Explication and Defense of Sampson’s Theory of Pathogenesis

A History of Endometriosis

10. Explication and Defense of Sampson’s Theory of Pathogenesis

Ronald E. Batt¹

(1)

State University of New York at Buffalo, Buffalo, New York, USA

Abstract

In May 1927, at the 52nd annual meeting of the American Gynecological Society at Hot Springs, Virginia, Sampson stated: “At the meeting of the American Gynecological Society in 1921, the writer presented a paper¹ on perforating hemorrhagic cysts of the ovary and their relation to pelvic adenomas of endometrial type…In view of the theories which have arisen to explain the origin of the peritoneal endometriosis associated with these cysts, the following quotation from that paper may be of interest.”² “The question naturally arises: in what way do the contents of the cyst or ovary cause the development of these adenomas? Is it due to some specific irritant present in the cyst contents which stimulates the peritoneal endothelium to a metaplasia with the development of endometrial tissue typical both in structure and function? Some may assert that dormant endometrial epithelium may be present in the tissues soiled by the contents of the cyst and this is stimulated to further growth. It seems to me that the conditions found in many of these specimens are analogous to the implantation of ovarian papilloma or cancer on the peritoneal surfaces of the pelvis from the rupture of an ovarian tumor containing these growths.”³

Peritoneal Endometriosis Due to Menstrual Dissemination

Immediately, Sampson made a critical statement, critical to medical historiography. “At that time I believed that the ovary was the principal, if not the only source of the peritoneal implantations which arose from endometrial tissue disseminated by the menstrual perforation of an endometrial cyst or by menstrual reaction of endometrial tissue on the surface of the ovary.”⁴

While Sampson’s theory of implantation endometriosis was intuitively reasonable and dramatically different from embryonic rests and coelomic metaplasia, he could not prove it. Furthermore, pathologists had demonstrated metaplasia in many tissues in the human body. One must remember the power of pathology in gynecology; pathology was the basic science – sine qua non – of gynecology before World War II. Among all of the theories put forth to explain the pathogenesis of extrauterine endometriosis, it seems that by 1927 the theory of serosal–coelomic metaplasia of Iwanoff and Meyer had gained increased acceptance,⁵ though the stimulus for metaplasia – inflammation or hormonal – was debated. Sampson believed the stimulus was hormonal, Meyer believed it was inflammation.⁶ Clinicians as well as gynecologic pathologists or for that matter anyone skilled in microscopy could see the transition from normal germinal layer to endometriosis and back to normal ovarian germinal epithelium. In uterine specimens, they could see also invasion of normal endometrium into underlying uterine myometrium. Cullen’s observation of direct invasion of the myometrium by endometrial mucosa was accepted virtually without criticism to explain the pathogenesis of diffuse adenomyosis – uterine endometriosis.⁷ In both microscopic demonstrations, seeing was synonymous with believing. The same held true, when in 1925, Sampson presented evidence for the malignant transformation of benign into malignant ovarian endometriosis.⁸ Every experienced microscopist could see the transition from benign to malignant. Both gynecologic surgeons and gynecologic pathologists were in complete agreement; there was no debate.⁹

However, in the case of Sampson’s original ruptured ovarian endometrioma theory and his second and refined retrograde menstruation theory to explain the pathogenesis of pelvic endometriosis, seeing and believing were not related to looking down the barrel of a microscope. Seeing and believing required active observation at surgery and pattern recognition skills honed by years of surgical experience. Seeing also required imagination to visualize, as a Goethe might have done, the transition from intact endometrium to living shed endometrium to transtubal transport to attachment to implantation and invasion of pelvic organs by the same living shed endometrium.¹⁰ Sampson’s first theory¹¹ of peritoneal implantation, implantation of endometrial fragments shed from ruptured hemorrhagic chocolate cysts of the ovary seemed to have been accepted by many clinicians if we can judge by the comments of Meigs¹² in 1922, Bailey¹³ in 1924, and Cullen¹⁴ in 1925. Investigators generally and rather rapidly accepted the term endometriosis. Hitherto, Cullen and the investigators, who were his contemporaries and predecessors, tended to see endometriosis as a tumor, an adenomyoma, an adenofibromyoma, or a cystoma. W. Blair Bell of London coined the terms endometriomyoma and endometrioma.¹⁵ The suffix “oma” denotes a neoplasm or tumor, an unfortunate designation for a benign disease process. This may explain the ready acceptance of the more general term “endometriosis” suggested by Sampson.¹⁶

Others acknowledged Sampson’s contribution toward defining the importance of endometrial cysts, but did not subscribe to his theory of implantation by cellular material spilled from chocolate cysts.¹⁷Sampson’s first and second theories of implantation endometriosis seemed logical and were supported by circumstantial evidence but he could not prove them.¹⁸ By 1927, the lines of debate were visible and palpable: some gynecologic surgeons saw and believed as Sampson did; most if not all gynecologic pathologists and embryologists – and some gynecologic surgeons – saw and believed as Robert Meyer and Emil Novak did. In short, the former favored one or both of Sampson’s theories of implantation; the latter chose to believe the evidence for the Iwanoff-Meyer theory of coelomic metaplasia was stronger.

By 1927, the honeymoon was over; Sampson was on the defensive. He could only continue to explain and defend his theory against mounting criticism from respected gynecologic pathologists, especially Robert Meyer and Emil Novak,¹⁹ who espoused the theory of coelomic metaplasia. Sampson defended his theories by employing inferences,²⁰ questions,²¹ “must” assertions,²² and qualifications such as “suggest…might escape…it might be assumed…not definitely demonstrated…but I believe…probably came from…if it can be shown…might infer…might become…should it be shown…we would have strong presumptive evidence…apparently did…may carry with it bits of endometrium.”²³ Sampson argued tirelessly and buttressed his arguments with carefully selected and annotated photomicrographic evidence. There are recurring glimpses of what in the following decade would be Sampson’s increasing reliance on analogy with uterine and ovarian endometrial cancer to support his implantation theory of the pathogenesis of endometriosis. The cancer analogy, though powerful and persuasive, was still argument by analogy and not proof.

One can sense Sampson’s frustration as he faced increasing acceptance of the theory of coelomic metaplasia.²⁴ He was open to the possible validity of the theory of coelomic metaplasia.²⁵ “I fully realize that the implantation theory does not account for all instances of ectopic endometrium-like tissue in the pelvis and that menstruation is only one means of disseminating that tissue.”²⁶ Sampson was a clinician, a skilled cancer surgeon, a natural scientist and a theorist, but not an experimentalist. And experimentation was needed, not analogy. True, animal experiments had been performed on rabbits and monkeys.²⁷ At this juncture in 1927, no one seems to have been able to conceive of an experiment other than transplantation of human shed menstrual endometrium into an animal or a human as Novak suggested. The needed clinical and laboratory experiments with human tissue would not be performed until after World War II, after Sampson’s death. Compared to Cullen’s simple demonstration of the pathogenesis of uterine endometriosis, proof of Sampson’s implantation theory for the pathogenesis extrauterine endometriosis would turn out to be unbelievably difficult. It would require a small army of basic scientists with sophisticated laboratory techniques; such was the brilliance of Sampson’s intuition and imagination.

