Talk:Book - Uterine and tubal gestation (1903) 2-1

From Embryology

Part II. The Essentials Of Tubal Gestation

Chapter 1. Processes Antedating Tubal Gestation.

Etiology. — In former years our views concerning the origin of ectopic gestation depended mainly on the discovery of pathological conditions macroscopically evident. Cases were reported with fibroma of the isthmus tubEe or with polyps at the uterine end of the tube. The growth of the ovum in a tubal diverticulum or in an accessory tube was considered to furnish a satisfactory etiology. In some cases the pressure of ovarian or abdominal tumors was supposed to obstruct the onward movement of the ovum. Abel and Freund found in a twisting of the tube and in a failure of development a satisfactory theory for the frequent occurrence of ectopic gestation. Since, in a majority of such cases, peritoneal adhesions are present, these were, and even yet are, considered to so alter the course of the tube's lumen as to prevent the entrance of the ovum into the uterus. Therefore visible inflammations were considered to be the important etiological element.

During my labors a colleague gave me a specimen of extrauterine gestation combined with a multilocular serous ovarian cyst, with the request that I should find, if possible, a Graafian follicle in the ovary, though he believed none to be present. I considered that this specimen would furnish a proof of external migration of the ovum, and that this migration might stand in an etiological relation to the ectopic gestation. Examination of the ovarian cyst showed no Graafian follicle. Therefore the ovum had come by external migration from the other ovary. In Case 3 the same absence of a Graafian follicle was noted.

The experiments of Leopold have shown that the ovum given off by one ovary may enter the tube of the other side. The cases are not so rare in which the tube of one side was closed or absent, and although the corpus luteum verum was found in the ovary of the same side, yet the ovum was found in the uterus. Schroder, Koblanck, and others have found a pregnancy in a rudimentary horn between which and the uterus no epithelial


connection existed. Manierre has collected 39 cases of pregnancy in rudimentary horns. The same is true of those cases in which the corpus luteum verum is on one side and the ovum has developed in the horn of a uterus unicornis of the opposite side. Kustner removed a right-sided extrauterine gestation sac "and a left-sided ovarian cyst. Shortly after a uterine pregnancy "took place.

, ■■External migration occurs frequently in tubal gestation. 'Although Kustner took note of the frequency of this event in only the la'st 25 of a series of 100 cases, he found it to have taken place in seven. Prochownik found that external migration had taken place in one case of eight which he had examined closely. Martin found the corpus luteum on the same side as the tubal gestation in thirty-seven cases, on the opposite side in four, and uncertain in thirty-six.

External migration of the ovum has been viewed by Sippel and others as the etiological factor. They believe that the ovum in its migration becomes too large to permit of its passage through the tube lumen. The examinations of Peters, however, show conclusively that no chorionic villi are present until the ovum has been nourished for a considerable time by the decidua in which it is embedded. In addition the Graafian follicle is in the majority of instances found in the ovary of the affected side, so that such an etiology would explain only the smaller number of cases.

This migration, however, calls our attention to the presence of a pathological condition in the mucous membrane of the opposite tube. While it calls our attention to the fact that the other tube is affected, it only proves that it is more affected than the tube in which the ovum is finally embedded, for some cilia must be present in the latter to influence the external migration of the ovum. Various experiments make it seem probable that in the perfectly normal tube no ovum can develop.

Kustner, in his experiments on rabbits, extirpated one ovary and extirpated or tied off the uterine horn of the other side. In these attempts no extrauterine gestation resulted. However, since such a pathological point of development of the ovum does occur in animals, a satisfactory explanation of the failure of these attempts could not be found until the experiments of Mandl and Schmit were published. In their work upon animals they found the following to be the case : After coitus, and after the lapse of time sufficient to permit union between the ovum


and the spermatozoa at the abdominal end of the tube, they tied off the tube at the uterine end. ' Their results were negative. No tubal gestation resulted. When, however, the uterine horn was tied off a pregnancy in this horn resulted, showing that the ligation was not the disturbing factor and that in all probability ova do not develop on a normal tubal mucosa.

In considering the history of those cases which have been closely noted, it is found that ectopic gestation occurs most frequently in multipara? and that a sterile period of varying length precedes this pathological development. Martin found that 65 multipara? were affected as compared with 20 nullipara?. In a series of 100 cases of Kiistner's, only 10 ectopic gestations occurred in nullipara?. The other 87 had borne children and 3 had aborted. In 24 cases it occurred five or more years after the last labor ; in 55 cases, from one to five years after ; and in 8, in less than twelve months. Veit found that in 52 cases of repeated ectopic gestation a sterile period of two to eleven years preceded the occurrence of this process. Between the two events was a period of six weeks to six years. This sterile period represents the time in which inflammatory changes in the mucosa may occur, either gonorrheal, puerperal, or tubercular. These changes naturally involve the uterine end of the tube more than the abdominal, and in the subsequent course of events, when healing does result, the uterine end improves slowly. What are, then, the pathological changes in the tubal mucosa which stand in an etiological relation to ectopic gestation?

In a series of 8 closely examined cases Prochownik found a gonorrheal history three times; in one of these cases there was an acute gonorrheal affection of the pregnant tube. Moskowicz found that of two cases tuberculosis was the etiological factor in one, and that in the other gonococci and staphylococci were present in the pyosalpinx of the non-pregnant tube. Median to the ovum Veit found microscopical changes which represent the results of pus inflammation. In two cases Diihrssen found cilia abdominal to the ovum, but none toward the uterine end. Kiistner observed very frequently a hemorrhagic tendency of the non-affected tube, showing that tube at least to have been abnormal. I found in three cases distinct changes in the mucosa median to the ovum.

Franz makes inflammatory changes in the tubes responsible for the occurrence of ectopic gestation. This is the more probable since inflammatory processes are so frequently found in the


other tube. Franz found such changes in eighty per cent of these cases in which a sterile period of two to seventeen years was noted.. In cases where a sterile period of less than two years was observed tubal changes of the other side were present in only 53 per cent. He comes to the conclusion that we must seek the etiology in those affections of the tubes which have run their course, and which, having for a long time prevented the moving of the ovum, have permitted a gradual and partial restoration to normal conditions.

