Emil Novak

Relation of endometrial hyperplasia to adenocarcinoma of the uterus

Abstract The evidence presented in this paper points to a relationship of some sort between hyperplasia of the endometrium and corporeal adenocar-cinoma. The material studied includes 804 cases of hyperplasia and 104 of adenocarcinoma, encountered in a review of 12,813 cases passing through our laboratory during an eleven-year period (Jan. 1, 1925, to Jan. 1, 1936). While in the overwhelming majority of cases hyperplasia is a very frankly benign lesion, a small minority (14 of our 804) reveals evidences of marked proliferative tendencies which may even simulate cancer. The histologic characteristics of benign hyperplasia present degrees and variations which are discussed in the paper, as are the proliferative and pseudomalignant pictures at times encountered (stratification, adenomatous proliferation, marked atypicalness of glands, syncytium-like epithelial proliferation, squamous metaplasia of gland or surface epithelium, etc.). Atypical gland proliferations, simulating adenocar-cinoma, are especially frequent in the polyps so often seen with hyperplasia. An interesting finding in this study was that hyperplasia is not rare in women long after the menopause (40 of 804 cases); and the etiology and significance of this are discussed. The occasional occurrence of hyperplasia with bleeding in elderly women lessens the significance of these findings as pointing to the probability of granulosa-cell carcinoma of the ovary, unless an ovarian tumor can actually be palpated. In the study of the 104 cases of adenocarcinoma, the most impressive result was the demonstration of a coexisting hyperplasia and adenocar-cinoma in fully 25 of the cases in which some of the noncancerous endometrium was available for study. Since the great majority of the adenocarcinoma cases (78 of 92 in which we have accurate age data) were beyond the menopausal age, this at once suggests that a postmenopausal hyperplasia, or, perhaps more accurately, the endocrine dysfunction responsible for it, must strongly predispose to the development of adenocarcinoma. Since a persistence and relative excess of estrin is accepted as the cause of hyperplasia, it would seem that it is this endocrine factor which must be suspected as the one predisposing to cancer genesis. It should be emphasized that the ordinary hyperplasia of the reproductive epoch is not only frankly benign from the histologic standpoint, but also that it has no apparent predisposing influence in the causation of adenocarcinoma during menstrual life. It would seem from our studies that it is the postmenopausal persistence of hyperplasia which is in some way bound up with the occurrence of the common postmeno-pausal type of adenocarcinoma. The question of the relation between estrogenic and carcinogenic substances, and the carcinogenic properties of estrogenic substances, is discussed in our paper. Whether the persisting estrin stimulation in cases of postmenopausal hyperplasia serves merely to keep up a form of chronic irritation, or whether its carcinogenic effects are more direct and fundamental, cannot be answered as yet, though the latter seems, in the light of recent experimental work, to be the more likely explanation. In our own cases of coexisting hyperplasia and adenocarcinoma, we have in some been able to show a definite transition of the benign to a border-line and then to an undoubtedly malignant pattern: so that, in the endometrium at least, it would seem that we are dealing with histologic intergrades between benign and malignant lesions. When such pictures are encountered in the endometrium obtained by diagnostic curettage, there is often a justification for the dictum, “Nicht Karzinom, aber besser heraus.” Since, on the basis of these observations, hyperplasia is to be classed as a precancerous lesion, in the sense of one predisposing to later adenocarcinoma, it would at first thought seem that we should all the more generally recommend abolition of ovarian function by x-ray or radium in the treatment of functional bleeding of the menopause. Whether or not this would be of any practical value, however, is questionable since it seems quite certain that other sources of estrogenic substance than the ovary must be reckoned with. One must be struck with the analogy of this problem to that of the breast, in which likewise the ovarian hormones can produce either very benign forms of cystic mastitis, or more markedly proliferative pictures which are scarcely or not at all distinguishable from cancer. In still an-other field, the cervix, we are now finding similarly puzzling borderline pictures, though we know almost nothing about their possible endocrine relationships. We refer particularly to leucoplakia and, even more, to the intraepithelial carcinoma or Bowens disease, in which the epithelial layer presents all the characteristics of cancer except for invasiveness, and which are quite similar in appearance to the growing margin of an actual cancer. Are such lesions in themselves benign, as would seem to be indicated by the fact that they have often been cured by the most conservative procedures? Or are they inevitable precursors or very early stages of cancer, which, as in Bowens disease of the skin, may not develop invasiveness and other malignant characteristics until the lapse of perhaps many years? Finally, may such lesions represent the response of the cervical epithelium to abnormal endocrine stimulation? In all three of these fields, cervix uteri, corpus uteri, and breasts, we have to deal with the possible roˆle of ovarian hormones. This seems significant in view of the growing opinion that the carcinogenic possibilities of estrogenic substances are most to be reckoned with in those organs in whose growth and activities estrin normally plays an important part.

