Book - Contributions to Embryology Carnegie Institution No.56-11

From Embryology

Chapter 11. Ovarian Pregnancy

Embryology - 19 Mar 2024    Facebook link Pinterest link Twitter link  Expand to Translate  
Google Translate - select your language from the list shown below (this will open a new external page)

العربية | català | 中文 | 中國傳統的 | français | Deutsche | עִברִית | हिंदी | bahasa Indonesia | italiano | 日本語 | 한국어 | မြန်မာ | Pilipino | Polskie | português | ਪੰਜਾਬੀ ਦੇ | Română | русский | Español | Swahili | Svensk | ไทย | Türkçe | اردو | ייִדיש | Tiếng Việt    These external translations are automated and may not be accurate. (More? About Translations)

Mall FP. and Meyer AW. Studies on abortuses: a survey of pathologic ova in the Carnegie Embryological Collection. (1921) Contrib. Embryol., Carnegie Inst. Wash. Publ. 275, 12: 1-364.

In this historic 1921 pathology paper, figures and plates of abnormal embryos are not suitable for young students.

1921 Carnegie Collection - Abnormal: Preface | 1 Collection origin | 2 Care and utilization | 3 Classification | 4 Pathologic analysis | 5 Size | 6 Sex incidence | 7 Localized anomalies | 8 Hydatiform uterine | 9 Hydatiform tubal | Chapter 10 Alleged superfetation | 11 Ovarian Pregnancy | 12 Lysis and resorption | 13 Postmortem intrauterine | 14 Hofbauer cells | 15 Villi | 16 Villous nodules | 17 Syphilitic changes | 18 Aspects | Bibliography | Figures | Contribution No.56 | Contributions Series | Embryology History

Historic Disclaimer - information about historic embryology pages 
Mark Hill.jpg
Pages where the terms "Historic" (textbooks, papers, people, recommendations) appear on this site, and sections within pages where this disclaimer appears, indicate that the content and scientific understanding are specific to the time of publication. This means that while some scientific descriptions are still accurate, the terminology and interpretation of the developmental mechanisms reflect the understanding at the time of original publication and those of the preceding periods, these terms, interpretations and recommendations may not reflect our current scientific understanding.     (More? Embryology History | Historic Embryology Papers)

Although the Carnegie Collection contains almost 3,000 specimens of abortuses and others from operations upon the tubes and uteri, it includes only 2 cases of ovarian pregnancy. The first of these (No. 550) was described by Mall and Cullen (1913) and the second (No. 1522) by Meyer and Wynne (1919). This is an incidence of only 1 in 1,500 miscellaneous accessions composed mainly of abortuses, but it is not at all unlikely that the near future will experience an increased frequency, if not among the accessions to the Carnegie Collection, at least in the cases reported. For, although the first case of ovarian pregnancy under that heading in the Index Medicus is that of Kouwer (1897 [van Tussenbroek, 1899]), careful scrutiny of the titles listed for the last decade reveals the fact that 5 cases of ovarian pregnancy were reported in 1908, 4 each in 1909 and 1910, 7 in 1911, 13 in 1912, 9 in 1913, 7 in 1914, 3 in 1915, 1 in 1916, and 5 in 1917. This makes a total of 58 apparent cases reported within this decade. Since the reports on some of the cases were published in three different journals, these were, of course, counted merely as one, and although the authenticity of 4 of the cases must be questioned on the basis of the titles alone, the series, nevertheless, is a large one in spite of these facts and of a marked decline in the number reported during the war. Since Norris (1909) stated that only 19 certain cases, approximately only one-third as many as all cases listed in the last decade, were reported in the decade between 1899 and 1909, it would seem that ovarian pregnancy is not only receiving increasing attention, but that a change in attitude as to what constitutes ovarian pregnancy is probably in progress. This conclusion would seem to be justified, even though a careful examination of the descriptions of the cases reported in the decade between 1908 and 1917 would reduce somewhat the number listed.


Lockyer (1917) accepted as authentic only 22 cases of those reported between 1910 and 1917, but his review is only a partial one. Even so, it shows that there is a decided increase in the number of cases which have been regarded as genuine from decade to decade. The marked increase in the number of genuine cases reported in recent decades becomes still more evident if one recalls that Williams (1910) found only 13 positive cases up to 1906, whereas Norris found 19 positive cases in the single decade between 1899 and 1909. That is, Norris found more positive cases reported in that decade than had been reported in all previous medical history up to 1906. This surely is striking.


The opinion that many, even if not all, cases of so-called hematocele, hematoma, apoplexy, blood-cysts, and rupture of the ovaries are probably nothing but cases of ovarian pregnancy in disguise has been held by various investigators for some time. Hence, if hematocele of the ovaries repeats the history of hematosalpinx, it is not unlikely that the near future will see a marked increase in the reported frequency of "a fact so curious and important in itself," as Granville aptly put it a century ago. This would seem to be true in spite of the fact reported by Norris and Mitchell (1908) that only a single case of ovarian pregnancy was found among 44 extrauterine specimens and 58 hemorrhagic cysts contained in the collection of 1,700 gynecological specimens at the hospital of the University of Pennsylvania. At any rate, a careful microscopic examination of all such cases would seem to be indicated in the future in order to determine, if possible, which cases are and which are not conceptual in origin.


Werth (1887) is said to have collected 12 cases, among which he regarded only that of Leopold (1882) as authentic. Leopold (1899) reported 14. Gilford (1901), in a splendid, succinct review of the literature, gave 28 cases, 16 of which he regarded as undoubted and 12 as probable. Roche (1902) accepted only 12 cases, Flith (1902) accepted 21. Kantorowicz (1904), using the criteria of Leopold (1899), together with a microscopic examination, as a basis, grouped the cases recorded in the literature as certain, probable, and uncertain. . He considered 17 as certain, 10 as probable, and 13 as uncertain. And to the 17 cases regarded as certain by him, Kantorowicz then added 2 of his own, basing his decision, however, mainly upon the presence of decidua in the ovaries, thus making 19 cases regarded as authentic by him. Freund and Thome (1906) regarded 23 of all the cases reported up to that time as certain. Norris and Mitchell (1908) considered 16 as positive, 15 as probable, and 9 as fairly probable. Warbanoff (1909) collected 34 cases and Norris (1909) regarded 19 of those contained in the literature of the previous decade as positive; but Williams (1910), from a critical review of the literature up to 1906, and upon the basis of the criteria of Spiegelberg, regarded only 13 as positive, 17 as highly probable, and 5 as probable. Mapes (1914) collected 30 cases, but wholly from secondary sources, and Lockyer (1917) 42, from the years 1910 to 1917. Of these cases Lockyer accepted 22 as authentic and 20 as questionable and undecided from the evidence available to him.