While in a certain number of cases no pathological microscopic changes are found in the tubal mucosa, it may be explained by the fact that so-called catarrhal conditions frequently show little microscopical change. Even during or after gonorrhea the tube may seem microscopically perfectly normal. Ahlfeld, in an experience of many years at the University of Marburg, met with so few cases of tubal gestation that he considers the relative freedom of his patients from gonorrhea, as compared with those in the larger cities, to be the only explanation.

Various inflammatory influences are etiological factors in that they destroy the cilia in whole or in part or diminish their functional activity. Besides, from the experiments made on animals we know that absence or early atrophy of the ovaries influences the muscular development and the functional activity of the uterine wall and the structure of the mucous membrane and the cilia. In cases of functional interference with the secretion of the ovary, or in the atrophy subsequent to labor or as the result of lactation or of constitutional disturbances or of failures of development, the activity of the cilia is diminished.

Naturally there must be activity to a certain extent on the part of the cilia, especially at the abdominal end of the tube, for extrauterine gestation occurs most frequently in the isthmus tubas. Prochownik found an ampullar location in only 3 cases out of 45. Mandl and Schmit found in 69 cases an ampullar situation in only 15. Diihrssen found an isthmic location in all of his 29 patients. In the 5 cases which I have examined closely the same is true. Although Mercier has collected 30 cases of interstitial gestation (with rupture), Leopold 10 cases of ovarian gestation ( ? ) , and although the growth of an ovum on the fimbriae, as well as tubo-ovarian pregnancies, occurs, yet the vast majority are found near the uterine end of the tube.

A further proof may be found in the fact that recurrences of tubal gestation take place but rarely in the same tube.


Patellani, in a tabulation of 36 cases, found that first one tube and then the other was the seat of development. Veit in 52 cases found that it recurred only three times in the same side. An additional point of importance is the occurrence of tubal gestation in either tube at the same time, of which Gebhardt mentions 9 cases. Further, Patellani has collected 37 instances of combined uterine and extrauterine gestation— a practical proof of an affection of one tube, and certainly excluding external migration.

I believe that in the so-called sterile period gonorrheal. puerperal, tubercular, and atrophic processes take place. The interval of years between the last labor and the ectopic gestation, the fact that the location is generally in the middle area of the tube, the fact that repeated gestations are observed and rarely in the same tube, the occurrence of an ectopic gestation on both sides at the same time, and the frequency of external migration together with a combination of extra- and intrauterine gestation, point certainly to an affection of one tube and probably, but to a lesser degree, of the other tube. The frequency with which, according to Kiistner, a hemorrhagic tendency of the nonaffected side occurs, as well as the microscopic discovery of catarrhal conditions, together with the history and the microscopical evidence of the presence of gonococci, point distinctly to a tubal affection. The observation of Duhrssen, who found cilia abdominal to the placental site and none median to it, and Veit's observation of the presence of inflammation median to the ovum, as well as the theory of congenital and acquired atrophy of the tube, especially subsequent to labor, lead us at the present day to seek in the microscopical changes of the tubal mucosa, the injury to the cilia, the etiological factor in tubal gestation.


The Decidua. — In the uterus the ovum descends by a centrifugal process into a well-developed decidua, which is gradually divided into a superficial compact layer and a deeper spongy glandular layer. Concerning the existence of a tubal decidua, however, opinions differ. Webster's view that a compacta and spongiosa are formed receives no support. Wyder found, in one case, decidua in the periphery of the region occupied by the ovum. Mandl describes decidua cells in the placental area, while Veit found decidua cells not only at the placental site but at a distance from it on the opposite wall. In his eight closely examined cases Prochownik describes decidua cells and in one a well-developed decidua. Abel found a well-marked decidua about an ovum, especially in the region of the placenta. His statement is positive, since he compared it microscopically with the decidua cast out from the uterus.

On the other hand, Aschoff found none at the placental site. Kiihne describes only a pseudo-decidua consisting of fibrin, connective tissue, and invading ectoderm cells. Though slight decidual changes may occur at the placental area, the cells which have been previously described as decidual are considered by him to be the cells of Langhans, and the same vieiv is shared by Aschoff and Ulesko-Stroganoiva. In most cases fibrin masses, blood, villi, and cells which at least look like decidua cells, but which many consider to be trophoblast cells, are to be found. It is a fact, however, that at the area of greatest pressure the socalled decidua is often thinned to a decided extent.

Veit in turn considers the so-called cells of Langhans to be, in the peripheral areas, decidual cells. Though he finds cells of the layer of Langhans in the maternal vessels, yet they are not connected with like cells about the vessels. There is an absence of like cells between the vessels and the serotinal surface. While granting that the resemblance is great, he considers those interstitial cells present in the peripheral areas about the ovum to be


decidual, and concludes that the connective tissue in the circumference of the intervillous space in tubal gestation undergoes decidual change, only that this layer is very much thinner than in the uterus.

Embedding of the Ovum, and the Reflexa or Capsularis. — A considerable difference of opinion exists as regards the reflexa. Wyder found none in his case and the ovum was only loosely attached. Abel found no capsularis and most of the villi were free and only few were adherent. "Werth described sections of a tube in which the ovum had settled at the height of a tubal fold. Werth described, in another case, a capside containing muscle fibres passing out at the base from the muscular fibres of the tube wall. The external covering of the capsule consisted of epithelium. According to Kreisch, the capside may consist of united folds. He described the presence of a reflexa in three cases, however, which was not composed of tubal folds. In a specimen examined eight days after conception a reflexa was found by him. The formation of a pseudo-reflexa readily takes place if the ovum is embedded in a thick system of folds. Others believe that in many cases a capsule consisting of longitudinal muscle fibres and mucosa, with no epithelium on the inner side, is of frequent occurrence.

Asehoff and Fiith make it evident that the ovum makes its way down into the mucosa, as is the case in the uterus, and that it extends even further deeply into the muscular wall. It is then covered laterally by the inner layers of the muscularis and by the mucosa, which are pushed over it, forming a capsularis and bounding the intervillous space. In several closely examined cases the ovum has been found almost under the entire thickness of the tubal wall, lying directly on the vessels of the ligamentum latum. Muscle fibres are absent, however, at the summit of this resulting capsule. In another case the ovum was nearly extratubal, lying between the muscularis of the tube and the muscle fibres of the ligamentum latum. The ovum was, of course, separated from the lumen by the mucosa and by muscle fibres. At the apex of this capsularis there was considerable thinning, but the epithelium upon the mucosa was intact. Such a development would explain the so-called pseudo-intraligamentous cases of Kiistner. The capsule has been found in some cases to be lined with the cells of Langhans.