Sampson stated explicitly that his purpose in writing this article was to present evidence in support of his theory of implantation.²⁸ In the very first photomicrograph, he presented an “endometrial cavity” filled with menstrual blood, situated within deeply invasive endometriosis of the vaginal vault. The ectopic endometrial cavity was “about” to rupture and menstruate into the vagina.²⁹ In the same case, Sampson had observed two similar endometrial cavities in the posterior vaginal fornix that actually ruptured into veins with “the embolic implantation of endometrial tissue.”³⁰ Sampson would refer to this case over and over again; it was his proof case of venous metastases.³¹Sampson seems to have adopted the term “endometrial cavity of the ovary” from Casler.³² As mentioned earlier, I believe Casler’s unique case influenced Sampson profoundly. In Casler’s case, a woman who had had a complete hysterectomy, with preservation of one ovary, menstruated from an endometrial cyst of that ovary through an ovarian–vaginal fistula into the vagina each month. The term “endometrial cavity of the ovary,” used as a metaphor for any ectopic endometrial cyst was heavily freighted with meaning favorable to Sampson’s theory and implied that ectopic endometrial cavities recapitulated the structure and function of uterine endometrial cavities. Clearly demonstrated in a series of photomicrographs,³³ many ectopic endometrial cavities were filled with menstrual detritus, blood, and fragments of endometrial tissue; powerful visual arguments working in Sampson’s favor. He used the term endometrial cavity repeatedly to describe cystic endometriotic lesions in various ectopic locations such as the posterior vaginal fornix, the broad ligament,³⁴ and the ovary.³⁵ In fact, he often labeled ovarian endometriomas within the actual photomicrographs “Endometrial Cavity.”³⁶ Clearly, Sampson was a master of the English language and used it skillfully to promote his theory.³⁷

Novak, in particular, challenged Sampson’s thesis that shed endometrial tissue was alive; he believed it was either “dead or dying.”³⁸ Novak suggested two experiments where positive results would bolster Sampson’s theory: first, successful growth of menstrual endometrium in tissue culture and a second experiment to prove the capacity of menstrual endometrium to grow in the peritoneal cavity or on the ovary of “the human being or perhaps even one of the lower animals.”³⁹ Sampson immediately referred to his proof-case of the ectopic endometrial cavity in the posterior fornix to argue that shed ectopic endometrium that metastasized into the surrounding veins “must have been alive.”⁴⁰ Following his proof-case argument, Sampson argued against coelomic metaplasia. He drew attention to two photomicrographs, right ovary⁴¹ and left ovary, surgical specimens from the same patient.⁴² The endometrial lesions were located on the lateral and under surfaces of the left ovary and on the lateral surface of the right ovary.⁴³ Sampson explained. “The lateral surface of the ovary normally lies against the posterior layer of the broad ligament or side of the pelvis and thus a crevice is formed which would retain any material lodging in it. The lesions are situated on the dependent portion of the ovary, namely, the bottom of the crevice.”⁴⁴ He repeated his “geographical distribution” argument for implantation endometriosis of the ovary.⁴⁵

Sampson followed with his “age differential” argument; the endometrial lesion on the right ovary was “evidently of more recent origin” than those on the left ovary. The difference in age favored the implantation theory and worked against the coelomic theory. “Should the endometrial tissue in this case have arisen from the differentiation of coelomic epithelium due to its stimulation by an ovarian hormone, we should expect that these lesions would all be of the same age.”⁴⁶ Earlier in this paper, Sampson had argued against differentiation of coelomic epithelium due to its stimulation by bacterial or malignant inflammation as the pathogenesis of endometriosis; here he accepted the differentiation of coelomic epithelium due to its stimulation by an ovarian hormone.

Still examining evidence for endometrial implantation or coelomic metaplasia, Sampson questioned the origin of an endometrial lesion situated in the floor of a peritoneal pocket beneath the ostium of an adherent left fallopian tube. “Is it an implantation of endometrial tissue or a metaplasia of the peritoneum arising from stimulation of the latter by some specific material escaping from the tube?”⁴⁷Examining an “older” lesion removed from the culdesac, older compared to the lesion in the peritoneal pocket, Sampson stated: “It must have arisen from an earlier implantation of endometrial tissue or an earlier specific stimulation of the peritoneal mesothelium.”⁴⁸ Sampson repeatedly left space for the theory of coelomic metaplasia all the while arguing for his implantation theory. Then, he switched back to the geographical distribution argument. In the very next illustration of a “patch” of peritoneal endometriosis situated “very close to the ostium of the patent left tube and directly over the ureter” and a similar patch involving the left uterosacral ligament, Sampson noted that both tubes were patent and both ovaries were normal.⁴⁹ To explain the pathogenesis of the patches of peritoneal endometriosis, Sampson asserted the seed and soil metaphor: “It is an implantation-like lesion in a situation easily soiled by material escaping from the patent left tube.”⁵⁰ He defended his first and second theories of endometrial implantation: “Peritoneal endometriosis occurs most frequently in situations in the pelvis easily soiled by material escaping from the tubes and ovaries. It would seem the tubes and ovaries are the chief distributing agents for the cause of pelvic peritoneal endometriosis. It is not peculiar to the pelvic peritoneum, as the appendix, cecum, small intestine and their mesenteries may be involved. The posture of mankind, whether standing, sitting, or lying down appears to be an important factor in determining the distribution of these lesions.”⁵¹

Gravity was as pivotal for Sampson’s theory of implantation endometriosis as it was for Galileo’s experiments at the Tower of Pisa or Newton’s apocryphal apple.⁵² The geographic distribution of intestinal endometriosis: on the appendix, the cecum, small intestine, rectum, sigmoid, and their mesenteries presented a strong argument for gravitational distribution of endometrial tissue shed from the fallopian tubes and ruptured ovarian endometrial cysts.⁵³ Except for the supine and knee–chest positions, the retrocervical area in the anterior rectovaginal pouch of Douglas is always the most dependent portion of the pelvic cavity in females and the ultimate repository for endometrial tissue shed retrograde through the fallopian tubes or from a ruptured ovarian endometrial cyst.

Next, Sampson asked the quintessential question; could shed endometrial fragments pass through the narrow caliber of the interstitial portion of the fallopian tube in order to gain access to the ovary, peritoneum, and other structures in the pelvis? Critics claimed the passage was too narrow and the transit time too long for retrograde menstrual fragments to pass and still be viable to implant. Sampson had already addressed the question of viability to the best of his ability. So, he addressed the problem of the diameter of the fallopian tube at its narrowest by returning to his uterine injection experiments published in 1918.⁵⁴ The diameter of the isthmus of the fallopian tubes, as viewed on roentgenograms, hysterosalpingographic x-rays of the uterus, and fallopian tubes of surgical specimens hardened by fixatives, varied so as to appear as a “mere thread” or “a relatively large canal.”⁵⁵ Sampson compared the diameter of the isthmus of fallopian tubes in two hysterosalpingograms. In the first, “the interstitial portions of the tubes appear as mere threads, but even so, small bits of endometrial tissue might be carried into the tubes by blood escaping from the uterine cavity during curettage and menstruation.”⁵⁶ In the second, he argued “The lumen of the interstitial portion of the tube is much greater than that of the preceding specimen, and therefore, larger bits of endometrial tissue could pass from the uterine cavity into the tubes than in the former.”⁵⁷ Sampson referred to an earlier publication of a Figure, in 1918, and stated that in the legend accompanying the figure, he had “suggested that menstrual blood, at times, might escape through the tubes into the peritoneal cavity.”⁵⁸ Sampson compared photomicrographs (10×) of a section of menstrual blood from the uterine cavity of one surgical specimen with the (10×) enlargement of “the very narrow threadlike interstitial portion of the tube” in a hysterosalpingogram x-ray. He concluded: “The smaller bits of the uterine mucosa in this blood would easily pass through the lumen of the tube and the larger pieces would readily pass through a tube of greater caliber as the tube of the roentgenogram. Blood escaping from the uterine cavity into the tubes during curettage and menstruation, at times, might carry with it bits of the uterine mucosa suspended in that blood.”⁵⁹

Thus far, Sampson had not demonstrated endometrial fragments within the interstitial portion or the ampullary portion of a fallopian tube; instead he compared the diameter of endometrial fragments and the diameter of fallopian tubes and concluded that small endometrial fragments might pass through the narrower isthmus and larger fragments though the tubes with a larger diameter isthmus. In other words, based on comparative measurements of passage and passenger, retrograde menstruation onto the ovary and pelvic peritoneum might be possible. The operative word was “might.” Based on the visual images so presented, the endometrial fragments were smaller than the diameter of the tube and seemingly could readily pass retrograde into the pelvis. Sampson presented photomicrographic evidence of endometrial fragments and fibrin “moulded” into a curvilinear “cast of the lumen of a narrow portion of the tube,” which he contended was formed during passage through the narrow tubal isthmus and constituted further evidence that the fragments originated in the endometrial cavity.⁶⁰ In five photomicrographs of the “healthiest” appearing shed endometrium from the vagina, uterus, and the aforementioned tubal “cast,” all at 250× magnification, Sampson contended that the tissue “seemed” alive based on histologic appearance, except perhaps for the central portion of the tubal “cast” specimen.⁶¹