This short summary suffices to show that there is as yet no consensus of opinion as to what constitutes an ovarian pregnancy. Although this fact finds its explanation partly in our lack of sufficient knowledge, it is due also to the meagerness of some of the reports. However, if complete disintegration and lysis of intraovarian conceptuses can occur, then it must always remain a question of opinion in the future whether some of the cases so reported really were or were not true ovarian pregnancies. This must remain true, no matter how thorough the microscopic examination, unless the clinical history or changes in the maternal organism can afford us crucial tests in such cases.


Anyone who reads far into the literature of ovarian pregnancy must also become aware of the fact that even very recently skepticism has been carried too far. Jacobson (1908), for example, placed the case of Kouwer-van Tussenbroek (a case which finally convinced Bland-Sutton) and that of Webster (1904) in the doubtful class! Furthermore, he also insisted upon the presence of an embryo or fetus as absolutely essential.


It must be emphasized, however, that even a liberal attitude on the part of a reviewer would not justify him in accepting all cases reported as genuine upon the basis of the reports themselves, for they especially the older ones are often too meager to enable one to form a reliable opinion. This is illustrated also by such recent reports as those of Garrard (1916), Martin (1917), Sweeney (1917), and Mills (1917). Although it must be remembered that from the very nature of things it is sometimes impossible to make a report which in itself carries conviction, it is regrettable that in a number of relatively recent cases, in which such a report apparently could have been made, this was not done. Mills 's case seems to have been an instance of ovarian implantation in a region other than the Graafian follicle, and hence recalls the first case of Granville and the cases of Franz (1902), Norris (1909), Paucot and Debeyre (1913), and perhaps also that of Kouwer (1897 [van Tussenbroek, 1899]).


Today it is no longer true, as stated by Freund and Thome (1906) and by Sencert and Aron as late as 1914, that authentic cases of ovarian pregnancy belong to the great rarities. Yet the fact that many of our States, as well as many large clinics, have not a single case on their records seems to suggest that the condition is still seldom recognized, a century after Granville observed his first case. Moreover, a number of continental gynecologists and obstetricians, for a quarter of a century, have regarded the occurrence of ovarian pregnancy as undoubted. Anderson (1917) stated that German writers began to report cases of ovarian pregnancy with some frequency after 1901, and Gilford (1901) also called attention to the fact that continental opinion had long accepted ovarian implantation not only as possible, but as proved. Gilford further referred to the oftenquoted opinion of Tait that ovarian pregnancy is as rare as "a blue lion or a swan with two necks," and in his article in 1899 also called attention to the opinion of Bland-Sutton that ovarian pregnancy not only has no existence, but that it is impossible. These opinions are particularly interesting in view of the careful reports made by Granville (1820 and 1834) in connection with the two cases which he then and which others since have regarded as cases of undoubted ovarian pregnancy, in spite of the absence of microscopic examination. In view of this lack, it is particularly fortunate that both of these reports of Granville are accompanied by splendid illustrations by Bauer, which also won him praise and admiration and which greatly strengthened his cases. It may be recalled in this connection that Werth (1901) accepted Granville's case recorded in 1820, but said nothing about his second more convincing instance reported in 1834.


Although there is as yet no agreement as to what constitutes an authentic case, a review of the literature justifies the growing and apparently well-founded belief that in the past too much emphasis has been laid on certain criteria which later experience has shown to be partly inapplicable. It is becoming clear that some cases, formerly excluded for reasons regarded as sufficient, with our present knowledge could no longer be rejected. Moreover, it dotes not seem at all improbable that some cases listed as tubal really were ovarian in origin. Nor must it be forgotten that not even the entire absence of remnants of the conceptus can positively exclude a case from the category of true ovarian pregnancy. In a number of cases in the literature, and also in our own cases, the clinical history and gross anatomic findings suggest the conclusion drawn by Scott (1901), on a priori grounds alone, that the conceptus may be completely resorbed. It may, of course, also be aborted or disintegrate completely. That such an assumption is justified is indicated by the lysis of the embryo or fetus in a large number of cases of ovarian pregnancy, and also by the very degenerated condition of some of the vesicles and of the surrounding ovarian stroma. The possibility of such an occurrence in the ovary is established also by similar events in uterine and tubal pregnancies discussed elsewhere (Meyer, 1919 e ). It probably is illustrated also by such cases as those of Anning and Littlewood (1901), in which no mention is made of an embryonic disk in a translucent conceptus the size of a "pea." Then, to be sure, there are the cases of unruptured ovarian pregnancies containing villi only, as well as the rare case, probably of double ovarian pregnancy, of Holland (1911). Although one can not be certain that embryonic tissue was removed from the left ovary with the blood-clot which was forcibly expelled at the time of operation in this case, it is not at all improbable that the small plasmodial masses found in the left ovary were the only remnants of the conceptus. We realize fully that the conclusion that young conceptuses may be wholly dissolved is fraught with great uncertainty, but it nevertheless appears to be justified by the facts, and that it therefore is in the direction of truth. It could only fail to be so if every ovum that becomes implanted within the ovaries were aborted or were removed by operation before lysis was possible.