Intervillous Space. — If the capsule be composed of folds, according to Kreisch, they may be so firmly united that an inter


villous space may exist. In Abel's case the space between the chorion frondosum and the decidua was larger than at any other point; adherent villi extended into the decidua, bnt most of the villi were found free in the space. The vessels of the decidua opened directly into these intervillous spaces, which contained blood. According to Aschoff, the centrifugal descent of the ovum makes the existence of an intervillous space easy. Veit, in a well-preserved non-interrupted case, found a well-developed intervillous space in which Mood circulated, and traced a vein containing villi into this common intervillous space. Ktihne, on the other hand, considers the invasion of the blood vessels by the villi to be pathological, and does not consider the existence of an intervillous space probable, because in eight cases he found no reflexa. Since Von Both collected 84 cases of ectopic gestation with viable fetus, an intervillous space must have been present.

Villi. — Abel, as stated before, finds a decidua vera in the peripheral areas of the ovum ; the villi, however, he found to extend only up to the muscle. Leopold and others, on the contrary, find the villi extending deeply into the muscular ivall of the tube. The villi are covered with two layers. At their ends are pillarlike groups of cells growing into the wall, between the muscle fibres, and representing extensions of the trophoblast. The cells of Langhans growing into the muscle and uniting with it form the so-called Saugeplatle. It is with this layer that the villi unite through the cell groups found at the ends. These complexes of cells of Langhans and syncytium speak for the fetal nature of the latter, and the pillar-like groups are simply groups of trophoblast not yet reached by the mesoderm.

Abel found most of the villi free and only some adherent to the decidua. As they did not enter into the muscle, the connection was not a close one. Wyder removed an ovulum in its entirety from a tube and later examination showed no villi extending through the decidua to the muscle. Kuhne, Ulesko-Stroganowa, and others find that the cells of Langhans extend actively into the submucous tissue and into the muscle bundles, and believe that these are the cells which are often mistaken for decidual cells. It is upon them, as upon the compacta of the uterus, that the villi grow. At the same time, they may extend entirely up to the serosa of the tube.

As just mentioned, the villi may extend up to the serosa, and perforations of the wall covered by thrombi have been frequently found. Kuhne describes cells which have entered the serotinai


veins. In the vessel spaces under the pseudo-decidua he finds free cells whose character proves them to be pseudo-decidual elements. He considers these cells to have entered through the vessel walls. Mandl found decidual cells in the vessels and believes that they have passed out of the surrounding tissue into the lumen. Other cells in the same location he views as proliferated endothelium. Cornil found syncytial elements in the vessels and considers them decidual cells which have entered through the vessel walls. Other investigators hold that the cells of Langhans grow directly into the submucous tissue and into the muscle bundles, and that these cells may penetrate the vessels and open them as in the uterine compacta. They describe the villi as following the course of these cells of Langhans into the muscularis, up to the serosa and through the serosa, which latter perforations are usually covered by thrombi. Aschoff, Ulesko-Stroganowa, Leopold, and others have described villi as perforating the vessel walls.

Deportation. — By Veit a quite different theory is given, however, to explain the presence of fetal cells and villi in the muscular wall of the tube. In discussing the relation of the ovum to the uterus, Heinz considers that the fetal tissues grow into the maternal structures and says that the maternal glands, vessel ivalls, and tissues are eaten up by the villi. Cornil considers that the ovum in an extrauterine gestation not only does the same, but follows a course resembling that of a uterine chorioma. AVhile Veit may grant the correctness of such a view in pathological conditions, he does not believe it justifiable to consider it the usual course.

In the case of a normal uterine gestation, villi do not project into the arterial openings of the intervillous space. They do, however, project normally, not only in the border sinuses but at all the serotinal points, into the veins alone. The villi are of normal structure and often so numerous that they occlude the vein lumina. They are frequently very long and follow the course of the veins to a considerable distance. In the uterus, villi are thus found in direct contact with the uterine muscle, or, better, in the veins of the uterine muscle. The same is true in tubal gestation. Whether this occurs in all cases or not cannot be positively stated (Veit).

Veit believes that cells in the periphery of the ovum enter at an early period into the peripheral veins. In normal gestation the uterus grows as much as the ovum needs, and the ovum


as much as the uterus permits. Later, as soon as the ovum in the uterus has attained a certain size and the edge of the placenta extends laterally, another process is added to that of general growth. This process is brought about by the villi which project into the serotinal veins. If a villus enters a vein it generally remains in contact with the periphery of the ovum and grows further in its length as well as in its circumference. In its growth it dilates the lumen of the vein so decidedly that its serotinal portion is everted and taken up into the general intervillous space. On either side endothelium then extends for a short distance upon the serotinal surface. In the uterus, normally, only cells of the ovum or of the villi are loosened and pass on into the veins, but the villi themselves, as a rule, remain in contact with the ovum. In uterine gestation the entire endometrium becomes deciduous, and, therefore, decidua cells are found in the periphery of these vessels which contain villi. In the peripheral region of the well-preserved tubal sac Veit found free villi in a vein. Series sections showed this vein to enter into the intervillous space. Therefore the villi were no longer connected with the ovum. This occurrence, whereby syncytium and villi, loosened from all connection with the ovum, are found in the veins, is given by Veit the name deportation.

Veit, then, explains the descriptions of Aschoff, Fiith, Cornil, and Ulesko-Stroganowa, who find a destruction of the tube wall and an invasion of vessels by the villi, on the theory of deportation through the veins.

Veit does not believe that the fetal cells possess the power of wandering, because he finds the cells of Langhans in a vessel but no such cells about it. Therefore, in his opinion, the cells have not perforated the vessel because none are present in the connective tissue. Veit, however, forgets that, in addition to deportation, a wandering of cells and a perforation of the vessels does occur.

Veit denies that the cells of the ovum wander into the connective tissue. We know that cell groups go gradually over into the cells of Langhans on the villi, while on the other side they are either separated from the decidua by fibrin, or else the cells pass gradually into the decidua with such a resemblance to the latter cells that it is difficult to differentiate them.