Sampson recalled eight patients in whom he saw menstrual blood dripping from the fimbriated extremity of the tubes and in which he found endometrial tissue in the histologic sections of the same tubes removed surgically. He also observed the blood dripping from the tubal ostia at laparotomy in non-menstruating patients.⁶² “Bits of endometrial tissue, apparently set free by the curette, were found in the lumina of tubes which had been removed and some of these pieces were larger than similar pieces of endometrial tissue set free by menstruation.”⁶³ He supported his clinical observations with solid comparative histologic evidence. A photomicrograph of a cross section of the ampulla of a fallopian tube excised following uterine curettage demonstrated endometrial fragments of larger diameter than found in cross section of the ampulla of a fallopian tube removed during menstruation.⁶⁴ On the basis of finding endometrial fragments in the fallopian tubes in the short time between uterine curettage and immediate laparotomy, Sampson deduced that the transit time from endometrium into the tubes “might be very short, not several days but a few moments.”⁶⁵

Spurred by the critique of Novak, Sampson considered five ways that endometrial fragments might be found in fallopian tubes removed at surgery. The endometrial fragments could be: (1) artifacts; (2) carried along with blood shed at curettage; (3) from menstruation of uterine mucosa “forming a part of the tubal mucosa”; (4) from entrapment of ectopic pelvic endometrial tissue by the tubal fimbriae⁶⁶similar to ovum pickup at ovulation time; and lastly, the endometrial fragments could be from retrograde transtubal menstruation.⁶⁷ Sampson eliminated the artifact possibility when observations were restricted to tubal contents of patients operated while menstruating.⁶⁸Analogous to Janney’s critique of embryologic evidence, Sampson demonstrated the possibility for retrograde menstruation through human fallopian tubes, but he had not demonstrated all steps in the chain of evidence from uterine endometrium to endometrial implant nor had he proved the viability of shed human endometrial fragments.

Recall Janney’s critique of embryologic evidence. “Embryology may be suggestive but hardly conclusive, for the reason that suggestive appearances in an embryo can never be proved to be the early stages of a condition found in adults unless all of the steps can be demonstrated, which seems unlikely in a condition of this rarity.”⁶⁹ Unlike embryonic rests; ovarian, peritoneal, and bowel endometriosis are common lesions. Given time, financing, sophisticated investigational tools, and resourceful laboratory scientists, Sampson’s theories of pathogenesis could be put to experimental proof.

Sampson stressed the presence of patent fallopian tubes in patients with endometriosis.⁷⁰ He demonstrated to his satisfaction that peritoneal irritation and exudate associated with gonococcal infection and malignancy did not induce metaplasia of the germinal epithelium of the ovary or of the pelvic peritoneum subsequently resulting in endometriosis. Placing three photomicrographs in juxtaposition, he demonstrated the reaction of germinal epithelium of the ovary to gonorrheal infection and argued that it never develops “into peritoneal carcinosis or true peritoneal endometriosis” under such circumstances.⁷¹ Furthermore, exudative reaction to malignant implants did not induce metaplastic transformation of parietal peritoneum into endometriosis. Instead, based on his comparative study of peritoneal endometriosis and peritoneal carcinosis, Sampson “demonstrated the peritoneal reaction is the same in both instances.”⁷² He also presented photomicrographs of peritoneal carcinosis implants resulting from spill of ovarian cancer into the peritoneal cavity.⁷³ To further emphasize their similarity, Sampson drew the parallel between cancer and benign endometrial tissue. Both metastasize through vascular channels.⁷⁴ The third photomicrograph showed an endometrial implant, one of several, on the lateral surface of an ovary. The implant “is enmeshed in an exudate somewhat similar” to that in the two photomicrographs shown in juxtaposition.⁷⁵ The surface epithelium [of the ovary] is intact on either side of the endometrial implant but otherwise has disappeared beneath the endometrial implant. Sampson concluded: “The endometrial tissue…may have been derived either from the implantation of similar tissue from other endometrial lesions of the ovary, from endometrial or tubal tissue escaping through the tubes (both were patent) or from localized metaplasia of the surface epithelium of the ovary.”⁷⁶ His conclusion diluted what seemed, at first reading, a strong argument in favor of his theory.

Sampson demonstrated by microphotographs that peritoneal endometriosis had “apparently” invaded the uterine musculature “causing a so-called adenomyoma” of the uterus (adenomyosis uteri interna).⁷⁷While Cullen had only to demonstrate continuity between uterine endometrium and underlying uterine adenomyoma, Sampson had the additional burden of convincing his peers that shed endometrium, transported through the fallopian tubes, had implanted on the uterine serosa and then invaded to produce the same adenomyoma as that produced by direct invasion of the endometrium from the endometrial cavity. Furthermore, Sampson was explicitly arguing that secondary endometriosis could invade the uterus and produce lesions identical with direct or primary endometriosis. In short, he was arguing that external implantation endometriosis could cause internal endometriosis, albeit in the periphery of the uterine musculature. Implicitly, Sampson was arguing for the identity of pathogenesis of primary (uterine) endometriosis and secondary (extrauterine) endometriosis. For the moment, all he had to do was to convince his peers to accept his theory of implantation endometriosis as a viable alternative to coelomic metaplasia. Sampson knew that, ultimately, general acceptance of his theory by the scientific world depended on experimental proof, not on circumstantial evidence no matter how compelling.

But not all circumstantial evidence was compelling. In the very next photomicrograph of the same specimen by which Sampson had argued that implantation endometriosis on the uterine serosa could invade and “apparently cause” a typical adenomyoma,⁷⁸ he pointed out “dead cells” and a “necrotic mass” of dead cells on the peritoneal surface of the uterus.⁷⁹ Adjacent to the necrotic mass, Sampson indicated a surface endometriosis and an underlying endometrial lesion embedded deeply in the uterine musculature. Sampson noted that “the greater portion of the tissue in this photomicrograph is dead, but some lived (that in contact with the surface of the uterus) with a resulting endometriosis.”⁸⁰

After presenting an impressive amount of circumstantial evidence, ardent assertions, and arguments; all with due deference to the principal competing theory of coelomic metaplasia, Sampson got to the crux of this paper: the integration of his first and second theories of pathogenesis of implantation endometriosis. Recall that his first theory, presented in 1921, explained the etiology of peritoneal endometriosis and accompanying endometriotic adhesive disease from perforating hemorrhagic (chocolate) cysts of the ovary.⁸¹ His second theory, initially presented in Boston in 1922, explained the etiology of ovarian endometriomas and some peritoneal endometriosis from implantation of endometrial fragments shed retrograde through the fallopian tubes during menstruation.⁸²

In explaining the etiology of endometrial-like tissue in the ovary, Sampson made reference to the variety of lesions of uterine and tubal origin characteristic of direct or primary endometriosis, which were his standard of comparison for both ovarian and peritoneal endometriosis. The spectrum of uterine lesions ranged from “typical endometrium with glands and stroma identical with that of the müllerian mucosa from which it came,” to dilated endometrial glands with little or no supporting endometrial stroma, to “extreme cases” that resembled mesothelium.⁸³

Sampson explained in detail. “Marked changes in the mucosa lining the uterine cavity often occur. Of particular interest are those in the mucosa over a submucous leiomyoma. This mucosa becomes thin, the glands disappear, the stroma becomes less and less and in extreme cases the submucous leiomyoma is covered by a mucosa not unlike the mesothelium covering of a subserous leiomyomas. The endometrial epithelium in unfavorable conditions may be very similar to the peritoneal mesothelium and the surface epithelium of the ovary. Both the peritoneal epithelium and the surface epithelium of the ovary, under the stimulation of any irritant, may become hypertrophied and resemble the epithelium of the uterine mucosa. We recognize the mucosal covering of a submucous leiomyoma to be of endometrial origin, even though it simulates the peritoneal mesothelium and the epithelial lining of a follicular cyst of the ovary. When we find patches of typical endometrial tissue in an ovarian cyst with a lining similar to the mucosal covering of a submucous leiomyoma, can it be truthfully said that the typical endometrial tissue represents a metaplasia of the epithelium of a follicular cyst of the ovary, just because the cyst is situated in the ovary and portions of its lining resemble that of a follicular cyst? Is it not more logical to claim that the entire cyst is lined by one kind of tissue and that the endometrial portions of which have failed to attain their full growth, and thus present a histologic picture identical with the linings of some of the atypical endometrial cavities of a direct endometriosis and mucosal covering of submucous leiomyomas?”⁸⁴ Sampson concluded his argument. “On account of the faculty of known endometrial epithelium to simulate peritoneal mesothelium and the surface epithelium of the ovary, it is often difficult to determine the origin of all misplaced endometrium-like tissue in the ovary.”⁸⁵ Sampson believed in his implantation theory and argued for its acceptance, but he did not believe it was the only mode of pathogenesis of ovarian endometriosis.⁸⁶ The hormone-driven theory of coelomic metaplasia provided considerable explanatory power and Sampson left ample room for it.