Cyclical and other variations in the tubal epithelium

Abstract While the tubal mucosa does not participate in the bleeding of the menstrual process, its epithelium exhibits a definite cyclical change comparable with that of the endometrium. It is, however, not nearly so conspicuous as the latter, being concerned more with microscopic changes in the cells rather than with the grosser changes in pattern seen in the endometrium. The present study, based upon the careful examination of the tubes, from 136 cases, in almost all of which the endometrium was also available for study, has demonstrated the following chief facts: 1. 1. The tubal epithelium consists of two chief types of cells, the ciliated and the nonciliated. The latter are often spoken of as “secretory” cells. A third type, the “peg” cells (“Stiftchenzellen,” “Schaltzellen”) are also described, but it is probable that these represent only a phase of the nonciliated cells. 2. 2. In the interval phase, the epithelium is uniformly tall, the ciliated cells being broad, with rounded nuclei near the free margin, while the nonciliated cells are rather narrower, the nuclei being more deeply placed and taking a deeper stain. 3. 3. In the premenstrual phase the ciliated cells become lower, so that the “secretory” cells project beyond them, giving the epithelial margin a ragged, uneven appearance. The “secretory” cells show a bulbous herniation into the lumen of the tube, often carrying the nucleus with it. This extrusion of nuclei is similar to that seen in many lower animals, but its significance is not known. In spite of the great loss of cells, mitoses are rarely seen in the tubal epithelium. 4. 4. During the stage of menstruation, the premenstrual changes are carried further, the epithelium becoming quite low. The ciliated cells, especially, remain broad and low, but the secretory cells also, having been emptied of their cytoplasm, are much lessened in height, the nuclei often being quite bare of cytoplasm. “Peg” cells are numerous, their appearance and distribution suggesting that they are merely emptied “secretory” cells. 5. 5. The postmenstrual phase is characterized first by a low epithelium, which quite rapidly, however, increases in height, so that by the third or fourth day after menstruation it is often almost as tall as during the interval phase. The cells are narrow, closely placed, and, after the first day or so, of uniform height. 6. 6. During pregnancy, the epithelium becomes even lower than in the menstrual stage, and in the later stages it may become almost flat in many places. Secretory changes are not seen at this time. 7. 7. Cilia can be demonstrated in all stages, especially through the examination of fresh tissues, by the technic described in the paper. They are also found in the tubal epithelium of young children and in women many years beyond the menopause. This suggests that they must have some other function than that of assisting in the propulsion of the ovum. Perhaps, as has been suggested by Hartman. their chief role may be that of keeping the tubal lumen cleansed of foreign particles of any kind. 8. 8. Efforts at differential staining of the secretion have thus far been unsuccessful. Neither glycogen nor mucin can be demonstrated, unlike the findings in the secreting endometrium. The various views as to the significance of this tubal secretion are discussed in the paper, although there is as yet no positive knowledge as to this point. 9. 9. The prepuberal tubal epithelium is rather low, but shows both chief types of cells. Cilia, however, are very sparse, and are usually not seen at all in fetal or very early postnatal life. 10. 10. The epithelium of the postmenopausal tube may remain quite high for a surprisingly long time, perhaps a number of years after the cessation of menstruation. Cilia, likewise, may persist for many years. Sooner or later, however, the tubal folds become rounded, of fibrous appearance, the epithelium becomes low or even quite flat, and cilia, of course, disappear. 11. 11. The tubal epithelium of tubes removed from patients suffering with hyperplasia of the endometrium was studied because the latter condition is unquestionably associated with a functional disturbance of the ovaries. Characteristically the epithelium was found to be high, uniform, compact, with narrow cells, most of which were ciliated. There was no evidence of secretory change. This bears out the view, for which there is other evidence, that the functional disturbance consists of an excess or persistence of the follicle stimulus, with an absence of the corpus luteum influence. 12. 12. The comparison of the tubal cycle in women with that of the lower animals, like the comparative study of the uterine and ovarian cycles, emphasizes the important differences, chronologic and histologic, which exist. For example, the estrus tube of the rodent resembles, not the menstrual or premenstrual tube of the human, but the interval phase. Since estrus in the lower type is undoubtedly due to the follicle hormone, it seems clear that in the human the maximum of follicle influence is reached during the interval phase, and that the later changes are due to the corpus luteum influence. To bear this out, the picture in the animal tube which resembles the human premenstrual tube is that seen in the metoestrum, during which stage the corpus luteum apparently plays the dominating role.