Veit does not believe that the cells of Langhans in the periphery of the tube look different from those which are near the villi. He found at the area of the tubal vera a few changed


connective-tissue cells and considers them to be a beginning stage of a decidua vera. In the periphery he saw cells about the vessels differing somewhat in appearance and stain from the cells of Langhans, and, since they are like the above-mentioned vera cells, he considers them decidual. In this connection he overlooks the power of the trophoblast.

Where the cell groups enter the decidua he sees a certain form of cells. At other points where, in series, sections, no villi enter the serotina he finds the same cells. Therefore either the cells of Langhans can wander in all directions or else a decidual change is present, and he believes the latter to be the case and states that a decidua exists, but much less than in the uterus.

Conclusions. — In the uterus there takes place a centrifugal descent of the ovum on a spot free of epithelium. The compact layer of the decidua forms the enveloping zone. There is an early connection between the maternal Mood and the fetal cells in the form of a trophoblast which is a product of the ectoderm. The trophoblast, at points, extends far into the compacta and the cells have a decided power of wandering. The trophoblast is changed to syncytium through the corrosive action of the blood. The opening of the maternal vessels occurs even before villi are formed, and at this and at all subsequent stages fetal cells enter the maternal circulation. The primary enveloping zone becomes the subsequent intervillous space. Neither the uterine epithelium nor the maternal endothelium play any part in the formation of the syncytium. No reflexa is formed, but through the descent of the ovum a capsularis results. The ovum, through its trophoblast cells and villi, invades and destroys the maternal decidua. The fibrin layer of Nitabuch and the thickness of the decidua prevent the extension of the villi too deeply into the wall.

In the tube we find no division into two layers, as is the normal in the uterus. In fact, no decidua is formed. There is an absence of the thin-walled decidual vessels, and the muscularis, if at all, is only slightly hypertrophied. We are only anticipating the results of our investigations when we say that, with the exception of the absence of the decidua and an enveloping zona composed of compacta, the processes of gestation in the tube are the same as those in the uterus, modified only, as we ivoidd naturally expect, by the absence of the decidua and the thinness of the tube wall.

The varying differences of opinion are due to five causes :


1. Ova of different states of preservation have given, quite naturally, different pictures.

2. Ova of very different ages have been described and compared without taking this fact into consideration.

3. As in the uterus, but even more so, the differentiation between fetal and maternal cells has been almost overlooked, so that trophoblast cells have been viewed as decidual. The ability of the fetal cells to wander, to invade the tube wall, and to perforate the vessels has been granted by only a few investigators.

4. The fact that there are several forms which the ovum follows in its embedding in the tube has almost entirely escaped attention.

5. "While Veit's deportation is an accepted fact, he leaves out of consideration the processes of "cell wandering" and "vessel invasion."

How nearly and how closely in all important details the histology of tubal gestation may resemble the course of uterine gestation, will be seen in the following pages.


Of the cases examined in series sections, these three are taken as types representing the important forms and stages.


The following specimen (Fig. 47) presents an oval body, measuring one-third of a centimetre in its greatest diameter, situated on the mucous folds of a tube removed because of the

_;._ ; . Coagulum

of ovum

Larw ■■ W7 ±— Tubal

half of — - — : • — ovum At. ".:





Fig. 47. — Very young ovum situated on the tip of the folds of the tubal mucosa in a case of ectopic gestation accompanied by an old intraperitoneal hemorrhage.

diagnosis of ectopic gestation (on operation the abdomen contained clotted blood). The tube was stained in toto, embedded in celloidin, and cut in series sections. Fifty slides from the middle area of the tube contain sections of this oval-shaped structure (Fig. 47a). A section through its greatest diameter shows it to be composed of two unequal halves : a smaller half consisting


of a blood coagulum, and a larger half composed of closely grouped cells, of protoplasmatic cells and masses containing one or more nuclei, and of villi. Villi are also found, but sparsely, about the coagulum (Fig. 48).

Sections through the smaller diameters show the coagulum diminishing in size and extent, with an actual and relative increase in the extent and area composed of closely grouped cells

i' > * TT

Fig. 47a. — Ovum of Fig. 47, showing division into two halves and an intimate connection with the tubal mucosa on the right. (Section through the greatest diameter.)

and villi, furnishing, then, a centre composed of cells and a periphery in which are villi (Fig. 49).

The more external sections are continually smaller (Fig. 50), and the central cell area diminishes in size, so that the small and most external sections present only villi (Fig. 51). In all the sections the villi are in the periphery of the centrally situated closely grouped cells.

By reconstruction, then, we obtain a solid, oval-shaped body possessing a covering composed of protoplasmatic cells and villi. Within this covering are closely grouped cells, and in the interior, although eccentrically situated, is the coagulum. We are dealing with an ovum (?) in which the amnion, umbilical vesicle, and the central cavity are obliterated by blood, while the trophoblast and villi are preserved. This ovum, as we may view it,



is adherent on one side to folds of mucosa which evidence a proliferation of their epithelial covering. On the epithelial

. ■ . >,m,



. - J

Fig. 48. — Another section through ovum of Fig. 47, showing in one half trophoblast cells, villi, and very dark syncytial masses.

covering, and between the folds, are rows of round, structureless bodies densely grouped, and the same are found at numerous points covering the ovum, except at the end opposite to the

■ %

Fig. 49. — A more peripheral section, showing trophoblast cells and villi and very little of the blood coagulum.

coagulum. In other words, on and between the mucosa folds in contact with the ovum is seen a plasmatic substance distinctly excreted by the epithelium, and in it are also red blood cells and leucocyte nuclei.


The tube wall facing the other end of the ovum presents mucosa folds, whose epithelium is interrupted at various points, while at other points are proliferations of epithelium and round cells. The stroma of many of the folds is absent, while others possess dilated capillaries and vessels, and still others are filled with blood extravasations. The tube wall itself is covered with epithelium, and in the wall are extravasations of blood, and dilated arteries and veins. At no point in the tube wall or in the mucosa is there any decidual change or any condition representing an entrance of trophoblast cells or of villi. The ovum is surrounded by a plasmatic substance which I have observed on uterine and cervical polyps covered with epithelium and in active growth. It is not a deposit from the blood, as such, but one actively secreted or produced. This same element is found on the mucosa folds in immediate contact with the ovum. The


Fig. 50. — A still more peripheral section, showing only villi.

adhesion of the ovum at these points is a firm one. It is well preserved and therefore in active contact with some nutritionproducing tissue. The shape and character of the oval body, together with its other characteristics, prove it beyond doubt to be an ovum. If not entirely surrounded by the tips ofx the mucosa folds, it was probably situated among them near the tube wall, but not on or in the wall.