Sampson began his explanation of the etiology of peritoneal endometriosis with a disclaimer. He could not prove that all lesions of peritoneal endometriosis that he had encountered in the past 5 years had been “instances of true peritoneal endometriosis.”⁸⁷ Nevertheless, I do believe he had faith that his implantation theories provided the best explanation for peritoneal endometriosis but not necessarily the more persuasive explanation for ovarian endometriosis. Consequently, he spent more time defending the etiology of ovarian endometriosis from tubal spill of shed endometrial tissue. He spent little time discussing the etiology of peritoneal endometriosis, commenting mostly on the histological standards for diagnosis.

Sampson held that typical endometrial histology provided the ideal standard of comparison for younger and more superficial peritoneal implants, while atypical endometrial histology characteristic of Cullen’s primary endometriosis (adenomyosis) provided the practical standard of comparison for older and deeper endometriotic lesions.⁸⁸

Realizing that his evidence was insufficient to convince skeptics, Sampson ended his explanation for the pathogenesis of peritoneal endometriosis with a second disclaimer. “These studies indicate that peritoneal endometriosis sometimes arises from the implantation of endometrial tissue disseminated by menstrual blood escaping into the peritoneal cavity.” The operative word in this qualified conclusion was “sometimes.” This disclaimer suggested a retreat from his earlier strong position, a loss of ground to coelomic metaplasia, and the need for further study.

In sum, when the gynecologic pathologists Robert Meyer, Emil Novak, and others scrutinized his implantation theories, Sampson realized objective circumstantial evidence garnered through his trained judgment⁸⁹ did not rise to the level of objective scientific proof, the benchmark for acceptance as fact.⁹⁰ He recognized the need for objective experimental studies using animal models as well as human subjects. Though he retreated from his earlier affirmative position, Sampson’s belief in the validity of his theories was undimmed as he continued his observational clinical research.

Sampson’s Implantation Theory

From 1928 until his death in 1946, Sampson wrote 13 papers, 1 related to medical education, 5 to cancer, and 6 to endometriosis with special emphasis on the fallopian tubes.⁹¹ Though interested in the pathogenesis and life history of all types of endometriosis, he realized he left many problems unsolved.⁹² In 1940, 1 year after he retired as Professor of Gynecology, Sampson distilled the observations of his scientific work on endometriosis in a memorable self-critical review entitled: The development of the implantation theory for the origin of peritoneal endometriosis.⁹³ Eighteen years had elapsed since Sampson had published the essentials of the implantation theory. Without stating the specific publication, Sampson must have been thinking of his presentation to the Harvard Medical Society at the Peter Bent Brigham Hospital in February of 1922 entitled: Ovarian hematomas of endometrial type (perforating hemorrhagic cysts of the ovary) and implantation adenomas of endometrial type, published that same year in The Boston Medical and Surgical Journal.⁹⁴ He elaborated on what he meant by the essentials of the implantation theory. What this writer has labeled Sampson’s first theory and second theory of implantation endometriosis, Sampson called his “first step [and] second step” in the development of the implantation theory.⁹⁵ What Sampson called “the third and final step” referred to his observations that a “secondary spread of endometriosis also could arise from foci in other situations than the ovary.”⁹⁶

In short, by approaching the history of endometriosis prospectively from early nineteenth century, the writer concluded that Sampson had developed two separate but complementary “theories” and did not recognize a third “theory.” However, to Sampson these were merely three steps in the development of his implantation theory of peritoneal endometriosis, not three separate theories.

In the introduction to this invited paper, Sampson wrote an autobiographical sketch that summarized the methodology and intellectual excitement as well as the integration of years of experience with cancer that characterized 10 years of intensive study of peritoneal endometriosis. “For over ten years I studied peritoneal endometriosis constantly and intensively, and since then intermittently according to the operative findings in individual cases. During the intensive study of this subject the distribution and character of its lesions were carefully noted at operation. Sketches were frequently made at that time. Great attention was paid to small implants. When feasible these were excised. Drawings, many in color, were made of all specimens of endometriosis before they left the operating room floor. All material was fixed intact in formalin. After fixation, I selected the exact portions of the specimens which I wished to study histologically. This tissue was embedded in celloidin, since it causes less unequal tissue shrinkage than paraffin. I supervised the mounting of the embedded tissue and instructed the technician how it should be cut. A small notebook was carried, in which I jotted down ‘inspirations’ before they vanished. Studies of the peritoneal implantation of cancer cells escaping from carcinoma of the ovary and of the body of the uterus and also studies of the spread of these tumors in other ways, were initiated by my desire to investigate more intelligently the spread of benign Müllerian mucosa. I enjoyed every bit of the study of endometriosis; there were an abundance of fresh material, excellent laboratory facilities, including well-trained technicians, an artist whose illustrations speak for themselves better than any words I might employ, a cooperative and skilled microphotographer and interested associates. My chief contribution was an insatiable curiosity which, stirred by difficulties and opportunities which were constantly arising, perpetuated my interest.”⁹⁷

Sampson chose the name endometriosis – which he intended to be an all inclusive term – because endometriosis expressed his conviction that uterine endometrial mucosa was the principal source of both internal and external endometriosis. Nevertheless, throughout this paper, as he had many times before, Sampson often reverted to the term müllerianosis – with its embryological implications that made it inclusive of all benign and malignant müllerian diseases – those identified such as adenomyosis, endosalpingiosis, and endometriosis, and those yet to be identified such as endocervicosis and developmental endometriosis.

In 1925, Sampson had described his dilemma regarding terminology and the reason for his choice of the term endometriosis. “The nomenclature of misplaced endometrial or müllerian lesions is a difficult one to decide upon. As they arise from tubal as well as uterine mucosa (probably much more frequently from the latter), the term müllerian would be inclusive and correct, but unfortunately it suggests an embryonic origin. A variety of lesions is produced by misplaced endometrial or müllerian tissue and it is difficult to classify all of them as true tumors. The term endometriosis müllerianosis would possibly be more correct, as applied to the entire subject, than ‘endometrioma and endometriomyoma,’ as suggested by Blair Bell, or ‘müllerianoma,’ as more recently suggested by Bailey. The term endometriosis is more descriptive than müllerianosis and is correct in the majority of instances, because we believe that the uterine mucosa is the chief source of these lesions.”⁹⁸

Sampson waited till his retirement before he defined the disease he had been studying for two decades: “The term endometriosis was introduced to indicate the presence of ectopic tissue which possesses the histologic structure and function of the uterine mucosa. It also includes the abnormal conditions which may result not only from the invasion of organs and other structures by this tissue, but also from its reaction to menstruation.”⁹⁹

At the end of his professional career, Sampson knew that he had provided “no positive proof” that epithelium escaping from endometriotic cysts or from the fallopian tubes implanted on the peritoneum to cause endometriosis.”¹⁰⁰ He had strong circumstantial evidence for his three-step implantation theory of peritoneal endometriosis, but not scientific proof.¹⁰¹ Sampson concluded this paper and his active academic career with an elegant disclaimer that expresses the ideal of scientific inquiry. “If bits of Müllerian mucosa carried by menstrual blood escaping into the peritoneal cavity are always dead, the implantation theory, as presented by me, also is dead and should be buried and forgotten. If some of these bits are even occasionally alive, the implantation theory also is alive. The viability of this theory is of secondary importance to me as compared with the pleasure and the increased knowledge of this and kindred subjects which I have gained in these studies and the resulting more intelligent treatment of patients who have peritoneal endometriosis.”¹⁰²

In 1943, James Robert Goodall, emeritus Professor of Clinical Gynecology and Obstetrics, McGill University, published a slender volume entitled A Study of Endometriosis, Endosalpingiosis, Endocervicosis, and Peritoneo-ovarian Sclerosis: A Clinical and Pathologic Study, which he dedicated to Dr. John A. Sampson “in recognition of his work as a pioneer in this field and as a token of gratitude for his generosity.”¹⁰³ In this volume, Goodall reported what he believed was a new benign müllerian disease, which he named endocervicosis. “Endocervicosis is a new disease, a recent discovery. It is characterized by a nonmalignant invasion of the deep cervical and paracervical tissues by the mucosa of the cervix uteri. Just as the endometrium may penetrate deeply into the uterine myometrium, and even extend into the parametrial tissues, so can the cervical mucosa under abnormal stimulation take on similar invasive properties, causing distortion and fixation of the organs involved. Only two cases have come under my observation, both in the past year. Doubtless others will be found when surgeons are familiar with the signs. The true nature of the disease was never suspected prior to operation, but at intervention the multiplicity of the cysts, their diminutive size, their opalescent tenacious contents, and their rigid walls led one to suspect their true nature.”¹⁰⁴