One can not rightly refuse to recognize the possibility of the spontaneous disappearance of an ovarian pregnancy. Since implantation in the ovary occurs under such abnormal conditions, it would seem that for this reason alone the great majority of such implantations inevitably must succumb. This would seem probable, wholly aside from considerations regarding the development of the corpus luteum, although lack of, or interference with, the development of the latter also would seem to condition early death of the conceptus if the results of the long series of experiments on rabbits by L. Fraenkel (1903, 1910 b ) are indicative of the role played by the corpus luteum in early implantations in man also. It surely is difficult, if not impossible, to see how implantation within the Graafian follicle, and especially the later development of the conceptus, can fail to interfere with the development of a normal corpus luteum. Cases in the literature, and also our case No. 1522, did not reveal the presence of any well-preserved or even true luteal cells at the time the pregnancy was terminated. Although this fact does not presuppose an entire lack of development of these cells in the earliest stages of the implantation, it undoubtedly does imply a defective development, which in itself may have become responsible for the death of the conceptus. Nor should the possible toxic effect of luteal cells upon the conceptus be forgotten in this connection.


It is not assumed that the clinical symptoms and signs alone should suffice finally to group a specimen as truly ovarian, but when these are indicative of the presence of an ectopic gestation, and when undoubted intrauterine decidual changes are present, in the absence of abdominal pregnancy or tubal involvement and a normal corpus luteum, and the presence of a blood-clot within the ovary, there would seem to be little reason for doubting the authenticity of the ovarian implantation, even in the absence of embryonic remnants. Since changes suggesting decidual reaction in the ovary have been reported so seldom, it is doubtful whether much emphasis can be laid on them. One seems justified in saying this in spite of the fact that the presence of decidua in the ovary formed the only anatomic evidence upon which Kantorowicz (1904) confidently classed his two rather advanced cases of ectopic, among authentic ovarian pregnancies. Moreover, if it be true, as stated by Webster (1904), that changes which can not be distinguished from true decidual changes not infrequently occur in the ovary in connection with normal uterine pregnancies, then the presence of islands of pseudo-decidual cells in an ovary surely can not be regarded as indicative of ovarian pregnancy. Perhaps, however, with modern histochemical methods, it would be possible to distinguish genuine decidual cells. In making this observation, we are fully aware that various criteria have been advanced from time to time by means of which to judge ovarian pregnancies, and that many of these have met with objection and have hence been modified. Such modifications would seem to be inevitable as long as there is progress in the solution of an unsettled question.


The absence of the fetus in many of the recorded cases in itself demonstrates the entire inapplicability of the criterion added by Jacobson (1908). Moreover, the histologic appearance of the ovarian tissue around certain portions of the blood-clot in the present, and also in some of the cases in the literature, would seem to suggest that it may be very difficult to find remnants of ovarian tissue at several points in a case of pregnancy which has advanced far. Hence this criterion of Spiegelberg (1878) can not be regarded as necessarily crucial. Whenever the implantation is developed at the outer instead of at the inner margin of a follicle, as in the case of Banks (1912), early destruction, even if not early rupture, of the overlying ovarian stroma and capsule would seem to be inevitable. Indeed, whenever the layer of ovarian stroma overlying the placenta is thin, very early death of the fetus would seem to be inevitable from defective nutrition alone. On the other hand, when placental development occurs in the region of the follicle directed toward the body of the ovary, great destruction of the ovarian stroma would seem to be unavoidable, even if something akin to normal decidual formation actually took place. In the case of Engelking (1913), for example, not a trace of an ovary was found in an ovarian pregnancy which had become interstitial. Even without assuming the complete authenticity of this rather equivocal case, it would seem highly probable that the presence of ovarian tissue later in the pregnancy probably is determined very largely by the location of the fertilized follicle within or by the exact location of the implantation upon the ovary.


From evidence contained in the literature, it is clear that further reports of single cases are not needed for the purpose of emphasizing the occurrence of ovarian pregnancy, yet such reports nevertheless may help in the determination of the relative frequency of this novel and sinister condition, and also throw further light upon its genesis and the finer relations of the implantations, as well as upon other matters. Moreover, the cases which are accompanied by a careful histologic examination, and which for this reason alone are wholly unequivocal from an anatomic standpoint, are still relatively few.

Specimen No. 550 of the Carnegie Collection is of interest to both the surgeon and the embryologist. It is also of great scientific value, for it shows conclusively that the ovum had lodged itself in the Graafian follicle, undoubtedly in the one from which it came, indicating that the sperm must have entered the follicle after it had ruptured. The fertilized ovum then found lodgment in the follicle, around which the corpus luteum developed. As in other cases which have been reported, no decidua was formed, showing that the decidua is not of embryonic origin.

This case illustrates well the advantage of cooperation in research. Under a special organization, with a properly equipped laboratory attached to a surgical clinic, specimens of great scientific value may be recognized and properly reported; and it is not necessary to have an entire medical faculty attached to each clinical laboratory in order to make progress in medico-biological science.


The specimen was sent to the Gynecological-Pathological Laboratory to be examined and the following record was made:


"The specimen consists of a tube and ovary from the right side. The tube at its outer extremity has been considerably mutilated. The portion received measures 6 cm. in length and is somewhat tortuous. It shows a few adhesions on the surface. Section through the middle portion of the tube shows the mucosa to be somewhat thickened and blood-tinged. Section through the distal portion gives a similar picture. There is no gross evidence of an extrauterine pregnancy. There is a portion of the fimbria present, but the portion of the tube between this and the middle is missing. There is nothing to suggest placental tissue. The mucosa appears normal in the sections.


"The ovary measures 5 by 4 by 3.5 cm. The surface shows a few old adhesions. On section the ovary shows a cyst 3 cm. in diameter. Clinging to the wall and bulging into the cavity is a blood-clot 2 cm. in width and 8 mm. in thickness. This on section appears to be corpus luteum. It is intimately connected with the walls of the cystic space. Further sections show villi in the clot attached to the inner surface of the space in the ovary."


The appearance of the specimen, with a section through the ovary, is shown well in figures 157 and 158. When received there came with it several sections from the Gynecological-Pathological Laboratory and these two drawings by Professor Brodel. These sections included the chorion, ovary, and uterine tube. The sections of the uterine tube appear normal, with a very extensive infolding of mucous membrane and occasional lymph-nodules in its walls. Doubtless the sections are from the distal or fimbriated end of the tube. The sections from the chorion are apparently at right angles to its main wall, as shown in the figure.


