition of child, 251 — diagnosis of pregnancy, 264 — fate of child, 251 — fertility, 244, 253 Pregnancy and tuberculosis, frequency, 244 — hysterotomy, 272 — influence of lactation, 259 — influence of pregnancy on the course of tuberculosis, 254 — influence of pregnancy upon lesions other than the lungs, 278 — influence of tuberculosis on the course of pregnancy, 253 — labor, 275 — lactation, 249 — laryngeal involvement, 258 — law regarding marriage, 261 — marriage, 261 — mortality, 245 — nursing, 278 — physiology of pregnancy, 245 — placenta, 248 — premature labor, indication for, 276 — prophylactic measures, 261 — puerperium, 248, 278 — results of abortion, 266 — sterilization of the tuberculous, 270 Pregnancy and tuberculosis, — susceptibility of the pregnant, 249 — treatment, 263 after fifth month, 274 hygienic, 263 indications for abortion, 265 pregnancy prior to fifth month, 265 — tubercle bacilli in maternal milk, 259 — tuberculin in, 260 — value of statistics, 251 Premature labor, indication in the tu- berculous, 276 Preoperative treatment, salpingitis, 216 Primary genital infection, 95 — by coitus, 97, 98, 99 — experimental, 99 Primary foci in secondary genital tu- berculosis, 105 Prognosis, cervicitis, 158 — external genitalia, 115 — pelvic inflammatory disease, 209 — pelvic peritonitis, 209 — salpingitis, 209 Pruritis, 113, See External Genitalia Pseudo-carcinoma, 40 Pseudoneoplasms, 227 Pseudotuberculous peritonitis, 338 INDEX 355 Pseudotumors in general peritonitis, 336 Puerperium in the tuberculous, 278 Pulmonary tuberculosis and anesthesia. See Anesthesia Pulmonary tuberculosis and pregnancy. See Pregnancy and Tuberculosis Pyometra, 230 Pyosalpinx, method of formation, 38 — pathology, 38 — rupture, 233 diagnosis, 234 into bladder, 235 into bowel, 235 symptoms, 234 treatment, 235 — torsion, 230 Routes of infection, summary, 102 Rupture, tuberculous adnexa, symp- toms, 234 Salpingitis, 192 — acute, 197 — age, 196 — anesthesia, 215 — and adenomyoma, 228 — and results, 211, 220 — carcinoma and, 224 — cervicitis with, 150 — chronic stage, 200 — development of secondary lesions after operation, 211, 220 — diagnosis, 204 — diagnostic use of tuberculin, 206 — differential diagnosis, 207 Salpingitis, etiologic factor in tubal pregnancy, 239 — frequency, 192 — general peritonitis with, 326, 327 — histologically suggesting carcinoma, 40 — isthimica nodosa, pathology, 36 — mistaken for nephritis, 236 — mortality, 211 — operative mortality, 220 — operative treatment, 218 — palliative treatment, 216 — pathology, 34, 36 — and periappendicitis, 240 — physical signs, acute stage, 199 — predisposition to, 195 — preoperative treatment, 216 Salpingitis, primary, 193 — prognosis, 209 — pseudoneoplasms, 227 — results of operative treatment, 211, 220 — rupture, 233 — secondary, 193 — spontaneous cure, 215 — torsion, 230 — treatment, 214 Sarcoma, differentiation from cervici- tis, 157 Scanty menstruation in the tubercu- lous, 291 Sclerosing mastitis, 315 Secondary genital infection, 101 Smega bacillus, 7 — differentiation from tubercle bacil- lus, 8 by staining, 9 Spinal anesthesia, 303 Sterilization of the tuberculous, 270 Streptococci, pelvic inflammatory dis- ease, differential diagnosis, 207 Suppurative general peritonitis, 331 Susceptibility, fetus, 50 Syphilis and tuberculosis, differential diagnosis, 239 Temperature, influence of menses on, 295 Tubal pregnancy, salpingitis, etiologic factor, 239 Tubercle bacillus, diagnostic demon- stration in leukorrhea, 11 — differential staining, 9 from bacillus leprae, 10 from other acid proof bacteria, 10 from smega bacillus, 8 — in decidua, historic, 58 — in fetus without histologic change, case reports, 81 — in fluid of general peritonitis, 339 — in the hypertrophic variety, exter- nal genitalia, 17 — in leucorrhea, 184 — in maternal milk, 259 — in placenta, 249 without histologic change, case reports, 81 — in ulcerative tuberculosis of the ex- ternal genitalia, 17 356 INDEX Tubercle bacillus, organisms likely to be mistaken for, 8 — variety causing general peritonitis, 333 Tuberculin, — diagnosis in salpingitis, 206 diagnostic use, 11 practical value in diagnosis, 12 in pregnancy and tuberculosis, 260 Tuberculosis and carcinoma, 225 — cancer and, 318 — genital as primary focus for spread, 237 — ■ in hernia, 237 — neoplasms and, 318 — non-malignant tumors of the geni- tal tract and, 226 — pregnancy and, 243 — syphilis and, 239 — wound infection, 239 Tumors, benign and tuberculous, 226 — differentiation from cervicitis, 157 — and tuberculosis, 224, 318 Ulcerative cervicitis, 155 — pathology, 20 Ulcerative endometritis, 183 — pathology, 24 Ulcerative external genitalia, 109, 113 Ulcerative tuberculosis of the external genitalia, pathology, 16 Ulcerative vaginitis, 142 — 'pathology, 18 Umbilical fistula in general peritonitis, 344 Urethra, See External Genitalia Urethritis, See External Genitalia Vagina, See Vaginitis Vaginitis, 140 — biopsy, 6 — case histories, 144 — diagnosis, 143 — etiology, 140 — frequency, 140 — historic, 140 — hypertrophic variety, 142 hypertrophic pathology, 19 — miliary variety, 142 pathology, 19 — ■ pseudoneoplasms, 227 — symptoms, 141 — treatment, 144 — ulcerative, 142 pathology, 18 — varieties, 142 pathology, 18 Vicarious menstruation in pulmonary tuberculosis, 284, 294 Vulva, See External Genitalia Vulvitis, See External Genitalia Vulvovaginal Gland, See External Genitalia Wound infection, 239 (1) COLUMBIA UNIVERSITY LIBRARIES This book is due on the date indicated below, or at the expiration of a definite period after the date of borrowing, as provided by the library rules or by special arrangement with the Librarian in charge. DATE BORROWED DATE DUE DATE BORROWED DATE DUE 'VPR ^ '3* 2 C28(251 ) lOOM 2002178753