The other end, with the numerous trophoblast cells, plasmodial elements, and villi, was probably surrounded, too, by mucosa folds, and here, too, an intervillous space, containing blood coming from the capillaries of the mucosa, was probably present. Although we shall see that this end was infiltrated by maternal blood, I cannot forbear, in this connection, to mention the possibility that the ovum obtained nutrition from the epithelium of the mucosa at the point of adhesion, which represents to me the reflexa or capsularis composed of tubal folds.



(In the placenta of many animals it is found that the blood extravasated from the uterine vessels is made use of by the fetus, in that the degenerating products of the extravasated blood cells are taken up and absorbed by the ectoderm cells. According to Strahl, in Galago, a Madagascar lemuride, numerous extravasations are found in the uterine mucous membrane, and their products lie in the connective tissue in the form of larger and smaller yellow granules. The epithelia of the glands near this extravasated blood are more or less filled with granules resembling these. He considers them to be rests of the blood cells which are taken up by the gland epithelium and made use of by the latter in furnishing an iron-containing secretion. The gland epithelia here perform a function which in other placenta? is performed by the fetal ectoderm. In Galago fetal villi are always


Fig. 51. — Most peripheral section, showing a villus only.

in contact with the uterine epithelium and obtain nourishment from it. Between the uterine epithelium and the epithelium of the villi are seen strands which represent the uterine milk taken up by the villi. Thus the fetus gains the greatest portion of its nourishment through the chorionic epithelium.)

The main cell body is composed of cells, without distinct cell boundaries, containing nuclei of varying form and staining darkly. At various points are found paler cells with a distinct nuclear membrane which unite in groups of three or more. More external, in the region of the villi, the cells are more distinct and the nuclear membrane is more clear and nucleoli can be distinctly seen (Fig. 52). The more external these cells are found the larger and the more swollen and the paler do they become, and a homogeneous intercellular substance is present. In the centre of those sections where the cells are so densely grouped (Fig. 49) are minute, darkly-staining granules whose identity at first was not distinct. In the more peripheral areas, however, their char


acter is evident ; for they are only found where red blood cells are present and they are therefore leucocyte nuclei. The nearer we approach the circumference the more do we find blood in the interstices between the cells, at first as isolated cells or groups, and later in larger spaces containing red blood cells. Wherever blood in small amount is present, the cell nuclei become dark and change their form, becoming long spindle-shaped, arranged in single strands or in longer parallel groups. "Where, however, blood in larger amounts is present, protoplasmatic masses containing numerous round or spindle nuclei may be observed. Near the

Youno <£&* villus *■'

Fig. 52. — An area from Fig. 48, showing trophoblast cells, their transition to the chorionic epithelium in the upper part, and to syncytial groups in the lower part.

circumference the number and extent of the round protoplasmatic groups and the number of the darker or longer spindleshaped nuclei increase, and the latter seem to invade the sections described from the exterior toward the interior. In the exterior of the sections are villi covered with a double layer of thin syncytium with flat nuclei. The centre is composed of embryonal connective tissue staining blue and containing branching cells (Fig. 52). Other villi, some of which are quite long, have a centre composed of round nuclei with nucleoli which are very closely grouped, and transitions from these cells to syncytium may be distinctly observed, as in Fig. 52. These villi, as well as the others, are found only in the external area or near the coagulum. Knobs of syncytium and so-called syncytial giant cells are pres



ent about these villi, connected by a pedicle or lying apart. In the central areas no connective tissue can be observed. In the external layers, however, as said before, it is, evident in the form of a structureless substance between the cells, giving these areas a pale appearance (Fig. 52). It is not present in the younger villi and is only evident in the fully formed ones.

Blood and fibrin at area of serotina

Dilated vessels of the tube wall


Tubal lining covered with fibrin

Fig. 53. — Section of tube, showing normal mucosa in the lower and left area, while above and to the right is the area of attachment of the ovum, giving a typical picture of the tube after tubal abortion.


A further stage of development of a tubal ovum is well represented in Fig. 53. The examination, in series sections, of this specimen, known clinically as tubal abortion, divulges numerous interesting features. One half of the circumference of the tube lumen is of a normal character; the other half is in a torn, infiltrated condition which involves not only the mucosa folds but also the submucosa up to the muscularis. Close examination evidences the fact that the ovum was situated on the tube wall, compressing and destroying the mucosa folds at the situation known as the serotina. On either side of this point the mucosa folds were evidently united about the ovum, forming a pseudoreflexa. The capillaries and large vessels at the serotinal point are filled with blood (Fig. 53), and many evidence larger or

Muscular fibres of tube wall



'Cell group." "* Jw&$

Trophoblast cells


Blood clot and fibrin

Fig. 54. — Portion of serotinal area of Pig. 53, showing pale trophoblast cells and cell groups and the darker syncytial cells.

smaller injuries and invasions of the wall. At the serotinal portion and for some distance on either side is observed in the submucosa a tissue resembling at first sight decidua.

On close examination we find oval, very closely grouped cells,

Fig. 54a. — Typical area in Pig. 54, commonly mistaken for decidua, showing pale trophoblast cells and dark syncytial cells.

consisting mainly of nuclei and possessing each a nucleolus (Fig. 54). They are extremely closely grouped and between them are found no capillaries or spaces. This tissue rests not only on the free surface of the tube wall, but invades it at many points to a considerable depth (Fig. 55). The invasion of the tube wall is not an even one, for at some points it forms a continuous deep layer like the decidua, at other points it enters the submucous tissue in irregular branches and projections, and at other points, through lateral infiltration, is separated from the free surface of




■ Muscularis



Dense group of trophoblast veils, representing a "cell group." as found at the tip of the villi

Fig. 55. — Section through tube wall at serotinal area, showing the paler character of the advanced trophoblast cells and the darker character of the densely grouped trophoblast cells near the ovum. Typical field formerly mistaken for decidua, but differentiated by the dark spindle-shaped syncytial cells. This drawing shows that in tubal abortion the "rests" are of the same nature as in uterine abortion.

the tube wall by a considerable space of normal submucous tissue.