In 1992, Clement and Young acknowledged that Goodall was the first to use the term endocervicosis “to refer to two cases in which there was deep infiltration of the cervical wall and paracervical tissues by benign-appearing endocervical glands.” Though Goodall had not illustrated his lesions, Clement and Young noted that, by the standards of late twentieth century pathology, “they would probably be interpreted as examples of minimal deviation adenocarcinoma (“adenoma malignum”).” Clement and Young then presented, for the first time, six cases of endocervicosis of the urinary bladder.¹⁰⁵

In 1925, Sampson had enumerated four benign müllerian diseases, three of which he had seen: adenomyosis, endosalpingiosis, and endometriosis; and he postulated a fourth, developmentally misplaced endometrial tissue, which he had not seen.¹⁰⁶ Why did Sampson include developmentally misplaced endometrium in his classification of the theories of pathogenesis for endometriosis? It was not like him to report that which he had not personally observed, nor include such as a possible mode of pathogenesis of the disease process he was studying so intently.

Although Sampson gave no reference to support inclusion of developmentally misplaced endometrial tissue, I believe he was referring to the case reported by W. W. Russell in 1899 as “aberrant portions of the müllerian duct found in an ovary.”¹⁰⁷ Russell had died in 1924 while Sampson was preparing this article: Heterotopic or misplaced endometrial tissue.¹⁰⁸ Russell’s case was forcibly brought to Sampson’s attention by the obituary for Russell written by his old chief and role model, Howard Kelly. In the last sentence of the obituary, Kelly wrote: “Without any doubt, his most important contribution was a carefully made objective study of a case, the first reported, of endometrial tissue in the ovary, far reaching in its consequences in view of the later studies of Thomas S. Cullen and John A. Sampson.”¹⁰⁹

By 1943, while Cullen and Sampson were still alive, three of the five benign müllerian diseases had been described. Rokitansky had described uterine endometriosis (adenomyosis) and extrauterine (ovarian) endometriosis in 1860 and Chiari had identified endosalpingiosis (salpingitis isthmica nodosa) in 1887. Possibly four of the five benign müllerian diseases had been identified, if one will allow that the demonstration by Robert Meyer of a developmental müllerian choristoma, “islands of endometrium in the uterine wall of a fetus of nine months,” constituted developmentally misplaced endometrial tissue, developmental adenomyosis;¹¹⁰ or the case report of Russell, or if one will accept the case report of endometriosis of the kidney in the absence of pelvic endometriosis reported by Victor F. Marshall.¹¹¹

In 1897 and 1898, Robert Meyer had demonstrated non-müllerian choristomas, adrenal rests – developmentally misplaced adrenal tissue – in “the broad ligament of both fetus and adult.”¹¹² Recall also that in 1901, William Welch had diagnosed a non-müllerian choristoma, a developmental anomaly: bone and embryonic connective tissue in a uterus, i.e., developmentally misplaced non-müllerian tissue within the musculature of the uterus.¹¹³ While the above are all examples suggesting developmentally misplaced tissue, in all likelihood, Sampson was aware only of Russell’s report – from conversations with Russell at Johns Hopkins during his residency years or later at the American Gynecological Society.

Having observed the historical emergence of acquired adenomyosis, endometriosis, endosalpingiosis, and developmentally misplaced endometrium by mid-twentieth century, we return to the question posed earlier in this work. Are they separate and distinct diseases or different phenotypic expressions of the same disease process? The answer to the question is both yes and no. It seems that all acquired benign müllerian tissues, normally associated with the mature human reproductive organs, possess a latent tendency to invade. Such is the case with benign endometrial and endosalpingeal mucosa, the human blastocyst, the placenta, as well as the hydatidiform mole. In contrast, all developmental benign müllerian choristomas – normal embryonic endocervical, endometrial, and/or endosalpingeal tissue: the developmentally misplaced müllerian tissue of Sampson – tend not to invade the normal organs in which they have been misplaced during organogenesis. Menstrual effusion from developmental choristomas containing endometrium is not considered invasive.

In sum, adenomyosis and endosalpingiosis are phenotypic expressions of direct or internal endometriosis. Endometriosis and peritoneal endosalpingiosis are phenotypic expressions of indirect or external endometriosis. All are invasive acquired müllerian diseases derived from mature müllerian tissue. In the sense that they are invasive as well as acquired müllerian diseases, adenomyosis, endosalpingiosis, and endometriosis differ from noninvasive, developmental müllerian diseases. On the other hand, both acquired müllerian diseases and developmental müllerian diseases may be considered as simply different phenotypic expressions of misplaced müllerian tissue and – in that all inclusive sense – the same disease process.

The research efforts of Cullen, Novak, Meyer, Grünwald, Goodall, and Sampson, and scores of other investigators illustrate once again the difficulty of defining pathology and pathogenesis of benign chronic disease in the interior of the human body at mid-twentieth century.

Footnotes

Sampson JA. Perforating hemorrhagic (chocolate) cysts of the ovary: their importance and especially their relation to pelvic adenomas of endometrial type (“Adenomyoma” of the uterus, rectovaginal septum, sigmoid, etc.) Archives of Surgery 1921;3:245–323.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;14:422–469:422.

Sampson JA. Am J Obstet Gynecol 1927;14:422–469:422.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;14:422–469:422.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:435. See also Bailey KV. The etiology, classification, and life history of tumors of the ovary and other female pelvic organs containing aberrant müllerian elements, with suggested nomenclature. J Obstet Gynaecol Brit Emp 1924;xxxi:539–573:541. “This serosal theory held good until quite recently for all known extra-uterine “adenomyomata” and also the sub-peritoneal variety…although Cullen stuck to his view…from congenital Müllerian relics.” Bailey believed that the serosal theory held good until Sampson’s revolutionary theory of 1921. By way of explanation, to demonstrate coelomic metaplasia, histologic observations can be made under high-power magnification and potentially ultrahigh magnification at leisure in the laboratory on specimens that can be cut into serial sections for further study. Tissue could be treated with differential stains derived from bacteriological studies. In sharp contrast, clinical observations are made on living pathology in surgery. Observations are limited several ways: by the time that the patient is under anesthesia; by the intensity of illumination of surface lesions, time for observation is limited by the ethics of prolonged anesthesia for research purposes, and finally observation is limited to surface of the living lesion under observation.

Bailey KV. The etiology, classification, and life history of tumors of the ovary and other female pelvic organs containing aberrant müllerian elements, with suggested nomenclature. J Obstet Gynaecol Brit Emp 1924;xxxi:539–573:540. “Meyer later upheld this serosal –coelomic metaplasia theory [of Iwanoff], and attributed the process to a primary inflammation with secondary epithelial heterotopy or displacement.” Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:443. See caption under Figure 25, page 443. For additional references to age differential, see Figure 19, page 439; Figure 15, page 443; Figures 27 and 28 on page 445; and Figure 30, page 447.

Sampson JA. Endometrial carcinoma of the ovary, arising in endometrial tissue in that organ. Arch Surg 1925; 10:1–72.

Where Sampson produced microscopic evidence of transformation – benign to malignant endometriosis, escape of endometrial fragments into the adjacent vein (at least, the first step in metastasis, he did not show the end organ that received the metastasis) to support his theory of venous metastases, he received support and not criticism.