The villi, which are irregular in arrangement, show attachment to the main wall of the chorion, while at their distal ends they invariably abut against the blood-clot (figure 162). In no instance is there any sign of the decidua, nor do the sections through the villi contain any of the adjacent ovarian tissue. The bloodclot is well organized, with strands of fibrin extending in all directions and without distinct red blood-corpuscles. Most of the villi have a fibrous mesenchyme; in some it is mucoid. Scattered through the mesenchyme of the chorion there are blood-islands, or rather groups of blood-cells, within the blood-vessels from the embryo. These are especially numerous where the villi are attached to the main wall of the chorion, showing that in its development the embryo must have been present at an earlier stage. The distal ends of the villi are apparently covered with a double layer of epithelial cells, which is as should be in normal development. However, a rich peripheral trophoblast is missing. In the trophoblast are numerous small masses of disintegrating cells. These appear to be pretty well intermingled with mesenchyme cells at the tips of the villi, as shown in the figure. Many polymorphonuclear leucocytes are present where the trophoblast comes in contact with the blood-clot. Among the leucocytes there are isolated cells of the trophoblast. At points the isolated cells are also embedded in the mesenchyme of the villi. Altogether, these processes are quite identical with those found in the villi of the uterine moles, where there is also every indication of degeneration of the villi and their trophoblast, due to either faulty implantation or to infection. In none of the sections is there any indication of the embryonic mass, nor do the sections which were sent show the character of the ovarian tissue adjacent to the clot containing the villi, although in a number of sections the chorionic wall is shown to be composed of two layers, which doubtless represent both the chorion and the amnion (figure 162, Am and Ch). In one section these two layers are blended for a short distance, and at this point there are numerous embryonic blood-vessels. The fact that the amnion, which is quite characteristic, is in close apposition with the chorion, and the presence of numerous blood-islands show quite conclusively that they are identical with an ovum which is sufficiently well advanced in development to contain an embryo about 15 mm. long.


When the specimen came to us it was composed of two pieces which were formed by cutting directly into it on the side of the rupture, as shown in figures 157 and 158. These are drawn natural size, and therefore give the dimensions of the ovary. A loose piece of clot was taken out and cut into serial sections, but upon close examination with a microscope no trace of the chorion could be found in any of them, so it appears as though we received only the clot and a small remnant of the chorion attached to it, which had possibly invaded the ovarian tissue and lodged itself freely within it. The ovary was then cut into slabs about 5 mm. thick, and at a distance from the cavity containing the villi a large corpus luteum, 10 mm. in diameter and entirely filled with blood, came into view (figure 159). New sections of the wall of the cavity were then made. These show that the cavity as a whole is lined with a smooth, grayish membrane barely 0.5 mm. in thickness. In the tissue between the corpus luteum and the main wall of the cavity there is an extravasation of blood which enters a few of the adjacent Graafian follicles. The arrangement of these follicles is well shown in figure 158, but the corpus luteum, which is filled with blood, is much nearer the proximate pole of the ovary, and is therefore not shown in this section. It is close to the point marked Adh (figure 157). Sections were then cut through the whole ovary, and give the entire wall of the cavity containing the ovum. The sections, showing most of the structures, are at right angles to the ovary (directly through the letter in the word ovary, figure 157). They were stained in a great variety of ways: hematoxylin, acid fuchsin, iron hematoxylin, orange G, and a number of connectivetissue stains. In general, they show that the ovary is active and not fibrous, apparently normal, containing numerous blood-vessels and a ring of large Graafian vesicles (figures 158 and 159), with an outside zone of small vesicles containing small ova. To all appearances this is as it should be in a y-oung ovary.

The wall of the cavity, containing the blood-clot and villi, is lined almost throughout with a layer of lutein cells (figure 161). This layer is quite uniform, ranging from 0.5 to 1 mm. in thickness. Between the lutein cells are numerous strands of blood-vessels, but on their inner side is a layer of fibrin before the bloodclot begins. On the outside the lutein cells form small islands of more compact cells which stain more intensely in hematoxylin (figure 161, Cl'). The section reminds one very much of a section of the adrenal. We have here a layer of lutein cells well spread out, possibly due to the distention of its cavity by the ovum and representing the corpus luteum, which, according to our conventions, is about as old as this ovum appears to be. In other words, it is clear that the ovum developed within the cavity of the Graafian vesicle to which it belongs. The corpus luteum, filled with blood, near the proximate pole of the ovary (figure 160), which at first appears to represent the one from which the ovum came, is considerably more advanced in development than the one containing the ovum; therefore it belongs, in all probability, to a previous ovulation. Had it not been for the additional sets of sections we made, it would have been necessary to interpret this specimen as Teacher, Bryce, and Kerr (1908) did theirs. The hemorrhage in the ovary between the older corpus luteum and the cavity containing the ovum could easily be viewed to indicate that the ovum invaded the ovarian tissue, as shown by the illustrations of the authors mentioned.


The older corpus luteum demonstrates once more very clearly that it is imperative to standardize anew the development of the corpus luteum. It is encircled by a very marked corpus fibrosum, which is wavy and forms a uniform sheet about 0.5 mm. in thickness beyond the clot (figures 159 and 160). There are few lutein cells within it. On the inner side of the corpus fibrosum there is a thick layer of degenerated blood, and in the center a large mass of well-defined red blood-corpuscles. Within the very center of the clot is a cleft which is curiously lined and filled with red corpuscles, staining somewhat differently from those of the rest of the clot. Also, at the periphery of this clot there is a curious vesicle lying immediately under the outer fibrous layer, which may indicate a more recent hemorrhage. At any rate, the lutein cells encircling the cavity containing the clot and ovum prove quite conclusively that the ovum did not wander from a distant Graafian vesicle and become implanted freely within the tissue of the ovary. This conclusion has also been reached by Serebrenikowa (L912) in a report of a case of ovarian pregnancy, which confirms fully that of van Tussenbroek.


Bryce, Teacher, and Kerr have given an excellent review of the literature on ovarian pregnancy and Serebrenikowa has presented it from another standpoint. Both of the papers demonstrate that in ovarian pregnancy no decidua is formed, showing that a true decidua can not arise from the tissues of the ovary.