These cells are found about the smaller and the larger vessels, invading and infiltrating their muscular walls up to and into the lumen of the vessel.

The character, arrangement, and structure of these cell masses 9


leave no doubt that they are trophoblast cells. In addition they may be identified by the fact that at almost all points, especially those points near the inner surface of the tube, they are accompanied by isolated, long, spindle-shaped protoplasmatic cells, by groups of polynuclear protoplasmatic masses, and by long, often parallel bands of the same character (Figs. 54a and 55a). These are the syncytial cells, absolutely identical with the same elements observed in all the previous specimens, especially those seen in the first tubal ovum.

These cells may be further identified as trophoblast cells by the fact that they are found, at the tip of some adherent villi,






Fig. 55a. — Another area from Pig. 55, showing the "advance guard" of infiltrating trophoblast and syncytial cells.

extending directly into the tube wall. These trophoblast cells enter into the submucosa and museularis, forming the tissue which so many investigators have called decidua. These complexes of trophoblast and syncytial cells, especially the groups found at the tips of the villi, are not yet filled out with mesoderm, but are to form future villi.

At no point do the connective-tissue cells of the tubal folds or of the submucosa evidence any change resembling, in the slightest degree, those changes occurring in the uterine mucosa which result in the formation of decidua cells.

At no point do the epithelial cells of the mucosa evidence any change of a so-called syncytial character.




Figure 56 represents a tubal gestation, containing a fetus, in the sixth or seventh week. The tube wall is preserved in its entire circumference, with the exception of two minute areas where the chorionic villi have actively perforated it. Projecting into the dilated lumen of the tube, and almost touching the mucous lining, is the outer covering of the fetal sac, the capsularis. One half of the circumference of the tube is invaded, infiltrated, and

Beginning of reflexa or capsularis

Perforation ■J covered by thrombus


Tube loall

Fig. 56. — Tubal gestation with fetus attached by umbilical cord, showing perforation by the growing villi closed by a thrombus.

stretched by the other half of the fetal sac wall, that is, the placenta. On both sides, at the junction between the infiltrated tube wall and the preserved but decidedly stretched tube wall which surrounds that half of the fetal sac which projects into the tube lumen, the mucosa of the latter passes over upon the fetal sae and can be followed upon it as a covering of the latter for a certain distance. The summit of the sac is covered by a tissue composed of fibrin, trophoblast cells, leucocytes, etc. The ovum possesses, then, a capsularis composed in part of fetal folds and submucosa. This capsularis is, strictly speaking, no reflexa. It must be called a pseudo-reflexa or capsularis.

The area of the tube circumference entirely filled by villi is.



the true placental site, and here is found a real intervillous space bounded on all sides by the point of union of the tube wall with the base of the capsularis. The remainder of the circumference of the fetal sac evidences villi, but the space between the capsularis and the inner lining of the sac is mainly filled with blood and blood crystals. At the real placental site almost no remnants of the tube wall can be found. At this half of the tube wall there are nothing but villi, trophoblast cells, syncytial cells, and a sea of maternal blood (Fig. 56a-).

The process involved in this invasion of the tube wall and the

Tube wall

Tube lumen

Tube ivall going over into the capsularis (reflexa)

Eroded tube ivall

Villi perforating tube wall

Pig. 56a. — Tubal ovum with perfectly developed capsularis. Tubal wall passes on both sides over into the capsularis. Major portion of placenta is at the lower border, where tubal tissue is eroded and perforated by the villi. This area is the intervillous space. Between the tube lumen and the central cavity of the ovum is the capsularis filled with blood and blood crystals.

process involved in the formation of the villi can, in each section, be extremely well judged on either side at the point of junction of the capsularis with the tube wall (Fig. 57). Here, on either side, there extend into the gradually thinning tissue of the tube villi, cell groups, and syncytial cells which invade the muscularis in what may be called a concentric path. We observe, at the tips of fully formed villi, the typical trophoblast cell groups. We see them extending into and between the muscular fibres, changing into and accompanied by syncytial



cells and masses of every kind and form yet described (Fig. 58). At no point is there the slightest evidence of any decidua or of any decidual reaction. The trophoblast cells invade the vessels of the tube wall, and the villi themselves enter the tube wall and, with the trophoblast groups at their tips, enter the vessel lumina.

At the main placental site are found trophoblast cells with the clearest and most distinct change into various forms of independent syncytial groups and into the syncytial covering of the

Tubal tissue

passing over into capsularis

Tubal tissue

Tube wall

Fig. 57. — Area XX of Fig. 56a magnified, showing the reflexa or capsularis and the infiltration and invasion of the tube wall by the trophoblast and villi.

villi (Fig. 58). They are likewise seen to pass over gradually into the layer of Langhans, which at all points resembles, the syncytial layer to such an extent that it might be said that the covering of the villi consists of a double layer of syncytial cells (Fig. 59). The villi have actively perforated the tube wall individually and in groups, so that the intervillous space communicated with the abdominal cavity at these points.

This stage is the exact macroscopic and microscopic counterpart of the stage seen in Fig. 39.



If no interruption take place, the capsularis unites with the mucosa of the enveloping tube wall in the same way that this processes exemplified in the uterus.

The various steps in the succeeding months of development of a tubal gestation may, so far as the development of the placenta, the fetal sac, and the fetus is concerned, be considered as identical with the same processes in the uterus; for over ninety cases of full-term ectopic gestation with viable fetus are recorded.

The various steps in the development of the tubal placenta,

Blood -form- if 't ., , J? . §

ing new t-^ti aj f ■

syncytium Syncytium TSyffij



Cell group

Stroma of villus

Syncytial mass


with nucleated




of villus

Fig. 58. — Area of advancing trophoblast cells of Fig. 57, showing transition of trophoblast cells to syncytial groups and to the syncytial covering of the villi.

depending as they do mainly on the cells of the ovum itself, the trophoblast cells, are the same as in the uterus. The difference is mainly in the different character of the base or trophospongia. We may draw the following conclusions :

(1) In tubal gestation no decidua or trophospongia develops. The mucous lining of the uterus is really a lymph tissue; the submucosa of the tube is not.