Benign trophoblastic lesions in Mathieu Chorionepithelioma Registry (hydatidiform mole, syncytial endometritis)☆

Abstract 1. 1. This is a report of 120 cases of benign hydatidiform mole, 34 cases of chorioadenoma destruens, and 27 of syncytial endometritis in the Mathieu Memorial Chorionepithelioma Registry. 2. 2. The gross and microscopic appearances of these two lesions are reviewed. 3. 3. Contrary to the findings of some authors, we have been unable to derive much help from the gross and microscopic appearance of evacuated molar tissue in predicting whether a given mole will or will not later develop malignant histologic or clinical characteristies. 4. 4. The lesion called syncytial endometritis is a residuum of normal pregnancy, abortion, or hydatidiform mole, and not an atypical variety of choriocarcinoma. Our follow-up of these cases supports this viewpoint. 5. 5. Once again we stress the frequency with which cases of hydatidifrom mole, chorioadenoma destruens, and syncytial endometritis are misdiagnosed as choriocarcinoma. The common sources of error are discussed and we have tried to show how most of these pitfalls may be avoided.

Choriocarcinoma of the uterus: Study of 74 cases from the Mathieu Memorial Chorionepithelioma Registry

Abstract This study is based on the rich material of the Mathieu Memorial Chorionepithelioma Registry. This includes 74 cases of authenticated choriocarcinoma, probably the largest group ever available for direct study by a single group of pathologists. It is chiefly as regards pathology, pathologic differentiation, dissemination, and prognosis that our material is most helpful, and it is these aspects which are selected for special emphasis in this paper. The question of nomenclature is discussed, with criteria for the classification we have used. The microscope is the final arbiter of diagnosis, and not the clinical course or the biologic behavior. The routes of dissemination are discussed and tabulated, with discussion of certain vagaries, such as the occurrence of metastasis in the usually only locally invasive chorioadenoma destruens or even benign mole, the possibility of spontaneous therapeutic regression of metastases, or of the primary uterine site with death from metastases, etc. We consider our survival figures more accurate than most others, because we believe all our cases are authentic choriocarcinoma. Of our 74 patients 13 (17.5 per cent) are living without recurrence one year or more, an adequate follow-up period for this rapidly fatal disease, which usually kills within 6 months. If we were to add 3 additional survivors living and well only 6 months, the survival rate would be increased to 21.6 per cent.

Atypical endometrial hyperplasia simulating adenocarcinoma.