Since the cavity containing the ovum in ovarian pregnancy does not always seem to be encircled by a layer of lutein cells, it is concluded that the ovum either invaded the ovary from its surface or that it burrowed from the Graafian vesicle after fertilization. Undoubtedly the latter is the case in the specimen recorded by Bryce, Teacher, and Kerr. In it the growing ovum broke through the layer of lutein cells and made for itself a cavity in the vascular stroma of the ovary. This conclusion could also have been drawn from our specimen had not a second set of sections been made which shows that a beautiful and characteristic layer of lutein cells is present. In the first set of sections the wall of the cavity was faulty, while the second was perfect. We do not wish to question the accuracy of other observers in this respect; we want only to record our own experience. At any rate, the possibility of a secondary attachment of the ovum to the ovary through its direct wandering from the Graafian vesicle into the adjacent tissue, or indirectly through a reinvasion from the surface of the ovary, can not be denied until it is shown that the ovum is invariably lodged in a Graafian vesicle and surrounded by a layer of lutein cells of the same age as that of the ovum. Before this is possible it will be necessary to standardize the corpus luteum in relation to the ovum and embryo, and also to present as evidence only well-preserved specimens of ovarian pregnancy.


Specimen No. 1522, which was donated to the Department by Dr. H. M. N. Wynne, is a firm, nodular, dark-colored mass, 26 by 16 by 11 cm., shown in figures 163 and 164. In the gross, it especially recalls the specimens of Freund and Thome" (1906), Giles (1914-1915), and Jaschke (1915), and Lockyer's (1917) second case. The exterior is smooth though bosselated and formed by a rather injected layer which is extremely thin, showing the blood-clot beneath, around the greater extent of the specimen. The surface layer is eroded over several small elevated areas in which the blood-clot underneath is exposed. Hence, the capsule may have been ruptured in several or only in one of the areas as noted at the time of the operation. Near the region of amputation through the mesovarium shown to the right in the figure and marked by the corrugations of the hemostat, the tissue overlying the clot becomes more opaque, thicker, and also more yellowish. Here it is studded with small cysts, the character of which in itself suggests ovarian tissue. The color of the area to the right is also suggestive, and the cysts later were found to contain a clear, viscid fluid, so characteristic of cysts of the ovary. The location of the main portion of the ovarian stroma shows that implantation occurred near the mid-point of the free convex or posterior border of the ovary, and that the stroma forming the sides gradually was forced apart, not by the growing conceptus, except perhaps at the beginning, but mainly by the hemorrhage itself.


The major portion of the surface of the divided specimen shown in figure 165 is composed of blood-clot, the presence of which confirms the "blood-clot crepitus" spoken of in the history. This clot contains an empty vesicle, the wall of which is formed for the greater part of its extent by a thin fibrous layer, except near the proximal or lower portion in the figure, where a thicker portion of ovarian tissue hoods the clot. Approximately only about one-third of the ovarian stroma seems to be preserved.


Near the exterior of the thicker portions of the latter, a small cyst with clear content is found directly beneath the surface, as indicated in figure 166. The distal or upper portion in the figure shows the clot to contain an empty, smoothwalled, degenerate chorionic vesicle, such as is frequently seen in tubal clots.


Examination of the cut surface with the binocular microscope shows the presence of only a few isolated, degenerated, and some hydropic villi scattered through the clot. Examination of the chorionic vesicle shows the presence of only a few attached villi and that the amnion is fused to the chorion. Since the entire specimen obtained at operation is covered still by a layer of ovarian tissue which is unbroken save in a few very small areas, it is evident that we are dealing with a very good example of an undoubted ovarian pregnancy in spite of the absence of an embryo. The latter undoubtedly did not escape through the small rupture in the capsule, for the chorionic vesicle seems entirely intact. Although the absence of a corpus luteum in the opposite or left ovary was not especially mentioned, lack of comment would seem to suggest that none was present, for the ovary very evidently was examined. Hence, this implantation probably took place within the Graafian follicle itself, and not in some other area of the ovary.


Celloidin sections of the excised portion show that the blood-clot contains no fibrin and that it is composed of relatively fresh and fairly well-preserved cells in the region near the main body of the ovarian stroma. The latter is quite normal, although decidedly vascular, and contains ova. The layer of the ovarian stroma which surrounds the clot becomes thinner and thinner the nearer the free border is approached. It also becomes more trabeculated, hemorrhaglc, and degenerate. No overlying layer of smooth muscle, as mentioned by Young and Rhea (1911) and also by Kantorowicz, was seen. Some infiltration with polymorphonuclear leucocytes is noticeable. Degenerate villi are scattered about in the blood and a few others are apparently still attached to the equally degenerate ovarian stroma. Trophoblast is absent on these, although some of the villi that lie isolated in the clot possess a very evident epithelium and also are associated with a few small masses of very degenerate syncytium. Only a few degenerate, nonvascular villi are still seen on the chorion. Very little evidence of epithelial proliferation is present on these, despite the fact that the blood in which the vesicle is embedded is not very degenerate. This seems to suggest that the hemorrhage which caused the rupture of the ovary was comparatively recent, although the conceptus had been dead for some time. Some of the villi scattered about in the blood-clot are outlined by degenerate syncytium only, and nothing but small degenerate masses of the latter are seen on the chorionic membrane or lying about isolated in the degenerate blood found in other places. Nevertheless, the epithelium of the chorionic vesicle is thickened at several points. The amnion is fused with the chorion and both membranes are very degenerate and destroyed amost completely in several places. The surrounding ovarian tissue, which is markedly vascular and degenerate, shows infiltration in places, especially where it is stretched over the large clot. No fibrous layer bounds the implantation cavity, as in the case reported by Seedorff. The ovarian stroma merely is slightly condensed here and there, and in places contains areas of hyaline degeneration, the exact origin of which could not be definitely determined. A few of these are found near the thin bounding layer of the ovarian stroma, but no lutein layer or even luteal cells could be recognized. The only objects seen which might be regarded as possibly luteal in origin are two microscopic rhomboidal areas which lie near a small depression upon the surface, indicated in figure 167. These areas, which were covered by a very thin layer of ovarian stroma only, were made up of parallel, degenerate, slightly separated cords consisting of a syncytium containing numerous rather pycnotic, unequal-sized nuclei. No pigment was seen in these areas, and were it not for the arrangement of the syncytial cords, one would not be reminded, even remotely, of a possible luteal origin. Although the germinal epithelium was wholly absent in the areas examined, these questionable areas nevertheless may have had such an origin. In the absence of lutein cells the present case corresponds to those of Freund and Thome and others, and stands in marked contrast to the cases of van Tussenbroek, Franz, Anning and Littlewood (1901), and Thompson (1902). As in the case of Seedorff and others, no decidua was present, and nothing suggestive of an attempt at decidual formation, as reported by Franz, Webster (1904, 1907), Kantorowicz (1904), and Caturani (1914), was seen in the portions examined.