(2) Fetal cells, in tubal gestation, may at any time enter the maternal circulation.

(3) As regards the troplioblast, the syncytium, the villi, the formation of blood by the trophoblast cells, and all other par


ticulars, the processes depending on and originating in the ovum are the same in both uterus and tube.

Certain investigators have described ova embedded in the tube wall, where a capsularis, was present containing muscle fibres passing out from the tube wall into the base of the capsularis. The ovum was found entirely under the entire thickness of the tube wall, lying on the vessels of the ligamentum latum— the socalled pseudo-intraligamentous form.


In the tube, the embedding of the ovum and the development of the placenta, then, is found to follow three fairly distinct forms: (1) the columnar, (2) the inter columnar, (3) the centrifugal.

Syncytium and cells of ,, „

Lunghans .jgjgj *

" *****»« 

Syncytium **> *&

Trophoblast cells




Fig. 59. — Drawing from serotinal area of Fig. 56a, showing trophoblast cells going over into isolated syncytial cells and into syncytial masses.

1. In the columnar type of development (Fig. 47) the ovum is surrounded by mucosa folds only. Here an invasion of the capillaries of the tubal mucosa occurs. Such a columnar situation makes abortion easy and of little danger. Very soon after the entrance of the ovum tubal bleeding may result; the ovum dies and further hemorrhage expels it. The tube may return to a normal state without any evidence of the previous condition, or else a hematosalpinx may be formed if the abdominal end of the tube is closed. The ovum may, theoretically, develop to a much further degree and press the folds against the tubal wall. If development continues the villi may extend into it, and the connection of the ovum and the villi with the surrounding tissue is a loose one, as in the case of Abel and of Wyder and in the specimen Fig. 47.


2. In the intercolumnar form (Fig. 53) the ovum may rest on the wall of the tube. Any tubal fold beneath it will be compressed, but epithelium may be present in a depression. Other folds may form a capsularis, which consists then of mucosa alone ; if such a capsularis be firm an intervillous space may develop. The villi at the placental site enter into the wall ; here a hemorrhage may result through this invasion of the wall and of the vessels and through an invasion of the capsularis by fetal cells ; or, since the capsularis does not undergo decidual change and is therefore less yielding, the capsularis may rupture. If it be torn, or if it be not closely adherent, the intervillous space is opened. Abortion, complete or incomplete — usually incomplete —is the general rule, but rupture might occur. If the abdominal end be closed a hematosalpinx or a tubal mole may represent the final outcome.

3. In the centrifugal form (Fig. 56a) the ovum sinks into the wall of the tube and an invasion of the wall and vessels by the villi may take place even up to the serosa. The capsularis is formed by muscularis and mucosa. It may rupture at its summit. The invasion of the vessels entering the intervillous space may cause hemorrhage. The villi which extend up to the serosa may cause bleeding, though their penetration is so gradual that these points are usually covered with thrombi. Finally a rupture may take place at the placental site through multiple perforations producing an arrosion. The ovum practically eats up the wall. Even though the tubal diameter be large enough to give sufficient room, this occurs. It is not the result of pressure, as may be seen in gestation at the fimbrian end, where rupture also can result. Villi which perforate the serosa may cause a very decided hemorrhage into the peritoneal cavity. "When no rupture has occurred and the abdominal end of the tube is closed, only the microscope may divulge the source of such an intraperitoneal bleeding. Such minute perforations, may cause collapse through hemorrhage, even though the opening be no larger than the head of a pin. Even after the death of the ovum the villi can grow, and an active tubal mole is found with continued bleeding. If they do not grow, hemorrhage continues, since no contraction can take place, as is the case in the uterus. The centrifugal form furnishes the majority of tubal ruptures. But the vast majority of these so-called tubal ruptures are either erosions or due to erosion by the perforating villi.


The theory that the tube ruptures because the ovum is too big is, as a rule, wrong for cases in the first three months. The various interruptions of ectopic gestation are all the result of hemorrhages primarily minute. The usual ending, clinically, of the gestation begins with bleeding in the tube. The invasion of the vessels of the mucosa and the tube wall and the invasion of the serosa furnish the causes for hemorrhage. The death of the fetus, as in the case of the uterus, brings about changes which result in bleeding. The primary cause is a lack of decidua. In a mucosa previously affected, when many large vessels are changed by the fetal cells and invaded by villi, an increase in tension through contraction of the tube walls furnishes an easy explanation of this hemorrhage. In the uterus the vessels are firmly embedded in the thick decidua and take a twisted course ; in the tube the vessels are straight and embedded in loose connective or fetal tissue. Bleeding on the part of the capsule is possible and of frequent occurrence, since it does not undergo decidual change and may be invaded by fetal cells. The contraction of the muscle fibres on either side of the capsularis renders the rupture of this pseudo-reflexa easy because of the absence of decidual changes, and the point of rupture is usually at the summit of the capsularis. If the capsularis be composed of tubal folds the intervillous space is easily involved. If the capsularis be composed of muscularis and mucosa a decided bleeding may result if only the summit of the capsularis be torn.

Rupture of the tube almost always takes place at the placental site, which is the seat of old and new hemorrhages. The hemorrhage and loosening of the ovum which represent the clinical ending of these cases is not the first bleeding, for older ones are usually present. The various processes depend upon the ovum, the condition of the tube before pregnancy, the character of the union of the ovum with the tube, the place of union, and trauma. The reaction of the tube is limited to the area of the ovum ; and in this we find the main difference between tubal and uterine


gestation. The uterus undergoes early independent growth, the tube does not. With the development of the ovum the uterus grows hand in hand, while in the tube the ovum makes room for itself and obtains its nourishment by the invasion of the tube walls. It may stretch the circumference of the tube so that its wall, as in the case of Abel, may be reduced to a layer of connective tissue so thin that rupture may result at any point.