Benign hyperplastic lesions of the endometrium have often been mistaken for adenocarcinoma. Areas of Swiss-cheese endometrium of considerable size may be found in normally functioning organs. The reaction of the endometrium to ovarian hormones is not always the same in all its parts. Young immature endometrium are highly responsive to the growth effect of estrogen and refractory to the differentiating hormone progesterone. The reverse is true in more mature endometrial elements. When squamous areas occur in a benign hyperplastic endometrium they are not necessarily evidence of malignancy. The gland characteristics should be used for diagnosis. Curettage or curettage plus radiologic induction of menopause was used in 18 described cases of uterine bleeding. All showed histologic changes which might have been mistaken for adenocarcinoma. 26 cases of atypical hyperplasia with subsequent hysterectomy without preliminary radiation are more extreme than the curettage cases therefore the simulation of cancer is more perfect. The absence of any gross lesion was incompatible with the malignancy diagnosis. There was no evidence of penetration of the basement membrane or of myometrial involvement. Cell characteristics are useful in some cases but not in others. It is impossible to decide whether or not a lesion is a cancer based on microscopic examination.

Endometrial carcinoma associated with feminizing ovarian tumors

Abstract Fifty cases of combined feminizing mesenchymomas of the ovary and carcinoma of the uterus were collected from the literature. The histories of 26 of these patients were critically examined. In a study of 66 feminizing mesenchymomas of the ovary in this laboratory, 4 cases of combined feminizing mesenchymomas and endometrial carcinoma were found. These 4 cases comprised 12 per cent of the postmenopausal group with feminizing tumors. The report of these cases brings the total number of recorded cases of combined tumors to 54. One case of combined granulosa-cell tumor and adenocarcinoma of the breast is also reported. Feminizing tumors of the ovary in combination with uterine carcinoma occur more frequently than is commonly realized. Various investigators have found that 15 to 27 per cent of postmenopausal women with feminizing tumors develop endometrial carcinoma. In the 54 cases of combined tumors, the thecoma occurred more often in combination with uterine carcinoma than did the granulosa-cell tumor, in spite of the much greater general incidence of the granulosal tumor. This suggests that the thecoma, by means of greater estrogen production, has the greater carcinogenic effect. It supports the concept that the thecal and not the granulosal cells are the sole or chief source of estrogen. The greatest carcinogenic response to tumor-produced estrogens occurs in the endometrium. Cervical and mammary carcinomas are seen only occasionally in combination with estrogen-producing tumors. The degree of carcinogenic response of these tissues to estrogens in postmenopausal life seems to parallel the degree of physiological response during menstrual life. A study of 26 case histories, still too few to be conclusive, indicates that, although feminizing tumors cause a greatly increased incidence of endometrial carcinoma, they do not seem to incite the appearance of this lesion at an age earlier than is noted in patients with endometrial carcinoma alone. Prolonged estrogen stimulation, rather than temporary intense estrogen stimulation, appears to be necessary for carcinogenesis. Several observations in these 26 cases support the concept that endometrial hyperplasia, in some predisposed postmenopausal women, occasionally is capable of transformation into carcinoma. Evidence is presented that, in the majority of cases of endometrial carcinoma, hyperestrogenism of varying degree is present. This hyperestrogenism seems to be the one added factor that sets off carcinogenesis in a postmenopausal woman already genetically predisposed to cancer. The cases of combined tumors are thought to represent the most extreme examples of this process. Therefore they offer an excellent opportunity for the study of the etiological role of estrogens in carcinoma of the endometrium. Clinical awareness, early preoperative diagnosis, and thorough endocrinological investigation of the combined tumors should throw additional light on this problem in coming years.