Daston, Lorraine and Peter Galison. Objectivity. New York, NY: Zone Books, 2007:44. As Daston and Galison described “exemplary personas,” Sampson was an intuitive expert with trained judgment. He was an “intuitive expert, who depends on unconscious judgment to organize experience into patterns in the very act of perception.” See also page 69. “The acute observer can intuit from cumulative experience, as Goethe ‘saw’ the Urpflanze,” the archetype of all flowers. Like Goethe, from cumulative experience, Sampson “saw” the archetypical theory of pathogenesis of extrauterine pelvic endometriosis; retrograde transtubal menstruation, and implantation.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:422. Sampson’s first theory of pathogenesis of peritoneal endometriosis was centered on the ovary; inspired by Casler’s unique case that had been presented by invitation before the American Gynecological Society. Shortly after he articulated his first theory, Sampson presented transtubal retrograde menstruation as his second theory of pathogenesis for endometriosis and endosalpingiosis. Sampson stated explicitly: “At the meeting of the American Gynecological Society in 1921, the writer presented a paper on perforating hemorrhagic cysts of the ovary and their relation to pelvic adenomas of endometrial type…[Sampson JA. Arch Surg 1921;iii, 245–323]…At that time I believed that the ovary was the principal, if not the only, source of the peritoneal implantations which arose from endometrial tissue disseminated by the menstrual perforation of an endometrial cyst or by the menstrual reaction of endometrial tissue on the surface of that organ.”

Meigs JV. Endometrial hematomas of the ovary. Boston Medical and Surgical Journal 1922;clxxvii:1–13:1. “In the Archives of Surgery for September 1921, Dr. John A. Sampson, of Albany, N.Y., published a paper which is the foremost contribution to gynecology and gynecological pathology in recent years.”

Bailey KV. The etiology, classification, and life history of tumors of the ovary and other female pelvic organs containing aberrant müllerian elements, with suggested nomenclature. J Obstet Gynaecol Brit Emp 1924;xxxi:539–573:541. “Sampson in 1921 revolutionized all preexisting theories as to the etiology of pelvic growths of “adenomyomatous” nature by pointing out their obvious relationship to the so-called “chocolate cysts” found in the ovaries, and described the gross and histological appearance of this ovarian condition. His work, which is well known, undoubtedly set the pathology of this pelvic condition on a sound basis.”

Cullen TS. Discussion following a Symposium on Misplaced Endometrial Tissue. Am J Obstet Gynecol 1925;10:732–733:733. “We are under a great debt to Sampson for the careful, painstaking, and brilliant work that he has done toward establishing the modes of origin of peritoneal adenomyomata.”

Bell WB. Endometrioma and endometriomyoma of the ovary. J Obstet Gynaecol Brit Emp 1922;xxix:443–446.

Sampson JA. Inguinal endometriosis (often reported as endometrial tissue in the groin, adenomyoma in the groin, and adenomyoma of the round ligament). Am J Obstet Gynecol 1925:462–503.

Bell WB. Endometrioma and Endometriomyoma of the ovary. J Obstet Gynaecol Brit Emp 1922;xxix:443–446:444–445. Referring to ovarian endometriomas, Bell states: “The credit for the discovery of this interesting pathological condition and its clinical importance is, however, entirely due to American investigators, and in particular to Sampson.” Sampson observed “that endometrium in the ovary is the cause of the so-called ‘chocolate cysts.” Bell continues: “It has been suggested, also, that the ‘cellular spill’ from such a cyst might become implanted in the pouch of Douglas.”

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469. Sampson did not mention the Casler case, which I believe was the proof case for his first theory of peritoneal implantation of endometriosis from ruptured ovarian endometriotic cysts.

Novak was a model critic; not only did he respect Sampson; he also offered constructive suggestions to Sampson that would strengthen his theory of implantation.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:437. “We might infer…might also”.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469. See pages 430, 432, and 439.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469. See pages 425, 433, and 439.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:428–431, 434, 436, 437; 458.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:425.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:448. Sampson repeatedly left space for the theory of coelomic metaplasia as he argued for his implantation theory. See figure 7, page 429; page 431; figure 31, page 448; figure 34, page 451; figure 35, page 452; figure 58, page 467. See also figure 31, page 448; this is the first instance where Sampson juxtaposed the theories of coelomic metaplasia and implantation to explain invasion of a structure (uterosacral ligament) by endometriosis. Hitherto, Sampson used coelomic metaplasia merely as an alternative explanatory model for the pathogenesis of endometriosis.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:425.

Jacobson VC. The autotransplantation of endometrial tissue in the rabbit. Arch Surg 1922;5:281–300. Jacobson VC. Further studies in autotransplantation of endometrial tissue in the rabbit. Am J Obstet Gynecol 1923;6:257–262. Jacobson VC. The intraperitoneal transplantation of endometrial tissue. Arch Path Lab Med 1926:1:169–174.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:424. “The purpose of this paper is to present the evidence indicating the origin of peritoneal endometriosis from the implantation of endometrial tissue disseminated by menstruation.”

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:432. “Many interesting endometrial lesions were present in an endometriosis of the posterior vaginal wall of the second case. An endometrial cavity filled with menstrual blood and containing bits of endometrial tissue had almost eroded the overlying vaginal mucosa and was about to rapture and discharge its menstrual contents into the vagina. From the study of this lesion, one could readily understand how a similar endometrial cavity in the ovary [Casler] or any other pelvic structure might rupture and disseminate its menstrual contents into the peritoneal cavity.” Interestingly, the only other time that Sampson addressed endometriosis in this area (the so-called rectovaginal septum) was in his first article on endometriosis published in 1921. Sampson JA. Perforating hemorrhagic (chocolate) cysts of the ovary: their importance and especially their relation to pelvic adenomas of endometrial type (“Adenomyoma” of the uterus, rectovaginal septum, sigmoid, etc.) Archives of Surgery 1921;3:245–323:245. See also: Bailey KV. The etiology, classification, and life history of tumors of the ovary and other female pelvic organs containing aberrant müllerian elements, with suggested nomenclature. J Obstet Gynaecol Brit Emp 1924;xxxi:539–573. See also Bailey KV. The etiology, classification, and life history of tumors of the ovary and other female pelvic organs containing aberrant müllerian elements, with suggested nomenclature. J Obstet Gynaecol Brit Emp 1924;xxxi:539–573:541. Bailey cites Lockyer work of 1917: “The prevailing view at the present time…is to regard adenomyoma of the recto-genital space as an inflammatory product and not a true neoplasm.”

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:423. See Figure 1.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:433. “I can see only one correct interpretation of the etiology of the embolic endometrial lesions in the veins about these endometrial cavities and that is they arose from the implantation of endometrial tissue, disseminated into the veins from the menstrual rupture of the walls of the endometrial cavities into these vessels. If so endometrial issue disseminated by menstruation in this instance must have been alive and capable of growing, when transferred to suitable situations.” See also page 437. See another reference to the proof case on page 443.

Casler DB. A unique, diffuse uterine tumor, really an adenomyoma, with stroma, but no glands. Menstruation after complete hysterectomy due to uterine mucosa in remaining ovary. Trans Am Gynecol Soc. 1919;44:69–84:78–79. “Microscopic examination at once reveals that we are dealing here with an ovarian cyst made up almost entirely of uterine tissue, the interior of the cyst corresponding to the uterine cavity and filled with blood while the walls contain many normal glands and others which show glandular dilatation. A pathological change has also occurred and we have an overgrowth of the interglandular stroma, much resembling that seen four years previously in the uterus….The entire cyst, or uterine cavity, as it really is, is lined throughout by a single layer of tall columnar epithelium of the uterine type, and in places cilia can be made out.”

Sampson used photomicrographs to serve as objective evidence. See Daston, Lorraine, and Peter Galison. Objectivity. New York, NY: Zone Books, 2007:164. Daston and Galison quoted Wilhelm His, the embryologist from Leipzig who held that drawings and photographs were complementary. His stated: “The photograph reproduces the object with all its particularities, including those that are accidental, in a certain sense as raw material, but which guarantees absolute fidelity.” As a youth, Sampson had trained himself to draw and it was he who drew the drawings in this paper. Again His: “In every sensible drawing, the essential is consciously separated from the inessential and the connection of the depicted forms is shown in the correct light, according to the view of the draftsman.” In other words, drawings were subjective and unretouched photographs were objective.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:431. See Figure 9.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469. For references to ectopic endometrial cavities, See Figure 1, page 423 for an example of “submucous endometrial cavity.” See also Figure 8, page 430; Figure 11, page 433, figure 13, page 435; and Figure 20, page 440 for examples of “endometrial cavity of the ovary.”

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:430. For an excellent example, see Figure 8, page 430.

Sampson JA. President’s Address before the American Gynecological Society, Hot Springs, Va., May 22, 1923. Fundamental elements in the advancement of medicine. Am J Obstet Gynecol 1923;6:1–11. This address is an autobiographical statement.