In describing his case, Seedorff declared that in some places of contact between the fetal and maternal tissues he could not discriminate between trophoblast and connective-tissue cells which looked like decidual cells and lutein cells. It is interesting that Seedorff also spoke of villi which were almost filled with Langhans cells, an observation which naturally makes one wonder whether by any possibility these could have been Hofbauer cells.


The preserved ovarian tissue which was found near the amputation stump contained hemorrhagic follicles, as observed also by Mall and Cullen (1913). A Graafian follicle 3 mm. in section, shown in figure 166, protruded above the rest of the stroma and was quite mature. The presence of this follicle might be taken as an evidence of the occurrence of ovulation during pregnancy, were it not for the fact that the presence of so degenerate a conceptus shows that as far as any effect upon the maternal organism was concerned, the pregnancy virtually had been terminated long before. That both ovulation and menstruation can and do return after the death of an ovarian conceptus, but before its removal from the ovary, is illustrated also by cases in the literature, especially by that of Norris (1909). One must assume, however, that few, if any, surviving fetal elements can be present under these circumstances. This conclusion also would seem to be confirmed by the remarkable case of Sencert and Aron (1919). These authors reported a case of ovarian pregnancy in which nothing but a portion of an umbilical cord 5 mm. long, containing Wharton's jelly, two arteries and a vein, and what was regarded as a placenta, remained. The latter was said to be composed of a narrow layer of plasmodium and a much thicker layer of trabeculated syncytial trophoblast containing blood between the trabeculse.


Because of the singular structure of this placenta and also because of the failure to find villi or any remnant of the membranes, the authors concluded that the chorionic vesicle therefore could not have reached the villous stage. How such a supposition can be reconciled with the survival of a portion of an umbilical cord entirely normal in structure it is difficult to see. The ovary concerned was brown, of the size of a "large fresh walnut," and contained a tumor, apparently the so-called placenta, which was 2 cm. in diameter. Although these fetal remnants had brought about not only almost complete amenorrhea for two years and also atrophy of the ovary and uterus, ablation of the affected ovary was followed not only by a return of the menses, but by a normal pregnancy within 7 months.


A second instance of ovarian pregnancy of special interest was that of -Giles. No fetus was found, although the pregnancy was unruptured, and Giles estimated that the conceptus had died in the third or fourth week. The operation was not done until 5 months after the onset of the pregnancy. What is particularly interesting in this case is that Giles speaks of the mucoid degeneration of the connective tissue of the villi. The latter were found to be large, much branched, and had ramified in the clot. Since Giles also spoke of one of the illustrations as showing a vesicular state of some of the villi, it seems possible that this was a case of hydatiform degeneration, even though there were no signs of activity of the syncytium. Since the fetal membranes were isolated in a blood-clot, very much degenerated, and the villi were without a Langhans layer, one scarcely could expect to find much evidence of epithelial proliferation so common (but not essential) in hydatiform degeneration. Giles estimated that 4 months had elapsed since death of the conceptus, and if this specimen really was a hydatiform degeneration, it is the first one observed in ovarian pregnancy and hence of particular interest for this reason alone.


Several features in the clinical history of our second case deserve comment. First among these is the menstrual age as compared with the size of the chorionic vesicle. Since the cross-sections of the latter measure 15 by 18 mm. and since it and the amnion are degenerated and devoid of an embryo, it is evident that the latter must have died a good while before the time of operation. Hence, the menstrual period reported for June 25, 1916, very evidently was not the last period before pregnancy supervened, but the first period which recurred after the death of the conceptus. Consequently, this pregnancy undoubtedly dates from near April 13, the time of the first omitted period. Moreover, the conceptus must have died long enough before June 25 to have made inhibition of the succeeding period impossible. It should be noted, however, that the original menstrual cycle apparently was broken, for with the customary intermenstrual period of 28 days, menstruation normally would have fallen due on June 7 instead of June 25. Hence the maintenance of the original cycle would have brought rupture of this ovarian pregnancy, as indicated by the symptoms, on July 7, in direct relation with the onset of menstruation. Nor does it seem unlikely that the hyperemia accompanying the return of menstruation on June 25, if such it really was, may have been partly responsible for the onset of a sufficiently large and persistent hemorrhage to cause the slight rupture indicated by the symptoms on July 7. It must also be remembered in this connection that cases of ovarian pregnancy have been reported in which menstruation was uninterrupted. But in the case of Chiene (1913), for example, the death of the conceptus may have occurred so early that the succeeding period was not inhibited, and the same thing may have been true in the case of Lea (1910).


Since the material from the curettage, done at the time of operation, showed the presence of a normal endometrium, the uterine decidua associated with this pregnancy must have been shed some time previously. Such a conclusion also would seem justified by the condition of the conceptus, which apparently was unable to prevent a return to the normal menstrual cycle. The absence of decidua at the time of operation also suggests that what was reported as a return of normal menstruation on June 25 may have been hemorrhage accompanying the expulsion of the decidua.

Since, in the present case, the chorionic vesicle was so degenerated and so completely isolated in a large clot, and especially since no well-implanted villi were found in the sections and gross portions examined, it is not at all probable that the hemorrhage that caused the rupture was due to a contemporaneous invasion of the vessels by the fetal trophoblast, such as occurs in uterine and tubal implantations, and as has been actually described also in ovarian implantations by Franz (1902) and by others. In the present and in similar cases in the literature it would seem that hemorrhage was made possible also by degenerative changes in the highly vascular stroma of the ovary, which had been greatly compressed and stretched by the proportionately large blood-clot, the organization of which would seem to have been precluded by its size alone.