The management of the menopause

Abstract In spite of a healthier attitude among women in general as to the significance of the menopause, there is still a considerable substratum of misconceptions on this point, and the physician must take cognizance of this in the management of climacteric women. The majority of women at this phase need no treatment at all, many require only reassurance and education, and in only a comparatively small proportion is ovarian endocrine therapy necessary. There is perhaps no gynecologic disorder in which the indication for organotherapy is more rational than in the treatment of typical climacteric symptoms, especially the vasomotor group. There is a definite field for both the parenteral and oral routes of administration of the estrogenic hormones, the former being much more effective when the symptoms are severe. On the other hand, it must be remembered that many symptoms frequently observed in menopausal women not directly due to the endocrine readjustments of this period, but that they are more logically explained as due to environmental and psychogenic factors of one sort or another. The physician who depends upon endocrine therapy alone will fall short of the requirements in many cases, and indiscriminate estrogenic therapy should certainly be frowned upon. The question of the possible hazard of inciting malignancy in cancer-susceptible individuals cannot be decided too arbitrarily in the present state of our knowledge, though it is fair to state that no impressive evidence of such a danger has as yet been adduced, after many years of employment of the method. Certainly it would at the present time be carrying conservatism and caution to an extreme to deprive the menopausal woman of proper estrogen therapy when this is otherwise indicated, merely on the basis of this slight theoretical possibility. Stilbestrol, because of its high degree of estrogenic activity, is very effective in the control of menopausal symptoms, but its use carries with it the disadvantage of toxicity in a considerable proportion of cases, in my own experience about 20 per cent. While these are practically always mild and while they disappear with cessation of the drug, their occurrence makes it inadvisable to release the preparation for general clinical use, especially in view of the uncertainty as to the possible effects of its long-continued use. The dosage should be kept at the lowest effective level, and it is only rarely necessary to employ more than 1 mg. daily.

Brenner tumors of the ovary

Abstract This paper is based upon the study of 17 cases of Brenner tumor of the ovary, including the 14 new cases herein reported. This brings the total of reported cases to 122, though new instances are being reported more and more frequently. The tumors are benign, and produce no characteristic symptoms. When small they are, therefore, likely to be found only accidentally in operations for other indications. They may, however, reach very large size, in which case they produce discomfort or pain, with perhaps the presence of a mass noticeable to the patient herself. The pathologic characteristics have been described in the paper. The essential elements are (1) the presence of nests or columns, often partially cystic, of rather uniform size and appearance embedded in (2) a matrix of fibromatous tissue which is sharply marked off from the surrounding ovarian stroma though there is no definite capsule. The tumors probably arise from the so-called Walthard islands of indifferent cells which may at times occur in the ovary, though other explanations have been suggested. The most interesting histologic characteristic is the frequently observed transition of the cells into a cylindrical type identical with that characterizing the ordinary pseudomucinous cystadenoma, so that large tumors of the latter variety may be produced, with only small nodular Brenner tumor vestiges in the wall to indicate their origin. There is logic, therefore, in the subdivision of Brenner tumors into the solid and cystic varieties. Three such tumors are included in our series. On the other hand the fibromatous reaction may be so striking as to produce large fibromas of the ovary. In such cases, of which 2 are included in our group of cases, the origin is indicated by the finding of the typical cell nests scattered either sparsely or richly throughout the tumor. There is little or no evidence to indicate that Brenner tumors exert any such endocrine effects upon sex characters as those which characterize granulosa cell carcinoma or arrhenoblastoma.

Pelvic endometriosis: Spontaneous rupture of endometrial cysts, with a report of three cases☆

Abstract After a review of the clinical aspects and a discussion of the etiology and histogenesis of pelvic endometriosis, three cases are reported, and a fourth cited from the literature, of spontaneous perforation of unusually large endometrial cysts, with the production of acute abdominal symptoms. In two these suggested acute appendicitis, in the third a presumptive diagnosis had been made of ovarian cyst with twisted pedicle. In Lees case the symptoms had suggested “ectopic pregnancy or acute peritonitis.” The possibility of this accident should be borne in mind when acute abdominal symptoms develop in cases in which the history and the pelvic findings suggest the probability of pelvic endometriosis.