Novak E. The significance of uterine mucosa in the fallopian tube, with a discussion of the origin of aberrant endometrium. Am J Obstet Gynecol 1926;xii:501, 503. See also Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:430.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:443. Emil Novak was a model critic. Not only did he respect Sampson, he also offered constructive suggestions that would strengthen Sampson theory of implantation. Late in the twentieth century, investigators tested the viability of cast off endometrium in tissue culture. See: Willemsen WNP, Mungyer G, Smets H, Rolland R, Vemer H, Jap, P. Behavior of cultured glandular cells obtained by flushing of the uterine cavity. Fertil Steril 1985;44:92. Kruitwagen RFPM, Poels IG, Willemsen WNP, de Ronde IJY, Jap PHK, Rolland R. Endometrial epithelial cells in peritoneal fluid during the early follicular phase. Fertil Steril 1991;55:297.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:433. See footnote 9 for the quotation.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:443. Figure 25.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:439. Figure 19.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469. For Geographic arguments that consistently located endometrial lesions of the ovary on the lateral and undersurface of the ovary see: See Figure 2, page 424; Figure 18, pages 438; Figure 19, page 439; Page 440; Figures 23 and 24 on page 442; Figure 25, page 443; Figure 26, page 444; Figure 29, page 446; Figure 32, page 449; Figure 43, page 458; Page 462; and Figure 54 on page 465.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:446. See Figure 29, page 446.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:462.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:443. See caption under Figure 25, page 443. For additional references to age differential, see Figure 19, page 439; Figure 15, page 443; Figures 27 and 28 on page 445; and Figure 30, page 447.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:451. See legend under Figure 34. See page 461. “It would seem that during the menstrual life of women some substance escapes from the tubes into the pelvis which plays an important role in the etiology of pelvic peritoneal endometriosis, including the development of endometrial tissue in the ovaries. This substance may be menstrual blood in some instances and tubal secretions in others. In either case epithelium may be present.”

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:452. See Figure 35.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:452. See Figure 36.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:452. See Figure 36. See also Figures 48 and 49. Implants of endometriosis on ovary and uterus “are in a situation readily ‘soiled’ by material escaping from the abdominal ostium of the patent left tube.” I believe I am familiar with the trees, soil, and landscapes from which Sampson drew his metaphors. For I have hiked, camped, or hunted the woodlands and meadows of New York State from the Adirondack State Park to the ravines that feed the Finger Lakes; and from the Great Bear Swamp in Western New York to the Mohawk and Hudson Rivers near Sampson’s Albany.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:462.

This is a mixed metaphor; Galileo’s experiments at Pisa were real, Newton’s apple is an allegorical story.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469. See pages 434–435; page 440 and page 462.

Sampson JA. The escape of foreign material from the uterine cavity into uterine veins. Am J Obstet Diseases of Women and Children 1918;lxxxii:161–175.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:444.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:456. Figure 40.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:456. Figure 41.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:456. Figures 40 and 41.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:457. Figure 42.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:460. Figure 46. “Should the endometrial tissue in the photomicrograph above it be forced into a narrow, curved tube, it would be moulded or fashioned into a mass like this. I believe that this moulded mass represents a cast of the lumen of a narrow portion of the tube and adds to the evidence already presented that the endometrial tissue in the tube or tubes was derived from the uterine cavity.” Again, Sampson qualified his statements that the evidence suggests but does not prove.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:461. See Figure 47.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:446. “I have often observed blood dripping from the abdominal ostia of the tubes in abdominal operations which had been preceded by a curettage of the uterus.” Sampson also observed the same phenomenon when he operated during the patient’s menstruation. See also page 458. “I fully realize that it is impossible definitely to state the origin of the blood in the lumen of the tubes in these eight cases.”

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:446.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:459. See Figures 44 and 45.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:446.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:449. “Novak has published photomicrographs of sections of tube showing endometrial tissue in their lumina and offers the theory that this tissue might have come from ectopic endometrial tissue in the pelvis and entered the tube through the abdominal ostium, just as the ovum and particles of cancer in peritoneal carcinosis are known to enter the tubes.”

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:447.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:451.

Janney JC. Report of three cases of a rare ovarian anomaly. Am J Obstet Gynecol 1922;Feb:173–187:187.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:449. For example of the association of endometriosis with patent fallopian tubes, see Figure 32, page 449. See also page 459. “One of the outstanding features of patients with peritoneal endometriosis is that the tubes are usually patent.” In a five year period, Sampson operated 293 patients with peritoneal endometriosis; both tubes “appeared to be patent” in 284 patients. In Sampson’s series, most patients were over 30 years old. Three patients had a unilateral hematosalpinx, four had a bilateral hematosalpinx, and only two had bilateral pyosalpinx.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:438. See Figure 16.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:462–463.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469. See Figures 21 and 22, page 441. See also Page 436: “cancer escaping into the peritoneal cavity sometimes becomes implanted on the surface of the peritoneum, causing lesions of peritoneal carcinosis.”

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:437.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:438–439. Fig. 18.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:438–439. Fig. 18.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:453. See Figure 37.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:453. Fig. 37.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469.:454. Fig. 38.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469. See Fig. 38, page 454 and Fig. 39 on page 455.

Sampson JA. Ovarian hematomas of endometrial type (perforating hemorrhagic cysts of the ovary) and implantation adenomas of endometrial type. Boston Medical and Surgical Journal 1922;186:445–456. After Sampson published his first theory in the American Journal of Obstetrics and Gynecology in 1921 and before he published his second theory in the Boston Medical and Surgical Journal in 1922, he annotated his first theory to announce the second, which was published separately in the Proceedings of the American Gynecological Society in 1921.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:463.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:463–464. Sampson reproduced two photomicrographs objectively illustrating his argument. See Figures 59 and 60 on page 468.

Sampson JA. Peritoneal endometriosis due to the menstrual dissemination of endometrial tissue into the peritoneal cavity. Am J Obstet Gynecol 1927;l4:422–469:464.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:465.

Sampson JA. Am J Obstet Gynecol 1927;l4:422–469:466.

Daston, Lorraine, and Peter Galison. Objectivity. New York, NY: Zone Books, 2007:344–346.

Daston, Lorraine, and Peter Galison. Objectivity, 36. “What is the nature of objectivity? First and foremost, objectivity is the suppression of some aspect of the self, the countering of subjectivity. Objectivity and subjectivity define each other, like left and right or up and down.” See also page 51. “Objective knowledge,” understood as “a systematized theoretical account of how the world really is, comes as close to truth as today’s timorous metaphysics will permit.”

Clement PB. History of gynecological pathology. IX. Dr. John Albertson Sampson. Int J Gynecol Pathol 2001;20:86–101. See pp. 99–100 references. References 2 through 68 all refer to Sampson’s articles arranged chronologically from 1902 to 1945.

Sampson JA. The development of the implantation theory for the origin of peritoneal endometriosis. Am J Obstet Gynecol 1940:40:549–557:557.

Sampson JA. The development of the implantation theory for the origin of peritoneal endometriosis. Am J Obstet Gynecol 1940:40:549–557:557. See also p. 549. “I greatly appreciate the appraisals of my observations and interpretations which have been made by others, However, it is my own critical evaluation of these observations and interpretations which I shall attempt to present in this review.”

Sampson JA. Ovarian hematomas of endometrial type (perforating hemorrhagic cysts of the ovary) and implantation adenomas of endometrial type. Boston Med Surg J 1922;186:445–456. See p. 448 where Sampson casually inserted an observation about lymphatic dissemination of adenomatous tissue, anticipating Halban’s publication by 2 years.

Sampson JA. The development of the implantation theory for the origin of peritoneal endometriosis. Am J Obstet Gynecol 1940:40:549–557:550, 553.

Sampson JA. Am J Obstet Gynecol 1940:40:549–557:555.

Sampson JA. The development of the implantation theory for the origin of peritoneal endometriosis. Am J Obstet Gynecol 1940:40:549–557:549.

Sampson JA. Heterotopic or misplaced endometrial tissue. Am J Obstet Gynecol 1925;10:649–664:649.

Sampson JA. The development of the implantation theory for the origin of peritoneal endometriosis. Am J Obstet Gynecol 1940:40:549–557:549.

100

Sampson JA. Am J Obstet Gynecol 1940:40:549–557:555.