Ampullar cases usually end -in abortion, generally with hematocele. There is no obstruction, unless decided adhesions are present, and the blood is generally poured out quickly into the pelvic peritoneum or into the sac of Douglas. Such an abortion may be complete or incomplete. Rupture in this situation occurs, but very rarely. The majority of tubal gestations are situated in the isthmus tuba? nearer the uterine end. In those cases we have (1) abortion without rupture, complete or incomplete, with bleeding from the abdominal end of the tube. Generally a hematocele is found at the abdominal end. The tubes are often so curved that it is difficult for the blood to make its way to the fimbriae, and the oozing is of a slow character. The blood extends rarely more than a very short distance toward the uterine end, because of the numerous short curves present here. ( 2 ) We may have single or multiple microscopic perforations of the tube wall by villi, causing even decided hemorrhage without apparent cause. (3) We may have macroscopic perforations or "erosions" of the tube wall, covered or not covered by thrombi, and causing great hemorrhage. (4) We may have abortion with rupture either into the free abdominal cavity with no hematocele at the abdominal end of the tube, or with partial encapsulation, in which event there may be hematocele at the abdominal end if the tube is open. (5) We may have an intraligamentous tear with hematocele at the abdominal end. In these latter cases the placental site is always on the inferior surface of the tube and the ovum has descended centrifugally to the vessels of the ligamentum latum. These are by far the most difficult cases surgically, and may require hysterectomy to remove the mass in toto.

Case I. (Fig. 47). — Twenty-eight years of age, married thirteen years. One labor eleven years ago. Divorced ten years ago, and operated vaginally at Mt. Sinai for pelvic abscess, since which time periodical attacks of pain on the right side every three months. Married again twenty months ago. One year ago, pain, fainting spells. Three months before admission, pain. Two months before, only stains instead of menstruation. One week


before, severe cramps, fainting spells, following a metrorrhagia of several weeks. Operation showed plenty of old blood in the pelvic cavity. Tube distended, clot at fimbrian end. Other tube normal. Microscope, columnar type, with evidences of a previous old ectopic gestation.

Case II. (Fig. 53). — Twenty-seven years old; four children, last nine months ago. Nursing. Four weeks of abdominal cramps and uterine hemorrhage. Well for one week and then symptoms returned up to admission, on which day she fainted three times on account of pain.

Operation.— Much blood; right tube distended and clot at fimbrian end. Ovarian cyst. Incomplete tubal abortion. Intercolumnar type, villi and cells of Langhans and syncytium in the wall. Keflexa consisted probably of folds involving one half the circumference of the tube. The other half of the lumen normal.

Case III. (Fig. 56). — Old blood clots in the abdominal cavity showed the cause of the increasing attacks of pain with intervals of relief to be due to bleedings from perforations of the tube wall which were closed at times by the formation of thrombi.

Case IV.— Patient 32 years old; two children, last three years ago. Skipped two menstrual periods. Two weeks later examined because of abdominal pain. Examination showed leftsided tumor, elastic, with pulsating vessels on the left side of the vagina. Uterus enlarged. Next day signs of hemorrhage, pale, pulse 120.

Operation showed much blood, dark clots, and bleeding from the tubal end. Incomplete tubal abortion, placental mole, slight involvement of the wall. No Graafian follicle in the ovary.

Case V. — Twenty-five years old; three children, last fifteen months ago. Three months ago first menstruation (probably nursing). Two months ago skipped menstrual period. For four weeks metrorrhagia and cramps lasting fifteen to twenty minutes and recurring as often as three to four times a day.

Operation.— Vaginal celiotomy. Little free blood. Microscope. Ovum, disintegrated by blood and surrounded by villi and tubal folds, found in the tube. Folds also present in the wall at many points. Organized clot adherent at one point. Intercolumnar and partly centrifugal, for muscularis and mucosa extend on either side partly over the mole. Organized tubal mole. No fresh bleeding.

Case VI. — Thirty-seven years old; married eighteen years; six


children, last three years ago. Last menstruation four months ago. Operated because of abdominal pains. No blood in the peritoneum. Tube and ovary free. Continued bleeding into the tube, which was full of fresh blood. Probably would have broken through the closed abdominal end. Incomplete tubal abortion of the type of hematosalpinx. Placental mole. Decided involvement of tubal wall. Large vessel in the periphery going over gradually into the intervillous space.

These cases, point to the decided danger from continued bleedings involved in tubal abortion. The general view is that tubal rupture gives much more pronounced symptoms and a much more decided hemorrhage than tubal abortion. When we consider that incomplete abortion means that villi are left in the tubal wall and that so-called complete tubal abortion means the retention of trophoblast cells, we may readily understand that bleeding may continue for an indefinitely long period. It is a fact, however, that even complete abortion may cause decided symptoms. Mandl reports two cases from the clinic of Schauta, accompanied by pronounced collapse and decided hemorrhage. In the first case no villi were found in the tube wall (see Fig. 53). In the second case, although villi were found in the blood clot in the tube, none were found in the tube wall. Like cases of tubal abortion, with symptoms as severe as are frequently the rule with tubal rupture, have been reported by Klein, Zedel, Piering, and others. It seems to me these histological and clinical evidences are of sufficient weight to destroy the view, prevailing in many minds, that tubal rupture should be treated by extirpation of the tube and that tubal abortion demands only conservative treatment. The proportion of tubal abortion to tubal rupture is probably 8 or 10 to 1. In this connection it is quite sufficient to mention the dangers arising from hematocele. The injury to the peritoneum, the adhesions which take place, and above all the by no means infrequent occurrence of subsequent purulent degeneration of such an accumulation of blood, are only some of the injurious results avoided by prompt removal.

The possibilities are represented by the processes of abortion, microscopic perforation, macroscopic perforation, rupture, hematosalpinx, and tubal mole. In 99 cases of interrupted tubal gestation in the clinic of Schauta, a hematocele was found 60 times — 55 after abortion, 5 times after rupture. If the bleedingbe very slow, the blood forms a capsule (due to peritoneal adhesions) into which the subsequent hemorrhages enter, the so


called secondary hematocele. If adhesions are present at the abdominal end of the tube they may form a portion of the capsule. The resulting hematocele after rapid bleeding furnishes the primary or diffuse form. The secondary hematocele occurs much more frequently than the primary. In the 60 hematoceles found among 99 cases in the clinic of Schauta, only 4 were diffuse. Of the abortions found in the same clinic, 75 were incomplete and 6 were complete.