Hyperplasia of the endometrium—A clinical and pathological study

Abstract Hyperplasia of the endometrium, a condition first described by Cullen as far back as 1900, is one of the most important of all endometrial lesions. Clinically its characteristic symptom is bleeding, while pathologically it is characterized by hyperplasia of both epithelial and stromal elements in varying degree and in varying proportion. The glands are of the “swiss cheese” pattern, large dilated glands being found side by side with glands which are small and narrow. The epithelium is at times considerably thickened, while the stroma is often over-abundant and may give evidence of proliferative activity by mitoses. Grossly the endometrium may be enormously increased in amount, and may present the polypoid picture which has so often been incorrectly described as “chronic polypoid endometritis.” In almost half of our cases, however, it was of normal thickness, and in over one-half it was smooth rather than polypoid. The hyperplasia may be localized in uterine polypi, in which case, unless associated with other lesions or with strangulation of the polyp, bleeding has not been a symptom. The hyperplasia pattern is not uncommon with adenomyoma, in either the endometrium of the surface or that of the islets in the musculature. This study comprises 66 cases of hyperplasia of the uterine endometrium. In 32 cases hysterectomy was performed and in the other 34 our only pathological material consists of uterine curettings. The cases studied were taken in chronological order and selected because of a definite endometrial hyperplasia and not on the basis of the clinical symptoms or signs of abnormal uterine bleeding. About one-half of our patients with hyperplasia were above 40 years of age while the remainder were in women under the age of forty. The occurrence of hyperplasia in girls near the age of puberty is, in our experience, not common (less than 5 per cent). Our study shows no definite abnormality in the menstrual history of these patients prior to the onset of the symptoms of hyperplasia. With the onset of the symptoms of hyperplasia there was, however, in almost every case excessive menstruation, manifested usually by an increase in both the amount and duration of the menstrual discharge. Menorrhagia or metrorrhagia may occur as symptoms of hyperplasia, the former being the more common of the two. The passage of blood clots is not unusual. Amenorrhea, as part of the symptom complex of hyperplasia, was noted in about one-sixth of our cases. Pregnancy occurring after the onset of hyperplasia is uncommon in our study, but did occur in two patients. About one-half of our patients had had one or more full term pregnancies prior to the onset of the symptoms which are associated with hyperplasia. Schroder believes that an absence of corpora lutea in the ovaries is a characteristic finding with hyperplasia. We have found some exceptions to this rule, so that the matter needs further investigation. The characteristic gland pattern is attributed by some merely to epithelial hyperplasia, and by some to simple cystic distention, while to us the evidence suggests that an enormous overgrowth of the basal layer of the endometrium and an absence of the superficial functioning layers is sufficient to explain the characteristic picture. The therapeutic measures to be considered for the relief of the uterine bleeding associated with hyperplasia are curettage, organotherapy, radiotherapy and hysterectomy. There are probably a certain number of mild cases which get well spontaneously. Curettage is necessary to make a diagnosis, and may bring about relief of symptoms. Recurrence of the bleeding within a comparatively short time is more common, and repeated curettage may be necessary in the case of young patients where both radiotherapy and hysterectomy are undesirable. At the menopausal age the bleeding can be checked by sufficient dosage of either x-ray or radium. Hysterectomy is reserved for intractable cases where radium is unavailable, or where associated lesions make laparotomy preferable to simple abolition of the menstrual function. The most important clinical group of hyperplasia cases—and the largest one—is that occurring at or near the menopausal age. The importance of the condition at this age lies in the fact that it is brought into diagnostic conflict with cancer. While it is still true that climacteric bleeding should be looked upon as of cancerous causation until this assumption is disproved, the fact remains that in many cases the cause will be revealed as hyperplasia, a benign condition with no tendency toward malignancy. This is an important point to impress upon those interested in educating the laity on the early recognition of cancer, for there can be no doubt that not a few women with climacteric bleeding delay seeking medical advice through dread of being told that cancer is the cause of their symptoms. Any elecment of justifiable hopefulness that can be injected into the situation will be of material aid in the cancer education of the public. In conclusion, the authors wish to express to Dr. Thomas S. Cullen, the head of the department, their appreciation of the privilege of making this study. To Mr. Herman Schapiro they are indebted for the excellent photomicrographs.