101

Sampson JA. Am J Obstet Gynecol 1940:40:549–557:555–556. Henry Campbell Black, Black’s Law Dictionary: Definitions of the Terms and Phrases of America and English Jurisprudence, Ancient and Modern. Sixth Edition by the Publisher’s Editorial Staff. [St. Paul, MN: West Publishing Co., 1990], 243. “Circumstantial evidence. Testimony not based on actual personal knowledge or observation of the facts in controversy, but of other facts from which deductions are drawn, showing indirectly the facts sought to be proved. People v. Yokum, 145 C.A.2n 245, 302 P.2d 406, 410. The proof of certain facts and circumstances in a given case, from which jury may infer other connected facts which, usually and reasonably follow according to the common experience of mankind. Foster v. Union Starch & Refining Co., 11 Ill.App.2d 346, 137 N.E.2d 499, 502. Evidence of facts or circumstances from which the existence or nonexistence of fact in issue may be inferred. Inferences drawn from facts proved. Process of decision by which court or jury may reason from circumstances known or proved, to establish by inference the principal fact. It means that existence of principal facts is only inferred from circumstances. Twin City Fire Ins. Co. v. Lonas, 255 Ky. 717, 75 S.W.2n 348, 350. The proof of various facts or circumstances which usually attend the main fact in dispute, and therefore tend to prove its existence, or to sustain, by their consistency, the hypothesis claimed. Or as otherwise defined, it consists in reasoning from facts which are known or proved to establish such as are conjectured to exist.” See also: Robert S. Hunter, Federal Trial Handbook: Civil. 4^th ed. [Danvers, MA:Thomson West, 2003], 29–1. “Admissibility of circumstantial evidence generally. Circumstantial evidence is evidence which tends to prove a disputed fact by proof of other facts which have a legitimate tendency to lead the mind to a conclusion that the fact exists which is sought to be established. It is legal evidence and the jury must act upon it as if it were direct when it is satisfactory beyond a reasonable doubt. Rumely v. U.S., 293 F. 532 (C.C.A. 2n Cir. 1923). As a legal matter, there is no distinction between direct and circumstantial evidence. Circumstantial evidence has probative value equal to that of testimonial evidence. The law does not belittle the value of circumstantial evidence by making a relative distinction between it and direct evidence. Rodella v. U.S., 286 F.2d 306 (9^th Cir. 1960).”

102

Sampson JA. The development of the implantation theory for the origin of peritoneal endometriosis. Am J Obstet Gynecol 1940:40:549–557:557.

103

James Robert Goodall, A Study of Endometriosis, Endosalpingiosis, Endocervicosis, and Peritoneo-ovarian Sclerosis: A Clinical and Pathologic Study [Philadelphia: JB Lippincott, 1943].

104

James Robert Goodall, A Study of Endometriosis, Endosalpingiosis, Endocervicosis, and Peritoneo-ovarian Sclerosis, 89. In 1982, the writer dedicated a monograph to James Robert Goodall. Ronald E. Batt, and John D. Naples, Conservative Surgery for Endometriosis in the Infertile Couple. Current Problems in Obstetrics and Gynecology, Vol. VI [Chicago: Year Book Medical Publishers, 1982], 1-98. See Leitch. Proc. R. Soc Med (Obst and Gyn Sect), July 1914:389. Cited by Cuthbert Lockyer, Fibroids and Allied Tumours (Myoma and Adenomyoma): Their Pathology, Clinical Features, and Surgical Treatment [London: Macmillan and Company, 1918], 330–331. Archibald Leitch described a “migratory adenomyoma” of the cervix that spread into the base of the left broad ligament. “The cervical epithelium had invaded the musculature of the cervix very deeply.” This description sounds like endocervicosis, a lesion that was not described and named as such until 1943 by Goodall. Leitch anticipated Cullen but did not recognize the significance of his observation. See Clement PB, Young RH. Endocervicosis of the urinary bladder. A report of six cases of a benign müllerian lesion that may mimic adenocarcinoma. Am J Surg Pathol 1992;16:633–42. On page 538 in a footnote, the authors’ noted: “The term “endocervicosis” was first used in 1943 by Goodall to refer to two cases in which there was deep infiltration of the cervical wall and paracervical tissues by benign-appearing endocervical glands. The pathology of the lesions was not illustrated, but by current criteria they would probably be interpreted as examples of minimal deviation adenocarcinoma (“adenoma malignum”). See also, Young RH, Clement PB. Endocervicosis involving the uterine cervix: a report of four cases of a benign process that may be confused with deeply invasive endocervical adenocarcinoma. Int J Gynecol Pathol 2000;19:322–8. “In adenoma malignum, the lesional glands originate from the mucosa and grow downward, whereas in cases of endocervicosis, the glands usually are confined to the outer aspect of the cervix. When there is a clear demarcation between normal endocervical glands and the more deeply seated glands of endocervicosis, with a zone of uninvolved cervical wall, it should be readily apparent that the deep glands cannot represent glands that have spread from the mucosa.”

105

Clement PB, Young RH. Endocervicosis of the urinary bladder. A report of six cases of a benign müllerian lesion that may mimic adenocarcinoma. Am J Surg Pathol 1992;16:533–542.

106

Sampson JA. Heterotopic or misplaced endometrial tissue. Am J Obstet Gynecol 1925;10:649–664:649–650. Of historical interest, in this publication, Sampson equated the following terms for adenomyosis: adenomyosis = direct endometriosis = primary endometriosis = müllerianosis. Thereafter, Sampson abandoned such usage for müllerianosis. The writer appropriated the abandoned term müllerianosis in 1990 and thereafter, to designate Sampson’s developmentally misplaced endometrial, endosalpingeal, and endocervical tissue. Ronald E. Batt, Richard A. Smith, Germaine M. Buck, Mark F. Severino, and John D. Naples, “Müllerianosis,” In Current Concepts in Endometriosis. Progress in Clinical and Biological Research, edited by D. Chadha and V. Buttram. [New York: Alan R. Liss, 1990], 323:413–426. (Proceedings of the Second International Symposium on Endometriosis, Houston, Texas, 1-3 May 1989). Ronald E. Batt, Smith RA, Buck Louis GM, Martin DC, Chapron C, Koninckx PR, Yeh J. Müllerianosis. Histology and Histopathology 2007; 22:1161–1166.

107

Russell WW. Aberrant portions of the müllerian duct found in an ovary. Bull Johns Hopkins Hospital 1899;10:8–10 plus 3 plates. Russell was the physician who displaced Cullen from the residency position promised by Howard Kelly; a displacement that lead to Cullen’s spending 3 years in pathology before he entered the gynecological residency.

108

Sampson JA. Heterotopic or misplaced endometrial tissue. Am J Obstet Gynecol 1925;10:649–664:649–650.

109

Kelly HA. William Wood Russell, M.D. 1866–1924. Trans Am Gynecological Society 1924;49:383–384.

110

Meyer R. Autobiography of Dr. Robert Meyer (1864–1947): A Short Abstract of a Long Life [New York: Henry Schuman, 1949], 33.

111

Marshall VF. Endometrial tissue in the kidney. J Urology 1943;50:652. This was the same Victor Marshall of the Marshall, Marchetti, Krantz operation for stress urinary incontinence in women. Certainly, Marshall was familiar with female pelvic pathology.

112

Robert Meyer, Autobiography of Dr. Robert Meyer (1864–1947): A Short Abstract of a Long Life [New York: Henry Schuman, 1949], 34. “In 1897 and 1898, I demonstrated some malformations and accessory adrenal tissue in the broad ligament of both fetus and adult. (The continuation of these studies is found in Zeitschr. f. Geb. u. Gyn. 41). Robert Meyer, Autobiography of Dr. Robert Meyer (1864–1947): A Short Abstract of a Long Life [New York: Henry Schuman, 1949], 86.” “Cysts in the Vaginal Wall of Fetuses of 3–5 months.” Arch. f. Gyn.151 [1932]. “Those cysts from the Müllerian epithelium lie in the median plane as do also the cysts in the uterus.”

113

William Henry Welch, Pathology, Preventive Medicine, vol. 1 of Papers and Addresses by William Henry Welch [Baltimore, MD: Johns Hopkins Press, 1920], 432–3. Johnston, George B. Osteo-Fibromyoma of the Uterus. Am Gynaec & Obst. J., N.Y., 1901, XVIII, 307–308.

Page

Contents

If you find an error or have any questions, please email us at admin@doctorlib.org. Thank you!