The fact that relatively few unruptured ovarian pregnancies are recorded suggests that the old tenet that rupture is less likely the more advanced the pregnancy becomes, probably is open to serious doubt, as suggested by Banks (1912), who believed that the tubes can accommodate themselves more readily than the ovary. Banks stated that in the majority of cases of ovarian pregnancy rupture occurred in the first two or three weeks, and Caturani (1914) also expressed doubts regarding the dictum that rupture of the ovary is less likely the more advanced the pregnancy. No one will deny, I presume, that the symptoms of rupture may have been totally absent, as reported in the cases of Norris (1909) and Grimsdale (1913), but this does not necessarily imply that the ovarian stroma or the germinal epithelium still surrounded the full-term conceptus. Such an occurrence would be possible only if the ovarian stroma and the overlying germinal epithelium underwent an astonishing hyperplasia. Although such a thing is conceivable, it is decidedly significant that no one has reported any such finding or observed the presence of mitotic figures.


Instead of undergoing hyperplasia, the ovarian stroma in this case is found invaded, stretched, compressed, and degenerate, and the germinal epithelium is entirely absent. The fact that several observers have seen what they took for the fibrin layer of Nitabuch also shows that degenerative changes i-n the ovarian implantations may be extensive. Hence, it would seem to follow that the absence of symptoms of rupture merely may mean that the ovarian stroma and epithelium which happened to overlie the fetal membranes have gradually died and degenerated before being forced apart by the expanding conceptus or the increasing hemorrhage. That such a sequence of events is possible would seem to be undoubted, and mere distention of the ovarian stroma until it completely surrounded a full-term pregnancy is hardly conceivable; whereas, the absence of pain upon the yielding of an exceedingly thin degenerate layer of ovarian stroma is quite conceivable.

That rupture may occur very early is exemplified also by the cases of Chiene (1913), Seedorff (1915), and especially by that of Aiming and Littlewood (1901) and of Holland (1911). In such curious instances as that of Grimsdale (1913) one can hardly assume that the ovarian tissue was preserved about the entire conceptus, and it is not at all unlikely that full-term ovarian pregnancies, which, according to Warbanoff, supplied a surprisingly large percentage of all cases collected by him, will form a far smaller percentage in the statistics of the near future. Indeed, they already form a far smaller percentage of those reported up to the present, and the advances in diagnosis alone make it very unlikely that in the future many cases of ovarian pregnancy will advance far before being detected.

A feature noticeable in both of these cases, and not heretofore described, we believe, is the presence of clubbing of some fibrous villi, as shown in figure 168. This is marked in the villi from case No. 550. It is less pronounced in case No. 1522. The villi and vesicles of both these specimens are so degenerate that one is almost led to surmise that these vesicles never became properly implanted, but depended very largely upon the surrounding blood for nutriment. As long as the fetal circulation was not established these conditions would seem to offer no specia) obstacles, for up to that time the conceptus necessarily is dependent upon other means of nutrition in uterine implantation also. Moreover, it may be doubted whether anything akin to true implantation can occur in the ovary or tube in the absence of decidual formation, and hence also of a capsularis. For even if the ovum buried itself in the ovarian stroma, the continued hemorrhage and the failure of a similar and proper response on the part of the ovarian stroma nevertheless would furnish decidedly abnormal conditions.


Clinical Histories

No. 550

The patient, who was 24 years old, came to the Johns Hopkins Hospital June 1 with a diagnosis of appendicitis, which later was changed to ovarian pregnancy. The diagnosis was made unusually difficult by the patient's misleading statements.

When admitted to the ward the patient was not complaining of acute pain, but only of general soreness in the abdomen. There were paroxysms of pain, general throughout the abdomen, with intermissions in which she was somewhat more comfortable. With difficulty the pain was localized in both left and right sides of the abdomen.

When examined, general soreness of abdomen was found, the pain being more acute along the left side, shooting up to the right shoulder. The pain had not changed in character or intensity. There were sharp attacks of pain, especially in the left side, when the patient tried to move. There was also difficulty in breathing.

During the afternoon the condition of the patient was very uncomfortable, with repetition of the symptoms just given. The pains became more acute after taking ice. A renewed onset occurred at 11 p. m., and this continued with some nausea and occasional vomiting.

The pain had been sharp (not crampy), and had apparently gone up under the right C. M. in the morning. There had been pain also under the shoulder. The patient said she had never had any similar attacks, and was not constipated previous to this one.

On June 3 patient said she had a similar attack of abdominal pain three years before. This was general at first and finally became more pronounced on the right side, accompanied by nausea and vomiting. From this attack she did not recover entirely for 10 days. She had a similar but milder attack several months later. Dr. Finney writes:

"When I saw the patient I did not think it was appendicitis, but the history of similar attacks, which I had reason to believe afterwards were fictitious, and the patient's misstatements as to the subjective signs, together with the fact that the patient was unmarried, misled us as to the true diagnosis. Upon opening the abdomen, however, it was found to be filled with blood. At once the diagnosis was clear. I looked for the tubes, but found both intact. The right ovary was the point of bleeding; it was swollen and appeared as you found it in the specimen. The whole process was so definitely confined to the ovary that it seemed, clinically, to be a definite case of ovarian pregnancy."


No. Template:CE1552


The patient, an Italian woman of 37 years, was admitted to the Gynecological Service of the Johns Hopkins Hospital July 12, 1916, complaining of a pain in the lower abdomen, nausea, and vomiting.

Family history. Negative.

Past history. General health good. She has never had any serious illness. For the past 5 years following a labor she has had recurring mild attacks of pain in the abdomen without nausea or vomiting.

Menstrual history. Always regular every month, except when pregnant or lactating. Duration 4 to 5 days; painless, moderate flow. Last period June 25, 1916. Last preceding period, March 16, 1916. No intermenstrual bleeding before present illness.

Marital Married 18 years; seven children, oldest 16, youngest born \y% years ago (died, 1915). Has had three miscarriages. History of labors and puerperia vague.

Present illness. Began 5 days ago (July 7, 1916) with sudden pain in lower abdomen, nausea, and vomiting. She has had marked dysuria and painful defecation. For 12 hours after onset there was rather profuse bleeding from the vagina and there has been a bloody vaginal discharge since.

(The patient does not understand English and her husband acted as interpreter.)

Physical examination. T. 101.6 degrees F. P. 96. W. B. C. 8400. Hbg. 46 per cent.

The patient lies in bed grunting with pain. The skin is pale. The lips and mucous membranes are quite pale. There is a systolic blow heard at the apex and increasing toward the base, being loudest over the pulmonic area.

A drop of clear fluid was expressed from the right breast.

The abdominal respiratory movements are limited, although she does not complain of pain on deep inspiration. The flanks bulge somewhat. There is no demonstrable movable dulness. There is tenderness all over the abdomen, most marked over the lower left quadrant. There is increased resistance over the lower abdomen, but no muscle spasm. No masses can be made out.

There is a profuse bloody vaginal discharge. The cervix is pushed up behind the symphysis by a soft, exquisitely tender mass, filling the cul-de-sac. No crepitus is made out. Rectal examination confirms the vaginal. The fundus of the uterus is not felt.

July 13, 1916. Ether examination.

There is a dark, bloody discharge from the vagina. There is no vaginal cyanosis. The cervix is lacerated, firm, and normal in size.


The fundus of the uterus is about normal in size and is in anteposition. A boggy mass fills the cul-de-sac and to the right of the uterus a fairly firm mass, the size of a small orange, which is somewhat movable, can be felt. Definite blood-clot crepitus can be felt on rectal examination.

Pre-operative diagnosis. Extra-uterine pregnancy, ruptured.

Operation. A free midline incision was made below the umbilicus. The peritoneum was blood-stained. The abdomen contained 200 to 300 c.c. of dark fluid blood and clots, and a large clot filled the cul-de-sac. Active bleeding had ceased. The left tube and ovary were normal and free of adhesions. The right tube, which was quite normal in appearance, lay over a mass which had replaced the right ovary. This mass was roughly spherical, 5 to 6 cm. in diameter and semi-solid in consistency. Over the surface there were six or eight nodular projections, about 1 cm. in diameter. At the top of one of these projections there was a very small opening, from which bloody fluid could be squeezed. The surface of this mass was white, with spots of bluish-black discoloration.


The appendix was normal, except at the tip, where it was adherent to a blood-clot.

The tumor was removed by clamping, tying, and cutting into the right infundibulo-pelvic ligament and the right utero-ovarian ligament. The right tube was not removed. The appendix was also removed and all blood and clots were cleaned out of the abdomen. The incision was closed without drainage. At the close of the operation the uterus was curetted. The uterine cavity measured 7.5 cm. in length.

A subcutaneous salt solution infusion was started on the table and continued on the ward until 2,000 c.c. had been absorbed.

The patient was in good condition at the end of the operation and made an uneventful recovery.

The urine on admission contained red bloodcells, white blood-cells, no casts, acetone, a trace of albumin, and no sugar. Several days after the operation it was negative, except for a faint trace of albumin.

August 2, 1916. Discharged in good condition.

Gyn. Path., No. 22346- Normal endometrium from curettage.

Description of Plates

Plate15

Mall Meyer1921 plate15.jpg


Fig. 157. Drawing of posterior view of pregnant ovary with tube. No. 550.

Fig. 158. Transverse section of pregnant ovary cut at point of rupture.

Fig. 159. Outline of transverse section of ovary taken in the region of the letter o in "Ovary" (fig. 157). Cl, lutein cells; Bel, Blood clot; Cf, corpus fibrosum; G, Graafian follicle. X2.

Fig. 160. Drawing of portion of the corpus luteum, indicated by small square to the left in fig. 159 (marked fig. 4 in drawing). Ot, tissue of ovary. X70.

Fig. 161. Drawing of a portion of the ovary, indicated by small square to the right in fig. 159 (marked fig. 5 in drawing). X70.

Fig. 162. Drawing of a section through free surface of clot, illustrated in section in fig. 158, showing wall of chorion and a villus surrounded by clot. Tr, trophoblast; L, leucocytes; Ch, chorion; V, villus; Coe, ccelom. X50.


Plate 15: Figs. 157+158 | Fig. 159 | Fig. 160 | Fig. 161 | Fig. 162 | Chapter 11 Ovarian Pregnancy

Plate 16

Mall Meyer1921 plate16.jpg


Fig. 163. External appearance of intact specimen. No. 1522. X0.75.

Fig. 164. External appearance of same specimen, showing where block was removed. X0.75.

Fig. 165. Appearance of cross section of pregnant ovary and tube, same specimen. X0.75.

Fig. 166. Photograph of a section taken from the thick portion of the ovarian stroma near the mesovarium, showing a well-developed Graafian follicle. Same specimen. X2.25.

Fig. 167. Section of part of the same specimen, showing the clot which contains the empty vesicle largely surrounded by ovarian stroma. XI. 9.

Fig. 168. Marked clubbing of villi of No. 550. X2.25.

Fig. 169. Cross section of decidua and conceptus. No. 698. (See Chapter XII.) X4.5.

Fig. 170. Section of conceptus, decidua, and muscularis. No. 970. (See Chapter XII.) X4.5.

Fig. 171. Section of decidua and conceptus. No. 962. (See Chapter XII.) X4.5.

Fig. 172. Section of a part of the conceptus, showing chorionic membrane, cyemic (?) rudiment (x) and yolk-sac. No. 1843. (See Chapter XII.) X51.5.

Fig. 173. External view of No. 2047, showing the distended amnion. (See Chapter XII.) X2.25.

Fig. 174. Section of the tube and the conceptus. No. 977. (See Chapter XII.) X3.3.

Uterine end

Clot containing cViorionic vilu Free fimbr. end Site of rupture Posterior view Fig. 161 -

Plate 16: Figs. 163 | Fig. 164 | Fig. 165 | Fig. 166 | Fig. 167 | Fig. 168 | Fig. 169 | Fig. 170 | Fig. 171 | Fig. 172 | Fig. 173 | Fig. 174 | Chapter 11 Ovarian Pregnancy